análisis genético y molecular de la inmunidad a la …análisis genético y molecular de la...
TRANSCRIPT
Análisis genético y molecular de la inmunidad a la bacteriosis
vascular en yuca (Manihot esculenta Crantz) mediante estrategias
de mapeo genético y transcriptómica por RNA-seq
Genetic and molecular analysis of the inmmunity to cassava
bacterial blight through genetic mapping and RNA-seq
approaches
Johana Carolina Soto Sedano
Universidad Nacional de Colombia
Facultad de Agronomía, Escuela de Postgrados
Bogotá, Colombia
2016
Análisis genético y molecular de la inmunidad a la bacteriosis
vascular en yuca (Manihot esculenta Crantz) mediante estrategias
de mapeo genético y transcriptómica por RNA-seq
Genetic and molecular analysis of the immunity to cassava
bacterial blight (Manihot esculenta Crantz) through genetic
mapping and RNA-seq approaches
Johana Carolina Soto Sedano
Tesis o trabajo de investigación presentada(o) como requisito parcial para optar al
título de:
Doctor en ciencias agrarias
Director:
PhD. Camilo Ernesto López Carrascal
Codirectora:
Ph.D. Adriana Jimema Bernal Giraldo
Línea de Investigación:
Genética y fitomejoramiento
Grupo de Investigación:
Manihot Biotec. Departamento de Biología
Universidad Nacional de Colombia
Facultad de Agronomía, Escuela de Postgrados
Bogotá, Colombia
2016
V
A Mauricio y María Alejandra por y para
ustedes
VI
Agradecimientos
Empiezo por agradecer a mi director de tesis, Dr. Camilo Ernesto López Carrascal por
toda su confianza, apoyo, dedicación, por sus enseñanzas tanto académicas como
personales y su comprensión en los momentos en que más lo necesité. Es usted un
formador en todo el sentido de la palabra. Me siento muy afortunada por haber
contado con su dirección y haber sido parte de su grupo de investigación.
A mi codirectora, Dra. Adriana Ximena Bernal y a mi comité doctoral, Dra. Liliana
López y Dra. Luz Stella Barreto, por la guía recibida, apoyo durante las actividades
académicas y por permitirme contar con su incondicional cooperación.
Un especial agradecimiento a la Dra. Teresa Mosquera, por su apoyo durantes estos
años tanto académico como personal, por contagiarme con su entusiasmo hacia la
ciencia, por su gestión para la realización de mi pasantía en Alemania y por acogerme
como una integrante más de su grupo de investigación.
A la Universidad Nacional de Colombia Facultad de Ciencias Agrarías, postgrado en
genética y mejoramiento por mi formación académica a través de excelentes
docentes, y por el apoyo recibido durante la duración del programa doctoral.
A la Universidad Nacional de Colombia y al Departamento Administrativo de Ciencia,
Tecnología e Innovación (COLCIENCIAS), por la financiación de esta investigación a
través del proyecto 521-2011 y por la financiación de mi formación doctoral, a través
de la convocatoria de doctorados Nacionales 528 de 2011.
Al instituto INRES “Institute of Crop Science and Resource Conservation”, Bonn,
Alemania, especialmente al Dr. Agim Ballvora, Dr. Boby Mathew y Dr. Jens Leon, por
permitirme vivir una de las mejores experiencias de mi vida y por el apoyo y guía
durante mi pasantía de investigación.
Al grupo Manihot Biotec, en especial a Paula, Juan Camilo, Mariana, Lina, Andrea,
Ruben, Paola y Zapata, gracias por su amistad y por tantas discusiones
enriquecedoras. A los estudiantes Rubén, Paola y Marly gracias por toda su
colaboración durante largas y agotadoras jornadas de trabajo de campo.
A la Universidad Nacional de Colombia sede Orinoquía y al señor Lisímaco López, por
permitirme desarrollar parte del trabajo de campo en sus instalaciones y por toda la
colaboración prestada.
VII
Al CIAT, especialmente al Dr. Luis Augusto Becerra por facilitarme el material vegetal
y a la Dra. María Cristina Duque por su instrucción inicial en el análisis de datos y por
unas interesantes tardes de discusión sobre QTLs.
Finalmente, al todopoderoso, a mi querida familia, a mis padres y hermanas Tata y
Polo, por sus enseñanzas y amor es que he logrado alcanzar las metas propuestas.
A mi amado esposo por su apoyo y por seguirme incondicionalmente con amor y
paciencia en todos mis sueños.
A mi hija María Alejandra, eres la más grande manifestación del amor de Dios en mi
vida, gracias por impulsarme a ser tu mejor ejemplo.
VIII
Resumen
La yuca es uno de los cultivos más importantes a nivel mundial. Una de las
enfermedades que más compromete su producción es la bacteriosis vascular (CBB
por sus siglas en Ingles) causado por Xanthomonas axonopodis pv. manihotis (Xam).
La mejor manera de controlar esta enfermedad es la siembra de variedades
resistentes, obtenidas a través de mejoramiento genético tradicional o por
transformación genética. No obstante, en la actualidad no hay reportes de la
clonación de genes relacionados con inmunidad para CBB. Uno de los pasos
importantes hacia el aislamiento de genes es la construcción de mapas genéticos
altamente densos que permitan la clonación posicional. Aquí se presenta el desarrollo
de uno de los mapas genéticos de yuca más densos. Este mapa se obtuvo a través de
la aplicación del enfoque de genotipificación por secuenciación (GBS por sus sigas en
Ingles), el cual permitió la obtención de miles de marcadores moleculares de
polimorfismo de un solo nucleótido (SNP por sus siglas en Ingles). Estos SNPs fueron
obtenidos y evaluados en la progenie de una población F1 segregante resultado del
cruce entre TMS30572 y CM2177-32. En total se identificaron 78,854 SNPs que
cubren el 87% (463.2 Mb) del genoma de yuca. El set completo de SNPs se evaluó
para parámetros de mapeo y aquellos SNPs de alta calidad se seleccionaron para
construir el mapa de ligamiento. El mapa cubre 2,571 cM distribuidos en 18 grupos
de ligamiento con una distancia promedio entre marcadores de 1.26 cM. Este mapa
fue usado para la detección de loci de un caracter cuantitativo (QTL por sus siglas en
Ingles) para la resistencia a CBB. La población de mapeo fue evaluada para la
resistencia a dos cepas de Xam (Xam318 y Xam681) en dos localidades en Colombia:
La Vega (Cundinamarca) y Arauca (Arauca). La evaluación se realizó durante las
épocas de lluvia y sequía. Una tercera evaluación fue realizada bajo condiciones de
invernadero. Adicionalmente, la población fue evaluada bajo condiciones naturales
de infección en Puerto López (Meta) durante una época de lluvia. A través de mapeo
de QTL se identificaron 18 QTL que explican entre el 10.9 y el 22.1% de la varianza
fenotípica. De estos QTL nueve mostraron estabilidad entre las épocas de evaluación.
Se detectaron interacciones QTL x ambiente significativas para diez de los QTL.
Dentro de los intervalos de los QTL se describió un repertorio de 151 genes
candidatos relacionados con defensa a CBB (CDRGs por sus siglas en Ingles), de los
cuáles trece corresponden a genes que codifican proteínas que contienen dominios
representativos de proteínas de inmunidad. Cuatro CDRGs mostraron expresión
diferencial durante la infección por Xam681 en el parental resistente TMS30572. El
repertorio de CDRGs que co-localizan con QTL representa una fuente de regiones
genómicas nuevas involucradas en la resistencia a CBB, que puede ser explorado y
validado para su uso futuro dentro de programas de mejoramiento de yuca.
IX
Palabras clave: Loci de un caracter cuantitativo, resistencia cuentitativa,
Xanthomonas axonopodis p.v. manihotis, genotipificación por secuenciación, mapa
genético.
Abstract
Cassava is one of the most important crops world-wide. One of the diseases
compromising its production is the cassava bacterial blight (CBB) caused by
Xanthomonas axonopodis pv. manihotis (Xam). The best way to control this disease is
growing resistant varieties obtained through traditional breeding or by genetic
transformation. Nevertheless, currently there are no reports of the cloning of
immunity related gene to CBB. One important step toward the isolation of genes is
the construction of high dense genetic maps allowing the positional cloning. Here we
present the development of one the highest dense genetic maps of cassava. This map
was obtained through the use of the Genotyping by Sequencing (GBS) approach
allowing the generation of thousands of Single Nucleotide Polymorphisms (SNPs).
These SNPs were evaluated in the F1 segregating progeny resulted from cross
between TMS30572 and CM2177-32. A total of 78,854 SNPs were identified covering
87% (463.2 Mb) of the cassava genome. The total set of SNPs was evaluated for
mapping parameters and high quality SNPs were selected to construct the linkage
map. The map covered 2,571 cM distributed in 18 linkage groups and includes 2,141
SNPs with an average distance of 1.26 cM between markers. This map was used to
perform QTL (Quantitative Trait Loci) detection for CBB resistance. The F1 mapping
population was tested for resistance to two Xam strains (Xam318 and Xam681) at
two locations in Colombia: La Vega (Cundinamarca) and Arauca (Arauca). The
evaluation was conducted during rainy and dry seasons. A third evaluation was
conducted on greenhouse conditions. Additionally, the population was evaluated
under natural infection conditions at Puerto López (Meta) during a rainy season.
Through QTL mapping, 18 strain-specific QTLs were detected, explaining between
10.9 and 22.1% of the phenotypic variance. From these QTL, nine showed stability
between the evaluated seasons. A significant QTL x Environment interaction was
detected for ten QTL. Within the QTL intervals were described a repertoire of 151
CBB candidate defense-related genes (CDRGs), from which thirteen correspond to
genes coding for proteins containing domains representative of the immunity
proteins. Four CDRGs show differentially expression during Xam681 infection in the
resistant parental TMS30572. The repertoire of CDRGs co-localizing with the QTL
reported here, represents a source of novel genomic regions involved in CBB
X
resistance to be explored and validated for its future use into cassava breeding
programs.
Keywords: Quantitative trait loci, quantitative disease resistance, Xanthomonas
axonopodis p.v. manihotis, genotyping by sequencing, genetic map.
XI
Table of contents
Pag.
Resumen ................................................................................................................................................... VIII
Abstract ........................................................................................................................................................ IX
List of figures ........................................................................................................................................... XIV
List of tables ............................................................................................................................................... XV
List of abbreviations ............................................................................................................................. XVI
Introduction .............................................................................................................................................. 19
Objectives ................................................................................................................................................... 22
CHAPTER 1 ................................................................................................................................................. 23
Review of related literature................................................................................................................. 24 Cassava classification and origin ..................................................................................................... 24 Biology and reproduction ................................................................................................................... 25 Diversity .................................................................................................................................................... 27 Global and national cassava production ....................................................................................... 28 Uses .................................................................................................................................................... 29 Cassava breeding programs............................................................................................................... 31 Cassava genome ...................................................................................................................................... 35 Pest and diseases .................................................................................................................................... 36 Cassava Bacteria Blight ........................................................................................................................ 39 The causal agent: Xanthomonas axonopodis pv. manihotis .................................................. 39 Etiology and disease incidence ......................................................................................................... 40 Xam diversity ........................................................................................................................................... 41 Xam genome ............................................................................................................................................. 42 ABC of plant immunity ......................................................................................................................... 43 Quantitative resistance ........................................................................................................................ 47 Molecular interaction cassava-Xam ................................................................................................ 49 Molecular basis of the pathogenecity ............................................................................................ 49 Molecular basis of resistance to CBB ............................................................................................. 50 Mapping the quantitative resistance to CBB ............................................................................... 51 Improving CBB resistance .................................................................................................................. 54 References ................................................................................................................................................. 55
CHAPTER 2 ................................................................................................................................................. 74
RNA-seq: herramienta transcriptómica útil para el estudio de interacciones planta patógeno ..................................................................................................................................................... 75
Resumen .................................................................................................................................................... 75 Abstract .................................................................................................................................................... 75 Introducción ............................................................................................................................................. 76 Tecnología RNA-seq .............................................................................................................................. 78 Plataformas y estrategias de secuenciación para RNA-seq .................................................. 79 Estrategias y consideraciones para experimentos RNA-seq ................................................ 83
XII
Aplicaciones enfocadas al estudio de interacciones planta patógeno ............................. 86 Antecedentes del uso de RNA-seq en interacciones planta patógeno ............................. 89 Conclusiones, retos y perspectivas ................................................................................................ 92 Referencias ............................................................................................................................................... 93
Unraveling the molecules hidden in the gray shadows ........................................................... 101 Abstract ................................................................................................................................................. 101 Introduction........................................................................................................................................... 102 The ABC of plant immunity ............................................................................................................. 103 Quantitative resistance enters into the game .......................................................................... 104 How to study complex traits and QDRs ..................................................................................... 105 A new era for QDR studies: phenotyping has the last word .............................................. 107 From theory to practice: QDR in breeding ................................................................................ 108 Molecular explanation of quantitative resistance ................................................................. 110 QDR as a continuous response that depends ongene expression intensity ................ 111 R weak alleles........................................................................................................................................ 113 Allelic variation .................................................................................................................................... 114 Kinases and signaling ........................................................................................................................ 115 Miscellaneous........................................................................................................................................ 116 Conclusions ............................................................................................................................................ 117 References .............................................................................................................................................. 118
CHAPTER 3 ............................................................................................................................................... 126
A genetic map of cassava (Manihot esculenta Crantz) with integrated physical mapping of immunity-related genes ................................................................................................................. 127
Abstract ................................................................................................................................................. 127 Introduction........................................................................................................................................... 128 Materials and methods...................................................................................................................... 131 Results ................................................................................................................................................. 134 Discussion............................................................................................................................................... 150 Acknowledgments............................................................................................................................... 153 References .............................................................................................................................................. 154 Supplementary data ........................................................................................................................... 161
CHAPTER 4 ............................................................................................................................................... 163
Novel genetic factors involved in cassava bacterial blight resistance detected through QTL analysis ............................................................................................................................................ 164
Abstract ................................................................................................................................................. 164 Introduction........................................................................................................................................... 165 Materials and methods...................................................................................................................... 168 Results ................................................................................................................................................. 171 Discussion............................................................................................................................................... 184 Acknowledgments............................................................................................................................... 189 References .............................................................................................................................................. 189 Supplementary data ........................................................................................................................... 197
CHAPTER 5 ............................................................................................................................................... 200
XIII
QTL identification for cassava bacterial blight resistance under natural infection conditions. ................................................................................................................................................ 201
Abstract ................................................................................................................................................. 201 Introduction .......................................................................................................................................... 202 Materials and methods ..................................................................................................................... 204 Results ................................................................................................................................................. 205 Discussion .............................................................................................................................................. 208 Acknowledgments .............................................................................................................................. 210 References .............................................................................................................................................. 211 Supplementary data ........................................................................................................................... 214
General discusion .................................................................................................................................. 216
General conclusions and perspectives ........................................................................................... 236
Publications and presentations ........................................................................................................ 238 Publications ........................................................................................................................................... 238 Oral presentations in scientific events ....................................................................................... 238 Poster presentations in scientific events .................................................................................. 239
XIV
List of figures
Pag Figure 2-1. Secuenciamiento masivo de ADNc, RNA-seq, por las tecnologías NGS Illumina y 454 81 Figure 2-2. Model in which the expression level of QDR genes is associated with the resistance phenotype 113 Figure 3-1. Cassava genetic map containing 2,141 markers 137 Figure 3-2. Summary of mapped annotated SNPs 138 Figure 3-3. Anchor markers showing co-linearity between different cassava genetic maps 139 Figure 3-4. Repertoire of genes coding for immune related proteins (IRPs) identified in the cassava genome 143 Figure 3-5. Orthology clusters between of the predicted immunity-related proteins in Manihot esculenta, Arabidopsis thaliana, Ricinus communis, Populus trichocarpa 147 Figure 3-6. The cassava genetic and physical map enriched with IRPs and QTLs for cassava disease resistance 149 Figure 4-1. Evaluation of parental responses to different bacterial Strains 173 Figure 4-2. Which Won Where/What graphic of GGE-Biplot analysis 179 Figure 4-3. QTL x environment interaction based on additive phenotypic effects (APE) 182 Figure 4-4. Gene expression of genes that co-localizes with QTLs in resistant parental against Xam681 183 Figure 5-1. Histogram of the distribution of the disease index values obtained in the field evaluation of the response to CBB 207 Figure 5-2. QTL detection for field resistance to CBB in linkage groups 4 and 8 by non-parametric interval mapping 207
XV
List of tables
Pag.
Table 3-1. Genetic map data summary 137
Table 3-2. Comparative analysis of cassava physical maps pag 142
Table 3-3. Relationships between genetic and physical maps,
representative for each linkage group and for the whole genome 144
Table 4-1. Codes for the localities, strain and seasons where the
inoculation and phenotyping was conducted 174
Table 4-2. Distribution of AUDPC values in the mapping population 177
Table 4-3. Pairwise Pearson correlation coefficients between
AUDPC values 177
Table 4-4. Better-parent heterosis 178
Table 4-5. Summary of QTL associated to CBB resistance 180
Table 5-1. Summary of QTL detected for field resistance to CBB 208
XVI
List of abbreviations
Abbreviation Meaning
AM Association mapping
APE Additive phenotypic effect
AUDPC Area under the disease progress
AFLPs Amplified fragment length polymorphism
BLUP Best linear unbiased prediction
CBB Cassava bacterial blight
CBSD Cassava brown streak disease
CCMD Cassava common mosaic disease
CDRG Candidate defense-related gene
CDS Coding DNA sequences
CEBiP Chitin elicitor-binding protein
CFSD Cassava frog-skin disease
CIAT Center for Tropical Agriculture
CIM Composite Interval Mapping
CMD Cassava mosaic disease
CVMD Cassava vein mosaic disease
DDPSC Donald Danforth Plant Science Center
DI Disease index
ECZs Edaphoclimatic zones
EF-Tu Elongation factor Tu
EST Expressed sequence tag
eQTLs Expression-QTLs
ETI Effector-triggered immunity
ETS Effector-triggered susceptibility
FAO The Food and Agriculture Organization
CBB Cassava bacterial blight
GBS Genotyping by sequencing
GEBV Genomic estimated breeding value
GO Gene ontology
GS Genomic selection
GWAS Genome wide association study
G x E Genotype x environment interaction
HR Hypersensitive response
ICGMC International Cassava Genetic Map Consortium
IITA International Institute of Tropical Agriculture
ILTAB International Laboratory for Tropical Agricultural Biotechnology
XVII
IM Interval Mapping
IP Pathogen invasion patterns
IPR IP receptors
IPTR IP-triggered response
IRG Immunity related genes
IRP Immunity-related proteins
LPS Lipopolysaccharides
LRR Leucine rich repeats
MAMP Pathogen associated molecular patterns
MAS Marker assisted selection
MAPK Mitogen-activated protein kinases
MARS Marker-assisted recurrent selection
miRNA microRNA
mQTLs Metabolite-QTLs
NBS-LRR Nucleotide binging site–Leucine reach repeats
NCBI National Center for Biotechnology Information
NGS Next generation sequencing
PAMP Microbe associated molecular patterns
PCR Polymerase chain reaction
pQTLs Protein-QTLs
PR proteins Pathogenesis-related proteins
PRR Pathogen recognition receptors
PTI PAMP- triggered immunity
QDR Quantitative disease resistance
QTL Quantitative trait loci
Q x E QTL x environment interaction
RGA Resistance gene analogues
RFLP Restriction fragment length polymorphism
RLP Receptor like proteins
RPKM Per kilobase of exon per million mapped reads
RILs Recombinant inbred lines
RLK Receptor like Kinase
ROS Reactive oxygen species
SAGE Serial analysis of gene expression
SAR Systemic acquired resistance
SNPs Single nucleotide polymorphism
SNV Single nucleotide variations
SSR Simple sequence repeats
TALE Transcription activator-like effectors
T3E Type III effector
XVIII
UTR Un-translated regions
WGS Whole genome shotgun
Xam Xanthomonas axonopodis pv. manihotis
Xoo Xanthomonas oryzae pv. oryzae
Xop Xanthomonas outer proteins
19
Introduction
Cassava, Manihot esculenta Crantz, is one of the most important crops worldwide and
is considered an essential crop for food security in developing countries. Cassava
represents the staple food for about 1,000 million people (FAO, 2013). Several
diseases compromise the production of cassava. One of the most limiting bacterial
diseases is cassava bacterial blight (CBB) caused by Xanthomonas axonopodis pv.
manihotis (Xam). CBB has been reported in all regions where cassava is grown
(López and Bernal, 2012). To provide options for effective protection to the cassava
crop it is imminent the requirement of the implementation of strategies aimed to
deepen the knowledge of the CBB and finding new ways to identify resistance
sources to this disease. The best alternative to control CBB is through the use of
resistant varieties, developed by traditional breeding or in the future by genetic
transformation. However, so far, any immunity related gene to CBB has been cloned
to be incorporated into breeding programs.
Cloning genes by positional mapping has been the main strategy used in both model
and crop plants (Jander et al., 2002). This strategy has been particularly valuable in
the cloning of resistance genes (Bent, 1996; Pflieger et al., 2001; Gebhardt et al.,
2007). Positional cloning requires the development of genetic maps (Collard et al.,
2005; Pflieger et al., 2001). Ideally these maps should contain the maximum number
of markers (high resolution). However, in the past, DNA-based molecular markers
technologies allowed the developing of low saturated genetic maps, usually with
large gaps between markers due the low coverage at genome level of the molecular
technologies. Also, usually these markers were anonymous with no knowledge of the
corresponding sequence. These facts hinder gene cloning because positional cloning
requires short intervals where there are located the loci responsible for the trait or
quantitative trait loci (QTL). By reducing the QTL interval, the strategy of fine
mapping can be implemented (evaluation of thousands of recombinant individuals
for the interest region), followed by physical mapping (Salvi and Tuberosa, 2005).
However, this represents a time consuming approach. Nevertheless, with the advent
of next generation sequencing new genotyping tools have been generated for
detection of thousands of single nucleotide polymorphism (SNPs) markers in
mapping populations, contributing in developing dense genetic maps, which carry
high number of non-anonymous molecular markers (Davey et al., 2011; Glaubitz et
al., 2014; Takagi et al., 2013). These maps have established close relationships
between markers and QTL (Davey et al., 2011), helping the subsequent identification
of genes involved in the trait of interest.
20
Resistance to CBB has been described as quantitative, with polygenic inheritance
(Jorge et al., 2001, 2000). Thus as in any other quantitative trait, the classical
procedure to study its genetics is through QTL mapping (Salvi and Tuberosa, 2005).
Several QTL for CBB resistance have been identified under greenhouse (Jorge et al.,
2000; Wydra et al., 2004; López et al., 2007) and field conditions (Jorge et al., 2001).
From these studies 31 QTL have been detected which explain 7.2% to 62% of the
resistance to Xam strains (Wydra et al., 2004; López et al., 2007). Some candidate
genes such as RXam1 and RXam2 have been identified from the QTL detected as key
elements in CBB resistance (López et al., 2003). RXam1 is a candidate gene that codes
a Receptor like Kinase (RLK) protein, associated to a QTL that explains 13% of the
resistance to Xam strain CIO136 (Lopez et al., 2007). While RXam2 is a gene coding
for a Nucleotide binging site–Leucine reach repeats (NBS-LRR) protein which co-
localizes with a QTL that explains the 62% of the resistance to Xam CIO151 strain.
The gene expression of both RXam1 and RXam2 has been tested during Xam infection
(López et al., 2007; Contreras and Lopez, 2008). A part of these two examples no
other candidate resistance gene to CBB have been identified so far. The generation of
a high dense cassava genetic map and the identification of new QTL associated to CBB
resistance will contribute in an important manner to identify this kind of genes in
cassava.
This thesis aims to contribute in identifying genomic regions associated with
resistance to CBB. The results obtained and the knowledge generated during the
development of this thesis are exposed here and presented under a four chapters-
format. In the first chapter, a theoretical framework regarding to the cassava crop,
the CBB, the mechanisms of the resistance to CBB and the status of cassava breeding
programs are shown. The second chapter presents a theoretical background of
different relevant topics for the development of the thesis. The first topic of this
second chapter consists in an exploration of the bases, applications and advantages of
the RNA-seq technology, as well as a discussion of the studies revealing the
importance and usefulness of this tool in the study of plant pathogen interactions.
This gave the conceptual basis for the results obtained during this thesis concerning
the differential gene expression analysis performed to a repertoire of candidate
genes putatively involved in CBB defense. This information was presented through a
review published in 2012 in the Fitosanidad journal. Also, in this chapter, it is
presented a critical review on the bases of quantitative resistance, its importance, the
recent efforts to elucidate the molecular bases and some studies exhibiting new plant
immunity models. This enriched the literature revision and deepened the current
status of quantitative resistance studies. This information is presented as a review
submitted to the Molecular Plant-Microbe Interactions journal.
21
The third chapter presents the development of a high-density genetic map of cassava
obtained through genotyping by sequencing (GBS). The first genetic and physical
mapping of a repertoire of genes related to immunity in cassava is presented in this
work. These results were valorized through a scientific publication in 2015 in the
journal BMC Genomics. The fourth chapter presents the effort to characterize novel
genetic factors involved in CBB resistance through QTL and RNAseq approaches. To
accomplish this goal a phenotyping evaluation was conducted in a segregant mapping
population after inoculation with two different Xam strains. This evaluation was
conducted under field conditions during rainy and dry seasons in two Colombian
locations and under greenhouse controlled conditions. The transcription profiles of
the candidate genes located within the QTL were obtained during Xam infection in
the resistant parental through RNA-seq approach. These results are part of a
manuscript in preparation to be submitted to Molecular Plant-Microbe Interactions
journal. Finally, as integral part of this chapter, is presented the identification of
novel QTL for CBB resistance detected under natural infection conditions. This study
was performed in Puerto Lopez (Meta), Colombia and the main results were
valorized in a scientific manuscript submitted to the journal Acta Biológica
Colombiana. At the end of this document a general discussion searching to integrate
the different results obtained, as well as to highlight the goals achieved and their
limitations is presented. This discussion includes also a general view of the
perspectives and the contribution of the knowledge generated through this thesis to
the breeding programs addressed to generate cassava varieties showing CBB
resistance.
22
Objectives
General objective
To develop a SNP-based high-density genetic map of cassava through next generation
sequencing approach and to identify genomic regions related with resistance QTL to
cassava bacterial blight.
Specific objectives
Determine contrasting defense responses in the parental of a cassava mapping
population against a collection of Xam strains.
Construct a SNP-based high-density genetic map of cassava through genotyping by
sequencing approach.
Identify resistant QTL to cassava bacterial blight.
23
CHAPTER 1
“The way to get started is to quit talking and begin doing”
– Walt Disney
24
Review of related literature
Cassava classification and origin
Cassava belongs to the Euphorbiaceae family, which includes more than 300 genera
and 8,000 species (Webster, 1994). The apomorphic (derived states) of the
Euphorbiaceae is characterized by being a flowering plant family, having laticifer cells
producing milky secretions (Aristizabal and Sanchez, 2007). It can be found as herbs
or shrubs (Webster, 1994). Several species of the genus Manihot have been described
as toxic due the high accumulation of cyanogenic glucosides, especially linamarin and
lotaustralin, compounds that can be found in leaves and roots (Ceped and Mattos,
1996). Additionally, to M. esculenta, in the Euphorbiaceae family can be found other
species of economic importance such as castor oil plant (Ricinus communis) and
rubber tree (Hevea brasiliensis).
The origin of Manihot esculenta Crantz (1766) has been extensively studied and
debated. Some authors recognize Brazil as the origin center where exist around 80
species of the Manihot genus. However, other authors consider that there is not
enough archaeological evidence to confirm this origin. Central America, especially
Mexico has been considered as possible origin center, where seventeen species of
Manihot exist (Clement et al., 2010). In addition in this country were found leaves
from cassava plants that were cultivated 2,500 years ago. Despite the debate,
nowadays is accepted the Amazonian basin as the origin center and Mexico as an
important center of diversity. M. esculenta is considered the most important food
crop that originated in the Amazonas (Clement et al., 2010). The common name of
this species is “cassava” term that comes from the Arawak word cazabi that means
bread (Lebot, 2009).
25
Some of the difficulties in establishing the origin of cassava are the large area of
distribution of wild subspecies and the fact that M. esculenta has been considered a
cultigen (plant species that has been selected artificially by humans) (Allem, 1994;
Nassar et al., 2008) which originated from several introgression events among wild
species. The species M. aesculifolia and M. carthaginensis, from Mexico and Colombia-
Venezuela respectively, were proposed as the wild relative of M. esculenta species.
However, some molecular studies propose that Manihot esculenta subsp. flabellifolia
is the wild progenitor of cassava (Olsen and Schaal, 1999, 2001). Another species that
has been considered as the wild relative of M. esculenta is Manihot glaziovii. Recently
through genome wide analysis of wild, cultivated and cassava-related species has
been revealed interesting facts regarding to the origin of this species (Bredeson et al.,
2016). Based on genome sequencing of cassava related species it was demonstrated
the high prevalence of haplotypes of M. glaziovii in the current African and South
America cassava cultivars even for different named accessions which are really near
clones and so far unknown (Bredeson et al., 2016).
Regarding to the domestication origin, it is believed that domestication of some
cassava cultivars occurred 6,000 B.C. in the Amazonian rainforests (Gibbons, 1990).
The cassava arrived to Africa from Brazil in the 16th century, through navigators that
travel to the west coast of Africa (Jones, 1959), and was in the 17th century when,
through African traders, initiated the cassava local expansion. Nowadays cassava is
for Africans an essential food and an important part of its culture. On the other hand,
in Asia the crop was introduced from South America by Spanish explorers, as a
famine reserve crop and a source of starch for commercial exploitation (Hershey et
al., 2000). Even today, in Thailand the Asian country with the highest cassava
production, the consumption as a food is very low with almost 90% of the production
destined to industry uses (Hershey et al., 2000; FAO, 2015).
Biology and reproduction
Cassava is an allogamous species (Da Silva et al., 2003), and as the majority of species
of the Manihot genus, is protogeneous, with pistillate flowers open before staminate
flowers of the same inflorescence (Nassar et al., 2008). This fact favors the cross-
pollination and thus the development of extremely heterozygous genotypes. Cassava
is a monoecious plant, with relatively small female flowers at the base of the
branched panicle while the male flowers are located at the tip. The flowering time as
well the fertilization is high dependent of both genetics and environment conditions
26
(Halsey et al., 2008). In Colombia the natural pollinator of cassava are wasps of the
genus Polistes, while in Africa is the honeybees Apis mellifera (Kawano, 1980).
Cassava has a diploid genome with 2n=36 and sexual reproduction, but for
production farmers use vegetative propagation by stem cuttings (De Carvalho and
Guerra, 2002; Raji et al., 2009; Sakurai et al., 2013). The buds initiate to sprout 5 to 8
days after planting (dap) and 10 to 13 dap appears the first true leaves (El-Sharkawy,
2004). The sprouting capacity and rootlets formation seems to be governed
genetically because is variety-dependent (El-Sharkawy and Cock, 1987).
The distribution of cassava is mainly throughout the tropics from Mexico to northern
Argentina (Rogers and Appan, 1973). The crop requires a warm weather with a day
temperature for optimum growth above 20°C. The maximum leaf photosynthesis rate
is obtained with temperatures between 25 to 35°C (El-Sharkawy et al., 1992).
However, there are some cassava cultivars which can be cultivated in tropical high
altitudes (>1800 m.a.s.l) (El-Sharkawy, 2004). Cassava also tolerates drought and has
the ability to grow in acid and low fertility soils (Aristizabal and Sanchez, 2007).
Molina and El-Sharkawy (1995), demonstrated that nutritional reserves contained in
the stem cuttings are more important than the fertility of the soil for an optimal
sprouting.
A special photosynthetic status can be found in cassava, which has a C4
photosynthetic cycle, however, due to the lack of the typical Kranz anatomy,
considerable carbon assimilation proceeds through the Calvin-Benson cycle (Cock et
al., 1987; El-Sharkawy, 2004). In cassava has been shown that the activities of
photosynthetic enzymes are considerable affected by water stress (El-Sharkawy,
2004), showing that under long exposure to water stress the activity of the C4
Phosphoenolpyruvate carboxylase is favored over the C3 Rubisco.
Cassava roots are rich in carbohydrates, containing between 250 to 300 kg/Tn and
very low in fat and protein content ranging 5 to 19 gr/kg of dry root matter (El-
Sharkawy, 2004). Cassava provides moderate vitamins and minerals, especially
potassium, magnesium, calcium and iron. The leaves, which are used for human
consumption in some countries of Africa and Asia, contain relatively high contents of
ascorbic acid and carotene (Diaz, 2012).
27
Diversity
The diversity of the genus Manihot is relatively high with 70 to 100 cassava wild-
related species. From these species farmers have developed new cassava varieties
over centuries trough traditional breeding (FAO, 2010). In South America exists
several diversity centers of cassava. The first one is Brazil, where around 80 species
are present. Mexico is the second most important with 17 species reported (Clement
et al., 2010) and according to the last report on the state of the world´s plant genetic
resources for food and agricultural (FAO, 2010), Asia and Africa are also important
diversity centers of cassava.
The use of few genotypes as parents in the African breeding programs has produced
a reduction of cassava´s genetic diversity (Bredeson et al., 2016). However, in recent
years big efforts, mainly led by International Institute of Tropical Agriculture (IITA),
have been done in order to improve and increase the germoplasm variability of
cassava. In Asia the cassava genetic diversity seems to be narrow compared with
Africa and Latin America. However, in Thailand the elite varieties, mainly for industry
uses, are more diverse than the African ones (Fu et al., 2014).
Most of the knowledge of the current status of the diversity of cassava comes from
the use of molecular marker techniques. Several studies using simple sequence
repeats (SSR), amplified fragment length polymorphism (AFLPs) and more recently
SNPs, have contributed with the determination of the genetic variability of cassava
materials. In South America the genetic diversity, based on the Shannon´s index,
seems to be relatively high ranging from 0.80 to 4.20 (Colombo et al., 2010; Mezette
et al., 2013). While in Africa ranges from 0.95 to 1.25 (Kawuki et al., 2013). This low
genetic diversity seems to be the reflection of a bottleneck effect in the African
cultivars.
In the 70s the International Center for Tropical Agriculture (CIAT), located in Cali,
Colombia, began an initiative to collect and conserve phylogenetic resources of
cassava from all over the world. Nowadays CIAT holds the largest cassava collection
with 5,709 accessions that represents the 17% of the cassava´s world resources,
which includes more than 32,000 accessions (FAO, 2010; https://ciat.cgiar.org/).
CIAT´s collection is considered the most important in the world, not only for the high
number of materials conserved, but also for the genetic diversity and geographic
28
regions that it represents (https://ciat.cgiar.org/). This collection includes entries
from Colombia (37.7%), Brazil (24.1%), and other South American countries
(21.2%), Central America and Caribbean (7.2%) and Asia (7.1%). The remaining
2.5% of accessions comes from other countries (https://ciat.cgiar.org/). This
collection is conserved in an in vitro condition and a back-up is maintained by the
International Potato Center in Lima (FAO, 2013). CIAT also maintains an especial
collection of 833 accessions that represents the wild species of the Manihot genus
(https://ciat.cgiar.org/). This collection is a potential source of genetic variability
useful to be employed in breeding programs. The Brazilian Agricultural Research
Corporation (Embrapa) and the IITA, located in Ibadan, Nigeria, also maintain two of
the most important collections of cassava, with 2,800 and 2,900 accessions
respectively. Other collections are also conserved in Benin, India, Indonesia, Malawi,
Nigeria, Thailand, Togo and Uganda (FAO, 2013).
Global and national cassava production
Cassava is grown in more than 100 countries (Taylor et al., 2012). In 2014, the global
area cultivated of cassava was 17.1 million hectares, with a production of 9.7 million
tons more than 2013 (FAO, 2015). The major production came from Nigeria, Brazil
and Thailand. In 2014 the production was more than 57 million tons, increasing
around 1 million tons compared to 2014. In Thailand, the Asia’s largest producer, the
total production for 2015 was more than 34 million tons, reaching record yields
(FAO, 2015). In Brazil the cassava production in 2014 was around 23 million tons,
maintaining the yield trend of 2012 (FAO, 2015).
Colombia is ranked eighteenth in the worldwide production and third in Latin
America after Brazil and Paraguay, with an average of annual production of 2.5
million tons (FAO, 2015). The total production of cassava in Colombia for 2014 was
2.3 million tons, from which 70.5 thousand corresponds to cassava growing for
industry purposes. From 2012 to 2014 the production increased 16.5%, an
outstanding growth compared to other crops (FAO, 2015). The total planted area was
223 thousand hectares. The departments with the largest planted areas and
production are Bolívar, Córdoba and Magdalena, with 17.4%, 13% and 9.5% of the
participation in National production, respectively. Sucre is the Colombian
department that leads the production of cassava for industry uses, with an area
planted of almost 4,000 hectares (http://www.agronet.gov.co/). Sucre is also the
department that leads the industrial cultivation of cassava with 3,644 hectares of
29
planted area, producing 70,540 tons in 2014, the 3% of national production
(http://www.agronet.gov.co/).
From 2012 to 2014 the global cassava productivity has been increasing by 4%, this
rate outstrips the majority of the staple crops and is more than the world population
growth (FAO, 2015). Despite the increasing trend in cassava production, the weather
adversities threat the world cassava production (FAO, 2015). For 2016 it is uncertain
for instance the effect that El Niño could have in cassava production (FAO, 2015).
Uses
Cassava is an essential crop for food security in developing countries; provides more
carbohydrate content than other crops such as potato and represents the staple food
for more than 1,000 million people in Africa, Asia and South America (FAO, 2013).
Human consumption is the major use of cassava. Paraguay is the first consumer of
fresh cassava in the world and in several African countries the human diet is mainly
based on it. Regarding to the taste, cassava can be divided in bitter and sweet; both
can be consumed as boiled, fried or roasted “rale” roots like potatoes (Steenkamp et
al., 2014), as flour for several preparations and several fermented foods and
beverages, including cassava beer and wine (Ray and Sivakumar, 2009) and the
popular Taiwanese bubble tea made from tapioca (cassava starch) pearls (Nicolau et
al., 2015). Most recently, in Uganda a new trade brand of beer was developed form
cassava ethanol (Impala) (Steenkamp et al., 2014). Some indigenous traditional foods
are made from both, cassava leaves and roots, especially in the Amazon region. These
are the “casabe” or “cassava bread” and “farinha de agua”, a toasted granular food,
which is called “garí” in Africa. For the preparation of casabe, the natives prefer bitter
varieties. Part of the tradition is the great technical skill they have to remove from
these varieties almost 97% of the cyanogenus compounds in a typical preparation of
casabe that takes 48 hours (Dufour, 2006). From cassava starch is produced several
bakery products such as “pandebonos”, “pandeyuca” (Colombia), “chipas” (Paraguay)
and bread “couac”, biscuits and cakes (Steenkamp et al., 2014). From cassava also it
can be produced different kind of chips such as the dried chips or sticks (Taiwo,
2006). Despite its high consumption cassava provides calories through its
carbohydrates but little nutrition, given as a result nutritional deficits when the diet
is exclusively based on it. The cassava roots offer a smaller portion of the daily
requirement of protein, and minerals such as iron, zinc and vitamin A (Sayre et al.,
2011; Talsma et al., 2016).
30
Even though the consumption of cassava leaves has been related with toxicity
because its high cyanogenic glucosides contents (Mlingi et al., 1992) and also that is
considered as a non-conventional food, at least for human consumption, since long
ago the cassava leaves have been recognized as a good nutritional alternative
(Lancaster and Brooks, 1983; Ufuan et al., 2005; Aletor, 2010). With an appropriated
cooked method, based on heat treatment, the potentially toxic components can be
removed up to 99% (Ufuan et al., 2005). Thus the cassava leaves can become an
important source of proteins (18% – 40%), vitamin A and B, as well as minerals
especially calcium, potassium, phosphorus and iron (Ufuan et al., 2005; Aletor, 2010).
Several efforts have been established in order to increase the nutritional value of
cassava through bio-fortification projects. In 2004, the HarvestPlus challenge
program arose with the objective to create varieties with high content of provitamin
A, protein, zinc, iron, as well as decreased the amount of cyanogens (linamarin) in
cassava, mainly through transgenic strategies (Sayre et al., 2011). Genetically
improved cassava has been obtained for several traits such as protein contain
(increase of fourfold with respect to the control) (Abhary et al., 2011), decrease of
cyanogen content (Siritunga and Sayre, 2003); high levels of carotenoid (up to 5.7
μg/fresh weight) (Failla et al., 2012) and increasing in zinc levels (Sayre et al., 2011).
The cassava crop also has important applications in the industry, mainly from its
starch which is used in a diverse range of products (Ospina and Ceballos, 2002) and it
is considered the cheapest one (FAO, 2013). The 80% of dried weight of a cassava
root is starch (Olomo and Ajibola, 2003). Some of the principal uses of cassava starch
are as a stabilizing agent for various foods in food industry; in pharmaceutical
industry for pill coating; in textile industry for the rubberizing of cotton-based cloths
(Aguilera, 2012), for biodegradable plastic and films (Larotonda et al., 2004; FAO,
2015) and in biofuel production. The ethanol from cassava is undoubtedly one of the
most attractive products for the energy industry. The demand of this biofuel (mainly
from Asia) boosted the growth of this crop worldwide. In average, from one ton of
cassava roots (30% percent of starch content) can be obtained around 280 liters of
96% pure bioethanol (FAO, 2015).
In Colombia, the principal cassava use is the human consumption. However since
2013 the government incentives the planting of the industrial cassava for the
production of flour, starches, and preparation of concentrate food for animal
consumption. This sector has been strengthened with a yield increase of more than
200% from 2012 (19,488 Ton/year) to 2014 (70,540 Ton/year)
(http://www.agronet.gov.co/).
31
Cassava breeding programs
The cassava as the majority of the food crops of economic importance has been
constantly selected from thousands of years by man for the improvement of their
genetic potentials. It is very likely that the first cassava plant breeders were the
aboriginal farmers, during the domestication of the crop, 6,000 B.C. in the Amazonian
rainforests (Gibbons, 1990). They should have chosen cassava materials, vegetative
propagated, with superior qualities such as the production of thick roots and high
number of roots by plant, and then preserved it for the next planting season. In fact,
the clonal selection seems to be the first strategy for cassava improvement applied by
man (Montaldo and Gunz, 1985).
During the 70s, the IITA located in Nigeria and CIAT in Colombia, became two of the
most important institutions where different breeding programs have been
implementing with successful results which allowed the generation of materials with
high starch content (Sanchez et al., 2009), high cooking quality (Asare and Safo-
Kantanka, 1997), low cyanogenic compounds in roots (Dixon et al., 1994), high
protein content in roots (Nassar and Sousa, 2007), low post-harvest deterioration
(Rudi et al., 2010), bio-fortification (Pfeiffer and McClafferty, 2007) and resistance to
pests and diseases (Hahn et al., 1979; Okogbenin et al., 2007).
As occurs in wheat and maize, the conventional cassava breeding schemes are based
on the production of F1 populations of full-sibs and half-sibs generated through
directed crosses between elite materials carrying desired traits or elite materials
with wild relatives. This genetic recombination by sexual crosses can be performed
by hand or through open pollination. The full-sib families usually are obtained by
through “controlled” manually pollinations, while by open pollinations are obtained
half-sib families which are produced in polycross nurseries.
In general, the cassava breeding scheme starts with the selection of varieties carrying
several interest traits as well as the selection of promising wild relative species
carrying a desirable trait that is absent in the cassava variety. These materials can be
act as parents for a first designed cross between them. From the donor parent (male)
it is obtained the pollen which is then transferred to the acceptor parent (female)
previously emasculated (removal of male reproductive structures). Once the
fertilization succeeds, the development of the seed takes around two months,
however the seed production in cassava seems to be low (Fregene et al., 2001), as
32
well as its germination rate (Hahn et al., 1973). In controlled pollinations can be
obtained from one to three (in trilocular ovary) viable seeds per fruit (Jennings,
1963; Ceballos et al., 2004). The seeds can become contaminated, making difficult its
germination. Without seed treatments the percentage of germination ranges from
10% to 40%, and takes from 2 to 4 months (Hahn et al., 1973). Some efforts have
been done in order to tackle this problem. The IITA for example has developed a
protocol based on insecticide and fungicide treatments, nursery cares, precise water
and temperature supply, increasing the germination of hybrid seed up to 80% (Hahn
et al., 1973).
The hybrids resulted of the F1 cross are grown in greenhouses and then transplanted
to the field where enters to a mass recurrent selection process based on their
phenotypic characteristics (Jennings and Iglesias, 2002; Ceballos et al., 2012). Then
multiple rounds of selection are performed. These consist in clonal evaluation trials
(preliminary, advanced and two regional yield trials). After these multiple rounds of
selection, those individuals carrying the major number of interesting and desirable
traits introgressed by the variety plus the desirable trait transferred by the wild
relative will be selected for a backcross with the elite cassava variety. It is expected
that some seedlings product of this cross have all desired traits (those from the elite
variety and the one from the wild relative). In order to fix the trait, it is possible to
perform multiple cycles of backcrosses (Nassar and Ortiz, 2010), although inbred
depression have been observed. Finally, those promising materials carrying the
desirable traits can be selected for its use in further breeding schemes as a parents or
evaluated them in stability tests, which ideally has to be done under a wide range of
environments and cultural practices (Hahn et al., 1973).
Despite that the conventional breeding scheme based on crosses between promising
materials has been the traditionally strategy followed by several food crops
improvement (Borlaug, 1983), it has not been an easy task in cassava. Some
challenges have to be faced within the breeding programs, mainly due to the biology
of the crop.
The first challenge that has to be solved in a cassava breeding program is the low
multiplication rate. From one plant can be obtained around eight stem cuttings
(Ceballos et al., 2012), thus to get the initial materials intended for the designed
crosses could be time consuming. Also, the flowering time in cassava is an issue. This
characteristic is highly genotype and environment-dependent. Even it has been
reported that some cassava materials never shown flowering. However, recently,
some strategies are under development to induce flowering in cassava
(http://nextgencassava.org/). In flowering varieties, once the plant blossom, the
33
branching appears. However, the varieties with erect and non-branching features are
those that are favored by farmers; thus incorporate elite varieties in crossing
schemes may become challenging due to the scarceness of flowers (Ceballos et al.,
2012). For those cassava clones that bloom, the lack of synchronic flowering is an
important issue. The flowering time in cassava has a wide range of time from 4 to 10
months after planting. This can increase the time to obtain seeds (Ceballos et al.,
2012).
In the cassava donator parental (male) also some challenges have been described.
The cassava male sterility has been highly reported for some varieties (Magoon et al.,
1968; Hahn et al., 1973). The pollen grains of cassava lose their viability some hours
after the pollen anthesis (Halsey et al., 2008). In fact, cassava breeders do the
pollinations no more than one hour after they collect the pollen, in order to increase
the chances of success in the fertilization process (P. Chavarriaga and N. Morante,
personal observation, 2005 in Halsey et al., 2008).
Another challenge that has to be overcome within cassava breeding programs is the
protogeneous nature of the crop, a mechanism which has been related to the
prevention of self-fertilization (Narbona et al., 2011). Despite that self-pollination in
cassava is possible, a high inbreeding depression has been reported (Fregene et al.,
2001; Rojas et al., 2009; de Freitas et al., 2016). This represents an important issue
for plant breeding schemes which commonly search for an increase in homozygosis.
In cassava, Ceballos et al (2015) considered that to get homozygous lines can take up
to 15 years. Some applications that could be applied in cassava through inbreeding
are the detection of useful and undesirable recessive traits, executing back-cross as
well as reciprocal recurrent selection schemes and get recombinant inbred lines
(RILs).
Over the last decade, the development of techniques in molecular biology has
contributed to the detection of loci and isolation of genes responsible for the most
economically important traits. Also molecular biology has assisted in the
identification of promissory materials carrying desirable loci for its future use in
breeding programs, and thus has helped to accelerate the crop genetic improvement.
The selection of plants carrying desirable traits in early plant stages, without the
necessity to wait until see the individual phenotype, can be achieved through the
detection of molecular markers developed from the sequence of the responsible
genes if it is known. Otherwise, the identification of molecular markers associated to
a particular trait can be employed. For example, to select resistant individuals to a
particular pathogen or abiotic stress, it is not necessary to expose the plant to the
34
stress because with the molecular analysis it is possible to anticipate the plant
response. This approach is known as marker assisted selection (MAS) and can reduce
considerably the time required for the selection of promissory materials within a
breeding program (Xu and Crouch, 2008).
Some successfully examples of the use of molecular markers within breeding cassava
programs are RME1 and NS158, which are associated with the CMD2 gene conferring
resistance to cassava mosaic disease (CMD). RME1 and NS158 have been used in MAS
strategies for CMD resistance selection in Latin American breeding programs
(Okogbenin et al., 2007, 2012). The first cassava variety that was selected using this
strategy was UMUCASS33, which was released in Africa in 2010 (Ferguson et al.,
2012). Also, the cultivars TMS 97/2205 and TMS 98/0505 selected for high CMD
resistance and stability in several regions of high CMD pressure represent other
examples of breeding lines selected by MAS (Okogbenin et al., 2012) with great
potential to become elite varieties.
Although the MAS strategy has been successful in some cases, for those traits
governed by multiple genes of small effect, this strategy has not been effective. In
consequence, an alternative approach, named genomic selection (GS), has been
developed recently. The GS allows selecting plant material carrying desirable genes
in early stages, but taking into account whole genome molecular markers associated
with the trait. The objective behind the GS is to foretell the phenotype of an individual
using a prediction model based on the genomic estimated breeding value (GEBV).
This GEBV is the criteria of selection and is obtained from whole genome molecular
markers that are associated to the interest trait and from the genotypic and
phenotypic data obtained in “training populations” (representative germplasm)
(Jannink et al., 2010). For the prediction of GEBV, several models have been
developed. The most representative models are the best linear unbiased prediction
(BLUP) (Henderson, 1984), Bayesian models (Gianola and Fernando, 1986) and
linear mixed models with pedigree data (Crossa et al., 2010) and previous QTL
(Barabaschi et al., 2016).
The implementation of GS in cassava breeding programs can increase the rate of
cassava genetic improvement by predicting the phenotype of materials before they
reach the field and accelerates the breeding cycle. The application of GS in cassava is
currently underway and is focus mainly on traits such as shorten the cassava
breeding cycle, to improved cassava flowering (nextgencassava.org), prediction of
shoot weight, dry matter content, fresh root yield, starch amylose content and starch
yield (de Oliveira et al., 2012). Preliminary results of the use of GS in cassava
improvement have shown high levels of accuracy in the predicting models; that range
35
from 0.67 for dry matter content to 0.83 for shoot weight (de Oliveira et al., 2012),
and highlights the reduction to almost half of the time that is required for these trait
selection compared to phenotype selection.
Cassava genome
The cassava genome size was studied for first time in 1994 by Awoleye et al., who
established that the amount of DNA in cassava is 1.67pg, corresponding to a haploid
genome size equivalent of 772 Mpb. Other authors had suggested an allopolyploid
(Umanah and Hartmann, 1973) and polyploid (Hahn et al., 1990) nature in cassava.
However, a karyotype analysis of 27 cassava accessions as well as the flow cytometry
analysis revealed the diploid status 2n=36 of the cassava genome (Awoleye et al.,
1994).
The initial attempt to get the complete cassava genome sequence started in 2003
within the Global Cassava Partnership (GCP- 21), a project led by the International
Laboratory for Tropical Agricultural Biotechnology (ILTAB) at the Donald Danforth
Plant Science Center (DDPSC), and CIAT in Palmira, Colombia. In 2009 a cassava
sequence was obtained using a whole genome shotgun (WGS) approach. The cassava
accession used was MCOL1505 (AM560-2), a partial inbred cultivar generated at
CIAT (Prochnik et al., 2012). A genome of 532.5Mb was obtained in 12,977 scaffolds,
showing the 68.9% of the haploid genome size. From the total scaffolds, 487
represent half of the assembled genome.
The current cassava genome draft version (v6.1) corresponds to Illumina based
sequencing from the same cassava accession used in the v4.1 (AM560-2). This
version has 2,001 unanchored scaffolds and 10,976 scaffolds anchored in 18
chromosomes, through the guide of a GBS-based high resolution genetic map
carrying 22,403 SNP markers (International Cassava Genetic Map Consortium
(ICGMC), 2015). The genome spans 582.28Mb and the 96.2% are represented in 317
scaffolds. A total of 33,033 loci correspond to protein coding transcripts and 8,348
alternative transcripts were described. More than 78,000 cassava ESTs (expressed
sequence tag), published in the National Center for Biotechnology Information
(NCBI), were mapped and annotated in this genome (https://phytozome.jgi.doe.gov).
Recently, several genomes of wild and cultivated cassava have been obtained,
annotated and compared (Wang et al., 2014; Bredeson et al., 2016). These studies
have given knowledge regarding the evolution and domestication of M. esculenta. The
36
majority of the draft cassava genomes seem to be a hybrid of M. esculenta and M.
glaziovii, with exception of the chromosomes 1 and 18 whose right arms proceed
exclusively from M. glaziovii. This fact shows the introgression of M. glaziovii into
several African and South American cassava accessions which should take place
during cassava domestication or within breeding programs (Bredeson et al., 2016).
The high heterozygosity of cassava, compared with crops such as potato (The Potato
Genome Sequencing Consortium, 2011), has been revealed based on the number of
single nucleotide variations (SNV) as well as insertions and deletions (InDels) located
in the genome. Comparative cassava genome analysis has shown the high percentage
of repetitive sequences, which ranges from 23% to 51% of the total assemblies
(Wang et al., 2014; Bredeson et al., 2016). The total of SNVs in the cassava genome
goes from 1.3 to 4.1 million, while the total of InDels ranges from 97 to 326 thousand
(Wang et al., 2014). The number of microRNA (miRNA) contained in the genome, has
also been estimated from 20 to 68 families (Zeng et al., 2009; Pérez et al., 2012;
Patanun et al., 2013; Wang et al., 2014). Recently, some bioinformatics analysis have
provided information regarding to the identification and locations of more than 120
clusters containing around one thousand immunity related genes within the cassava
genome (Lozano et al., 2015; Soto et al., 2015).
Pest and diseases
Cassava, as any other crop, is affected by several pests and diseases affecting yield
drastically. There are several reported diseases caused by bacteria, fungus,
phytoplasm and virus, as well as mite and insect pests.
The viral diseases are positioned as the most devastating in cassava. The main viral
diseases affecting this crop are the Cassava Mosaic Disease (CMD) and Cassava
Brown Streak Disease (CBSD). Both diseases can be transmitted through the use of
infected planting material or through whiteflies vectors, especially by the species
Bemisia tabaci (Homoptera: Aleyrodidae) (Legg, 2009). CMD is a disease caused by a
virus member of the Geminiviridae family, the Begomovirus (Legg and Threshb,
2003). It was first described in 1894 in Tanzania (Warburg, 1894 in Legg, 1999).
CMD has been reported in countries of East, West and Central Africa, as well as in
India and Sri Lanka. The typical symptoms of CMD are leaf chlorosis, leaves with
abnormal shape, mottling and mosaic (Legg and Thresh, 2003). However these
symptoms can vary depending on the virus strain, the environmental conditions and
the cassava variety (Legg and Thresh, 2003). Several reports have been made
regarding to yield losses up to 80% of plantations by CMD in Kenya and Uganda
37
(FAO, 2013), as well as in India and Sri Lanka (Dutt et al., 2005). CBSD is caused by
the Ipomovirus, a member of the Potyviridae family (Monger et al., 2001). It was first
reported in Tanzania in the 30s (Storey, 1936). The most significant symptom of this
disease is the dry, brown, necrotic lesions in the root tissue (Nichols, 1950). Despite
several symptoms can be present in leaves and stems these are not always evident
(FAO, 2013). CBSD has an important economic impact due to its devastating
behavior. In 1999 it was reported CBSD incidences up to 90% to 100% in
Mozambique fields (Hillocks et al., 2002). It has been also reported in Tanzania
(Mtunda et al., 2003) and Kenya (Njeru and Munga, 2003) with high yield losses.
Neither CMD nor CBSD have been reported in the Americas. However, viruses such as
the Cavemovirus the causal agent of the Cassava Vein Mosaic Disease (CVMD) and
Potexvirus causing Cassava Common Mosaic Disease (CCMD) have been reported as
pathogens producing important yield losses, especially in South America (Calvert and
Thresh, 2002). CVMD has incidence in Brazil (Calvert et al., 1995) while CCMD has
been reported also in Colombia, Paraguay and even in some African and Asian
regions (Chen et al., 1981; Carvajal-Yepes et al., 2014). The typical symptoms of
CCMD are green patches, chlorosis and mosaic in the leaves. While the most common
CVMD symptoms are vein chlorosis, which become ring-spots and leaf deformation
(Calvert and Thresh, 2002). For these diseases there have not been described
transmission vectors, thus the mechanical transmission seems to be the most
probable mechanism of spreading.
For both, CMD and CBSD it is recommended a strict implementation of quarantine
protocols during international exchange of cassava germplasm, and the achievement
of appropriate cultural practices, including virus-free planting material, and the use
of resistant cultivars (FAO, 2013). Despite that several efforts have been
accomplished to identify the loci governing resistance to both CMD (Akano et al.,
2002; Okogbenin et al., 2012; Rabbi et al., 2014) and CBSD (Rabbi et al., 2012), until
now the best alternative to protect cassava fields for these diseases is the use of
resistant varieties.
In the 70s in Cauca, Colombia a devastating disease appeared in cassava storage
roots. It was named cassava frog-skin disease (CFSD), because of the appearance the
roots (Álvarez et al., 2015). Since then, this disease has had prevalence in other South
American countries such as Brazil, Panama, Peru, Venezuela and Paraguay (Calvert et
al., 2008; Chaparro-Martinez and Trujillo-Pinto, 2001; Álvarez et al., 2015; Téllez et
al., 2016). Particularly in Colombia, incidences of up to 90% of CFSD have been
described in Valle del Cauca, Cauca, Meta and the North Coast (Pineda et al., 1983).
This disease generates plants with thin roots that accumulate little or no starch.
38
However, affected plants usually look healthy in the aerial parts (Cuervo et al., 2010).
Despite that for years the pathogen causing this important disease was unknown, a
phytoplasma belonging to a 16SrIII-A Subgroup has been linked to CFSD (De Souza
and Da Silva, 2014). The phytoplasma-like structures and phytoplasma 16SrIII was
detected in cassava infected tissue. This detection was possible by using fluorescent
microscopy and real-time polymerase chain reaction (PCR) analysis (Valverde, 2015).
Despite the evidence of the 16SrIII-A phytoplasma as a putative causal agent of CFSD,
recently, three viruses of the Secoviridae, Alphaflexiviridae and Luteoviridae families,
have been found in cassava plants showing root symptoms of CFSD (Carvajal-Yepes
et al., 2013). These results demonstrate that complex viral infections in cassava are
common but also show that more studies have to be carried out in order to establish
the pathogen(s) causing CFSD.
Several fungi have caused diseases in cassava and although they are not considered
limiting pathogens in some cases can generate important crop losses. The most
important fungi causing diseases in cassava are Colletotrichum gloeosporioides
(Anthracnosis) (Fokunang et al., 1997); Armillaria mellea (Armillaria root rot)
(Lozano and Terry, 1976) Scytalidium sp. (Black root and stem rot) (Bejarano et al.,
1991); Cercospora vicosae (Blight leaf spot) (Teri et al., 1978) and Cercosporidium
henningsii (Brown leaf spot) (Sugawara et al., 1991). Less devastating but with high
incidence rate are the root rots caused by several fungal genera such as
Macrophomina phaseolina, Fusarium sp. and Botryodiplodia theobromae
(Bandyopadhyay et al., 2006). This disease is a typical condition in all regions was
cassava grows and it is presented during intense rainy periods due to poorly drained
soils (FAO, 2013). The root rots is caused by several fungal and/or bacterial
pathogens, causing leaves losses, dieback and root deterioration, often resulting in
considerable decrease of the harvest. The Food and Agriculture Organization (FAO)
has implemented a strategy called “Save and Grow” to help the cassava small farmers
to face this sanitary problem through the use of a compendium of “clean” cultural
practices.
The pests affecting cassava usually are not crop-specific. However, insects that belong
to the genus Phyllophaga sp., Cyclocephala sp., and Anomala sp. feed on cassava roots
(Bellotti and van Schoonhoven, 1978) given way to secondary infections. Also, some
worms (larvae) such as Agrotis ipsilon, Spodoptera ornithogalli, Spodoptera frugiperda
cut the cassava buds and new stems (Alvarez et al., 2002) damaging the plant.
Crickets can also affect the cassava crop. The most common species found are Gryllus
assimilis (grillo negro) and Gryllotalpa sp. (grillo topo) (Bellotti, 1983). In addition,
termites such as Heterotermes tenuis can feed on stem cuttings, roots and even on
adult plants producing yield losses (Batista-Pereira et al., 2004).
39
Some pests directly affect the cassava leaves. This is the case of the mites
Mononychellus tanajoa, M. caribbeanae, Tetranychus urticae, T. cinnabarinus and
Oligonychus peruvianus, as well as the green mites Mononychellus tanajoa and M.
caribbeanae (Bellotti and van Schoonhoven, 1978). The typical symptoms of mite
attack in cassava are in leaves similar to the mosaic appearance, embryonic leaves
grow with deformations and the presence of spider webs. Finally, the whiteflies
Aleurotrachelus socialis Bondar, Bemisia tuberculata, B. tabaci, Trialeurodes variabilis,
Aleurodicus disperses and Aleurothrixus aepin (Gold et al., 1991), even if they are not
the most important pests in cassava for the damage in the crop, represent an
important vector for virus transmission. Particularly A. socialis has been described as
the most economically important pest in cassava caused by flies in Colombia. This
pest can cause plant tissue deformation and a yellowish green spotted in apical and
intermediate leaves.
Bacterial diseases can be a serious limiting factor under special circumstances (FAO,
2013). Some common bacterial diseases are Bacterial stem gall (Agrobacterium
tumefaciens), Bacterial stem rot (Erwinia carotovora subsp. carotovora), Bacterial
wilt (Erwinia sp) (Hillocks and Wydra, 2002) and cassava bacterial necrosis
(Xanthomonas cassavae) (Onyango et al., 1980). However, the principal disease
caused by bacteria is cassava bacterial blight (CBB).
Cassava Bacteria Blight
The causal agent: Xanthomonas axonopodis pv. manihotis
The causal agent of CBB is the bacillus Xanthomonas axonopodis pv. manihotis (Xam).
Xam is a vascular and foliar pathogen which belongs to the phylum Proteobacteria,
class Gammaproteobacteria, order Xanthomonadales and Xanthomonadaceae family.
Is a gram-negative rod-shaped bacterium of 1.0-1.75 x 0.28-0.6 µm, motile with a
single polar flagellum. The individual colonies become visible after 24 h of incubation
at 28°C in nutrient agar plates. The colonies are white-cream in color, convex, smooth
and shiny with entire edges (Maraite and Meyer, 1975).
Xam has a worldwide distribution, with the exception of Europe. Some quarantine
measures exist in all the places were cassava grows, especially regarded to the
40
movement of stem cuttings (Addoh, 1972). Artificial infection with Xam in species
belonging to different genus of the Euphorbiaceae family, such as poinsettia
(Euphorbia pulcherrima) and Pedilanthus tithymaloides seems to present similar
symptoms to those seen on cassava (Dedal et al., 1980).
Etiology and disease incidence
CBB was first described in Brazil in 1912 (Lozano, 1986), since then it has been
described as a very high destructive disease, causing losses between 12% up to 100%
in affected areas (Lozano, 1986; López and Bernal, 2012). CBB has been reported in
all regions where cassava is grown (López and Bernal, 2012). It is also described as a
potentially devastating disease and its causal agent is considered as a quarantine
organism (Mansfield et al., 2012; FAO, 2013). The effect of CBB in crop yield may vary
depending on factors such as the crop location, variety, climate and quality of initial
seed material (FAO, 2013). In 1974, CBB generated losses up to 50% of large
plantations in Brazil (FAO, 2013). Also in the 70s this disease was the cause of a
famine in Zaire (now Democratic Republic of Congo) and Nigeria (Moses et al., 2007;
Strange and Gullino, 2010). Thus CBB could be considered as the second most
devastating disease after CMD in epidemic episodes, causing losses of 90% of the
cassava yield (Lozano, 1986).
The principal way of transmission of CBB is the planting of infected material through
vegetative seed or cuttings, or by the use of infected crop tools. Likewise, the
transmission of the disease can occur from plant to plant by the rain splashing, by the
transit of personnel, machinery or animals infected to healthy fields. The youngest
leaves are initially infected by Xam and then the pathogen invades the vascular
tissues of the plant and finally causes death. The general described CBB symptoms
are blight, angular lesions in leaves; brown spots, production of gum exudates in
stem, wilting and defoliation (Mansfield et al., 2012; Lozano, 1986). Initially, the
symptoms appear in young leaves as angular leaf spots and blight. The spots become
brown (eventually with yellow halos), before they are transformed in necrotic areas
and finally suffer defoliation (Lozano, 1986). The presence of CBB symptoms in roots
is not common. However, this tissue can be affected in very susceptible varieties,
showing dry roots and vascular strands surrounded by rotted spots (Lozano, 1986).
High relative humidity favors Xam growing. The incidence of CBB has been reported
to be higher in warm and wet weather. In the 90s Fanou (1999), remarked the
importance of rainfall and high relative humidity, in humid forest, under the
41
epidemiology of CBB. Moreover, high CBB severities had been reported in the forest
zone, where is typical the rainfall and relative elevated humidity (Banito et al., 2001),
as well as deforested high rainfall areas in Nigeria (Wydra and Verdier, 2002).
Climate conditions also have been reported as an important factor for the disease
cycle of CBB (Fokunang et al., 2000). In Africa, especially in areas where there are
dissimilar rainy and dry seasons, the disease involves two phases. The first one
consists of angular leaf spots which begin and continue during the rainy season. The
second one, is the epiphytic phase in which the availability of moisture favors the
increase of the pathogen, and with it the effects in susceptible varieties of wilting and
defoliation of infected leaves, tip die-back and plant death (Lozano, 1986; Fokunang
et al., 2000).
Xam diversity
The knowledge on Xam diversity is an indispensable requirement to achieve an
integrated disease control of CBB. In the last decade several studies have been
focused on understanding the population dynamics of this pathogen (Verdier et al.,
1994; Ogunjobi et al., 2007; Ogunjobi, 2006; Verdier et al., 2004)
The Edaphoclimatic Zones (ECZs) were defined in the 80s by CIAT, as a way to study
different aspects of the crop and direct the breeding programs, including CBB
improvement. The ECZs were classified according to the importance of cassava
production, climatic conditions, soil type and predominant pest and disease problems
in the region. The first studies on Xam diversity carried out in Colombian showed a
high genetic variability within and among populations in different ECZs (Restrepo et
al., 2004). From 1995 to 1999, 96 Xam strains were isolated from infected cassava
stems or leaves in four ECZs in Colombia. Based on restriction fragment length
polymorphism (RFLPs) analysis, 45 haplotypes were detected (Restrepo et al., 2004).
In addition it was shown that Colombian Xam populations are highly dynamic and
changing in time. Evidence of this was the rapid change of haplotypes frequencies
detected in different ECZs. This dynamism is typical of migration events, which in this
case may be due to the exchange of infected cassava cuttings between growing areas,
which is a common behavior among farmers (Restrepo et al., 2000).
Recent studies on different Colombian regions, such as Cienaga de Oro, Chinú,
Palmitos, San Jacinto, Tolú Viejo, Meta and Casanare, have shown the current
situation of the Xam populations in the Caribbean region and Eastern Colombia. From
42
2008 to 2010, more than one hundred isolates of Xam were characterized using
AFLPs; the results shown that populations remain highly dynamic and highly diverse.
Also the haplotype composition of the isolates allowed identifying migratory
processes in the populations; as well as showing regions such as Chinú that
represents a diversity reservoir evidenced by the extensive genetic distances and
high diversity indices observed in this zone compared to those estimated in other
locations (Trujillo et al., 2014).
Some studies had initially shown that Xam African populations are highly
homogenous, contrary to the American ones (Verdier et al., 1994). However, in the
last decade, it was identified greater variability in populations in African Xam strains
(Ogunjobi, 2006; Ogunjobi et al., 2007; Verdier et al., 2004). Although it exists
knowledge of the diversity of this pathogen in Africa, the current status of the Xam
populations in this continent remains unexplored.
Xam genome
The genome sequence of several bacterial species belonging to the genus
Xanthomonas have been obtained, including Xam (Bart et al., 2012; Arrieta-Ortiz et
al., 2013). Through the study of these pathogen genomes have been attained not only
knowledge of the genome structure but also localization of pathogenicity factors
(Bart et al., 2012).
A draft genome sequence of sixty-five Xam strains originating from South America,
Africa and Asia was obtained (Bart et al., 2012). The genome sizes of these drafts
ranged from 4.50Mb to 5.12Mb in length. Through a phylogenetic analysis using
polymorphic SNPs markers present in the genomes, it was found a strong clustering
by country of origin, as well as a common ancestor between the Brazilian, Colombian
and African clades (Bart et al., 2012).
Recently a more complete genome sequence of Xam was obtained (Arrieta et al.,
2013). This genome corresponds to Xam CIO151 strain, which had a 5.15Mb genome.
This version of the genome has 36 scaffolds, 65.1% of G-C content, rRNA operons,
tRNA for all residues and more than four thousand coding DNA sequences (CDS)
which were automatic and manually annotated based on previous bacteria and
Xanthomonas genomes (Arrieta et al., 2013).
High similarities and small inversions were found when compared the CIO151
genome structure with those from Xanthomonas citri pv. citri and Xanthomonas
43
euvesicatoria. The 12.4% of the genome seems to be related to horizontal gene
transfer events. Within these regions, four pathogenicity islands were found. Also, the
gene annotation and proteome comparisons with other Xanthomonas revealed more
than 50 Xam-specific proteins (Arrieta et al., 2013).
ABC of plant immunity
Plants are an important source of nutrients such as carbohydrates, proteins and
metabolites for a wide range of microorganisms. Through the co-evolution, plants
and microorganisms have developed mechanisms to defend and access to these
resources respectively. Currently, the molecular understanding of plant-pathogen
interactions has allowed developing a model of the function and evolution of plant
immunity, known as zig-zag (Jones and Dangl, 2006). This model is based on the
knowledge generated through the study of the well-known pathosystems. According
to this model, the first event is the ability to recognize conserved molecules in
microorganisms known as PAMPs or MAMPs (Pathogen/Microbe Associated
Molecular Patterns). Among the molecules recognized, the most studied are flagellin
(flg22) originally described in Pseudomonas syringae pv. tabaci in the interaction with
tomato, tobacco, potato and Arabidopsis (Felix et al., 1999; Chinchilla et al., 2006).
The other best well known PAMP is the elongation factor Tu (EF-Tu) characterized in
Agrobacterium tumefaciens in the interaction with Arabidopsis (Zipfel et al., 2006).
Lipopolysaccharides (LPS) and chitin have also been described as PAMPs recognized
by Arabidopsis and rice respectively (Kaku et al., 2006; Miya et al., 2007). From the
genus Xanthomonas it has been described the PAMP eMAX in the interaction with
Arabidopsis (Jehle et al., 2013). However, this has not been purified and thus its
chemical nature is still unknown.
Recognition of PAMPs depends on the presence of plant proteins called Pathogen
Recognition Receptors (PRRs), which are proteins located generally in the plasmatic
membrane of the plant cell (Gómez-Gómez and Boller, 2000; He et al., 2007; Boller
and He, 2009; Zipfel, 2008; Thomma et al., 2011). Several PRRs have been described.
The first to be identified and one of the most studied is FLS2, which is a receptor that
recognizes flagellin, originally identified in Arabidopsis (Gomez-Gomez and Boller,
2000). This PRR is a transmembrane RLK with a rich extracellular domain of LRR and
serine/threonine intracellular kinase domain, possibly responsible for signal
transduction (Gómez-Gómez and Boller, 2000; Asai et al., 2002). Several homologues
of FLS2 have been found in tomato and rice (Takai et al., 2008). Another PRR widely
studied is EFR receptor that recognizes the N terminal region of EF-Tu (Kunze et al.,
2004). EFR also belongs to the family of RLK receptors (Kemmerling et al., 2011). The
44
EFR receptor is present only in members of the Brassicaceae family. The first PRR for
chitin to be identified was the chitin elicitor-binding protein (CEBiP) of rice, a
membrane protein with two LysM extracellular motives and one transmembrane
domain (Kaku et al., 2006). The CeBiP homologue in Arabidopsis was also identified
and named CERK1 / LysM-RLK (chitin elicitor receptor kinase) (Miya et al., 2007). It
seems that this PRR is capable of recognizing molecules present in the walls of gram
negative bacteria like peptidoglycan, which has characteristics similar to chitin, such
as an oligosaccharide. Recently, it has been described two membrane proteins in rice
which have LysM, LYP4 and LYP6, that act as PRRs inducing immunity responses
when recognize the bacterial peptidoglycan of Xanthomonas oryzae pv. oryzae (Xoo)
and chitin of Magnaporthe oryzae (Liu et al., 2012).
A particular case of PRR is Xa21. This protein is a receptor kinase (RK) and was
initially characterized as a protein encoded by a resistance gene against Xoo, the
causal agent of bacterial blight in rice (Song et al., 1995). However, after several years
of efforts to identify the corresponding Avr (AvrXa21), it was not until 2009 when
this protein was characterized. It seemed that RaxX protein was recognized as a
PAMP for the activation of the Xa21-mediated immunity (Pruitt et al., 2015).
The first response based on the interaction between PAMPs-PRRs is referred to
PAMP- Triggered Immunity (PTI). In this type of immunity, the infection usually
stops before the microorganism multiplication starts and is sufficiently effective
against non-adapted pathogens (Chisholm et al., 2006). Furthermore, during
evolution, a particular group of pathogens developed a special type of proteins called
effectors that when injected into plant cells block PTI, achieving a successful
colonization (Jones and Dangl, 2006). Such interactions are called compatible, and
the result for the plant is the disease development. According to the zig-zag model,
such interactions fall under the concept of effector-triggered susceptibility (ETS).
There are several mechanisms by which some adapted pathogens suppress or escape
from the detection mediated by the PTI. The best studied are the bacterial type III
effector (T3E) proteins which are translocated to the host cytoplasm by the T3SS
(Coburn et al., 2007). These effectors belong to different families which differ by their
functions. Some of the T3Es can induce the degradation of PRRs, such as the effector
AvrPtoB from Pseudomonas syringae pv. tomato (Pst) DC3000 (Göhre et al., 2008).
Other targets are the Mitogen-activated protein kinases (MAPK) which are active
after pathogen recognition. Effectors as AvrPto and AvrPtoB from Pst block these
pivotal proteins (Abramovitch et al., 2006; Rasmussen et al., 2012). Some effectors as
Hopl1 from Pst and P. syringae pv. maculicola (Psm) interfere with cell organelles
45
such as the chloroplast, promoting the reduction of acid salicylic production (Jelenska
et al., 2007).
One of the best known families of T3Es is the Transcription Activator-Like effectors
(TALEs) (Boch et al., 2014) which includes the AvrBs3 effector family from
Xanthomonas campestris pv. vesicatoria (Xcv) (Bonas et al., 1989). These effectors are
restricted to the genus Xanthomonas and have the ability to induce the disease by
selectively bind to the DNA promotor sequence of target susceptibility (S) genes
activating its transcription (Boch and Bonas, 2010). The gene induction of S genes is
TAL effector-dependent and has direct consequences on disease symptoms (Boch et
al., 2014). On the other hand, plants evolved to activate resistance genes mediated by
TALs, which are named executor (E) genes (Bogdanove and Voytas, 2011; Tian et al.,
2014; Zhang et al., 2015). All proteins members of the TAL family have an N-terminal
where is located the T3SS-mediated translocation signal; a central domain containing
a repetitive domain of 33 to 35 amino acids where the amino acids 12 and 13 are
polymorphic and are called repeat-variable di-residues (RVDs) (Boch et al., 2014;
Zhang et al., 2015). These RVDs determinates the binding specificity of the TALs to
specific promoter regions of S or E genes.
In order to face the ETS, plants developed a second branch of immunity based on the
recognition of pathogen effector proteins, which correspond to the third phase of the
zig-zag model called ETI (Effector-Triggered Immunity). The ETI depends on the
presence of resistance (R) proteins that can recognize directly or indirectly effectors,
triggering a reaction of incompatibility or resistance (Chisholm et al., 2006) (Jones
and Dangl, 2006) (Dodds and Rathjen, 2010). The ETI corresponds to the classic
concept of race specific resistance or gene by gene theory proposed by Flor in the
50s. This recognition is governed by the presence of an R gene in the plant and for the
existence of a specific avirulence gene Avr in particular strains of a species of
pathogen (Flor, 1955; Dangl and Jones, 2001). Thus, some varieties of a crop will be
resistant to some strains of the pathogen, but susceptible to others. This differential
response depends on the races which usually carry a repertory of effectors or
differential Avr genes. Most R proteins have an NBS (Nucleotide Binding Site) domain
and a LRR domain at its N-terminal. In the carboxy terminal can have a TIR domain
(Toll/Interleukin-1 receptor) or a coiled coil (CC) domain (Bent, 1996; Hammond-
Kosack and Kanyuka, 2007).
The recognition of pathogen effectors by R proteins can be direct or indirect. In direct
recognition, the R protein acts as a receptor that interacts with the pathogen Avr
protein, which acts as a ligand (Dodds et al., 2006). In many cases it has not been
demonstrated an R-Avr interaction. This led to argue that the recognition could be
46
mediated by a third protein, known as pathogenicity target (guarded protein). This
model has been called "guard gene model", and suggests that R protein detects a
change effector-induced in the pathogenicity plant target protein (Dangl and Jones,
2001). Thus, alteration of the target in the host will confer pathogen colonization in
susceptible host genotypes, but it will induce ETI response in hosts carrying the
corresponding R protein (Dangl and Jones, 2001; Caplan et al., 2008). It has been
considered that in this model the guarded protein would be facing opposing selective
forces, due to the dependence of the presence of the R protein. Thus the guarded
protein will be selected against mutations leading the lack of the effector recognition
if the R protein is not present. On the other hand, if the R protein is present, there will
be a selection favoring increase the recognition by the effector. Given this
discrepancy, the "decoy" model was proposed. In this model, the "decoy" protein
represents an effector target but it is not involved in plant immunity and have any
function neither in resistance or susceptibility, nor contributing to the pathogen
fitness in the absence of R protein (van der Hoorn and Kamoun, 2008; Dodds and
Rathjen, 2010).
Despite PTI and ETI have been described in several pathosystems (Thomma et al.,
2011b), there are a number of concerns around these separated branches (Boller and
Felix, 2009; Pritchard and Birch, 2014). As an alternative to the zig-zag model, the
invasion model has arisen, which is based on the pathogen invasion patterns (IPs). An
IP can be either an external or a modified host ligand that is perceived by host IP
receptors (IPRs), indicating invasion. The defense response that is activated by this
perception is called IP-triggered response (IPTR) (Cook et al., 2015). The invasion
model considers that several ligands and receptors could be acting at the same time
and in consequence, the defense response will be the result of the interaction
between all of them. The description of this model in pathosystems is still under
investigation.
Although initially it was considered that the responses triggered by the PTI and ETI
were different, recent studies have shown that these overlap considerably (Schulze-
Lefert and Panstruga, 2011). Once the perception of PAMPs or effectors is given by
the PRRs or R proteins, a signaling cascade is triggered mainly mediated by MAPK
(Göhre et al., 2008; Colcombet and Hirt, 2008; Beckers et al., 2009) which culminate
with the activation of transcription factors and leads the reprogramming of gene
expression. Additionally, these responses are associated with events of opening of
ionic channels in the membrane, production of ROS (Reactive Oxygen Species) and
fortification of cell walls through the deposition of callose (Zipfel, 2008).
47
The reprogramming of gene expression during PTI, ETS and ETI has been studied in
pathosystems models mainly using transcriptomic approaches. The responses
between compatible interactions (virulent bacteria), incompatible (avilurent bacteria
carrying Avr proteins recognized by plant R proteins) and non-host are similar and
the differences found in the gene expression are quantitative in terms of intensity
and the speed with which these occur (Tao et al., 2003). However, it seems that the
plant response in incompatible interaction is robust in terms of the high level of
resistance gene expression, and is not affected by environmental conditions (Tao et
al., 2003).
During the plant responses several genes have been identified whose products are
pathogenesis-related proteins (PR proteins) (Van Loon and Van Strien, 1999). Some
PRs have antimicrobial activities such as chitinase and glucanase. These PRs have
been classified into 17 families and although some of them were initially found in
tobacco and Arabidopsis, its induction by pathogens has been reported in several
plant species (Van Loon and Van Strien, 1999; Sudisha et al., 2012). The recognized
defense activity of these proteins has made that its gene expression is an indication
(marker) of the activation of defense responses (Sudisha et al., 2012; Zhang et al.,
2012).
Quantitative resistance
Unlike PTI or ETI, whose explanatory models are based on the resistance mediated
by a single gene, the quantitative resistance is governed by multiple genes each
contributing differentially with a given percentage in the total resistance (Poland et
al., 2009; Kou and Wang, 2010). Quantitative resistance is also known as field
resistance, polygenic, incomplete resistance, broad spectrum or horizontal resistance.
It has been considered that the quantitative resistance does tend to be durable over
time, mainly for two reasons. The first one is that is governed by multiple genes, thus,
the likelihood of simultaneous mutations in several Avr genes, which allows
successfully escape recognition, is very low. In addition, this type of resistance is not
specific to a particular race; conversely it is broad spectrum, which means that the
plant is resistant to different races, strains or variants of the same pathogen species,
and even to different host species (Poland et al., 2009; Kou and Wang, 2010).
As any other quantitative trait, quantitative resistance shows a continuous variation
of phenotypes, therefore a single phenotypic classification of resistance or
48
susceptibility is not easy. Thus usually the phenotype is determined by the use of
scales of symptoms, ranging from any or few affectations to full disease. This
phenotypic variability present in quantitative resistance typically has an important
environmental component making its study a challenge.
One of the most widely used approaches for the study of quantitative resistance and
thus the identification of the loci that governs this resistance, has been the QTL
mapping (Young, 1996); and more recently the genome wide association study
(GWAS) analysis (Zhu et al., 2008; Brachi et al., 2011). The detection of QTL is based
on i) the variation of DNA, through polymorphic markers which allow the
construction of a genetic map and ii) the observed phenotype in the segregating
population generated to construct the genetic map (Mackay et al., 2009). Combining
this information, it is possible to associate some markers with a particular
phenotype. On the other hand, the GWAS strategy generates a direct association
between the trait with the polymorphic markers taking advantage of the historical
recombination events at population level (Zhu et al., 2008) 1
The molecular basis of quantitative resistance has not been elucidated in detail.
However, it has been proposed that both the qualitative and quantitative resistance
are controlled, at least partially by classic R genes (Poland et al., 2009; Kou and Wang,
2010; Lopez, 2011). This hypothesis is based on the observation of co-localization of
genes encoding proteins with NBS domains with QTL associated to pathogen
resistance (López et al., 2003a; Ramalingam et al., 2003)
Through QTL mapping it has been possible the identification and cloning of genes
involved in quantitative resistance. These genes code proteins related to signal
pathways and are: i) a wheat kinase-START (WKS) (Fu et al., 2009); ii) an ATP
binding cassette (ABC) protein transporter (Krattinger et al., 2009); iii) a protein rich
in proline residues that is associated with the transport of metals with motifs for
protein-protein interactions (Fukuoka et al., 2009); iv) an atypical kinase which lacks
several domains for the kinase catalytic (Huard-Chauveau et al., 2013) and v) a
receptor like kinase (RLK) (Hurni et al., 2015). The cloning of these genes has
contributed to a better understanding of the molecular nature of the quantitative
resistance.
1These concepts will be further developed in a scientific manuscript that is part of Chapter 1.
49
Molecular interaction cassava-Xam
Molecular basis of the pathogenecity
As was described above, during the co-evolution between plant and pathogens,
pathogenic bacteria has developed ways to suppress the plant defense by the
translocation of virulence factors or effectors into the host cells and thus facilitating
the successful establishment of disease.
The availability of the draft genome for 65 Xam strains from diverse geographical
origins (Bart et al., 2012) and the genome of the strain Xam CIO151 (Arrieta et al.,
2013), has allowed the identification of more than 25 T3Es gene families. This
repertoire includes the effector AvrBs2, a group of Xanthomonas outer proteins
(Xop) and TALE effectors (Bart et al., 2012).
AvrBs2 is an effector widely described in the genus Xanthomonas containing a
conserved glycerolphosphodiesterase (GDE) domain pivotal for virulence (Kearney
and Staskawicz, 1990). For several pathosystems it has been demonstrated its
virulence contribution and the capacity to suppress plant immunity (Kearney and
Staskawicz, 1990; Li et al., 2015). In Xam it was identified a protein similar to AvrBs2
(Arrieta et al., 2013). A recent study reported the generation of an AvrBs2 mutant in
Xam. This mutant showed a significant reduction in virulence and aggressiveness to
cassava (Medina et al., 2016 unpublished results).
In Xam the role and/or function of different predicted effectors still remains to be
elucidated. However, a recent study of gene mutagenesis has revealed that some
members of the Xop family have a pivotal role in virulence and suppression of PTI
and ETI in cassava (Medina et al., 2016 unpublished results). For example, it was
demonstrated that additional to AvrBs2, the effector XopAO1 was important for
virulence. While the mutation of the effectors XopR, XopQ, XopE4, XopN and XopV not
compromised the aggressiveness of Xam, suggesting a redundancy function between
different effectors.
Concerning to the TAL family effectors present in Xam, the first member described
was TALE1Xam (Castiblanco et al., 2013), which was previously known as pthB. This
TAL has 13.5 tandem repeats and it was demonstrated have an important role in
virulence (Castiblanco et al., 2013). Through transcriptomic analyses it was
demonstrated that TALE1Xam can activates the transcription of several genes in
cassava cells being one of the predict targets a heat shock transcription factor B3
50
(HsfB3) (Muñoz-Bodnar et al., 2014). Recently it was established that TAL20Xam668
promotes virulence through the induction of the expression of the sugar transporter
MeSWEET10a (Cohn et al., 2014). More than 50 virulence targets for TALE effectors
have been described in cassava, specifically for the Xam strain Xam668 (Cohn et al.,
2016), including proteins related to cell wall-modifying, proteases and immunity
related proteins such as LRR-kinases.
The analysis of more than 180 Colombian Xam strains has revealed the presence of
several TALs in the Xam genome ranging from two to five (Zarate et al., 2015,
unpublished results). The identification of S and E target genes for some of these
TALs has been predicted using bioinformatics tools and their validation is currently
in progress (Mora et al., 2016 unpublished results; Ramirez et al., 2016, unpublished
results).
Molecular basis of resistance to CBB
Despite that in the pathosystem cassava-Xam the zig-zag model has not been
determined; it is possible to expect that both the PTI and ETI are present. In Xam
some effectors have been described (Bart et al., 2012; Arrieta et al., 2013), as well as
some repertoires of immunity related genes (IRG) (Leal et al. 2013; Lozano et al.,
2015; Soto et al., 2015)2.
Taking advantage of the presence of conserved domains (NBS, TIR, LRR) in the R
proteins several RGAs (Resistance Gene Analogues) have been amplified and
identified by the use of degenerate primers designed based on these conserved
domains (López et al., 2003). More recently, taking advantage of the available cassava
genome it was possible to design bioinformatics tools to identify genes coding for
proteins containing the typical conserved domains present in R proteins. Two
repertoires of more than 550 RGAs have been described (Lozano et al., 2015; Soto et
al., 2015).
2These concepts will be developed further in a scientific manuscript that is part of Chapter 1)
Using primers designed from the resistance gene Xa21, which confers resistance to
Xanthomonas oryzae in rice (Song et al., 1995), it was amplified a fragment of cassava
51
genome having a high degree of similarity to the gene. This fragment co-localized
with a QTL explaining 13% of the resistance to Xam strain CIO136 and has been
called RXam1 (resistance to Xam 1) (Jorge et al., 2000). The gene RXam1 encodes for
a RLK protein. In resistant cassava plants inoculated with CIO136 Xam strain, RXam1
is induced at five days post inoculation (López et al., 2007). The similarity between
Xa21 and RXam1 suggests that it may have a function of resistance in cassava and
several experiments are being conducted to validate its function (López, 2011).
López et al (2007) mapped a set of defense-related genes and BACs-containing RGAs
into 11 linkage groups on the cassava genetic map. Within these linkage groups were
identified two QTL that explains the 21.4% and 61.6% of the resistance to Xam
strains CIO121 and CIO151, respectively. The last QTL co-localizes with a BAC
containing an RGA, which was called RXam2 (Resistance to Xam 2) and encodes a
protein with a NBS domain and a LRR domain in the C-terminus (López et al., 2003;
López et al., 2007). This gene is also being validated through different approaches to
confirm its role in CBB resistance.
Several important efforts have been conducted in order to identify proteins involved
in pathogen recognition and to study the reprograming gene expression during
cassava-Xam interaction. Thus for example, employing a cassava resistant variety it
was established important gene expression changes during cassava immune
responses. The comparison of some selected genes between resistant and susceptible
cassava varieties show that these genes are also induced but stronger and faster in
the resistant variety (Lopez et al., 2005). Moreover, aspects such as the identification
of non-coding microRNAs (miRNAs) (Pérez et al., 2012) and trans-acting small
interfering RNAs (ta-siRNAs) induced and repressed in cassava during Xam infection
(Quintero et al., 2013) have been reported. Several targets of these non-coding RNAs
were predicted and include genes coding for transcription factors and LRR-
containing proteins (Pérez et al., 2012; Quintero et al., 2013). Also the plant immune
response involves a complex network of protein-protein interactions some of which
have been studied in cassava-Xam through experimental assays (Román et al., 2014)
or through in silico predictions (Leal et al., 2013).
Mapping the quantitative resistance to CBB
52
Cloning genes through positional mapping has been the most widely used method for
resistance gene cloning (Bent, 1996; Pflieger et al., 2001; Gebhardt et al., 2007). This
approach involves the development of genetic maps (Collard et al., 2005), ideally
with low average distance in cM between markers.
By definition, a genetic map is a descriptive diagram of the position and relative
genetic distance between markers either morphological or molecular and loci or
genes along linkage groups (Paterson, 1996). The markers are positioned relative to
each other based on their recombination frequencies and representing relative
distances between them. It has been established that the distances are indicated in
centi-Morgan units (cM), where 1cM corresponds to a 1% probability that a
crossover occurs between two loci during meiosis (Wu et al., 2007). Such maps are
usually called linkage maps, as they determine the position of genes or markers
"linked" within a single chromosome.
In order to develop a genetic map a segregating population must be generated
through targeted crosses. It is a necessary condition to use contrasting and/or
phylogenetically distant parental for the interest trait, in order to increase the
probability of finding loci and / or polymorphic markers. The segregation of these
markers has to be tested in the progeny through the Mendelian expected segregation,
according to the population type (Wu et al., 2007). Once constructed the genetic map
and knowing the relative distances between the molecular markers within linkage
groups, these can be associated with the phenotypic value of the trait of interest,
through QTL mapping for the case of complex traits (Wu et al., 2007).
For cassava it has been built several genetic maps based on different types of
molecular markers and some of these have been used in QTL mapping for important
agronomical traits including disease resistance. The first genetic map of cassava was
developed from an intraspecific cross between the Nigerian variety TMS30572,
developed by IITA, with the Colombian elite variety CM21772. The resulting F1
progeny of this cross consisted of 150 genotypes, highly heterozygous (Fregene et al.,
1997). This map was developed with 132 RFLPs, 30 RAPDs, three microsatellites and
three isozymes, with a resolution of 8 cM. Later, this map was increased with 172
microsatellite markers but with lower resolution (Mba et al., 2001). Chen et al,
(2010), using another cassava population (SC6 x Mianbao) obtained a genetic map
with 231 AFLPs markers, 41 microsatellites, 48 SRAPs (Sequence-related amplified
polymorphism) and 35 ESTs, with a resolution of 4.8 cM. Since then, the cassava
maps have increased the number of molecular markers. Kunkeaw et al. (2010, 2011)
develop two genetic maps; the first one was obtained from the genotyping of 58 F1
progeny resultant from the cross between Rayong 90 x Rayong 5. This map consisted
53
in 110 AFLPs markers and 19 SSRs with a resolution of 8 cM. The second map was
obtained from the genotyping of 100 F1 progeny from the cross between Hauy Bun
60 x Hanatee, and consisted in 56 EST-SRRs and 155 SRR, with a resolution of 5.6 cM.
In 2012, Rabbi et al. obtained a map with 434 SNPs and 134 SSRs with a resolution of
3.4 cM, which was obtained though the genotyping of 130 F1 full sibs from the cross
between Namikunga x Albert (Rabbi et al., 2012).
With advances in high-throughput genotyping, the cassava genetic maps integrate
thousands of SNP markers. Rabbi et al developed two cassava genetic maps using
genotyping by sequencing (GBS) approach. These maps were developed using the F1
progeny from the crosses between MME9 x TMS30571 (182 full sibs) and
TMS011412 x IITA-TMS-4(2)1425 (180 full sibs); these maps had 772 and 6,756
SNPs, with resolutions of 2 cM and 0.52 cM respectively (Rabbi et al., 2014a, 2014b).
Recently Soto et al (2015) has obtained a genetic map of 2,141 SNP, with a resolution
of 1.26 cM, which was obtained though the genotyping of 132 F1 full sibs from the
cross between TMS30572 and CM2177-2 (K family).3
Some studies focused on the detection of QTL mapping for resistance to CBB have
been performed based on the resistance phenotypic evaluation in the high
segregating population “K family” as well as through the use of some of the genetic
maps presented above. These studies have been performed under natural and
controlled conditions. The latter is the case of the study conducted in 2000 by Jorge
et al, where 12 QTLs were identified, eight located on linkage groups B, D, L, N, and X
of the linkage map from the parental TMS30572 and four were located in linkage
groups G and C of the linkage map from the parental donor CM1477-2. These QTL
were obtained through simple regression analysis, and explained from 9 to 27% of
the phenotypic variance of the response to Xam strains CIO-84, CIO-1, CIO-136, CIO-
295 and ORST X-27. In this study transgressive segregants were also identified,
showing possible gene dominance. QTL that explain the variance of resistance to the
CIO-84 strains and ORST X-27, apparently are introgressions of a wild Manihot.
3 The development of this map will be further presented in a scientific manuscript that is part of Chapter 2
These QTLs are located in the linkage group D of the TMS30572 map, which has a
large number of polymorphic markers and shows a low recombination frequency
compared to the rest of the genome. The nature of this linkage group suggests that it
is a vestige of the genome of M. glaziovii (Jorge et al., 2000).
The phenotyping response to Xam has also been tested under natural high pathogen
pressure conditions. Jorge et al (2001) conducted the evaluation of the same K family
54
derived from TMS30572 x CM1477-2 in two growth cycles (years) under high field
Xam pressure. QTL detection was also performed by simple regression analysis and
revealed eight QTLs explaining between 7.2% and 18.2% of the resistance to Xam.
From cycle to cycle were detected changes in the QTL, suggesting that these are not
stable or that there are changes in the pathogen population. Only one QTL located on
linkage group D, was detected in the two years of evaluation. Also, Wydra et al (2004)
using a population product of the backcross (TMS30572 x CM1477-2) x TMS30572,
detected nine QTLs that explain from 16% to 33% of the resistance to four African
Xam strains. The phenotypic evaluation was performed through inoculation in leave
and stems under greenhouse controlled conditions.
Strategies such as the mapping of several defense-related genes to CBB and BACs
carrying RGAs have contributed with the detection of QTL (López et al., 2007), as was
described above. Despite of all these efforts, until today any resistant gene to the CBB
disease has been identified yet.
Improving CBB resistance
The development of resistant varieties to CBB through traditional breeding has been
one of the objectives within some breeding programs in countries where the disease
has an impact (Russell, 2013).
It has been considered that the source of resistance to CBB in the current cassava
varieties comes from the introgression of its wild relative M. glaziovii (Hahn et al.,
1980) Mahungu et al., 1994). The successful development of clones and varieties
promissory for CBB resistance from crosses between M. glaziovii and M. esculenta
support this hypothesis (Hahn et al., 1974). The cross between these two species has
also been useful to break the linkage between genes responsible for resistance and
undesirable productivity genes, commonly found in the wild species (IITA, 1977).
Breeding schemes to improve CBB resistance starts with the search for materials
with the highest resistance to CBB but also with high genetic variability for important
agronomical traits (Mahungu et al., 1994). The sources of these materials are the
bank germoplasm, local varieties or some of the commercial varieties. Two
evaluations to CBB resistance are usually performed. The first evaluation is done in
pre-selection of seedlings. The second one is conducted three to five months after
planting. In both evaluations the inoculation is performed artificially through the
stem puncture inoculation method and the disease is scored using: i) scale of
55
symptoms, ii) the total number of plants affected, iii) the disease incidence (scored in
percentages) and iv) the disease severity (Mahungu et al., 1994).
It is necessary to conduct extra evaluations for selected materials, under different
environmental conditions, as is done in some African programs, where the
evaluations have showed high stability of resistance to CBB (IITA, 1976). As the
environment conditions have a high influence in the incidence of CBB and contributes
to accelerate the development of the symptoms (Leu, 1978; Banito et al., 2000, 2001;
Wydra and Verdier, 2002; Restrepo et al., 2004), it is imperative to conduct
evaluations under different environments within the breeding schemes, in order to
obtain cassava materials with durable resistance and adapted to broad or specific
environmental conditions.
The estimation of low values for broad sense heritability of CBB resistance has also
highlighted the important effect of the environmental conditions under the
development of the disease. In cassava breeding programs, values of CBB broad
heritability ranged from 24% to 48%, which has been estimated in half-sib families
and clones, respectively (Hahn et al., 1974). In experimental populations the
heritability values reported ranged from 10% to 69% (Hahn et al., 1998, Jorge et al.,
2000; Fregene et al., 2001; Ly et al., 2013).
Due to the fact that CBB is influenced by environmental conditions and show low
heritability, some considerations have to be taken into account in order to guarantee
the development of durable resistance to the disease. Multi-environment evaluations
are highly advisable, as well as the evaluation of the stability of the resistance during
several seasons and crop cycles. Also, as it is known that Xam populations are
variable in space and time, it is important to perform evaluations under high disease
pressure, in different localities and/or using different strains. On the other hand, it is
advised that within breeding programs the dynamics and diversity of Xam
populations should be analyzed in the fields of evaluation but also in the regions
where the improved varieties will be established.
References
Abhary, M., Siritunga, D., Stevens, G., Taylor, N. J., and Fauquet, C. M. 2011. Transgenic biofortification of the starchy staple cassava (Manihot esculenta) generates a novel sink for protein. PLoS One. 6:e16256
56
Abramovitch, R. B., Anderson, J. C., and Martin, G. B. 2006. Bacterial elicitation and evasion of plant innate immunity. Nat. Rev. Mol. Cell Biol. 7:601–611
Addoh, P. G. 1972. Nouvelle maladie bacterienne du manioc au Nigeria. Circular of the "Conseil Phytosanitaire Interafricain", CIRIOUAICSTRICPI. 72:18
Aguilera, M. 2012. La yuca en el Caribe colombiano: De cultivo ancestral a agroindustrial. Banco de la República de Colombia.
Akano, A., Dixon, A. G. O., Mba, C., and Fregene, M. 2002. Genetic mapping of a dominant gene conferring resistance to cassava mosaic disease. TAG Theor. Appl. Genet. 105:521–525
Aletor, O. 2010. Comparative, nutritive and physico-chemical evaluation of cassava (Manihot esculenta) leaf protein concentrate and fish meal. J. Food, Agric. Environ. 8:39–43
Allem, A. C. 1994. The origin of Manihot esculenta Crantz (Euphorbiaceae). Genet. Resour. Crop Evol. 41:133–150
Alvarez, E., Bellotti, A., Arias, B., Cadavid, L. F., and Llano, G. 2002. Guía práctica para el manejo de las enfermedades, las plagas y las deficiencias nutricionales de la yuca. CIAT.
Álvarez, E., Pardo, J. M., Mejía, J. F., de Oliveira, S., Alves, S., Zacher, M., Cardozo, L., and Gómez, Y. 2015. Manejo del Cuero de sapo enfermedad limitante de la yuca. CIAT.
Aristizábal, J., Sánchez, T., Mejia-Lorio, D. J., and others. 2007. Guía técnica para producción y análisis de almidón de yuca. Organización de las Naciones Unidas para la Agricultura y la Alimentación.
Arrieta-Ortiz, M. L., Rodriguez-R, L. M., Pérez-Quintero, Á. L., Poulin, L., Diaz, A. C., Rojas, N. A., Trujillo, C., Benavides, M. R., Bart, R., Boch, J., and others. 2013. Genomic survey of pathogenicity determinants and VNTR markers in the cassava bacterial pathogen Xanthomonas axonopodis pv. manihotis strain CIO151. PLoS One. 8:e79704
Asai, T., Tena, G., Plotnikova, J., Willmann, M. R., Chiu, W.-L., Gomez-Gomez, L., Boller, T., Ausubel, F. M., and Sheen, J. 2002. MAP kinase signalling cascade in Arabidopsis innate immunity. Nature. 415:977–983
Asare, E., and Safo-Kantanka, O. 1997. Improvement of cassava cooking quality through mutation breeding. Improv. basic food Crop. Africa through plant breeding, Incl. use Induc. Mutat. Int. At. Energy Agency, Vienna, IAEA-TECDOC-951. :19–24
Awoleye, F., van Duren, M., Dolezel, J., and Novak, F. J. 1994. Nuclear DNA content and in vitro induced somatic polyploidization cassava (Manihot esculenta Crantz) breeding. Euphytica. 76:195–202
57
Bandyopadhyay, R., Mwangi, M., Aigbe, S. O., and Leslie, J. F. 2006. Fusarium species from the cassava root rot complex in West Africa. Phytopathology. 96:673–676
Banito, A., Kpémoua, K. E., Wydra, K., and Rudolph, K. 2001. Bacterial blight of cassava in Togo: its importance, the virulence of the pathogen and the resistance of varieties. Pages 259–264 in: Plant Pathogenic Bacteria, Springer.
Barabaschi, D., Tondelli, A., Desiderio, F., Volante, A., Vaccino, P., Valè, G., and Cattivelli, L. 2016. Next generation breeding. Plant Sci. 242:3–13
Bart, R., Cohn, M., Kassen, a., McCallum, E. J., Shybut, M., Petriello, a., Krasileva, K., Dahlbeck, D., Medina, C., Alicai, T., Kumar, L., Moreira, L. M., Neto, J. R., Verdier, V., Santana, M. a., Kositcharoenkul, N., Vanderschuren, H., Gruissem, W., Bernal, a., and Staskawicz, B. J. 2012. PNAS Plus: High-throughput genomic sequencing of cassava bacterial blight strains identifies conserved effectors to target for durable resistance. Proc. Natl. Acad. Sci. 109:E1972–E1979
Batista-Pereira, L. G., Dos Santos, M. G., Corrêa, A. G., Fernandes, J. B., Arab, A., Costa-Leonardo, A. M., Dietrich, C. R. R. C., Pereira, D. A., and Bueno, O. C. 2004. Cuticular hydrocarbons of Heterotermes tenuis (Isoptera: Rhinotermitidae): analyses and electrophysiological studies. Zeitschrift fur Naturforsch. C. 59:135–139
Beckers, G. J. M., Jaskiewicz, M., Liu, Y., Underwood, W. R., He, S. Y., Zhang, S., and Conrath, U. 2009. Mitogen-activated protein kinases 3 and 6 are required for full priming of stress responses in Arabidopsis thaliana. Plant Cell. 21:944–953
Bejarano, C. A., LABERRY, R., Orrego, A., and LOZANO, J. C. 1991. La pudricion negra de la yuca (Manihot esculenta) causada por Scytalidium sp. Ascolfi Informa. 5:44–46
Bellotti, A. 1983. Descripción de las plagas que atacan la yuca (Manihot esculenta Crantz) y características de sus daños. CIAT.
Bellotti, A., and van Schoonhoven, A. 1978. Mite and insect pests of cassava. Annu. Rev. Entomol. 23:39–67
Bent, A. F. 1996. Plant disease resistance genes: function meets structure. Plant Cell. 8:1757
Boch, J., and Bonas, U. 2010. Xanthomonas AvrBs3 family-type III effectors: discovery and function. Phytopathology. 48:419
Boch, J., Bonas, U., and Lahaye, T. 2014. TAL effectors--pathogen strategies and plant resistance engineering. New Phytol. 204:823–832
Bogdanove, A. J., and Voytas, D. F. 2011. TAL effectors: customizable proteins for DNA targeting. Science (80). 333:1843–1846
Boller, T., and Felix, G. 2009. A Renaissance of Elicitors: Perception of Microbe-Associated Molecular Patterns and Danger Signals by Pattern-Recognition Receptors. Annu. Rev. Plant Biol. 60:379–406
58
Boller, T., and He, S. Y. 2009. Innate immunity in plants: an arms race between pattern recognition receptors in plants and effectors in microbial pathogens. Science. 324:742–744
Bonas, U., Stall, R. E., and Staskawicz, B. 1989. Genetic and structural characterization of the avirulence gene avrBs3 from Xanthomonas campestris pv. vesicatoria. Mol. Gen. Genet. MGG. 218:127–136
Borlaug, N. E. 1983. Contributions of conventional plant breeding to food production. Science (80). 219:689–693
Brachi, B., Morris, G. P., and Borevitz, J. O. 2011. Genome-wide association studies in plants: the missing heritability is in the field. Genome Biol. 12:1
Bredeson, J. V, Lyons, J. B., Prochnik, S. E., Wu, G. A., Ha, C. M., Edsinger-Gonzales, E., Grimwood, J., Schmutz, J., Rabbi, I. Y., Egesi, C., and others. 2016. Sequencing wild and cultivated cassava and related species reveals extensive interspecific hybridization and genetic diversity. Nat. Biotechnol. 34:562–570
Calvert, L. A., Cuervo, M., Lozano, I., Villareal, N., and Arroyave, J. 2008. Identification of three strains of a virus associated with cassava plants affected by frogskin disease. J. Phytopathol. 156:647–653
Calvert, L. A., Ospina, M. D., and Shepherd, R. J. 1995. Characterization of cassava vein mosaic virus: a distinct plant pararetrovirus. J. Gen. Virol. 76:1271–1278
Calvert, L., and Thresh, J. M. 2002. The Viruses and Virus Diseases of Cassava. Pages 237–260 in: Cassava: Biology, Production and Utilization, R.J. Hillocks, J.M. Thresh, and A. Bellotti, eds. CABI Publishing, New York.
Caplan, J., Padmanabhan, M., and Dinesh-Kumar, S. P. 2008. Plant NB-LRR Immune Receptors: From Recognition to Transcriptional Reprogramming. Cell Host Microbe. 3:126–135
Carvajal-Yepes, M., Olaya, C., Lozano, I., Cuervo, M., Castaño, M., and Cuellar, W. J. 2014. Unraveling complex viral infections in cassava (Manihot esculenta Crantz) from Colombia. Virus Res. 186:76–86
De carvalho, R., and Guerra, M. 2002. Cytogenetics of Manihot esculenta Crantz (cassava) and eight related species. Hereditas. 136:159–168
Castiblanco, L. F., Gil, J., Rojas, A., Osorio, D., Gutiérrez, S., Muñoz-Bodnar, A., Perez-Quintero, A. L., Koebnik, R., Szurek, B., Lopez, C., and others. 2013. TALE1 from Xanthomonas axonopodis pv. manihotis acts as a transcriptional activator in plant cells and is important for pathogenicity in cassava plants. Mol. Plant Pathol. 14:84–95
Ceballos, H., Iglesias, C. A., Pérez, J. C., and Dixon, A. G. O. 2004. Cassava breeding: opportunities and challenges. Plant Mol. Biol. 56:503–516
Ceballos, H., Kulakow, P., and Hershey, C. 2012. Cassava breeding: current status, bottlenecks and the potential of biotechnology tools. Trop. Plant Biol. 5:73–87
59
Chaparro-Martinez, E. I., and Trujillo-Pinto, G. 2001. First report of frog skin disease in cassava (Manihot esculenta) in Venezuela. Plant Dis. 85:1285
Chen, X., Xia, Z., Fu, Y., Lu, C., and Wang, W. 2010. Constructing a genetic linkage map using an F1 population of non-inbred parents in cassava (Manihot esculenta Crantz). Plant Mol. Biol. Report. 28:676–683
Chen, С. T., KO, N. C., and Chen, M. J. 1981. Electron microscopy of cassava common mosaic in Taiwan. Rep. Taiwan Sugar Res. Inst. 9:20–27
Chinchilla, D., Bauer, Z., Regenass, M., Boller, T., and Felix, G. 2006. The Arabidopsis receptor kinase FLS2 binds flg22 and determines the specificity of flagellin perception. Plant Cell. 18:465–476
Chisholm, S. T., Coaker, G., Day, B., and Staskawicz, B. J. 2006. Host-Microbe Interactions: Shaping the Evolution of the Plant Immune Response. Cell. 124:803–814
Clement, C. R., de Cristo-Araújo, M., Coppens D’Eeckenbrugge, G., Alves Pereira, A., and Picanço-Rodrigues, D. 2010. Origin and domestication of native Amazonian crops. Diversity. 2:72–106
Coburn, B., Sekirov, I., and Finlay, B. B. 2007. Type III secretion systems and disease. Clin. Microbiol. Rev. 20:535–549
Cohn, M., Bart, R., Shybut, M., Dahlbeck, D., Gomez, M., Morbitzer, R., Hou, B.-H., Frommer, W., Lahaye, T., and Staskawicz, B. 2014. Xanthomonas axonopodis virulence is promoted by a transcription activator like (TAL) effector mediated induction of a SWEET sugar transporter in cassava. Mol. Plant. Microbe. Interact. 27:1186–1198
Cohn, M., Morbitzer, R., Lahaye, T., and Staskawicz, B. J. 2016. Comparison of gene activation by two TAL effectors from Xanthomonas axonopodis pv. manihotis reveals candidate host susceptibility genes in cassava. Mol. Plant Pathol.
Colcombet, J., and Hirt, H. 2008. Arabidopsis MAPKs: a complex signalling network involved in multiple biological processes. Biochem. J. 413:217–226
Collard, B. C. Y., Jahufer, M. Z. Z., Brouwer, J. B., and Pang, E. C. K. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: the basic concepts. Euphytica. 142:169–196
Colombo, C., Second, G., and Charrier, A. 2000. Diversity within American cassava germplasm based on RAPD markers. Genet. Mol. Biol. 23:189–199
Contreras, E., and López, C. E. 2008. Expresión de dos genes candidatos a resistencia contra la bacteriosis vascular en yuca. Acta Biológica Colomb. 13:175–188
Cook, D. E., Mesarich, C. H., and Thomma, B. P. H. J. 2015. Understanding Plant Immunity as a Surveillance System to Detect Invasion. Annu. Rev. Phytopathol. 53:541–563
60
Crossa, J., de Los Campos, G., Pérez, P., Gianola, D., Burgueño, J., Araus, J. L., Makumbi, D., Singh, R. P., Dreisigacker, S., Yan, J., and others. 2010. Prediction of genetic values of quantitative traits in plant breeding using pedigree and molecular markers. Genetics. 186:713–724
Cuervo, M., Moreno, M. G., Flor, N. C., Ramirez, J. L., Medina, C. A., and Debouck, D. G. 2010. Handbook of procedures of the Germplasm Health Laboratory. Health certification of cassava germplasm.
Dangl, J. L., and Jones, J. D. G. 2001. Plant pathogens and integrated defence responses to infection. Nature. 411:826–833
Davey, J. W., Hohenlohe, P. A., Etter, P. D., Boone, J. Q., Catchen, J. M., and Blaxter, M. L. 2011. Genome-wide genetic marker discovery and genotyping using next-generation sequencing. Nat. Rev. Genet. 12:499–510
Dedal, O. I., Palomar, M. K., Napiere, C. M., and others. 1980. Host range of Xanthomonas manihotis Starr. Ann. Trop. Res. 2:149–155
Dixon, A. G. O., Asiedu, R., and Bokanga, M. 1994. Breeding of cassava for low cyanogenic potential: problems, progress and prospects. Pages 153–162 in: International Workshop on Cassava Safety 375.
Dodds, P. N., Lawrence, G. J., Catanzariti, A.-M., Teh, T., Wang, C.-I., Ayliffe, M. A., Kobe, B., and Ellis, J. G. 2006. Direct protein interaction underlies gene-for-gene specificity and coevolution of the flax resistance genes and flax rust avirulence genes. Proc. Natl. Acad. Sci. 103:8888–8893
Dodds, P. N., and Rathjen, J. P. 2010. Plant immunity: towards an integrated view of plant--pathogen interactions. Nat. Rev. Genet. 11:539–548
Dufour, D. L. 2006. “Bitter” Cassava: Toxicity and Detoxification. Cassava Improv. to Enhanc. Livelihoods Sub-Saharan Africa Northeast. Brazil.
Dutt, N., Briddon, R. W., and Dasgupta, I. 2005. Identification of a second begomovirus, Sri Lankan cassava mosaic virus, causing cassava mosaic disease in India. Arch. Virol. 150:2101–2108
El-Sharkawy, M. A. 2003. Cassava biology and physiology. Plant Mol. Biol. 53:621–641
Failla, M. L., Chitchumroonchokchai, C., Siritunga, D., De Moura, F. F., Fregene, M., Manary, M. J., and Sayre, R. T. 2012. Retention during processing and bioaccessibility of β-carotene in high β-carotene transgenic cassava root. J. Agric. Food Chem. 60:3861–3866
Fanou, A. A. 1999. Epidemiological and ecological investigations on cassava bacterial blight and development of integrated methods for its control in Africa. Cuvillier.
FAO. 2010. The Second Report on the state of the world’s plant genetic resources for food and agriculture. Food Agric. Organ. United Nations, Rome.
61
FAO. 2015. Food Outlook, Biannual report on global food markets. Food Agric. Organ. United Nations, Rome
FAO. 2013. Save and Grow: Cassava. A Guide to Sustainable Production Intensification. Food Agric. Organ. United Nations, Rome.
Felix, G., Duran, J. D., Volko, S., and Boller, T. 1999. Plants have a sensitive perception system for the most conserved domain of bacterial flagellin. Plant J. 18:265–276
Ferguson, M., Rabbi, I., Kim, D.-J., Gedil, M., Lopez-Lavalle, L. A. B., and Okogbenin, E. 2012. Molecular markers and their application to cassava breeding: past, present and future. Trop. Plant Biol. 5:95–109
Flor, H. H. 1955. Host-parasite interaction in flax rust-its genetics and other implications. Phytopathology. 45:680–685
Fokunang, C. N., Ikotun, T., Dixon, A. G. O., and Akem, C. N. 1997. First report of Colletotrichum gloeosporioides f. sp. manihotis, cause of cassava anthracnose disease, being seed-borne and seed-transmitted in cassava. Plant Dis. 81:695
Fokunang, C. N., Ikotun, T., Dixon, A. G. O., Akem, C. N., and others. 2000. Field reaction of cassava genotypes to anthracnose, bacterial blight, cassava mosaic disease and their effects on yield. African Crop Sci. J. 8:179–186
Fregene, M., Angel, F., Gomez, R., Rodriguez, F., Chavarriaga, P., Roca, W., Tohme, J., and Bonierbale, M. 1997. A molecular genetic map of cassava (Manihot esculenta Crantz). TAG Theor. Appl. Genet. 95:431–441
Fregene, M., Okogbenin, E., Mba, C., Angel, F., Suarez, M. C., Janneth, G., Chavarriaga, P., Roca, W., Bonierbale, M., and Tohme, J. 2001. Genome mapping in cassava improvement: Challenges, achievements and opportunities. Euphytica. 120:159–165
de Freitas, J. P. X., da Silva Santos, V., and de Oliveira, E. J. 2016. Inbreeding depression in cassava for productive traits. Euphytica. 209:137–145
Fu, D., Uauy, C., Distelfeld, A., Blechl, A., Epstein, L., Chen, X., Sela, H., Fahima, T., and Dubcovsky, J. 2009. A kinase-START gene confers temperature-dependent resistance to wheat stripe rust. Science (80). 323:1357–1360
Fu, Y. B., Wangsomnuk, P. P., and Ruttawat, B. 2014. Thai elite cassava genetic diversity was fortuitously conserved through farming with different sets of varieties. Conserv. Genet. 15:1463–1478
Fukuoka, S., Saka, N., Koga, H., Ono, K., Shimizu, T., Ebana, K., Hayashi, N., Takahashi, A., Hirochika, H., Okuno, K., and Yano, M. 2009. Loss of function of a proline-containing protein confers durable disease resistance in rice. Science. 325:998–1001
Gebhardt, C., Li, L., Pajerowska-Mukthar, K., Achenbach, U., Sattarzadeh, A., Bormann, C., Ilarionova, E., and Ballvora, A. 2007. Candidate gene approach to
62
identify genes underlying quantitative traits and develop diagnostic markers in potato. Crop Sci. 47:S–106
Gebhardt, C., and Valkonen, J. P. T. 2001. Organization of genes controlling disease resistance in the potato genome. Annu. Rev. Phytopathol. 39:79–102
Gianola, D., and Fernando, R. L. 1986. Bayesian methods in animal breeding theory. J. Anim. Sci. 63:217–244
Glaubitz, J. C., Casstevens, T. M., Lu, F., Harriman, J., Elshire, R. J., Sun, Q., and Buckler, E. S. 2014. TASSEL-GBS: a high capacity genotyping by sequencing analysis pipeline. PLoS One. 9:e90346
Göhre, V., Spallek, T., Häweker, H., Mersmann, S., Mentzel, T., Boller, T., de Torres, M., Mansfield, J. W., and Robatzek, S. 2008. Plant pattern-recognition receptor FLS2 is directed for degradation by the bacterial ubiquitin ligase AvrPtoB. Curr. Biol. 18:1824–1832
Gold, C. S., Altieri, M. A., and Bellotti, A. C. 1991. Survivorship of the cassava whiteflies Aleurotrachelus socialis and Trialeurodes variabilis (Homoptera: Aleyrodidae) under different cropping systems in Colombia. Crop Prot. 10:305–309
Gómez-Gómez, L., and Boller, T. 2000. FLS2: an LRR receptor-like kinase involved in the perception of the bacterial elicitor flagellin in Arabidopsis. Mol. Cell. 5:1003–1011
Hahn, S. K., Bai, K. V, and Asiedu, R. 1990. Tetraploids, triploids, and 2n pollen from diploid interspecific crosses with cassava. Theor. Appl. Genet. 79:433–439
Hahn, S. K., Howland, A. K., and Terry, E. R. 1973. Cassava Breeding at IITA. IITA,[sd].
Hahn, S. K., Howland, A. K., and Terry, E. R. 1980. Correlated resistance of cassava to mosaic and bacterial blight diseases. Euphytica. 29:305–311
Hahn, S. K., Terry, E. R., Leuschner, K., Akobundu, I. O., Okali, C., and Lal, R. 1979. Cassava improvement in Africa. F. Crop. Res. 2:193–226
Halsey, M. E., Olsen, K. M., Taylor, N. J., and Chavarriaga-Aguirre, P. 2008. Reproductive Biology of Cassava (Crantz) and Isolation of Experimental Field Trials. Crop Sci. 48:49–58
Hammond-Kosack, K. E., and Kanyuka, K. 2007. Resistance genes (R genes) in plants. eLS.
He, P., Shan, L., and Sheen, J. 2007. Elicitation and suppression of microbe-associated molecular pattern-triggered immunity in plant--microbe interactions. Cell. Microbiol. 9:1385–1396
Henderson, C. R. 1984. Estimation of variances and covariances under multiple trait models. J. Dairy Sci. 67:1581–1589
63
Hillocks, R. J., Thresh, J. M., and Bellotti, A. 2002a. Cassava: Biology, production and utilization. CABI.
Hillocks, R. J., Thresh, J. M., Tomas, J., Botao, M., Macia, R., and Zavier, R. 2002b. Cassava brown streak disease in northern Mozambique. Int. J. Pest Manag. 48:178–181
Hillocks, R. J., and Wydra, K. 2002. Bacterial, Fungal and Nematode Diseases. Pages 261–280 in: Cassava: Biology, Production and Utilization, R.J. Hillocks, J.M. Thresh, and A. Bellotti, eds. CABI Publishing, New York.
van der Hoorn, R. A. L., and Kamoun, S. 2008. From guard to decoy: a new model for perception of plant pathogen effectors. Plant Cell. 20:2009–2017
Huard-Chauveau, C., Perchepied, L., Debieu, M., Rivas, S., Kroj, T., Kars, I., Bergelson, J., Roux, F., and Roby, D. 2013. An Atypical Kinase under Balancing Selection Confers Broad-Spectrum Disease Resistance in Arabidopsis. PLoS Genet. 9
Hurni, S., Scheuermann, D., Krattinger, S. G., Kessel, B., Wicker, T., Herren, G., Fitze, M. N., Breen, J., Presterl, T., Ouzunova, M., and others. 2015. The maize disease resistance gene Htn1 against northern corn leaf blight encodes a wall-associated receptor-like kinase. Proc. Natl. Acad. Sci. 112:8780–8785
Jander, G., Norris, S. R., Rounsley, S. D., Bush, D. F., Levin, I. M., and Last, R. L. 2002. Arabidopsis map-based cloning in the post-genome era. Plant Physiol. 129:440–450
Jannink, J.-L., Lorenz, A. J., and Iwata, H. 2010. Genomic selection in plant breeding: from theory to practice. Brief. Funct. Genomics. 9:166–177
Jehle, A. K., Fürst, U., Lipschis, M., Albert, M., and Felix, G. 2013. Perception of the novel MAMP eMax from different Xanthomonas species requires the Arabidopsis receptor-like protein ReMAX and the receptor kinase SOBIR. Plant Signal. Behav. 8:e27408
Jelenska, J., Yao, N., Vinatzer, B. A., Wright, C. M., Brodsky, J. L., and Greenberg, J. T. 2007. AJ domain virulence effector of Pseudomonas syringae remodels host chloroplasts and suppresses defenses. Curr. Biol. 17:499–508
Jennings, D. L. 1963. Variation in pollen and ovule fertility in varieties of cassava, and the effect of interspecific crossing on fertility. Euphytica. 12:69–76
Jones, J. D. G., and Dangl, J. L. 2006. The plant immune system. Nature. 444:323–329
Jorge, V., Fregene, M. A., Duque, M. C., Bonierbale, M. W., Tohme, J., and Verdier, V. 2000. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). Theor. Appl. Genet. 101:865–872
Jorge, V., Fregene, M., Vélez, C. M., Duque, M. C., Tohme, J., and Verdier, V. 2001. QTL analysis of field resistance to Xanthomonas axonopodis pv. manihotis in cassava. Theor. Appl. Genet. 102:564–571
64
Kaku, H., Nishizawa, Y., Ishii-Minami, N., Akimoto-Tomiyama, C., Dohmae, N., Takio, K., Minami, E., and Shibuya, N. 2006. Plant cells recognize chitin fragments for defense signaling through a plasma membrane receptor. Proc. Natl. Acad. Sci. U. S. A. 103:11086–11091
Kawuki, R. S., Herselman, L., Labuschagne, M. T., Nzuki, I., Ralimanana, I., Bidiaka, M., Kanyange, M. C., Gashaka, G., Masumba, E., Mkamilo, G., and others. 2013. Genetic diversity of cassava (Manihot esculenta Crantz) landraces and cultivars from southern, eastern and central Africa. Plant Genet. Resour. 11:170–181
Kearney, B., and Staskawicz, B. J. 1990. Widespread distribution and fitness contribution of Xanthomonas campestris avirulence gene avrBs2. Nature. 346:385–386
Kemmerling, B., Halter, T., Mazzotta, S., Mosher, S., and Nürnberger, T. 2011. A genome-wide survey for Arabidopsis leucine-rich repeat receptor kinases implicated in plant immunity. Front. Plant Sci. 2:88
Kou, Y., and Wang, S. 2010. Broad-spectrum and durability: understanding of quantitative disease resistance. Curr. Opin. Plant Biol. 13:181–185
Krattinger, S. G., Lagudah, E. S., Spielmeyer, W., Singh, R. P., Huerta-Espino, J., McFadden, H., Bossolini, E., Selter, L. L., and Keller, B. 2009. A putative ABC transporter confers durable resistance to multiple fungal pathogens in wheat. Science. 323:1360–1363
Kunkeaw, S., Tangphatsornruang, S., Smith, D. R., and Triwitayakorn, K. 2010. Genetic linkage map of cassava (Manihot esculenta Crantz) based on AFLP and SSR markers. Plant Breed. 129:112–115
Kunkeaw, S., Yoocha, T., Sraphet, S., Boonchanawiwat, A., Boonseng, O., Lightfoot, D. A., Triwitayakorn, K., and Tangphatsornruang, S. 2011. Construction of a genetic linkage map using simple sequence repeat markers from expressed sequence tags for cassava (Manihot esculenta Crantz). Mol. Breed. 27:67–75
Kunze, G., Zipfel, C., Robatzek, S., Niehaus, K., Boller, T., and Felix, G. 2004. The N terminus of bacterial elongation factor Tu elicits innate immunity in Arabidopsis plants. Plant Cell. 16:3496–3507
Lancaster, P. A., and Brooks, J. E. 1983. Cassava leaves as human food. Econ. Bot. 37:331–348
Larotonda, F. D. S., Matsui, K. N., Soldi, V., and Laurindo, J. B. 2004. Biodegradable films made from raw and acetylated cassava starch. Brazilian Arch. Biol. Technol. 47:477–484
Leal, L. G., Perez, Á., Quintero, A., Bayona, Á., Ortiz, J. F., Gangadharan, A., Mackey, D., López, C., and López-Kleine, L. 2013. Identification of immunity-related genes in Arabidopsis and Cassava using genomic data. Genomics. Proteomics Bioinformatics. 11:345–353
65
Lebot, V. 2009. Tropical root and tuber crops: cassava, sweet potato, yams and aroids. CABI.
Legg, J. P. 2009. Bemisia tabaci: The whitefly vector of cassava mosaic geminiviruses in Africa: An ecological perspective.
Legg, J. P. 1999. Emergence, spread and strategies for controlling the pandemic of cassava mosaic virus disease in east and central Africa. Crop Prot. 18:627–637
Legga, J. P., and Threshb, J. M. 2003. Cassava virus diseases in Africa.
Li, S., Wang, Y., Wang, S., Fang, A., Wang, J., Liu, L., Zhang, K., Mao, Y., and Sun, W. 2015. The type III effector AvrBs2 in Xanthomonas oryzae pv. oryzicola suppresses rice immunity and promotes disease development. Mol. Plant-Microbe Interact. 28:869–880
Liu, B., Li, J.-F., Ao, Y., Qu, J., Li, Z., Su, J., Zhang, Y., Liu, J., Feng, D., Qi, K., He, Y., Wang, J., and Wang, H.-B. 2012. Lysin Motif-Containing Proteins LYP4 and LYP6 Play Dual Roles in Peptidoglycan and Chitin Perception in Rice Innate Immunity. Plant Cell. 24:3406–3419
Van Loon, L. C., and Van Strien, E. A. 1999. The families of pathogenesis-related proteins, their activities, and comparative analysis of PR-1 type proteins. Physiol. Mol. Plant Pathol. 55:85–97
López, C. 2011. Descifrando las bases moleculares de la resistencia cuantitativa. Acta Biologica Colomb. 16:3
López, C. E., and Bernal, A. J. 2012. Cassava Bacterial Blight: Using Genomics for the Elucidation and Management of an Old Problem. Trop. Plant Biol. 5:117–126
López, C. E., Quesada-Ocampo, L. M., Bohorquez, A., Duque, M. C., Vargas, J., Tohme, J., and Verdier, V. 2007. Mapping EST-derived SSRs and ESTs involved in resistance to bacterial blight in Manihot esculenta. Genome. 50:1078–1088
López, C. E., Zuluaga, A. P., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2003. Isolation of resistance gene candidates (RGCs) and characterization of an RGC cluster in cassava. Mol. Genet. genomics. 269:658–671
López, C., Soto, M., Restrepo, S., Piégu, B., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2005. Gene expression profile in response to Xanthomonas axonopodis pv. manihotis infection in cassava using a cDNA microarray. Plant Mol. Biol. 57:393–410
Lozano, J. 1986. Cassava Bacterial Blight: a Manageable Disease. Plant Dis. 70:1089–1093
Lozano, J. C., and Terry, E. R. 1976. Cassava diseases and their control. Proc. Symp. Int. Soc. Trop. Root Crop. 4th. J. Cook, R. McIntyre, M. Graham, eds. CIAT, Cali, Colomb. :156–160
Lozano, R., Hamblin, M. T., Prochnik, S., and Jannink, J.-L. 2015. Identification and distribution of the NBS-LRR gene family in the Cassava genome. BMC Genomics. 16:1
66
Ly, D., Hamblin, M., Rabbi, I., Melaku, G., Bakare, M., Gauch, H. G., Okechukwu, R., Dixon, A. G. O., Kulakow, P., and Jannink, J.-L. 2013. Relatedness and genotype$\times$ environment interaction affect prediction accuracies in genomic selection: a study in cassava. Crop Sci. 53:1312–1325
Magoon, M. L., Krishnan, R., and Bai, K. V. 1969. Morphology of the pachytene chromosomes and meiosis in Manihot esculenta Crantz. Cytologia (Tokyo). 34:612–626
Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., Dow, M., Verdier, V., Beer, S. V., Machado, M. a., Toth, I., Salmond, G., and Foster, G. D. 2012. Top 10 plant pathogenic bacteria in molecular plant pathology. Mol. Plant Pathol. 13:614–629
Maraite, H., and Meyer, J. A. 1975. Xanthomonas manihotis (Arthaud-Berthet) Starr, causal agent of bacterial wilt, blight and leaf spots of cassava in Zaire. PANS. 21:27–36
Mba, R. E. C., Stephenson, P., Edwards, K., Melzer, S., Nkumbira, J., Gullberg, U., Apel, K., Gale, M., Tohme, J., and Fregene, M. 2001. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: Towards an SSR-based molecular genetic map of cassava. Theor. Appl. Genet. 102:21–31
Mezette, T. F., Blumer, C. G., and Veasey, E. A. 2013. Morphological and molecular diversity among cassava genotypes. Pesqui. Agropecuaria Bras. 48:510–518
Miya, A., Albert, P., Shinya, T., Desaki, Y., Ichimura, K., Shirasu, K., Narusaka, Y., Kawakami, N., Kaku, H., and Shibuya, N. 2007. CERK1, a LysM receptor kinase, is essential for chitin elicitor signaling in Arabidopsis. Proc. Natl. Acad. Sci. U. S. A. 104:19613–8
Mlingi, N., Poulter, N. H., and Rosling, H. 1992. An outbreak of acute intoxications from consumption of insufficiently processed cassava in Tanzania. Nutr. Res. 12:677–687
Monger, W. A., Seal, S., Isaac, A. M., and Foster, G. D. 2001. Molecular characterization of the Cassava brown streak virus coat protein. Plant Pathol. 50:527–534
Montaldo, A., and Gunz, T. 1985. La yuca o mandioca. IICA.
Moses, E., Akrofi, S., and Mensah, G. A. 2007. Characteristics and control of a new basidiomycetous root rot of cassava (Mannihot esculenta) in Ghana. Pages 307–311 in: Proceedings of the 13th ISTRC Symposium,
Mtunda, K. J., Muhanna, M., Raya, M. D., and Kanju, E. E. 2003. Current status of Cassava brown streak virus disease in Tanzania. Cassava Brown Streak Virus Dis. Past, Present. Futur.
Muñoz-Bodnar, A., Perez-Quintero, A. L., Gomez-Cano, F., Gil, J., Michelmore, R., Bernal, A., Szurek, B., and Lopez, C. 2014. RNAseq analysis of cassava reveals
67
similar plant responses upon infection with pathogenic and non-pathogenic strains of Xanthomonas axonopodis pv. manihotis. Plant Cell Rep. 33:1901–1912
Narbona, E., Ortiz, P. L., and Arista, M. 2011. Linking self-incompatibility, dichogamy, and flowering synchrony in two euphorbia species: Alternative mechanisms for avoiding self-fertilization? PLoS One. 6
Nassar, N. M. A., and Sousa, M. V. 2007. Amino acid profile in cassava and its interspecific hybrid. Genet. Mol. Res. 6:292–297
Nassar, N., and Ortiz, R. 2010. Breeding cassava to feed the poor. Sci. Am. 302:78–84
Nichols, R. F. W. 1950. The brown streak disease of cassava: distribution, climatic effects and diagnostic symptoms. East African Agric. J. 15:154–160
Nicolau, A. I. 2015. Safety of Fermented Cassava Products. Regul. Saf. Tradit. Ethn. Foods. :319
Njeru, R. W., and Munga, T. L. 2003. Current status of cassava brown streak disease in Kenya. Cassava Brown Streak Virus Dis. Past, Present Futur. :12
Ogunjobi, A. A. 2006. Molecular variation in population structure of Xanthomonas axonopodis pv manihotis in the south eastern Nigeria. African J. Biotechnol. 5
Ogunjobi, A. A., Dixon, A. G. O., and Fagade, O. E. 2007. Molecular genetic study of cassava bacterial blight casual agent in Nigeria using random amplified polymorphic DNA. Electron. J. Environ. Agric. Food Chem. 6:2364–2376
Okogbenin, E., Egesi, C. N., Olasanmi, B., Ogundapo, O., Kahya, S., Hurtado, P., Marin, J., Akinbo, O., Mba, C., Gomez, H., and others. 2012. Molecular marker analysis and validation of resistance to cassava mosaic disease in elite cassava genotypes in Nigeria. Crop Sci. 52:2576–2586
Okogbenin, E., Porto, M. C. M., Egesi, C., Mba, C., Espinosa, E., Santos, L. G., Ospina, C., Marin, J., Barrera, E., Gutiérrez, J., and others. 2007. Marker-assisted introgression of resistance to cassava mosaic disease into Latin American germplasm for the genetic improvement of cassava in Africa. Crop Sci. 47:1895–1904
de Oliveira, E. J., de Resende, M. D. V., da Silva Santos, V., Ferreira, C. F., Oliveira, G. A. F., da Silva, M. S., de Oliveira, L. A., and Aguilar-Vildoso, C. I. 2012. Genome-wide selection in cassava. Euphytica. 187:263–276
Olomo, V., and Ajibola, O. 2003. Processing factors affecting the yield and physicochemical properties of starch from cassava chips and flour. STARCH-STUTTGART-. 55:476–481
Olsen, K. M., and Schaal, B. A. 1999. Evidence on the origin of cassava: phylogeography of Manihot esculenta. Proc. Natl. Acad. Sci. U. S. A. 96:5586–5591
Olsen, K. M., and Schaal, B. A. 2001. Microsatellite variation in cassava (Manihot esculenta, Euphorbiaceae) and its wild relatives: further evidence for a southern Amazonian origin of domestication. Am. J. Bot. 88:131–142
68
Onyango, D. M., Mukunya, D. M., and others. 1980. Distribution and importance of Xanthomonas manihotis and X. cassavae in East Africa. in: Root crops in Eastern Africa: Proceedings of a workshop held in Kigali, Rwanda, 23-27 November 1980.,
Ospina, B., and Ceballos, H. 2002. La yuca en el tercer Milenio: Sistemas Modernos de produccion, procesamiento, utilizacion y comercialización. CIAT.
Patanun, O., Lertpanyasampatha, M., Sojikul, P., Viboonjun, U., and Narangajavana, J. 2013. Computational identification of microRNAs and their targets in cassava (Manihot esculenta Crantz.). Mol. Biotechnol. 53:257–269
Pérez-Quintero, Á. L., Quintero, A., Urrego, O., Vanegas, P., and López, C. 2012. Bioinformatic identification of cassava miRNAs differentially expressed in response to infection by Xanthomonas axonopodis pv. manihotis. BMC Plant Biol. 12:1
Pfeiffer, W. H., and McClafferty, B. 2007. Biofortification: breeding micronutrient-dense crops. Breed. major food staples. 61–91
Pflieger, S., Lefebvre, V., and Causse, M. 2001. The candidate gene approach in plant genetics: a review. Mol. Breed. 7:275–291
Pineda, B., Jayasinghe, U., and Lozano, J. C. 1983. La enfermedad “cuero de sapo” en yuca (Manihot esculenta Crantz). ASIAVA (Colombia). 4:10–12
Poland, J. A., Balint-Kurti, P. J., Wisser, R. J., Pratt, R. C., and Nelson, R. J. 2009. Shades of gray: the world of quantitative disease resistance. Trends Plant Sci. 14:21–29
Pritchard, L., and Birch, P. R. J. 2014. The zigzag model of plant--microbe interactions: is it time to move on? Mol. Plant Pathol. 15:865–870
Prochnik, S., Marri, P. R., Desany, B., Rabinowicz, P. D., Kodira, C., Mohiuddin, M., Rodriguez, F., Fauquet, C., Tohme, J., Harkins, T., Rokhsar, D. S., and Rounsley, S. 2012. The Cassava Genome: Current Progress, Future Directions. Trop. Plant Biol. 5:88–94
Pruitt, R. N., Schwessinger, B., Joe, A., Thomas, N., Liu, F., Albert, M., Robinson, M. R., Chan, L. J. G., Luu, D. D., Chen, H., and others. 2015. The rice immune receptor XA21 recognizes a tyrosine-sulfated protein from a Gram-negative bacterium. Sci. Adv. 1:e1500245
Rabbi, I., Hamblin, M., Gedil, M., Kulakow, P., Ferguson, M., Ikpan, A. S., Ly, D., and Jannink, J.-L. 2014a. Genetic mapping using genotyping-by-sequencing in the clonally propagated cassava. Crop Sci. 54:1384–1396
Rabbi, I. Y., Hamblin, M. T., Kumar, P. L., Gedil, M. A., Ikpan, A. S., Jannink, J.-L., and Kulakow, P. A. 2014b. High-resolution mapping of resistance to cassava mosaic geminiviruses in cassava using genotyping-by-sequencing and its implications for breeding. Virus Res. 186:87–96
69
Rabbi, I. Y., Kulembeka, H. P., Masumba, E., Marri, P. R., and Ferguson, M. 2012. An EST-derived SNP and SSR genetic linkage map of cassava (Manihot esculenta Crantz). Theor. Appl. Genet. 125:329–342
Raji, A. A. J., Anderson, J. V, Kolade, O. A., Ugwu, C. D., Dixon, A. G. O., and Ingelbrecht, I. L. 2009. Gene-based microsatellites for cassava (Manihot esculenta Crantz): prevalence, polymorphisms, and cross-taxa utility. BMC Plant Biol. 9:1
Ramalingam, J., Vera Cruz, C. M., Kukreja, K., Chittoor, J. M., Wu, J.-L., Lee, S. W., Baraoidan, M., George, M. L., Cohen, M. B., Hulbert, S. H., and others. 2003. Candidate defense genes from rice, barley, and maize and their association with qualitative and quantitative resistance in rice. Mol. Plant-Microbe Interact. 16:14–24
Rasmussen, M. W., Roux, M., Petersen, M., and Mundy, J. 2012. MAP kinase cascades in Arabidopsis innate immunity. Front. Plant Sci. 3:169
Ray, R. C., and Sivakumar, P. S. 2009. Traditional and novel fermented foods and beverages from tropical root and tuber crops: Review. Int. J. Food Sci. Technol. 44:1073–1087
Restrepo, S., Duque, M. C., and Verdier, V. 2000. Characterization of pathotypes among isolates of Xanthomonas axonopodis pv. manihotis in Colombia. Plant Pathol. 49:680–687
Restrepo, S., Velez, C. M., Duque, M. C., and Verdier, V. 2004. Genetic structure and population dynamics of Xanthomonas axonopodis pv. manihotis in Colombia from 1995 to 1999. Appl. Environ. Microbiol. 70:255–261
Rojas, M. C., Pérez, J. C., Ceballos, H., Baena, D., Morante, N., and Calle, F. 2009. Analysis of Inbreeding Depression in Eight S Cassava Families. Crop Sci. 49:543–548
Román, V., Bossa-Castro, A. M., Vásquez, A., Bernal, V., Schuster, M., Bernal, N., and López, C. 2014. Construction of a cassava PR protein-interacting network during Xanthomonas axonopodis pv. manihotis infection. Plant Pathol. 63:792–802
Rudi, N., Norton, G. W., Alwang, J., Asumugha, G., and others. 2010. Economic impact analysis of marker-assisted breeding for resistance to pests and post-harvest deterioration in cassava. Afr J Agric Resour Econ. 4:110–122
Russell, G. E. 2013. Progress in Plant Breeding—1. Elsevier.
Sakurai, T., Mochida, K., Yoshida, T., Akiyama, K., Ishitani, M., Seki, M., and Shinozaki, K. 2013. Genome-wide discovery and information resource development of DNA polymorphisms in cassava. PLoS One. 8:e74056
Salvi, S., and Tuberosa, R. 2005. To clone or not to clone plant QTLs: present and future challenges. Trends Plant Sci. 10:297–304
Sánchez, T., Salcedo, E., Ceballos, H., Dufour, D., Mafla, G., Morante, N., Calle, F., Pérez, J. C., Debouck, D., Jaramillo, G., and others. 2009. Screening of starch quality traits in cassava (Manihot esculenta Crantz). Starch-St{ä}rke. 61:12–19
70
Sayre, R., Beeching, J. R., Cahoon, E. B., Egesi, C., Fauquet, C., Fellman, J., Fregene, M., Gruissem, W., Mallowa, S., Manary, M., Maziya-Dixon, B., Mbanaso, A., Schachtman, D. P., Siritunga, D., Taylor, N., Vanderschuren, H., and Zhang, P. 2011. The BioCassava Plus Program: Biofortification of Cassava for Sub-Saharan Africa. Annu. Rev. Plant Biol. 62:251–272
Schulze-Lefert, P., and Panstruga, R. 2011. A molecular evolutionary concept connecting nonhost resistance, pathogen host range, and pathogen speciation. Trends Plant Sci. 16:117–125
Da Silva, R. M., Bandel, G., and Martins, P. S. 2003. Mating system in an experimental garden composed of cassava (Manihot esculenta Crantz) ethnovarieties. Euphytica. 134:127–135
Siritunga, D., and Sayre, R. T. 2003. Generation of cyanogen-free transgenic cassava. Planta. 217:367–373
Song, W.-Y., Wang, G.-L., Chen, L.-L., Kim, H.-S., and others. 1995. A receptor kinase-like protein encoded by the rice disease resistance gene, Xa21. Science (80). 270:1804
Soto, J. C., Ortiz, J. F., Perlaza-Jiménez, L., Vásquez, A. X., López-Lavalle, L. A. B., Mathew, B., Léon, J., Bernal, A. J., Ballvora, A., and López, C. E. 2015. A genetic map of cassava (Manihot esculenta Crantz) with integrated physical mapping of immunity-related genes. BMC Genomics. 16:190
de Souza, A. N., da Silva, F. N., Bedendo, I. P., and Carvalho, C. M. 2014. A phytoplasma belonging to a 16SrIII-A subgroup and dsRNA virus associated with cassava frogskin disease in Brazil. Plant Dis. 98:771–779
Steenkamp, V., and McCrindle, C. M. 2014. Production, consumption and nutritional value of cassava (Manihot esculenta, Crantz) in Mozambique: An overview. J. Agric. Biotechnol. Sustain. Dev. 6:29–38
Storey, H. H. 1936. Virus diseases of East African plants. VI-A progress report on studies of the disease of cassava. East African Agric. J. 2:34–39
Strange, R. N., and Gullino, M. L. 2010. The role of plant pathology in food safety and food security. Springer.
Sudisha, J., Sharathchandra, R. G., Amruthesh, K. N., Kumar, A., and Shetty, H. S. 2012. Pathogenesis related proteins in plant defense response. Pages 379–403 in: Plant defence: biological control, Springer.
Sugawara, F., Strobel, S., Strobel, G., Larsen, R. D., Berglund, D. L., Gray, G., Takahashi, N., Coval, S. J., Stout, T. J., and Clardy, J. 1991. The structure and biological activity of cercosporamide from Cercosporidium henningsii. J. Org. Chem. 56:909–910
Taiwo, K. A. 2006. Utilization potentials of cassava in Nigeria: the domestic and industrial products. Food Rev. Int. 22:29–42
71
Takagi, H., Abe, A., Yoshida, K., Kosugi, S., Natsume, S., Mitsuoka, C., Uemura, A., Utsushi, H., Tamiru, M., Takuno, S., and others. 2013. QTL-seq: rapid mapping of quantitative trait loci in rice by whole genome resequencing of DNA from two bulked populations. Plant J. 74:174–183
Takai, R., Isogai, A., Takayama, S., and Che, F. 2008. Analysis of flagellin perception mediated by flg22 receptor OsFLS2 in rice. Mol. Plant-Microbe Interact. 21:1635–1642
Talsma, E. F., Brouwer, I. D., Verhoef, H., Mbera, G. N. K., Mwangi, A. M., Demir, A. Y., Maziya-Dixon, B., Boy, E., Zimmermann, M. B., and Melse-Boonstra, A. 2016. Biofortified yellow cassava and vitamin A status of Kenyan children: a randomized controlled trial. Am. J. Clin. Nutr. 103:258–267
Tao, Y., Xie, Z., Chen, W., Glazebrook, J., Chang, H.-S., Han, B., Zhu, T., Zou, G., and Katagiri, F. 2003. Quantitative nature of Arabidopsis responses during compatible and incompatible interactions with the bacterial pathogen Pseudomonas syringae. Plant Cell. 15:317–330
Téllez, L. C., Pardo, J. M., Zacher, M., Torres, A., and Álvarez, E. 2016. First Report of a 16SrIII Phytoplasma Associated with Frogskin Disease in Cassava (Manihot esculenta) in Paraguay. Plant Dis.
Teri, J. M., Thurston, H. D., Lozano, J. C., Brekelbaum, T., Bellotti, A., and others. 1978. The Cercospora leaf diseases of cassava. Pages 101–116 in: Proceedings, cassava protection workshop, CIAT, Cali, Colombia, 7-12 November, 1977.,
Thomma, B. P. H. J., Nurnberger, T., and Joosten, M. H. A. J. 2011. Of PAMPs and Effectors: The Blurred PTI-ETI Dichotomy. Plant Cell. 23:4–15
Tian, D., Wang, J., Zeng, X., Gu, K., Qiu, C., Yang, X., Zhou, Z., Goh, M., Luo, Y., Murata-Hori, M., White, F. F., and Yin, Z. 2014. The Rice TAL Effector-Dependent Resistance Protein XA10 Triggers Cell Death and Calcium Depletion in the Endoplasmic Reticulum. Plant Cell.
Trujillo, C. A., Ochoa, J. C., Mideros, M. F., Restrepo, S., López, C., and Bernal, A. 2014. A Complex Population Structure of the Cassava Pathogen Xanthomonas axonopodis pv. manihotis in Recent Years in the Caribbean Region of Colombia. Microb. Ecol.
Ufuan Achidi, A., Ajayi, O. A., Bokanga, M., and Maziya-Dixon, B. 2005. The use of cassava leaves as food in Africa. Ecol. Food Nutr. 44:423–435
Umanah, E. E., and Hartmann, R. W. 1973. Chromosome numbers and karyotypes of some Manihot species. J Am Soc Hortic Sci. 98:272–274
Valverde Gamboa, W. D. 2015. Metodologías de diagnóstico temprano de la enfermedad“ Cuero de sapo” en yuca (Manihot esculenta Crantz).
Verdier, V., Boher, B., Maraite, H., and Geiger, J.-P. 1994. Pathological and molecular characterization of Xanthomonas campestris strains causing diseases of cassava (Manihot esculenta). Appl. Environ. Microbiol. 60:4478–4486
72
Verdier, V., Restrepo, S., Mosquera, G., Jorge, V., and López, C. 2004. Recent progress in the characterization of molecular determinants in the Xanthomonas axonopodis pv. manihotis-cassava interaction. Plant Mol. Biol. 56:573–584
Wang, W., Feng, B., Xiao, J., Xia, Z., Zhou, X., Li, P., Zhang, W., Wang, Y., Møller, B. L., Zhang, P., Luo, M.-C., Xiao, G., Liu, J., Yang, J., Chen, S., Rabinowicz, P. D., Chen, X., Zhang, H.-B., Ceballos, H., Lou, Q., Zou, M., Carvalho, L. J. C. B., Zeng, C., Xia, J., Sun, S., Fu, Y., Wang, H., Lu, C., Ruan, M., Zhou, S., Wu, Z., Liu, H., Kannangara, R. M., Jørgensen, K., Neale, R. L., Bonde, M., Heinz, N., Zhu, W., Wang, S., Zhang, Y., Pan, K., Wen, M., Ma, P.-A., Li, Z., Hu, M., Liao, W., Hu, W., Zhang, S., Pei, J., Guo, A., Guo, J., Zhang, J., Zhang, Z., Ye, J., Ou, W., Ma, Y., Liu, X., Tallon, L. J., Galens, K., Ott, S., Huang, J., Xue, J., An, F., Yao, Q., Lu, X., Fregene, M., López-Lavalle, L. A. B., Wu, J., You, F. M., Chen, M., Hu, S., Wu, G., Zhong, S., Ling, P., Chen, Y., Wang, Q., Liu, G., Liu, B., Li, K., and Peng, M. 2014. Cassava genome from a wild ancestor to cultivated varieties. Nat. Commun. 5:5110
Webster, G. L. 1994. Classification of the Euphorbiaceae. Ann. Missouri Bot. Gard. :3–32
Wydra, K., and Verdier, V. 2002. Occurrence of cassava diseases in relation to environmental, agronomic and plant characteristics. Agric. Ecosyst. Environ. 93:211–226
Wydra, K., Zinsou, V., Jorge, V., and Verdier, V. 2004. Identification of Pathotypes of Xanthomonas axonopodis pv. manihotis in Africa and Detection of Quantitative Trait Loci and Markers for Resistance to Bacterial Blight of Cassava. Phytopathology. 94:1084–1093
Xu, Y., and Crouch, J. H. 2008. Marker-assisted selection in plant breeding: from publications to practice. Crop Sci. 48:391–407
Young, N. D. 1996. QTL mapping and quantitative disease resistance in plants. Annu. Rev. Phytopathol. 34:479–501
Zeng, C., Wang, W., Zheng, Y., Chen, X., Bo, W., Song, S., Zhang, W., and Peng, M. 2009. Conservation and divergence of microRNAs and their functions in Euphorbiaceous plants. Nucleic Acids Res. :gkp1035
Zhang, J., Yin, Z., and White, F. 2015. TAL effectors and the executor R genes. Front. Plant Sci. 6:1–9
Zhang, J.-Y., Qiao, Y.-S., Lv, D., Gao, Z.-H., Qu, S.-C., and Zhang, Z. 2012. Malus hupehensis NPR1 induces pathogenesis-related protein gene expression in transgenic tobacco. Plant Biol. 14:46–56
Zhu, C., Gore, M., Buckler, E. S., and Yu, J. 2008. Status and Prospects of Association Mapping in Plants. Plant Genome J. 1:5
Zipfel, C. 2008. Pattern-recognition receptors in plant innate immunity. Curr. Opin. Immunol. 20:10–16
73
Zipfel, C., Kunze, G., Chinchilla, D., Caniard, A., Jones, J. D. G., Boller, T., and Felix, G. 2006. Perception of the bacterial PAMP EF-Tu by the receptor EFR restricts Agrobacterium-mediated transformation. Cell. 125:749–760
74
CHAPTER 2
"Always let your conscience be your guide"
-The Blue Fairy. The Adventures of Pinocchio, 1940
RNA-seq: herramienta transcriptómica útil para el estudio de
interacciones planta patógeno
1Johana Carolina Soto Sedano y 2Camilo Ernesto López Carrascal
1, 2. Manihot Biotec Laboratory, Biology department, Universidad Nacional de Colombia, Bogotá,
Colombia.
Published in Fitosanidad 16(2) August (2012) 101-113
Resumen
El conocimiento del transcriptoma y su regulación es fundamental para la
interpretación articulada de los diversos constituyentes moleculares que integran la
red de respuesta génica ante un determinado evento inductor, como los que se
presentan en interacciones planta patógeno. La actual tecnología de secuenciación ha
llevado al desarrollo del RNA-seq como herramienta transcriptómica que permite el
secuenciamiento masivo de ADNc o ARN y hace posible obtener perfiles de expresión
génica de las respuestas de defensa, lo cual ofrece posibilidades para profundizar en
el entendimiento de los mecanismos que se activan durante las respuestas inmunes
en plantas. El RNA-seq está cambiando la manera de cómo se estudian los
transcriptomas y ha permitido identificar y cuantificar nuevos y conocidos
transcriptos relacionados con defensa vegetal. Aquí se presenta el principio,
aplicaciones y ventajas del RNA-seq, además se discuten trabajos recientes que
revelan la importancia y utilidad de esta herramienta en estudios de interacciones
planta patógeno.
Palabras clave: secuenciación, ADNc, fitopatología
Abstract
The knowledge gained from transcriptome and its regulation is essential to articulate
the constituents that make up the molecular network of response gene induction for
a certain event, such as those occurring in plant pathogen interactions. The current
sequencing technology has led to the development of RNA-seq as a tool that enables
mass sequencing of cDNA or RNA, making possible to obtain gene expression profiles
for defense responses, which offers great potential to deepen the understanding of
mechanisms that are activated during immune responses in plants. The RNA-seq is
76
changing the way of how we used to study the transcriptome and has helped identify
and quantify new and known plant defense related transcripts. Here, we present the
basis, applications and advantages of RNA-seq, also we discuss recent studies that
have revealed the importance and usefulness of this tool for studies of plant pathogen
interactions.
Key words: cDNA, sequencing, phytopathology
Introducción
Celularmente, la información genética cifrada en el ADN y contenida en los genes se
expresa través de los mecanismos de transcripción y traducción a partir del cual se
producen moléculas de ARNm y proteínas, respectivamente. Eventos celulares tales
como la replicación, la diferenciación y la división celular y otros caracteres
macroscópicos tales como rasgos fenotípicos, morfológicos, funcionales y de
respuesta ante estímulos son producto de la expresión diferencial de genes. En
plantas el control de la respuesta frente a estados de estrés biótico y abiótico está
dado por la actividad transcripcional de activación o represión de genes (Proudfoot
et al., 2002). La transcripción es un proceso nuclear cuya activación depende de
estímulos intra o extracelulares que activan cascadas de señalización para
determinar cuáles genes deben expresarse o reprimirse de acuerdo al tipo de
estímulo inicial.
La regulación de la transcripción depende de la unión de activadores o represores en
los elementos del promotor ubicados en la región 5’ de la secuencia codificante. Los
activadores o represores dictaminan la tasa de síntesis de ARNm que debe producir
la maquinaria basal de transcripción, la cual está constituida por los factores de
transcripción generales (GTFIIs) y la ARN polimerasa II (Proudfoot et al., 2002).
La cantidad de moléculas producidas de determinado ARNm depende de la función
que éste tenga en un proceso celular específico. Así, cuando se requiera dar respuesta
a una condición determinada en la cual un gen juega un papel importante, más
moléculas de este transcrito se producirán. De manera similar bajo ciertas
circunstancias particulares hay genes que permanecen apagados pero un estímulo
hace que se expresen iniciándose entonces la transcripción. De esta manera la
determinación de dónde, cómo y cuándo es generado un transcripto, bajo una
condición dada, es fundamental para el entendimiento de la actividad biológica de un
gen. Más aún, los niveles de ARNm pueden dar no sólo una visión clara de patrones
77
de expresión sino también cuantificaciones altamente correlacionadas entre cambios
en la abundancia de ARNm con cambios en la abundancia de proteínas (Lockhart y
Winzeler, 2000). En conjunto, todos los transcritos derivados de genes que se
producen en una célula en un momento y bajo una condición fisiológica determinada
se denomina transcriptoma.
El estudio y análisis del transcriptoma es esencial para el entendimiento de la función
de genes. De manera general se puede establecer que si un gen se expresa en una
condición o célula determinada es porque cumple allí una función. El estudio global
del transcriptoma permite también establecer patrones de regulación génica
coordinada lo que contribuye no sólo a dilucidar la función y agrupamiento de varios
genes bajo un estímulo o condición específica sino también a identificar elementos
promotores comunes a varios genes. En la década de los 90´s, los northern blots, los
microarreglos de ADNc (ADN complementario obtenido por transcriptasa inversa a
partir de ARNm), los cDNA-AFLPs y el análisis serial de expresión de genes SAGE (del
inglés serial analysis of gene expression) entre otras técnicas, permitieron el
desarrollo y generación de conocimiento en transcriptómica, al estudiar la expresión
de genes relacionados con respuestas a estímulos o a condiciones particulares, así
como para determinar cambios en los patrones de expresión génica en tratamientos y
cinéticas de expresión (Shalon et al., 1996; Schena et al., 1998; Meyers et al., 2004;
Marguerat y Bahler, 2010). Sin embargo, estas estrategias resultan limitantes al estar
basadas en hibridación, tener baja cobertura y en algunos casos necesitar algún
conocimiento previo de la secuencia del genoma para su implementación (Ward et
al., 2012).
Actualmente y gracias a los avances en las técnicas de secuenciación del ADN, a
través de tecnologías de nueva generación, NGS (del inglés Next Generation
Sequencing), se han revolucionado campos como los de la genómica y la
transcriptómica. Estas tecnologías han permitido no sólo generar información con
altos rendimientos y a bajo costo, sino también, abrir nuevos horizontes para el
entendimiento detallado y global de procesos de expresión génica (Mochida y
Shinozaki, 2011; Schneeberger y Weigel, 2011; Ward et al., 2012).
La caracterización completa y el análisis global de la expresión génica en una célula o
tejido, aún sin ninguna información genómica previa, es ahora posible a través de la
implementación de la secuenciación de ADNc o más recientemente de la
secuenciación directa de ARN, tecnología conocida como RNA-seq (Wang et al., 2009;
Garber et al., 2011; Egan et al., 2012; Ward et al., 2012). Esta herramienta
transcriptómica está cambiando la manera como se analizan y comprenden los
transcriptomas (Wang et al., 2009). Además, el RNA-seq da una cobertura completa
78
de transcriptos, genera información no sólo de la secuencia, sino también de la
estructura de exones y posibles eventos de splicing alternativo (Lister et al., 2009;
Gulledge et al., 2012). La información de esta manera puede ser integrada e
interpretada constituyéndose de gran utilidad para vislumbrar procesos biológicos y
mecanismos de coexpresión.
En el poco tiempo que esta tecnología se encuentra disponible se han desarrollado un
grupo relativamente amplio de investigaciones dirigidas a caracterizar y a cuantificar
transcriptomas así como a comprender los mecanismos de la variación de la
expresión génica. Las aplicaciones de RNA-seq se han llevado a cabo en especies
eucariotas tales como Saccharomyces cerevisiae, Schizosaccharomyces pombe,
Drosophila melanogaster, el ratón y el humano (Nagalakshmi et al., 2008; Mortazavi
et al., 2008; Maher et al., 2009; Pickrell et al., 2010; Gan et al., 2010; Daines et al.,
2011; Peng et al., 2012), lo que ha demostrado la alta aplicabilidad que el RNA-seq ha
tenido en estudios de especies modelo.
Recientemente la tecnología de RNA-seq también se ha implementado para estudios
de transcriptomas vegetales. Los estudios de RNA-seq en plantas ha permitido por
ejemplo la identificación de genes expresados frente a tratamientos de vernalización
y respuesta a giberelinas en remolacha azucarera, (Mutasa-Gottgens et al., 2012) y en
respuesta a estrés hídrico y calor, (Gulledge et al., 2012). La información generada
por RNA-seq también se ha explotado para la identificación de SNPs (del inglés single
nucleotide polimorphism) en uva y arroz (Zenoni et al., 2010; Lu et al., 2010) y para
la detección de variantes de splicing alternativo durante el desarrollo del fruto de uva
y en Arabidopsis (Zenoni et al., 2010; Gulledge et al., 2012). También se ha
constituido en un herramienta fundamental para ayudar en la anotación de genes (Lu
et al., 2010). Lo anterior muestra la utilidad y posibilidades que esta herramienta
transcriptómica tiene y su aplicación bajo diferentes enfoques investigativos.
Tecnología RNA-seq
El RNA-seq es una herramienta transcriptómica actual que está fundamentada en la
secuenciación de ADNc basada en los desarrollos NGS. En esta tecnología, se captura
el ARN total o ARNm, el cual es fragmentado y convertido en una librería de ADNc.
Uno de los pasos fundamentales es la obtención de un ARN de buena calidad que
represente todos los transcritos que se están produciendo en la condición y tejido de
estudio. Para el aislamiento del ARN con frecuencia se emplean kits de extracción de
ARNm que aplican la captura a partir de la cola poly(A) (Ward et al., 2012). La
fragmentación del ARN o del cDNA se realiza o bien por nebulización, por digestión
79
con enzimas de restricción o a través del uso de cationes divalentes bajo condiciones
de presiones elevadas (Wang et al., 2009). Generalmente, el fraccionamiento se
realiza posteriormente a la síntesis de ADNc. Esta síntesis se realiza con
procedimientos estándar bien establecidos para la mayoría de organismos haciendo
uso de la enzima transcriptasa reversa. Una vez obtenido el ADNc se ligan
adaptadores de tal forma que cada fragmento generado contendrá un adaptador
ligado en sus extremos 3´y 5´. Las secuencias de estos adaptadores se conocen y
serán necesarias para que cada fragmento pueda ser secuenciado y en algunos casos
pueden ser empleados para diferenciar otros grupos de fragmentos obtenidos a
partir de muestras de ADNc diferentes. Sin embargo, no en todos los casos se
requiere la ligación de adaptadores, lo cual dependerá de la plataforma de
secuenciación a emplear. Los adaptadores se pueden ligar directamente a la muestra
de ARN, previa síntesis de ADNc (Core et al., 2008; Marguerat y Bahler, 2010), o
alternativamente se pueden adicionar directamente a la cadena sencilla de ADNc,
(Maher et al., 2009; Marguerat y Bahler, 2010).
En cuanto a la cantidad y concentración del ARNm que se requiere para la tecnología
RNA-seq, el rango esta actualmente entre 5μg y 10μg, con una concentración
alrededor de 500ng/μl, (Ryan Kim, UC Davis Genome Center; Nong Chen, Business
Development Director, BGI Americas, comunicación personal).
Por otro lado, dentro de las aplicaciones y ventajas que tiene la tecnología RNA-seq
esta que da una cobertura completa de transcriptos, genera información tanto de la
secuencia como de la estructura de exones y sitos de splicing alternativo (Lister et al.,
2009; Gulledge et al., 2012). Así mismo, los datos arrojados por RNA-seq tienen una
alta precisión con respecto a los niveles de expresión génica que se obtienen a través
de PCR (del inglés polimerase chain reaction) cuantitativa (qPCR) (Nagalakshmi et
al., 2008; Wang et al., 2009; Ward et al., 2012). Además, también se ha mostrado que
los resultados son altamente reproducibles (Wang et al., 2009).
Plataformas y estrategias de secuenciación para RNA-seq
La tecnología RNA-seq actualmente está disponible comercialmente en las compañías
Roche/454, Solexa/Illumina, SOLiD/Life Technologies y Helicos/BioSciences. Sin
embargo, de las tecnologías de NGS disponibles las mas aplicadas son Roche/454 y
Solexa/Illumina (Strickler et al., 2012). No obstante, estas compañías y otras no cesan
en la búsqueda de mayores rendimientos de secuenciación, obtención de lecturas
80
más largas que se lleven a cabo en tiempo real y cada vez a costos más bajos
(Metzker, 2010; Mardis, 2011).
454
Roche/454 es una tecnología que emplea el secuenciador 454-Genome-Sequencer
FLX, desarrollado en el 2005. Este fue el primer sistema comercial de NGS. Su
principio se basa en la piro-secuenciación o detección de pirofosfato descrita en la
década de los 80´s (Nyrén y Lundin, 1985). En esta tecnología se incorporan
adaptadores a los extremos de los fragmentos de cadena simple de ADN o ADNc, los
cuales serán adheridos a microperlas que contienen en la superficie oligonucleótidos
complementarios a las secuencias de los adaptadores. Posteriormente se lleva a cabo
una PCR en emulsión para la amplificación de los fragmentos. El objetivo de esta
amplificación es la obtención de un gran número de moléculas idénticas que
producirán altas intensidades de señal en cada lectura (figura 1) (Ansorge, 2009;
Mardis, 2011; Egan, 2012).
Al finalizar la amplificación se lleva a cabo una denaturación y las microperlas son
transferidas a pozos en un chip de fibra óptica, en donde son incubadas con las
enzimas ADN polimerasa, ATP sulfurilasa, luciferasa y apirasa, así como con los
sustratos luciferina y adenosin-5-fosfosulfato (APS). Posteriormente sobre el chip se
adiciona un deoxinucleótido particular (dNTPs), así cuando la ADN polimerasa
incorpore el dNTP a la cadena naciente se liberará pirofosfato. Este pirofosfato
proviene de la formación del enlace fosfodiester y es convertido a adenosin trifosfato
(ATP), en presencia de la APS. El ATP así producido reaccionará con la enzima
luciferasa en presencia de luciferina para generar oxiluciferina produciéndose luz en
proporciones equivalentes a las cantidades de ATP producidas. La emisión de luz será
detectada por una cámara CCD (dispositivo de carga acoplada, del inglés charge
coupled device). Finalmente, la apirasa removerá el ATP y dNTPs no incorporados.
Una vez realizado esto, se repite el ciclo con un nuevo dNTP (Nyrén, 2007; Ansorge,
2009; Egan, 2012).
81
Figure 2- 1. Secuenciamiento masivo de ADNc, RNA-seq, por las tecnologías NGS
Illumina y 454. Tecnología de secuenciación 454, basada en el principio de PCR en
emulsión y piro-secuenciación. b. Tecnología de secuenciación Illumina basado en el
principio de amplificación en puente y uso de marcaje por fluorescencia de
terminadores reversibles
Actualmente, Roche/454 ofrece el servicio de secuenciación para RNA-seq con su
más reciente secuenciador, el sistema SG FLX+. Con esta tecnología es posible
obtener lecturas de hasta 1000 nucleótidos. Sin embargo, la cobertura es baja,
alrededor de 2.5 millones de lecturas o reads por corrida (454/Roche sequencing
contact, Comunicación personal).
82
Illumina
Por su parte, Solexa/Illumina, se basa en el principio de amplificación en puente y el
uso de marcaje por fluorescencia de nucleótidos modificados como terminadores
reversibles (figura 1) (Metzker, 2010). En esta tecnología, uno de los adaptadores de
los extremos de los fragmentos de ADN o ADNc se ligan complementariamente a
oligonucleótidos adheridos a una superficie sólida o “flow cell”. Estos
oligonucleótidos harán las veces de primers o cebadores sentido o antisentido, y
crean puentes que favorecen la amplificación. Los amplicones permanecerán
adheridos y luego de una denaturación formarán otro puente para permitir la
amplificación. Estos pasos se repiten hasta generar millones de grupos o clusters de
un fragmento determinado. Una vez formado estos grupos se desnaturalizaran
nuevamente para iniciar la polimerización o síntesis de cada fragmento, y se
introduce esta vez en la mezcla de síntesis cuatro nucleótidos marcados 3′-O –
azidometil ó terminadores reversibles (Ju et al., 2006; Gu et al., 2008; Metzker, 2010).
Los terminadores reversibles (dideoxinucleótidos) detendrán la síntesis de ADN una
vez la ADN polimerasa integre a la cadena naciente el nucleótido correspondiente.
Seguido de la síntesis, los fluoróforos de los terminadores reversibles integrados a la
cadena naciente son activados por un laser. La emisión de luz será diferencial de
acuerdo al nucleótido incorporado. La información será registrada y almacenada. Una
vez hecha la detección, un lavado retira los dideoxinucleótidos no integrados y
enzimáticamente es cortado el terminador para que así un nuevo ciclo permita la
incorporación del siguiente nucleótido (Metzker, 2010; Egan et al., 2012).
Cada “flow cell” de Illumina contiene ocho carriles o lanes. Cada uno de ellos en la
actualidad tiene un rendimiento de 150 millones de lecturas o reads. La longitud de
las lecturas generadas es pequeña, del rango de 50 -100 nucleótidos, las cuales
pueden ser lecturas sencillas desde un sólo extremo, SE (del inglés single end
sequencing reads) o bien lecturas desde ambos extremos PE (del inglés pair end
sequencing reads) (Wang et al., 2009; Garber et al., 2011). Aunque las longitudes
sean cortas, el alto número de lecturas generadas incrementarán la cobertura y
permitirá extender la secuencia hasta poder, en algunos casos, obtener la secuencia
de todo el transcripto. Además, cada lane tiene una capacidad para 24 muestras o
librerías de ADNc. Sin embargo, entre mayor el número de muestras que se ubiquen
en cada lane de Illumina, los millones de lecturas por muestra disminuirán. En el caso
en el que se desee potencializar la técnica a través de la obtención de un mayor
número de lecturas por muestra, el uso de replicas biológicas es adecuado (Ryan Kim,
UC Davis Genome Center y Veridiana Cano, Illumina®, comunicación personal).
83
El número de millones de lecturas deseables dependerá de varios factores. Uno de
ellos es la disponibilidad de un genoma de referencia que facilitará el ensamblaje de
las lecturas. Otro factor importante es la información sobre el número de genes de la
especie, pues lo que se busca es tener representado en miles de lecturas cerca de la
totalidad de genes expresados.
Secuenciación directa de ARN – Tecnología Helicos
Helicos/BioScience, fue el primer sistema comercial en disponer de una tecnología de
secuenciamiento por síntesis capaz de secuenciar una molécula sencilla (Harris et al.,
2008; Thompson y Milos, 2011). Más aun, esta compañía ha desarrollado un método
de secuenciación directo de ARN ó DRS (del inglés direct RNA sequencing), sin
necesidad de la previa conversión de ARN a ADNc, método que fue reportado por
primera vez en transcriptos de Saccharomyces cerevisiae y que aun esta en
perfeccionamiento (Ozsolak et al., 2009; Lipson et al., 2009; Levin, 2010; Ozsolak y
Milos, 2011). La secuenciación directa de ARN podría disminuir alguno de los
inconvenientes que se pueden presentar durante la conversión de ARN a ADNc. Uno
de los problemas más frecuente es la obtención de ADNc quimérico, en donde la
cadena naciente se puede disociar del molde ARN y posteriormente realinear a una
cadena diferente de ARN con una secuencia similar, e iniciar la síntesis de nuevo
(Mardis, 2011; Ozsolak y Milos, 2011). Sin embargo, hasta que la tecnología DRS no se
establezca ampliamente, la obtención de librerías ADNc para RNA-seq seguirá
liderando.
Estrategias y consideraciones para experimentos RNA-seq
Independiente de la plataforma de secuenciación a usar, es recomendable que cada
muestra o librería contenga patrones de reconocimiento llamados códigos de barra o
barcodes, lo que permitirá maximizar el número de muestras a ubicar en un sólo lane.
Esta estrategia es conocida como barcoding o multiplex. Cada barcode es una
secuencia corta, alrededor de cinco o seis nucleótidos, que caracteriza cada muestra o
librería de ADNc. Esta secuencia generalmente está ubicada contigua a la secuencia
del adaptador en el fragmento de ADNc (Strickler et al., 2012). Los barcodes son muy
útiles ya que permiten no sólo diferenciar las muestras sino también que hacen más
eficiente el uso de cada lane.
Los barcodes deben diseñarse uno por cada muestra o librería de ADNc. Aunque las
compañías de secuenciación ofrecen secuencias prediseñadas y probadas, el diseño
84
debe ser cuidadoso. Algunas de las recomendaciones es que los dos primeros
nucleótidos no pueden ser iguales de un barcode a otro, además un barcode no puede
ser la secuencia reversa complementaria del otro, ni tampoco pueden ser secuencias
palíndromas. Al hacer uso del barcoding se debe tener en cuenta que el tamaño de
cada lectura de los transcriptos se reducirá en aproximadamente 18 nucleótidos,
pues en el ensamblaje de estas se eliminaran tanto las secuencias de los adaptadores
como las de los barcode (Strickler et al., 2012).
Una vez revisadas de manera general las tecnologías de secuenciación para RNA-seq,
surge la pregunta de ¿cómo escoger la plataforma de secuenciación apropiada para
un proyecto RNA-seq? Partiendo de la premisa de similitud en cuanto a rendimiento,
precisión y capacidad de las lecturas desde ambos extremos del ADNc (3´y 5´), y que
actualmente las tecnologías con mayor demanda son Roche/454 y Solexa/Illumina, el
punto crítico para la respuesta dependerá de cuál es la longitud de las lecturas que se
desea obtener, así como si se cuenta o no con un genoma referencia e
indudablemente del presupuesto con el que se disponga, pues en términos generales,
el costo por base es menor con la tecnología Illumina. Un lane de Solexa/Illumina
para RNA-seq en la actualidad tiene un costo cercano a los USD 3000.
Como se mencionó anteriormente Roche/454 produce lecturas más largas con
respecto a Solexa/Illumina, Sin embargo, cuando el propósito es realizar ensamblaje
de Novo, es decir sin genoma de referencia, lo más recomendable es obtener lecturas
más largas, con el fin de facilitar el ensamblaje. En este caso Roche/454 sería de gran
utilidad. No obstante, es necesario considerar que el número de lecturas obtenidas no
será muy alto. Si se dispone de un presupuesto cuantioso se puede realizar un
elevado número de lecturas lo que facilitará el ensamblaje. Alternativamente, si se
cuenta con un genoma referencia, las lecturas cortas obtenidas con Illumina pueden
ser la mejor opción ya que no es necesario ensamblar el transcriptoma completo y el
alto número de lecturas que se generan darán mayor confiabilidad a los datos para la
cuantificación de la expresión y para otros fines como la detección de polimorfismos
(Ozsolak y Milos, 2011).
Para el ensamblaje de las lecturas existe una amplia diversidad de programas
bioinformáticos disponibles. Hay programas especializados en la eliminación de
barcodes, eliminación de ruido, etc. Si se trata de un ensamblaje de Novo, las lecturas
serán ensambladas a partir de contigs, es decir en secuencias sobrelapadas, de allí la
importancia de tener secuencias más largas. Algo importante de este tipo de
ensamblaje es que es posible encontrar transcritos que pueden no estar
representados en el genoma, como por ejemplo eventos de splicing alternativo o
genes que no se expresan bajo la condición de estudio (Strickler et al., 2012).
85
Dentro de los programas bioinformáticos aplicados al ensamblaje y análisis de RNA-
seq se encuentra por ejemplo Scriptuture, desarrollado en el Massachusetts Institute
of Technology. Este se ha empleado para el estudio del transcriptoma de ratones.
(Guttman et al., 2010). Trinity es otro programa ampliamente usado el cual fue
desarrollado por Broad Institute y la Hebrew University of Jerusalem, el cual permite
la reconstrucción de transcriptos, reconoce eventos de splicing alternativo y es
especializado en análisis de muestras que no tienen genoma de referencia (Grabherr
et al., 2011). El programa R ya ha diseñado un paquete estadístico, denominado
DEGseq, dirigido al análisis de la expresión diferencial entre muestras y tratamientos
(Wang et al., 2010). Además de estos actualmente, existen muchos otros programas
útiles, probados y de acceso gratuito (Ward et al., 2012).
Uno de los objetivos al emplear la tecnología de RNA-seq no es sólo identificar la
presencia de transcritos sino la de cuantificar el nivel de expresión de cada transcrito.
En este sentido aquellas lecturas que se encuentren en alta proporción representarán
niveles altos de expresión de determinado gen y aquellos transcritos ausentes o con
un bajo número de lecturas serán aquellos que o no se expresan o lo hacen a niveles
muy bajos (Schenk et al., 2012). En algunos casos se realizan normalizaciones
químicas en las librerías ADNc con el fin de igualar la abundancia de transcritos. De
esta manera aquellos transcriptos altamente expresados no serán los únicos para los
que se obtengan lecturas en la secuenciación (Ward et al., 2012). Para este fin existen
comercialmente diferentes kits y generalmente están asociados con cada plataforma
de secuenciación. En otras situaciones lo que se desea es comparar el perfil de
expresión génica entre diferentes muestras que pueden corresponder a tratamientos,
tejidos o condiciones.
Para los análisis de las lecturas generadas por RNA-seq es muy frecuente el uso de
varios parámetros estadísticos. Uno de ellos es el RPKM (del inglés reads per kilobase
of exon per million mapped reads). Con este parámetro es posible cuantificar niveles
de transcritos, y facilita la comparación entre muestras (Mortazavi et al., 2008). Otro
parámetro útil y muy usado es el fold change de las lecturas que corresponde a la
división del numero de lecturas generadas para un gen particular en una muestra Vs.
el de la otra muestra. Al estimar este parámetro es posible correlacionar la expresión
de un gen en dos condiciones distintas así como establecer radios de expresión
génica diferencial entre tratamientos (Auer y Doerge, 2010).
86
Aplicaciones enfocadas al estudio de interacciones planta patógeno
Dentro de los retos de la fitopatología están el detallar cómo los patógenos son
reconocidos por sus hospederos y cómo se establecen interacciones de resistencia y
susceptibilidad. En este sentido, la biología molecular y la bioinformática, así como el
estudio de la expresión génica en eventos de patogenicidad han contribuido de
manera importante (Verhage et al., 2010; Lodha y Basak, 2012; Schenk et al., 2012).
Actualmente, el entendimiento molecular que se tiene de las interacciones planta
patógeno ha permitido desarrollar un modelo de la función y evolución de la
inmunidad vegetal (Jones y Dangl, 2006). Según este modelo denominado “zigzag”, el
primer evento o fase de la respuesta de resistencia consiste en la capacidad de
reconocimiento de moléculas conservadas en los microorganismos conocidas como
PAMPs (del inglés pathogen associated molecular patterns) o MAMPs (del inglés
microbe- associated molecular patterns). Dentro de las moléculas reconocidas de este
tipo, las más estudiadas son la flagelina (flg22), (Felix et al., 1999; Chinchilla et al.,
2006), el factor de elongación Tu (EF-Tu), (Zipfel et al., 2006), los lipopolisacáridos
(LPS) y la quitina (Kaku et al., 2006). El reconocimiento de PAMPs depende de
proteínas denominadas PRRs (del inglés pathogen recognition receptors), las cuales
son proteínas de reconocimiento ubicadas generalmente en la membrana plasmática
de la célula vegetal (Gomez-Gomez y Boller, 2000; He et al., 2007; Boller y Yang 2009;
Zipfel, 2009; Thomma et al., 2011).
El primer PRR identificado y más estudiado es el receptor de flg22, denominado FLS2
el cual es una proteína transmembranal con un dominio extracelular rico en
repeticiones de leucinas (LRR, leucine rich repeats) y un dominio serina/treonina
kinasa intracelular, posiblemente encargado de la comunicación de la señal. La
proteína FLS2 es clasificada como un receptor con actividad kinasa (RLK, Receptor-
Like Kinase) (Gomez-Gomez y Boller, 2000; Asai et al., 2002). No fue hasta el 2006,
cuando fue posible la clonación del receptor EFR que reconoce al PAMP EF-Tu, el cual
también es de tipo RLK (Zipfel et al., 2006). Esta primera respuesta basada en la
interacción entre PAMPs-PRRs es denominada como inmunidad mediada por PAMPs
o PTI (PAMP triggered immunity). Este tipo de inmunidad generalmente detiene la
infección antes que el microorganismo comience su multiplicación y es
suficientemente efectiva contra patógenos potenciales no adaptados (Chisholm et al.,
2006).
87
Una vez se da la percepción del ligando por parte de los PRRs, se desencadena una
cascada de señalización mediada principalmente por MAP Kinasas (del inglés
mitogen associated protein kinase) (Gohre y Robatzek, 2008; Colcombet y Hirt, 2008;
Beckers et al., 2009), que finaliza con la activación de factores de transcripción lo que
conlleva a una reprogramación en la expresión génica. Por ejemplo en Arabidopsis se
ha establecido que la inducción de la expresión de los genes FRK, At2g17740 y
WRKY22/29 es un criterio diagnóstico de la activación de la PTI (Asai et al., 2002;
Shan et al., 2008). Adicionalmente estas respuestas están asociadas con eventos de
apertura de canales iónicos en la membrana plasmática, producción de especies
reactivas de oxígeno y fortificación de paredes celulares a través de la deposición de
callosa (Buchanan et al., 2000; Zipfel, 2008, 2009).
Durante la evolución, un grupo particular de patógenos desarrolló un tipo especial de
proteínas denominadas efectoras que al ser inyectada al interior de las células de las
plantas hospederas bloquean la PTI, y logran una colonización exitosa (Jones y Dangl,
2006). Este tipo de interacciones se denominan compatibles y el resultado para la
planta es el desarrollo de la enfermedad. Según el modelo zigzag propuesto, este tipo
de interacciones cae bajo el concepto de ETS (del inglés effector triggered
susceptibility). Sin embargo las plantas frente a esta situación desarrollaron una
segunda rama de la inmunidad basada en el reconocimiento de proteínas efectoras
del patógeno, tercera fase del modelo zigzag. Esta inmunidad es denominada como
ETI (del inglés effector triggered immunity). La ETI depende de la presencia de
proteínas de resistencia que pueden reconocer de manera directa o indirecta
efectores, este reconocimiento desencadena una reacción de incompatibilidad o
resistencia (Chisholm et al., 2006; Jones y Dangl, 2006; Dodds y Rathjen, 2010). De
manera similar a la PTI, parte de las respuestas desencadenadas durante la ETI
incluyen la reprogramación de la expresión génica. El estudio de esta reprogramación
permite conocer los mecanismos moleculares de respuesta y es susceptible de ser
estudiada mediante la técnica RNA-seq.
La inducción de varios genes importantes para la defensa vegetal se ha identificado
durante las respuestas ETI. Producto de la expresión de estos genes se encuentran las
proteínas PR (del inglés pathogenesis-related proteins) (Loon y Strien, 1999).
Algunas proteínas codificadas por estos genes PR tienen actividad quitinasas y
glucanasas. En la cascada de señalización que lleva a la transcripción de estos genes,
el ácido salicílico SA (del inglés salycilic acid) y el etileno, desempeñan una función
reguladora y sinérgica. Así mismo, se ha encontrado que la activación de grupos de
genes PR puede ser mediada por patógenos a través de un mecanismo llamado
resistencia sistémica adquirida o SAR (del inglés systemic acquired resistance). Para
que este mecanismo ocurra, la infección inicial debe resultar en una lesión necrótica,
88
como parte de una respuesta hipersensible o HR (del inglés hypersensitive response).
La activación de SAR produce una reducción de los síntomas de la enfermedad ante
un siguiente ataque del patógeno e incluso algunos otros tipos de patógenos no
relacionados con el de la primera infección (Durrant y Dong, 2004).
Las proteínas PR son una gran familia y aunque muchas de ellas fueron inicialmente
encontradas en tabaco y en Arabidopsis, su inducción por patógenos se ha reportado
en diversas especies vegetales (Loon y Strien, 1999). Más aun, su reconocida
actividad de defensa ante eventos de patogénesis ha convertido a estos genes PR en
genes marcadores de defensa y ha hecho que la medición de los niveles de su
expresión sea ampliamente utilizada en estudios de interacciones planta patógeno
(Slaughter et al., 2012; Zhang et al., 2012).
La reprogramación de la expresión génica durante las reacciones de PTI, ETS y ETI ha
sido bastante estudiada en diferentes patosistemas a través de la aplicación de
microarreglos o cDNA-AFLP (del inglés amplified fragment length polymorphism). En
el modelo Arabidopsis, a través del estudio de perfiles de expresión de ARNm con
microarreglos y frente a interacciones con P. syringae, fue demostrado que las
respuestas entre interacciones compatibles (bacteria virulenta), incompatibles
(bacteria avilurenta) y no hospedero (P. syringae pv phaseolicola), son similares y
que las diferencias en expresión encontradas son cuantitativas en cuanto a la
intensidad y en la rapidez con la que se producen. Sin embargo, la respuesta vegetal
en la interacción incompatible es robusta en términos de altos niveles de expresión
de los genes de resistencia RPS2 y RPM1, los cuales activan la respuesta mediada por
los genes avr avrRpt2 y avrB, avrRpm1 de P. syringae (Tao et al., 2003). Así mismo, el
empleo de microarreglos contribuyó a la identificación del receptor EFR que
reconoce el PAMP Ef-Tu (Zipfel et al., 2006). También recientemente esta técnica
permitió profundizar en el entendimiento de la actividad de las citoquininas en la
inmunidad vegetal, y puso en evidencia, que este fitorregulador, regula positivamente
el incremento de las respuestas de defensa mediada por acido salicílico y más aun,
que este último tiene una retroalimentación negativa al inhibir la señalización de
citoquininas (Argueso et al., 2012).
De manera más general los microarreglos han permitido el estudio de la
reprogramación génica en Arabidopsis en respuesta a virus (Whitham et al., 2003) a
hongos (Ramonell et al., 2002) e incluso a nematodos (Puthoff et al., 2003). También
se han empleado microarreglos de papa para el estudio de las interacciones
incompatibles con Phythophtora (Avrova et al., 1999) pero también durante la
interacción compatible (Restrepo et al., 2005). En arroz los perfiles de expresión por
89
microarreglos se han estudiado frente a infección por patógenos como Xanthomonas
oryzae pv. oryzae y Magnaporthe oryzae (Li et al., 2006).
En cuanto al uso de cDNA-AFLP para análisis de transcriptomas vegetales en
patosistemas, esta ha sido aplicada fundamentalmente para la búsqueda de perfiles
polimórficos relacionados con la expresión de genes frente al ataque por patógenos
(Birch y Kamoun, 2000; Durrant et al., 2000). También a través de cDNA-AFLP se han
estudiado perfiles de expresión relacionados con respuesta hipersensible en tabaco y
tomate (Vandenabeele et al., 2003; Gabriëls et al., 2006) así como en resistencia
sistémica adquirida en Arabidopsis (Maleck et al., 2000). Sin embargo, como se
mencionó antes, estas técnicas o bien requieren de pasos de hibridación como los
microarreglos, o poseen una baja cobertura de transcriptos, esto hace que la
alternativa del uso de RNA-seq sea llamativa y muy prometedora.
Antecedentes del uso de RNA-seq en interacciones planta patógeno
Los primeros trabajos de secuenciación de ADNc en plantas se llevaron a cabo en
Medicago truncatula, Arabidopsis thaliana y Zea mays con la tecnología 454. En estos
reportes, se identificaron más de 17 mil genes de Arabidopsis, 25 mil secuencias
genómicas en maíz que no estaban anotadas ni tenían similitud alguna con otras
especies y 400 SSR (del inglés simple sequence repeats) en M. truncatula (Cheung et
al., 2006; Emrich et al., 2007; Weber et al., 2007). Desde entonces el RNA-seq mostró
ser altamente sensible y prometedor para el análisis profundo de transcriptomas en
plantas. Dado lo novedoso de la técnica, los estudios en interacciones planta patógeno
en donde se hace uso de RNA-seq son escasos. A la fecha se han reportado estudios
en Arabidopsis thaliana, algodón y soya, pero en otras especies como yuca y maíz las
investigaciones están en desarrollo.
Uno de los trabajos más recientes enfocado al análisis de múltiples genomas y
transcriptomas en Arabidopsis usando RNA-seq, reveló que la variación en los niveles
de expresión génica es mucho mayor en genes que están involucrados en respuestas
a estrés biótico. Así mismo, que los genes de resistencia de las subfamilias NB-LRR
(del inglés nucleotide binding leucine rich repeat), coiled-coil, receptores Toll
interleuquina-1 y genes relacionados con defensa, codifican para proteínas mas
variables, que aquellas codificadas por genes del metabolismo basal (Xiangchao et al.,
2011).
90
En el patosistema algodón-Verticillium dahliae se reportó inicialmente, mediante
microarreglos, cambios transcriptómicos en las respuestas de defensa de 211 genes
así como la activación de la respuesta de defensa basal, correspondiente a la PTI, y a
la rápida producción de fitoalexinas, terpenoides y felipropanoides (Cui et al., 2000).
No obstante este conocimiento, el entendimiento de las respuestas de defensa del
algodón ante V. dahliae era limitado. Recientemente, a través de la aplicación de RNA-
seq y la plataforma Illumina se logró obtener el primer análisis global del
transcriptoma de defensa en algodón. En este estudio se pudo monitorear los perfiles
de expresión en raíces a las 4, 12, 24 y 48 horas post inoculación (hpi), se detectó
expresión diferencial en más de 3 mil genes, lo que permitió enriquecer el
entendimiento de cómo los genes involucrados en actividad enzimática,
especialmente en la ruta metabólica fenilpropanoide, están involucrados en eventos
de respuesta (Xu et al., 2011). También se reportaron niveles de lignificación y
actividad enzimática contrastante, así como expresión génica diferencial en plantas
resistentes y susceptibles al hongo.
En soya, con RNA-seq, fueron mapeados más de 43 mil genes con el genoma
referencia de esta especie y se estudiaron expresiones diferenciales de más de mil
genes en líneas casi isogénicas a las 0, 6 y 12 hpi con Xanthomonas axonopodis pv.
glycines (Xag), agente causal de la BPL (del inglés bacterial leaf pustule) de la soya.
Así mismo, bajo este enfoque, se detectó la sobreexpresión de PRRs y genes que son
inducidos por estos en líneas resistentes a Xag (Kim et al., 2011). En esta
investigación se unieron tres réplicas biológicas de cada tiempo 0, 6 y 12 hpi, se
corrió un lane de Illumina para cada uno de ellos y se obtuvo más de 125 millones de
reads. También en soya y con la tecnología Illumina, pero en un estudio de la roya,
otra de las enfermedades más limitantes de este cultivo, recientemente se analizaron
patrones de expresión de genes, con el objetivo de dilucidar los eventos moleculares
que ocurren tras la infección por parte del hongo Phakopsora pachyrhizi (Tremblay et
al., 2011). En plantas susceptibles y en etapas avanzadas de infección, un alto
porcentaje de genes involucrados en el metabolismo de síntesis de aminoácidos,
carbohidratos y lípidos fueron detectados con regulación negativa. Por el contrario,
muchos otros genes relacionados con rutas metabólicas implicadas en defensa se
sobre-expresaron en etapas iniciales de la infección. De acuerdo con los autores,
muchos de los genes encontrados en este trabajo han dado luces para el desarrollo de
un programa de mejoramiento genético enfocado a lograr resistencia amplia contra
la roya de la soya a través de sobreexpresión o silenciamiento génico (Tremblay et al.,
2011).
El RNA-seq también ha sido aplicado para análisis de perfiles transcriptómicos en
fitopatógenos. Es el caso de Phytophthora phaseoli, agente causal del mildeo velloso
91
en lima bean, Phaseolus lunatus. Poco se conoce de la base molecular de la interacción
de este oomycete con su hospedero. En esta investigación con la tecnología Illumina
para RNA-seq, se compararon los transcriptomas de tres tratamientos: el primero P.
phaseoli creciendo bajo condiciones de cultivo, el segundo P. phaseoli infectando a P.
lunatus 3 días post inoculación (dpi) y el tercero a los 6 dpi. Se trabajaron dos
replicas biológicas por tratamiento y un total de 150 millones de reads fueron
obtenidas. Los resultados mostraron similitud en 10.427 genes de P. phaseoli con el
genoma de P. infestans, de los cuales 318 son genes putativos de virulencia, y que
mostraron ser sobre expresados en los transcriptomas del patógeno en interacción
con la planta (Kunjeti et al., 2011). Este reporte muestra que esta herramienta
también tiene un gran potencial en investigaciones direccionadas hacia el estudio de
transcriptomas en fitopatógenos.
Sin duda, la implementación de RNA-seq está dentro de las actuales y futuras
proyecciones de investigación en diversos patosistemas. A modo de ejemplos, está el
maíz Zea mays - Aspergillus flavus y la yuca Manihot esculenta- Xanthomonas
axonopodis pv. manihotis (Xam). En maíz, se planea encontrar genes expresados
diferencialmente entre líneas susceptibles y resistentes, así como integrar esta
información del transcriptoma con la presencia de QTL (del inglés quantitative trait
loci) ya reportados de resistencia a la acumulación de aflatoxinas producidas por el
patógeno (Kelley et al., 2010).
En yuca, actualmente se desarrolla una estrategia en donde se combinará la
información arrojada por RNA-seq de líneas resistentes y susceptibles a Xam y el
genoma referencia, con el fin de asignar funciones putativas a muchos de los 35.000
genes de la yuca. Así mismo, se busca identificar los genes que se inducen o reprimen
frente al ataque de la bacteria y finalmente realizar identificación de SNPs, con el fin
de establecer relaciones entre estos marcadores y QTL para resistencia (López y
Bernal, 2012). Este enfoque contribuirá de manera determinante no sólo en el mejor
entendimiento de la interacción molecular de este patosistema, sino también
aportará información para incrementar el número de marcadores presentes en el
mapa genético de yuca lo que facilitará la clonación de genes R contra la bacteriosis
vascular de yuca.
De igual forma, hoy por hoy es factible la posibilidad de profundizar en preguntas
como ¿cuáles son aquellos genes que son expresados en el hospedero ante el ataque
de un fitopatógeno, incluso, cuáles son aquellos genes que son expresados en una
planta dentro de un sistema heterólogo no hospedero?, ¿cuáles son aquellos genes
que se expresan en el patógeno frente a la interacción con su hospedero?, ¿cuál es el
nivel de esta expresión?, ¿en qué momento después de iniciada la interacción ocurre?,
92
¿de manera temprana, tardía? ¿cuáles son esos nuevos transcriptos que no pudieron
ser identificados con previas técnicas para análisis de transcriptoma?, ¿cómo en
determinado patosistema, la estructura génica en cuanto a exones y sitios de splicing
alternativo, están relacionados con activaciones de respuestas de defensa?. Las
respuestas a estos y otros cuestionamientos, pueden encontrarse bajo el enfoque de
RAN-seq.
Conclusiones, retos y perspectivas
El secuenciamiento masivo de ADNc permite ampliar los rangos de detección de
transcritos que se lograron en su momento bajo la herramienta de microarreglos. Es
evidente el gran potencial que tiene el RNA-seq en cuanto a reconocimiento de genes
relacionados con defensa, factores de transcripción involucrados en activaciones de
respuesta génica y determinación de expresión diferencial, incluso cuando el
conocimiento del genoma es escaso o incluso nulo.
Así mismo, el conocimiento generado por el RNA-seq contribuirá de manera
importante dentro de programas de fitomejoramiento. Con la información generada
será posible la identificación de nuevos genes blancos para su uso en transformación
genética, en búsqueda de expresión génica que se traduzca en el desarrollo de plantas
resistentes a patógenos. Con seguridad, en el futuro cercano la literatura enfocada a
interacciones planta patógeno contendrá un sin número de reportes alrededor de
esta herramienta transcriptómica. Más aún, las tecnologías de NGS avanzaran y en
esa medida, la herramienta se volverá cada vez más poderosa, mejorarán sus
rendimientos, se podrán obtener lecturas más largas, en un menor tiempo y por
supuesto a menores costos.
El secuenciamiento directo de ARN es uno de los mayores retos que presenta la
técnica, sin embargo, las investigaciones, desarrollos y comprobaciones al respecto
están en marcha y pronto la alta oferta de esta tecnología será una realidad (Ozsolak
et al., 2009; Ozsolak y Milos, 2011). Así mismo, el reto computacional es grande, el
análisis de millones de lecturas que generan archivos de tamaños enormes, requiere
del desarrollo de herramientas bioinformáticas cada vez con mayor capacidad de
análisis, rendimientos más altos, procesos de ensamblaje más sencillos, así como con
capacidad de caracterización precisa de estructura y dinámica de transcriptoma.
Estos son otros de los retos que tiene esta poderosa herramienta transcriptómica
para los próximos años.
93
El reto para la fitopatología es explotar la utilidad de esta herramienta
transcriptómica y a través de ella profundizar en el entendimiento de las complejas
interacciones planta patógeno, así como revelar eslabones moleculares involucrados,
que hasta hoy no habían podido ser revelados.
Referencias
Ansorge, W. J. 2009. Next-generation DNA sequencing techniques. N. Biotechnol. 25:195–203
Argueso, C. T., Ferreira, F. J., Epple, P., To, J. P. C., Hutchison, C. E., Schaller, G. E., Dangl, J. L., and Kieber, J. J. 2012. Two-component elements mediate interactions between cytokinin and salicylic acid in plant immunity. PLoS Genet. 8:e1002448
Asai, T., Tena, G., Plotnikova, J., Willmann, M. R., Chiu, W.-L., Gomez-Gomez, L., Boller, T., Ausubel, F. M., and Sheen, J. 2002. MAP kinase signalling cascade in Arabidopsis innate immunity. Nature. 415:977–983
Auer, P. L., and Doerge, R. W. 2010. Statistical design and analysis of RNA sequencing data. Genetics. 185:405–416
Avrova, A. O., Stewart, H. E., De Jong, W., Heilbronn, J., Lyon, G. D., and Birch, P. R. J. 1999. A cysteine protease gene is expressed early in resistant potato interactions with Phytophthora infestans. Mol. plant-microbe Interact. 12:1114–1119
Beckers, G. J. M., Jaskiewicz, M., Liu, Y., Underwood, W. R., He, S. Y., Zhang, S., and Conrath, U. 2009. Mitogen-activated protein kinases 3 and 6 are required for full priming of stress responses in Arabidopsis thaliana. Plant Cell. 21:944–953
Birch, P. R. J., and Kamoun, S. 2000. Studying interaction transcriptomes: coordinated analyses of gene expression during plant-microorganism interactions. A Trends Guid. 12:77–82
Boller, T., and He, S. Y. 2009. Innate immunity in plants: an arms race between pattern recognition receptors in plants and effectors in microbial pathogens. Science (80). 324:742–744
Buchanan, B. B., Gruissem, W., and Jones, R. L. Biochemistry and molecular biology of plants. 2000. Am. Soc. Plant Physiol. Rockville, Maryl.
Cheung, F., Haas, B. J., Goldberg, S. M. D., May, G. D., Xiao, Y., and Town, C. D. 2006. Sequencing Medicago truncatula expressed sequenced tags using 454 Life Sciences technology. BMC Genomics. 7:1
Chinchilla, D., Bauer, Z., Regenass, M., Boller, T., and Felix, G. 2006. The Arabidopsis receptor kinase FLS2 binds flg22 and determines the specificity of flagellin perception. Plant Cell. 18:465–476
94
Chisholm, S. T., Coaker, G., Day, B., and Staskawicz, B. J. 2006. Host-Microbe Interactions: Shaping the Evolution of the Plant Immune Response. Cell. 124:803–814
Colcombet, J., and Hirt, H. 2008. Arabidopsis MAPKs: a complex signalling network involved in multiple biological processes. Biochem. J. 413:217–226
Core, L. J., Waterfall, J. J., and Lis, J. T. 2008. Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science (80). 322:1845–1848
CUI, Y., BELL, A. A., Joost, O., and MAGILL, C. 2000. Expression of potential defense response genes in cotton. Physiol. Mol. Plant Pathol. 56:25–31
Daines, B., Wang, H., Wang, L., Li, Y., Han, Y., Emmert, D., Gelbart, W., Wang, X., Li, W., Gibbs, R., and others. 2011. The Drosophila melanogaster transcriptome by paired-end RNA sequencing. Genome Res. 21:315–324
Dodds, P. N., and Rathjen, J. P. 2010. Plant immunity: towards an integrated view of plant--pathogen interactions. Nat. Rev. Genet. 11:539–548
Durrant, W. E., and Dong, X. 2004. Systemic acquired resistance. Annu. Rev. Phytopathol. 42:185–209
Durrant, W. E., Rowland, O., Piedras, P., Hammond-Kosack, K. E., and Jones, J. D. G. 2000. cDNA-AFLP reveals a striking overlap in race-specific resistance and wound response gene expression profiles. Plant Cell. 12:963–977
Egan, A. N., Schlueter, J., and Spooner, D. M. 2012. Applications of next-generation sequencing in plant biology. Am. J. Bot. 99:175–185
Emrich, S. J., Barbazuk, W. B., Li, L., and Schnable, P. S. 2007. Gene discovery and annotation using LCM-454 transcriptome sequencing. Genome Res. 17:69–73
Felix, G., Duran, J. D., Volko, S., and Boller, T. 1999. Plants recognize bacteria through the most conserved domain of flagellin. Plant J. 18:265–276
Gabriëls, S. H. E. J., Takken, F. L. W., Vossen, J. H., de Jong, C. F., Liu, Q., Turk, S. C. H. J., Wachowski, L. K., Peters, J., Witsenboer, H. M. A., de Wit, P. J. G. M., and others. 2006. cDNA-AFLP combined with functional analysis reveals novel genes involved in the hypersensitive response. Mol. Plant-Microbe Interact. 19:567–576
Gan, Q., Chepelev, I., Wei, G., Tarayrah, L., Cui, K., Zhao, K., and Chen, X. 2010. Dynamic regulation of alternative splicing and chromatin structure in Drosophila gonads revealed by RNA-seq. Cell Res. 20:763–783
Gan, X., Stegle, O., Behr, J., Steffen, J. G., Drewe, P., Hildebrand, K. L., Lyngsoe, R., Schultheiss, S. J., Osborne, E. J., Sreedharan, V. T., and others. 2011. Multiple reference genomes and transcriptomes for Arabidopsis thaliana. Nature. 477:419–423
Garber, M., Grabherr, M. G., Guttman, M., and Trapnell, C. 2011. Computational methods for transcriptome annotation and quantification using RNA-seq. Nat. Methods. 8:469–477
95
Göhre, V., and Robatzek, S. 2008. Breaking the barriers: microbial effector molecules subvert plant immunity. Annu. Rev. Phytopathol. 46:189–215
Gómez-Gómez, L., and Boller, T. 2000. FLS2: an LRR receptor-like kinase involved in the perception of the bacterial elicitor flagellin in Arabidopsis. Mol. Cell. 5:1003–1011
Grabherr, M. G., Haas, B. J., Yassour, M., Levin, J. Z., Thompson, D. A., Amit, I., Adiconis, X., Fan, L., Raychowdhury, R., Zeng, Q., and others. 2011. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat. Biotechnol. 29:644–652
Gulledge, A. A., Roberts, A. D., Vora, H., Patel, K., and Loraine, A. E. 2012. Mining Arabidopsis thaliana RNA-seq data with Integrated Genome Browser reveals stress-induced alternative splicing of the putative splicing regulator SR45a. Am. J. Bot. 99:219–231
Guo, J., Xu, N., Li, Z., Zhang, S., Wu, J., Kim, D. H., Marma, M. S., Meng, Q., Cao, H., Li, X., and others. 2008. Four-color DNA sequencing with 3′-O-modified nucleotide reversible terminators and chemically cleavable fluorescent dideoxynucleotides. Proc. Natl. Acad. Sci. 105:9145–9150
Guttman, M., Garber, M., Levin, J. Z., Donaghey, J., Robinson, J., Adiconis, X., Fan, L., Koziol, M. J., Gnirke, A., Nusbaum, C., and others. 2010. Ab initio reconstruction of cell type-specific transcriptomes in mouse reveals the conserved multi-exonic structure of lincRNAs. Nat. Biotechnol. 28:503–510
Harris, T. D., Buzby, P. R., Babcock, H., Beer, E., Bowers, J., Braslavsky, I., Causey, M., Colonell, J., DiMeo, J., Efcavitch, J. W., and others. 2008. Single-molecule DNA sequencing of a viral genome. Science (80). 320:106–109
He, P., Shan, L., and Sheen, J. 2007. Elicitation and suppression of microbe-associated molecular pattern-triggered immunity in plant--microbe interactions. Cell. Microbiol. 9:1385–1396
Jones, J. D. G., and Dangl, J. L. 2006. The plant immune system. Nature. 444:323–329
Jones, K. A., Kadonaga, J. T., Rosenfeld, P. J., Kelly, T. J., and Tjian, R. 1987. A cellular DNA-binding protein that activates eukaryotic transcription and DNA replication. Cell. 48:79–89
Ju, J., Kim, D. H., Bi, L., Meng, Q., Bai, X., Li, Z., Li, X., Marma, M. S., Shi, S., Wu, J., and others. 2006. Four-color DNA sequencing by synthesis using cleavable fluorescent nucleotide reversible terminators. Proc. Natl. Acad. Sci. 103:19635–19640
Kaku, H., Nishizawa, Y., Ishii-Minami, N., Akimoto-Tomiyama, C., Dohmae, N., Takio, K., Minami, E., and Shibuya, N. 2006. Plant cells recognize chitin fragments for defense signaling through a plasma membrane receptor. Proc. Natl. Acad. Sci. U. S. A. 103:11086–11091
Kelley, R. Y., Gresham, C., Harper, J., Bridges, S. M., Warburton, M. L., Hawkins, L. K., Pechanova, O., Peethambaran, B., Pechan, T., Luthe, D. S., and others. 2010. Integrated
96
database for identifying candidate genes for Aspergillus flavus resistance in maize. BMC Bioinformatics. 11:1
Kim, K. H., Kang, Y. J., Kim, D. H., Yoon, M. Y., Moon, J.-K., Kim, M. Y., Van, K., and Lee, S.-H. 2011. RNA-Seq analysis of a soybean near-isogenic line carrying bacterial leaf pustule-resistant and-susceptible alleles. DNA Res. 18:483–497
Kunjeti, S. G., Evans, T. A., Marsh, A. G., Gregory, N. F., Kunjeti, S., Meyers, B. C., Kalavacharla, V. S., and Donofrio, N. M. 2012. RNA-Seq reveals infection-related global gene changes in Phytophthora phaseoli, the causal agent of lima bean downy mildew. Mol. Plant Pathol. 13:454–466
Levin, J. Z., Yassour, M., Adiconis, X., Nusbaum, C., Thompson, D. A., Friedman, N., Gnirke, A., and Regev, A. 2010. Comprehensive comparative analysis of strand-specific RNA sequencing methods. Nat. Methods. 7:709–715
Li, Q., Chen, F., Sun, L., Zhang, Z., Yang, Y., and He, Z. 2006. Expression profiling of rice genes in early defense responses to blast and bacterial blight pathogens using cDNA microarray. Physiol. Mol. Plant Pathol. 68:51–60
Lipson, D., Raz, T., Kieu, A., Jones, D. R., Giladi, E., Thayer, E., Thompson, J. F., Letovsky, S., Milos, P., and Causey, M. 2009. Quantification of the yeast transcriptome by single-molecule sequencing. Nat. Biotechnol. 27:652–658
Lister, R., Gregory, B. D., and Ecker, J. R. 2009. Next is now: new technologies for sequencing of genomes, transcriptomes, and beyond. Curr. Opin. Plant Biol. 12:107–118
Lockhart, D. J., and Winzeler, E. A. 2000. Genomics, gene expression and DNA arrays. Nature. 405:827–836
Lodha, T. D., and Basak, J. 2012. Plant--pathogen interactions: what microarray tells about it? Mol. Biotechnol. 50:87–97
Van Loon, L. C., and Van Strien, E. A. 1999. The families of pathogenesis-related proteins, their activities, and comparative analysis of PR-1 type proteins. Physiol. Mol. Plant Pathol. 55:85–97
López, C. E., and Bernal, A. J. 2012. Cassava Bacterial Blight: Using Genomics for the Elucidation and Management of an Old Problem. Trop. Plant Biol. 5:117–126
Lu, T., Lu, G., Fan, D., Zhu, C., Li, W., Zhao, Q., Feng, Q., Zhao, Y., Guo, Y., Li, W., and others. 2010. Function annotation of the rice transcriptome at single-nucleotide resolution by RNA-seq. Genome Res. 20:1238–1249
Maher, C. A., Kumar-Sinha, C., Cao, X., Kalyana-Sundaram, S., Han, B., Jing, X., Sam, L., Barrette, T., Palanisamy, N., and Chinnaiyan, A. M. 2009. Transcriptome sequencing to detect gene fusions in cancer. Nature. 458:97–101
Maleck, K., Levine, A., Eulgem, T., Morgan, A., Schmid, J., Lawton, K. A., Dangl, J. L., and Dietrich, R. A. 2000. The transcriptome of Arabidopsis thaliana during systemic acquired resistance. Nat. Genet. 26:403–410
97
Mardis, E. R. 2011. A decade’s perspective on DNA sequencing technology. Nature. 470:198–203
Marguerat, S., and Bähler, J. 2010. RNA-seq: from technology to biology. Cell. Mol. life Sci. 67:569–579
Metzker, M. L. 2010. Sequencing technologies—the next generation. Nat. Rev. Genet. 11:31–46
Meyers, B. C., Galbraith, D. W., Nelson, T., and Agrawal, V. 2004. Methods for transcriptional profiling in plants. Be fruitful and replicate. Plant Physiol. 135:637–652
Mochida, K., and Shinozaki, K. 2011. Advances in omics and bioinformatics tools for systems analyses of plant functions. Plant Cell Physiol. 52:2017–2038
Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. 2008. Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods. 5:621–628
Mutasa-Göttgens, E. S., Joshi, A., Holmes, H. F., Hedden, P., and Göttgens, B. 2012. A new RNASeq-based reference transcriptome for sugar beet and its application in transcriptome-scale analysis of vernalization and gibberellin responses. BMC Genomics. 13:1
Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., Gerstein, M., and Snyder, M. 2008. The transcriptional landscape of the yeast genome defined by RNA sequencing. Science (80). 320:1344–1349
Nowrousian, M. 2010. Next-generation sequencing techniques for eukaryotic microorganisms: sequencing-based solutions to biological problems. Eukaryot. Cell. 9:1300–1310
Nyrén, P. 2007. The History of Pyrosequencing®. Pyrosequencing® Protoc. :1–13
Nyrén, P., and Lundin, A. 1985. Enzymatic method for continuous monitoring of inorganic pyrophosphate synthesis. Anal. Biochem. 151:504–509
Ozsolak, F., and Milos, P. M. 2011. RNA sequencing: advances, challenges and opportunities. Nat. Rev. Genet. 12:87–98
Ozsolak, F., Platt, A. R., Jones, D. R., Reifenberger, J. G., Sass, L. E., McInerney, P., Thompson, J. F., Bowers, J., Jarosz, M., and Milos, P. M. 2009. Direct RNA sequencing. Nature. 461:814–818
Peng, Z., Cheng, Y., Tan, B. C.-M., Kang, L., Tian, Z., Zhu, Y., Zhang, W., Liang, Y., Hu, X., Tan, X., and others. 2012. Comprehensive analysis of RNA-Seq data reveals extensive RNA editing in a human transcriptome. Nat. Biotechnol. 30:253–260
Pickrell, J. K., Marioni, J. C., Pai, A. A., Degner, J. F., Engelhardt, B. E., Nkadori, E., Veyrieras, J.-B., Stephens, M., Gilad, Y., and Pritchard, J. K. 2010. Understanding mechanisms underlying human gene expression variation with RNA sequencing. Nature. 464:768–772
98
Proudfoot, N. J., Furger, A., and Dye, M. J. 2002. Integrating mRNA processing with transcription. Cell. 108:501–512
Puthoff, D. P., Nettleton, D., Rodermel, S. R., and Baum, T. J. 2003. Arabidopsis gene expression changes during cyst nematode parasitism revealed by statistical analyses of microarray expression profiles. Plant J. 33:911–921
Ramonell, K. M., Zhang, B., Ewing, R. M., Chen, Y., Xu, D., Stacey, G., and Somerville, S. 2002. Microarray analysis of chitin elicitation in Arabidopsis thaliana. Mol. Plant Pathol. 3:301–311
Ren, X.-Y., Vorst, O., Fiers, M. W. E. J., Stiekema, W. J., and Nap, J.-P. 2006. In plants, highly expressed genes are the least compact. Trends Genet. 22:528–532
Restrepo, S., Myers, K. L., Del Pozo, O., Martin, G. B., Hart, A. L., Buell, C. R., Fry, W. E., and Smart, C. D. 2005. Gene profiling of a compatible interaction between Phytophthora infestans and Solanum tuberosum suggests a role for carbonic anhydrase. Mol. plant-microbe Interact. 18:913–922
Schena, M., Heller, R. A., Theriault, T. P., Konrad, K., Lachenmeier, E., and Davis, R. W. 1998. Microarrays: biotechnology’s discovery platform for functional genomics. Trends Biotechnol. 16:301–306
Schenk, P. M., Carvalhais, L. C., and Kazan, K. 2012. Unraveling plant--microbe interactions: can multi-species transcriptomics help? Trends Biotechnol. 30:177–184
Schneeberger, K., and Weigel, D. 2011. Fast-forward genetics enabled by new sequencing technologies. Trends Plant Sci. 16:282–288
Severin, A. J., Woody, J. L., Bolon, Y.-T., Joseph, B., Diers, B. W., Farmer, A. D., Muehlbauer, G. J., Nelson, R. T., Grant, D., Specht, J. E., and others. 2010. RNA-Seq Atlas of Glycine max: a guide to the soybean transcriptome. BMC Plant Biol. 10:1
Shalon, D., Smith, S. J., and Brown, P. O. 1996. A DNA microarray system for analyzing complex DNA samples using two-color fluorescent probe hybridization. Genome Res. 6:639–645
Shan, L., He, P., Li, J., Heese, A., Peck, S. C., Nürnberger, T., Martin, G. B., and Sheen, J. 2008. Bacterial effectors target the common signaling partner BAK1 to disrupt multiple MAMP receptor-signaling complexes and impede plant immunity. Cell Host Microbe. 4:17–27
Slaughter, A., Daniel, X., Flors, V., Luna, E., Hohn, B., and Mauch-Mani, B. 2012. Descendants of primed Arabidopsis plants exhibit resistance to biotic stress. Plant Physiol. 158:835–843
Strickler, S. R., Bombarely, A., and Mueller, L. A. 2012. Designing a transcriptome next-generation sequencing project for a nonmodel plant species1. Am. J. Bot. 99:257–266
Tao, Y., Xie, Z., Chen, W., Glazebrook, J., Chang, H.-S., Han, B., Zhu, T., Zou, G., and Katagiri, F. 2003. Quantitative nature of Arabidopsis responses during compatible
99
and incompatible interactions with the bacterial pathogen Pseudomonas syringae. Plant Cell. 15:317–330
Thomma, B. P. H. J., Nurnberger, T., and Joosten, M. H. A. J. 2011. Of PAMPs and Effectors: The Blurred PTI-ETI Dichotomy. Plant Cell. 23:4–15
Thompson, J. F., and Milos, P. M. 2011. The properties and applications of single-molecule DNA sequencing. Genome Biol. 12:1
Tremblay, A., Hosseini, P., Alkharouf, N. W., Li, S., and Matthews, B. F. 2012. Gene expression in leaves of susceptible Glycine max during infection with Phakopsora pachyrhizi using next generation sequencing. Sequencing. 2011
Vandenabeele, S., Van Der Kelen, K., Dat, J., Gadjev, I., Boonefaes, T., Morsa, S., Rottiers, P., Slooten, L., Van Montagu, M., Zabeau, M., and others. 2003. A comprehensive analysis of hydrogen peroxide-induced gene expression in tobacco. Proc. Natl. Acad. Sci. 100:16113–16118
Verhage, A., van Wees, S. C. M., and Pieterse, C. M. J. 2010. Plant immunity: it’s the hormones talking, but what do they say? Plant Physiol. 154:536–540
Wang, L., Feng, Z., Wang, X., Wang, X., and Zhang, X. 2010. DEGseq: an R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics. 26:136–138
Wang, Z., Gerstein, M., and Snyder, M. 2009. RNA-Seq: a revolutionary tool for transcriptomics. Nat. Rev. Genet. 10:57–63
Ward, J. A., Ponnala, L., and Weber, C. A. 2012. Strategies for transcriptome analysis in nonmodel plants. Am. J. Bot. 99:267–276
Weber, A. P. M., Weber, K. L., Carr, K., Wilkerson, C., and Ohlrogge, J. B. 2007. Sampling the Arabidopsis transcriptome with massively parallel pyrosequencing. Plant Physiol. 144:32–42
Whitham, S. A., Quan, S., Chang, H.-S., Cooper, B., Estes, B., Zhu, T., Wang, X., and Hou, Y.-M. 2003. Diverse RNA viruses elicit the expression of common sets of genes in susceptible Arabidopsis thaliana plants. Plant J. 33:271–283
Xu, L., Zhu, L., Tu, L., Liu, L., Yuan, D., Jin, L., Long, L., and Zhang, X. 2011. Lignin metabolism has a central role in the resistance of cotton to the wilt fungus Verticillium dahliae as revealed by RNA-Seq-dependent transcriptional analysis and histochemistry. J. Exp. Bot. 62:5607–5621
Zenoni, S., Ferrarini, A., Giacomelli, E., Xumerle, L., Fasoli, M., Malerba, G., Bellin, D., Pezzotti, M., and Delledonne, M. 2010. Characterization of transcriptional complexity during berry development in Vitis vinifera using RNA-Seq. Plant Physiol. 152:1787–1795
Zhang, Z., Wu, Y., Gao, M., Zhang, J., Kong, Q., Liu, Y., Ba, H., Zhou, J., and Zhang, Y. 2012. Disruption of PAMP-induced MAP kinase cascade by a Pseudomonas syringae effector activates plant immunity mediated by the NB-LRR protein SUMM2. Cell Host Microbe. 11:253–263
100
Zipfel, C. 2009. Early molecular events in PAMP-triggered immunity. Curr. Opin. Plant Biol. 12:414–420
Zipfel, C. 2008. Pattern-recognition receptors in plant innate immunity. Curr. Opin. Immunol. 20:10–16
Zipfel, C., Kunze, G., Chinchilla, D., Caniard, A., Jones, J. D. G., Boller, T., and Felix, G. 2006. Perception of the bacterial PAMP EF-Tu by the receptor EFR restricts Agrobacterium-mediated transformation. Cell. 125:749–760
101
Unraveling the molecules hidden in the gray shadows
1Andrea Ximena Vasquez Chacón, 2Johana Carolina Soto Sedano and 3Camilo Ernesto
López Carrascal
1, 2, 3. Manihot Biotec Laboratory, Biology department, Universidad Nacional de Colombia,
Bogotá, Colombia.
Review submitted to Plant Molecular Pathogen Interaction
Abstract
One of the most challenging questions in plant breeding and molecular plant
pathology research is what are the genetic and molecular bases of quantitative
disease resistance (QDR). The scarce knowledge of how this type of resistance works
has hindered plant breeders to fully take advantage of it. To overcome these
obstacles methodologies for the study of quantitative traits have been developed.
Approaches such as genetic mapping, identification of quantitative trait loci and
association mapping, including candidate gene approach and genome wide
association studies, have been historically employed to dissect quantitative traits and
therefore to study QDR. Additionally, great advances in quantitative phenotypic data
collection have come on the scene to improve these analyses. Recently, genes
associated to QDR have been cloned, opening new hypothesis concerning the
molecular bases of this type of resistance. In this review we present the more recent
advances and application of QDR, which have allowed postulating new ideas that can
help to construct new QDR models. Some of the hypotheses presented here as
possible explanations for QDR are related to the expression intensity and alternative
splicing of some defense-related genes, the action of “weak alleles” of R genes, the
presence of allelic variants in genes involved in the defense response and a pivotal
role of kinases or pseudokinases. With the information recapitulated in this review it
is possible to conclude that the division between qualitative and quantitative
resistance corresponds to an anthropic vision because indeed both share important
components.
Key words: quantitative disease resistance (QDR), plant immunity, quantitative trait
loci (QTLs).
102
Introduction
Understanding the genetics and molecular bases of phenotypic traits is one of the
more important challenges for scientists in this new era. The study of traits that show
simple inheritance has been the focus of most genetic research from the beginning
and it has come a long way, reaching historical milestones. The association between
phenotypic traits and single genes started with Mendel’s experiments, the foundation
of modern genetics. Since then, genes for thousands of monogenic traits have been
characterized in organisms that belong to almost all of the taxonomic groups in
nature. The study of these traits is straight forward because the phenotype reveals
the underlying genotype without ambiguity (St. Clair, 2010). The fact that
researchers have described genes that control single traits so broadly that is
impossible to summarize gives the false impression that most phenotypic traits
follow single inheritance. However, the phenotypic variation observed in natural
populations is governed mainly by multiple genes and, to a lesser extent, by single
genes, indicating that the complex inheritance of traits is the rule rather than the
exception. In model plants, as well as in agronomically important crops, although
single genes that control morphology, productivity, yield, food quality and disease
resistance have been described elsewhere, the real genetic bases of these traits, in
most cases, depends on the concerted and simultaneous action of multiple genes.
The study of the genetic bases of plant resistance has not escaped the above-
mentioned oversimplification. The response phenotypes of individuals with
qualitative resistance have a discrete (categorical) distribution and the genes
involved segregates following the expected Mendelian ratios (St. Clair, 2010). The
association of plant resistance with single genes was first proposed by Flor with the
well-known gene-by-gene model (Flor, 1955). Since then, a lot has been
accomplished in understanding how these genes control the response to pathogens.
The number of single genes associated with plant immunity that have been cloned
and characterized is large enough that it has resulted in a broad view of molecular
mechanisms that control plant immunity, but quantitative resistance has not been
considered. In this review, we try to reposition the importance of this type of
resistance by discussing the efforts that have been made to elucidate the molecular
bases. We collected recent studies that are not only worthy of being included in new
immunity models, but also are helpful in understanding how quantitative genetics
have been studied. It is important to note the lack of field implementation of
knowledge on genomic regions that explain quantitative disease resistance (QDR).
We also want to explain how quantitative resistance works at the molecular level
103
based on QDR genes that have been cloned and functionally validated with the aim of
highlighting shared characteristics.
The ABC of plant immunity
Since plants are continuously threatened by different kind of pathogens, it is
imperative to develop new strategies in order to control plant diseases. The most
favorable and environmental friendly strategy is to exploit the natural mechanisms
that plants have evolved to control invading pathogens. The activation of an effective
immune response depends on the ability of plants to recognize pathogens. Based on
the knowledge on molecules from pathogens and their recognition by hosts, a scheme
known as the zig-zag model has outlined how to describe immunity systems in plants
(Jones and Dangl, 2006). This model state that plants have evolved immune receptors
that are able to recognize pathogen-associated molecular patterns (PAMPs) or
specialized effector proteins that are present in particular races/strains of pathogens.
The recognition of PAMPs depends on the pattern recognition receptors (PRRs) that
constitute the first line of molecular defense, known as PAMP triggered immunity
(PTI) (Zipfel, 2014). Adapted pathogens translocate effector proteins into plant cells
to manipulate host components or suppress PTI (Cui et al., 2015). Plants can
recognize pathogen effectors through R proteins and the immunity they activate is
known as effector triggered immunity or ETI (Chisholm et al., 2006). Although this
resistance is even higher and more specific than PTI, it can be easily overcome by
point mutation in effectors that escape plant recognition (Houterman et al., 2009).
The ETI is the molecular explanation of the gene-by-gene model proposed by Flor.
According to this model, a plant is resistant when the interaction with the pathogen is
incompatible. On the other hand, when the plant is susceptible, the interaction is
compatible. In this case, there are only two possible phenotypes, resistant or
susceptible, and the intermediates are not considered.
The zig-zag model does not include several host-pathogen dynamics, such as
intermediate phenotypes and, consequently, is somewhat artificial. A novel “invasion
model” was recently proposed in which the classification of the immunity response is
based on the pathogen invasion patterns (IPs). These IPs are a large spectrum of
molecules that indicate invasion and are perceived by invasion pattern receptors
(IPRs). The functions of IP scan vary from microbiological physiology to host defense
suppression and, consequently, they trigger a large spectrum of continuous defense
responses. Plant responses can be either symbiotic or not, depending on the ability of
the plant to recognize the IPs with IPRs and activate an IPTR (IP-triggered response).
104
This new model also takes into account the fact that there is a complex interaction of
multiple receptors and ligands at the same time and that the output is the
combination of all of them. The invasion model was developed as an alternative to
the adopted classifications that separate PTI from ETI and in which the PAMPs are
defined from the host perspective while the effectors are considered from the
pathogen perspective(Cook et al., 2015). In its application, the invasion model
emphasizes the identification and understanding of molecules produced by the
pathogen. Therefore, it is necessary to develop a model that uses the idea of defense
as a continuum of responses, with a synergy and interaction between components
from the invasion model, but that also shows the plant perspective of the model and
its application.
Quantitative resistance enters into the game
In plant populations, when the response to a pathogen is a continuous phenotypic
value, varying from highly susceptible to highly resistant individuals, it is considered
quantitative resistance (Huard-Chauveau et al., 2013).QDR is controlled by several
genes, each one contributing to a different degree, to the reduction of the disease (St.
Clair, 2010).The term polygenic, or oligogenic, resistance is frequently associated
with QDR because of its inherent genetic architecture(Mackay et al., 2009; Niks et al.,
2015). Although the concept of QDR is well-defined, it is also widely used and
sometimes misunderstood, misused or exchanged. Partial resistance is the most
widely used concept in literature to describe the intermediate phenotype when the
resistance is not complete (Niks et al., 2015). It is important to emphasize that this
definition must be used to just describe a phenotype and not to define the genetic
basis underlying the response(Niks et al., 2015).QDR is often called field resistance
because it has been evaluated frequently in polycyclic field conditions (Niks et al.,
2015); however, it should not been used as a synonym because QDR has also been
observed and assessed under greenhouses and controlled conditions.
Traditionally, QDR has been associated with two important concepts: broad spectrum
and durable resistance; however, it is important to stress that these two concepts are
not strictly exclusive to QDR. Durable resistance is a concept that was defined by
Johnson in 1981to refer to the resistance that retains its effectiveness in crops that
are widely cultivated in an environment that is favorable to the pathogen, but this
does not mean that it has to be permanent (Johnson, 1983). QDR has been considered
more durable than qualitative resistance and, consequently, more reliable. However,
experiment evidence for this assumption is scarce (St. Clair, 2010). On the other
105
hand, broad spectrum resistance occurs when the defense is effective against two or
more types of pathogen species or to several strains or isolates of the same species of
pathogen (Kou and Wang, 2010).The term is often used as a synonym for QDR, but it
is important to note that there are several examples of strain-specificquantitative
resistance loci (QRLs) (Poland et al., 2009). On the other hand,PRRs can confer a
broad spectrum defense (Zipfel, 2014) and single R genes can also mediate broad-
spectrum resistance (Xiao et al., 2001; Zhao et al., 2004; Narusaka et al., 2009)or can
be engineered to achieve this type of resistance (Segretin et al., 2014;
Giannakopoulou et al., 2015).
How to study complex traits and QDRs
Continuous traits do not exhibit single phenotype inheritance, as it occurs in
progenies that segregate according to the Mendelian rules (e.g. 3:1 or 15:1).
However, this is not a consequence of a different genetic mechanism per se.
Conversely, the variation in the phenotype observed for a quantitative trait is the
result of a multiple genotypic expression of segregating alleles. In addition, it is highly
influenced by the environment. As a consequence, it is not possible to clearly
discriminate a phenotype of a particular genotype. For this reason, the Mendelian
mechanisms in the study of quantitative traits is masked despite the fact that each
gene could be segregated in a Mendelian mode (Griffiths, 2005).
For the above reasons, how to study quantitative traits has represented a new
challenge to classical geneticists. The first studies on complex quantitative traits in
plants included the association between the size of the seed (a quantitative trait) and
the seed coat color (a qualitative trait)(Sax, 1923). The initial idea of quantitative loci
mapping was first proposed by Thoday (1961), based on the observation that
segregating single gene markers could be linked with loci associated with complex
traits. Later, in 1982, the term quantitative trait loci (QTL) was used for first time to
name the different loci that determine several quantitative traits in tomato (Tanksley
et al., 1982).
The logic behind QTL detection is to determinate the relationship between DNA
variation, as captured by DNA-based markers, and the observed phenotypic variation
(Mackay et al., 2009). The identification of QTL starts with the construction of a
genetic map, where a large group of molecular markers are positioned in linkage
groups (chromosomes) based on recombination frequencies. Once obtained, these
106
markers are associated with the phenotypic trait of interest. This is accomplished
with the principle that the genes responsible for a particular trait segregate via
chromosomal recombination during meiosis (Collard et al., 2005).
In the past, the challenge was to increase the number of molecular markers present
in genetic maps for QTL mapping purposes. During the 90s, the development of DNA-
based markers revolutionized the ability to detect DNA variations (Phillips and Vasil,
2013). Molecular markers, such as Restriction Fragment Length Polymorphism
(RFLP), Random Amplified Polymorphic DNA (RAPD), Amplified fragment length
polymorphism (AFLP) and Simple Sequence Repeat (SSR), contributed significantly
to the development of high-dense genetic maps, allowing for the dissection of
qualitative and quantitative traits. Despite significant efforts, the genetic maps
obtained through the use of these markers were generally low-saturated because of
the lack of markers representing the complete set of recombination events. Thus, the
first versions of maize and tomato maps contained only 50 RFLP markers each
(Helentjaris et al., 1988) and the first potato map had 135 RFLP markers (Bonierbale
et al., 1988). Moreover, at that time, the QTL intervals were large, usually ranging
from 10 to 30 cM (Glazier et al., 2002). These limitations were overcome by massive
sequencing technologies (Ansorge, 2009). These new genotypification technologies
allowed for the high throughput identification of Single Nucleotide Polymorphism
(SNPs) (), which have become the most widely used molecular marker. Nowadays,
hundreds or thousands of widely distributed SNPs and the positions in the genome (if
the reference genome is accessible) can be identified in a relatively short period of
time and at a low cost. Thereby, in recent years, the number of molecular markers
and, therefore, map resolutions have increased, which ultimately leads to the
reduction of QTL interval lengths to a few cM (Gautami et al., 2012;; Soto et al., 2015).
The association mapping (AM) approach arrived at the beginning of the XXI century
as an alternative to the linkage mapping approach for QTL identification. In this case,
the analysis is based on the phenomenon of linkage disequilibrium (LD) andthe
exploration of the historical recombination events at the population level (Zhu et al.,
2008). The AM take advantage of the explosion of new genome-scale data, allowing
for a higher resolution, as compared with linkage mapping (Zhu et al., 2008). In this
case, two strategies for the dissection of complex traits can be followed. The first is
the candidate gene approach and the second is genome wide association studies
(GWAS) (Brachi et al., 2011). While in the QTL linkage mapping approach, the
quantitative candidate genes are located at an interval; in the AM, a direct association
between complex traits and the polymorphic markers, usually SNPs, is achieved
(Rafalski, 2002). However, both approaches seem to be complementary in the sense
that their ultimate goal is the detection of the genes underlying the quantitative
107
complex trait for further cloning. Several examples of the use of the association
mapping approach in QDR studies for the more limiting diseases can be found in
recent literature (Benson et al., 2015; Gutiérrez et al., 2015; Iquira et al., 2015;
Arruda et al., 2016; Olukolu et al., 2016; Turuspekov et al., 2016). However, despite
the broad use of this approach, nogenes detected by AM for plant QDR have been
cloned so far.
Recently, two approaches had been proposed for studies on quantitative traits. First,
there is extreme-phenotype GWAS (XP-GWAS), a new variant combining bulk
segregant analysis (BSA) and GWAS (Yang et al., 2015). The second approach takes
advantage of the use of clustered regularly interspaced short palindromic repeats
(CRISPR-Cas9) by obtaining targeted mitotic recombination events without needing
to develop directed crosses (Sadhu et al., 2016). Through this approach, high
frequency double strand breaks (DSB) are induced in regions of interest in mitotic
cells. Then, the intrinsic cell reparation by homologous recombination (HR),
generates recombination events that lead to the formation of a recombinant. Thus,
the high efficiency of CRISPR-Cas9 mediating recombination events within 20 kb of
the targeted site has been demonstrated. Comparing this rate of recombination with
that obtained by random meiotic segregation, the later would require more than
seven thousand individuals (Sadhu et al., 2016). The application of XP-GWAS and
CRISPR-Cas9 approaches and their potential scopes in QDR is promising.
In recent years, the concept of the set of all the information supported
experimentally, no matter the methodology followed, about the QTL and its allele
variations for a trait in one species, has received the name QTLome (Salvi and
Tuberosa, 2015). Beyond constructing a QTLome, it is necessary to find a way to
integrate and give a global sense to all the high volume of QTL information. This is the
challenge of statistical QTL meta-analyses. The detection of common QTLs and the
identification of co-location resistance candidate genes from different experiments
and populations have been recently achieved using QTL meta-analyses in maize to
find resistance genes for virus diseases (Wang et al., 2016), leaf rust in wheat
(Soriano and Royo, 2015) and verticillium wiltin cotton (Zhang et al., 2015).
A new era for QDR studies: phenotyping has the last word
The greatest aim for QDR studies in the past century was to increase molecular
markers in mapping populations to capture all the allelic variants of genes that
govern complex traits. Advances in high-throughput sequencing technologies have
overcome this limitation, at least partially. The current challenge is to produce
108
quality phenotype data, increasing molecular information and representing the
bedrock of a new era of plant quantitative trait studies that will contribute to a better
understanding of QDR (Basu et al., 2015).
Advances in automated precision phenotyping or high-throughput phenotyping
apply technologies principally based on image, thermal, spectra and digital sensors,
from which quantitative phenotypic information can be generated (Araus and Cairns,
2014). There are several advantages of these approaches. First, the reduction of
subjectivity in the determination of disease incidence and symptoms during a
particular plant-pathogen interaction. Second, the increase in the number of plants
that can be evaluated. Finally, the increase in the reproducibility and the possibility
to collect data at numerous time points (Mutka and Bart, 2015).
Some of the more sophisticated technologies for high-throughput phenotyping
applied in QDR studies are hyper-spectral imaging, chlorophyll fluorescence imaging
and thermal imaging (Mutka and Bart, 2015). Plant diseases produce different
spectral reflectance patterns and plants suffering biotic stresses display changes in
chlorophyll fluorescence emission (Baker, 2008). It has been shown that pathogens
can change plant tissue temperature during the infection process. With these recent
technologies, all of these parameters can be measured, even in the early plant
phenological stages (Mutka et al., 2015). Wheat and sugarcane are some crops where
these technologies have been used for detection and study of QDR (Mahlein et al.,
2012; Bauriegel and Herppich, 2014; Mutka et al., 2016). Despite the fact that these
techniques require a large number of previous evaluations in order to set the
parameters for each disease, their potential in phenotyping plant disease is
undeniable.
Phenotype has also been studied from an “omics” view (Salvi and Tuberosa, 2015).
This new phenotyping method includes transcripts, proteins and metabolites, such as
elements directly related to the phenotype, which have led to approaches such as
expression-QTLs (eQTLs), protein-QTLs (pQTLs) and metabolite-QTLs (mQTLs),
respectively. eQTLs and mQTL are the more used given their progress in collection,
automation and analysis of data (Salvi and Tuberosa, 2015).
From theory to practice: QDR in breeding
109
The complexity of QDR represents a challenge and an opportunity for plant breeding.
A breeding scheme focused on obtaining qualitative disease resistance is relatively
simple. It would be enough to introduce a single R gene into a susceptible plant
background to confer resistance. On the other hand, in the case of QDR, the
introduction of a gene from a QTL can confer a reduction, but not absence of disease.
Thus, it would not be possible to get complete resistance until all of the resistance
responsible loci are identified. In addition, in contrast to the current relative large
repertoire of isolated R genes, the isolation of genes governing QDR for future use in
breeding programs has not been an easy task.
For decades, marker-assisted selection (MAS) (Xu and Crouch, 2008) and gene
pyramiding (Brun et al., 2010) efforts have been directed toward the identification of
QTLs with major effects, explaining more than 20% of phenotypic variance, for
introduction into plant resistance breeding programs (Collard et al., 2005). Some
examples with great success in achieving high levels of resistance are found in rice
(Bustamam et al., 2002), common bean (Miklas et al., 2006) and pearl millet (Sehgal,
2016), but unfortunately this has not been the case for most crops, including staple
crops, such as cassava.
In plant breeding programs focused on QDR, one of the limitations to be considered is
the linkage disequilibrium between the genes conferring resistance and closely
linked undesirable genes, a phenomenon called linkage drag (Summers and Brown,
2013). Undesirable genes may affect the commercially accepted gene pool and,
therefore, modify the quality and crop yield. If linkage drag is not eliminated or
decreased, the use of the QTL in the program will be impractical. The MAS strategy
has counteracted this phenomenon. Through high throughput genotyping and the use
of haplotype analysis of the introgressed region (QTL), the linkage drag in seedlings
can be detected and tracked in order to subsequently backcross these individuals to
resistant varieties lacking drag. This strategy was applied to detect and remove the
linkage drag around the Rpv12 gene and confer resistance to powdery mildew in
wine grapes (Vitis vinifera L.)(Venuti et al., 2013). Alternatively, the marker-assisted
recurrent selection (MARS), combined with genomic selection (GS)(Heffner et al.,
2009), can also contribute to solving the linkage drag problem for QDR (Summers
and Brown, 2013). The GS selects plant material carrying whole genome molecular
marker that are associated with resistance to a specific pathogen through the
prediction of the phenotype using GEBV (Falconer and Mackay, 1996). These GEBVs
are obtained by the compilation of molecular marker scores, phenotypic data
evaluation of several germplasm and populations under a range of environmental
conditions and (if it is available) pedigree information. Thus, MARS would increase
110
the frequency of insertion of the desirable gene, decreasing the incorporation of
undesirable ones and speeding up the detection of resistance loci with GS.
Another limiting factor in exploiting QTL with the aim of generating new varieties is
the effect of the environment on the QTL. Multi-environment analyses in QDR studies
offer the opportunity to detect the QTL x environment interaction (Q x E), conditional
QTLs (El-Soda et al., 2014) and QTL stability during seasons and crop cycles. Special
attention should be given to an eventual Q x E interaction in plant quantitative
resistance that is widely influenced by the environment and in which heritability
values are usually low (Ntare and Williams, 1998). On the other hand, the functional
validation of candidate genes is an important part of QDR studies, which can be
carried out by overexpressing or down regulating the candidate gene by applying
genetic engineering (Mittler and Blumwald, 2010) or exploiting the mutant
collections (Cavanagh et al., 2008).
In a large number of QDR studies the phenotypic evaluation (host response to the
pathogen) is done after an artificial inoculation, employing a particular strain or a
group of strains, allowing for the detection of QTL associated with these strains and
leaving aside other genomic regions involved in resistance to other strains. When
these QTLs are introgressed in particular varieties and evaluated under naturally
diseased fields, where different pathogen strains or races can be found, it is possible
to obtain unsatisfactory results. For this reason, it is mandatory that a breeding
program starts with the knowledge on the dynamics and diversity of the pathogen
populations.
Molecular explanation of quantitative resistance
Although important progress in understanding and analyzing complex traits has been
accomplished in recent years, knowledge on the molecular basis of the QDR is still
scarce. Several hypotheses have been generated to explain the function of the genes
that control the QDR (Poland et al., 2009). Five genes have been cloned from QTLs,
which has enriched the proposed models. The rice Pi21 gene, which encodes for a
protein that has a heavy metal–transport/detoxification domain, confers resistance
to several races of Magnaporthe oryzae (Fukuoka et al., 2009). The QDR genes from
wheat Yr36 (resistance to Puccinia striiformis f. sp. tritici) and Lr34 (resistance to
Puccinia striiformis, P. triticina and to Blumeria graminis) encode for a Kinase-START
protein and a pleiotropic drug resistance subfamily of ABC transporters, respectively
(Fu et al., 2009; Krattinger et al., 2009). In addition, the RKS1 gene, which encodes for
111
an atypical kinase identified in the model plant Arabidopsis thaliana, confers
resistance to most Xanthomonas campestris races, and to the pathovar sraphani,
incanae or armoraciae (Huard-Chauveau et al., 2013). Finally, the receptor-like
protein coded by the ZmWAK gene of maize that confers resistance to Sphacelotheca
reiliana was cloned recently (Zuo et al., 2015). Considering these discoveries and the
gaps in QDR knowledge, different explanations of how it works are plausible.
QDR as a continuous response that depends on gene expression intensity
The expression level of genes involved in plant resistance can play important roles on
the intensity of the final output response. Transcriptomic analyses have allowed for
the identification of global changes in the expression profiles of genes related to plant
immunity. Through expression analysis, it was possible to demonstrate that, during
incompatible, compatible and non-host interactions, gene expression profiles were
almost the same and that differences were seen in the intensity and speed of their
induction (Tao et al., 2003). Other studies support the overlap between PAMP and
ETI at the gene expression level (Navarro et al., 2004; Bozsó et al., 2009; Bozso et al.,
2016). For QDR, several studies have reported a direct relationship between the
expression level of some genes and the degree of resistance response. The maize
ZmWAK gene is induced after pathogen inoculation. This gene is highly expressed in
the mesocotyl and, at a lesser extent in the coleoptile of resistant lines, and the
expression level of ZmWAK can be associated with the degree of pathogen growth
restriction in mesocotyl and coleoptiles (Zuo et al., 2015). A similar situation is seen
for the RKS1 gene, whose expression is correlated with the resistance level in
different Arabidopsis accessions to Xanthomonas campestris pv. campestris (Xcc)
strain 568 (Huard-Chauveau et al., 2013). Finally, following the same rationality, but
in a contrary sense, the increase in the expression of susceptibility genes can also
increase the susceptibility of plants, as has been demonstrated for the susceptibility
Pi21 gene from rice. In this case, transgenic plants showing higher expression of this
gene were more susceptibility to a virulent race of Magnaporthe oryzae (Fukuoka et
al., 2009).
QDR can be explained not only as a consequence of the intensity of the gene
expression, but also by the specificity or pattern expression of these genes. The wheat
Lr34 gene, which was recently cloned from a QTL, is expressed at the relatively same
level in resistant and susceptible plants. However, its expression is lower in wheat
seedlings than in high adult plants where it is more effective Krattinger et al., 209. In
112
this case, the quantitative response depends not on the genetic background of the
plant, but on the control of gene expression exerted by other genes, whose
expression can be controlled by different factors, such as the development of the
plant, tissue or organ specificity, etc.
The above examples suggest a relationship between the transcription level of QDR
genes and the degree of resistance. We hypothesized that the higher the expression of
a gene involved in defense is, the higher the resistance level will be (Figure 2-2).
However, this relationship should not necessarily be linear. This is consistent with
the results obtained by Huard-Chauveau et al (2013), who found that the expression
of RKS1-L in natural accessions was negatively correlated with the disease index. The
question is: which is the factor that determines the intensity in the induction of gene
expression? During the plant response, it has been considered that the induction of
gene expression is a consequence of the activation of a signal pathway, which in turn
is dependent on the pathogen recognition. According to these ideas, the level of gene
expression could be conditioned depending on the ability, specificity and strength of
the interaction between pathogen-derived molecules and plant receptors.
Another aspect related to the transcriptional control and level of resistance is the
requirement for the presence of alternative spliced transcripts. The classical example
is the N gene from tobacco, conferring tobacco mosaic virus (TMV) resistance.
Through alternative splicing, NS and NL transcripts are produced. During the initial
phase of infection, the NS version, coding for the full-length N protein, is more
abundant, but, 4 hr after inoculation, the relationship is inverted. If only one of the
two variants is present, complete resistance is lost (Dinesh-Kumar and Baker, 2000).
For QDR, the above mentioned RKS1 gene, two transcripts were identified, with
differences in length between resistant and susceptible Arabidopsis accessions
(Huard-Chauveau et al., 2013). Another example is the Yr36 gene from wheat, which
can have up to six alternative transcript variants; however, only one of them codes
for a protein containing a complete START domain. This transcript is differentially
regulated by temperature and is the only one that is up-regulated after inoculation
with the fungus Pucciniastriiformis f. sp. tritici (Fu et al., 2009).
Figure 2-2. Model in which the expression level of QDR genes is associated with
the resistance phenotype.
113
To summarize, the transcription level differences shown by the QDR genes suggest
that it is necessary to incorporate the information on gene expression into the DNA
variation data in order to achieve a systematic genetic approach and, thus, gain a
better understanding of the molecular bases of the quantitative response (Mackay et
al., 2009).
R weak alleles
As mentioned before, the first step in the activation of plant immunity is pathogen-
derived molecules recognition. In the ETI, a specific, strong and direct or indirect
interaction between R proteins and the corresponding effector (named Avr) conducts
the activation of a signaling pathway, leading to immunity, which, in most cases, is
associated with an HR response. In ETI, this Avr-R interaction has only two
alternatives: it happens or it does not, generating two phenotypes, R or S. Several
studies on QTLs, have demonstrated the presence of the typical qualitative R genes,
coding for NB-LRR in QTLs, suggesting that the molecular bases of the pathogen
114
recognition can be shared with ETI. Supporting this hypothesis is the fact that both
responses, ETI and QDR, share some molecular components (Roux et al., 2014b). In
this model, the cause of the QDR is the presence of “weak alleles” of R genes (Roux et
al., 2014b). How to explain that an R protein confers only partial resistance? The R
protein is responsible for the recognition of a specific effector or the activity of it on a
pathogenicity target protein. A weak allele of the R protein can correspond to a
protein that is able to interact with an effector (or pathogenicity target), but its
affinity is not high enough to induce a strong response.
Allelic variation
The above hypothesis of weak R alleles can be the extreme case of allelic variation
related to QDR. However, other QDR genes have shown allelic variation. There is
enough evidence to indicate that QDR is associated with the allelic variation of genes
that differ in structure from canonical R proteins and that are important for plant
defense. For example, the recently cloned QDR gene, ZmWAK, exhibits seven
substitutions and a deletion between the resistant and susceptible maize lines.
Although, in this case, polymorphisms affecting protein function were not found(Zuo
et al., 2015), these polymorphisms could affect the interaction with other molecules
or could prevent protein complexes formation. A similar situation was observed for
the RKS1 gene, which, even if it is present in both resistant and susceptible
accessions, has several SNPs that have been found to be associated with both
phenotypes. These SNPs are located in the coding region and in the 5’ and 3’
regulatory regions. In fact, one of the identified susceptible alleles corresponds to a
stop codon in RKS1. The authors suggested that mutations could be associated with
susceptibility as a consequence of altered RKS1 long transcript expression (Huard-
Chauveau et al., 2013). Some polymorphisms have been also found between resistant
and susceptible wheat plants that are located in the Lr34 gene; two polymorphisms
were located in exons and one in an intron. The polymorphisms located in exons are
present in the resistance cultivar and correspond to a deletion of 3 bp and an SNP
that changes the amino acid tyrosine for histidine andaffect the first transmembrane
domain of the ABC transporter (Krattinger et al., 2009). Resistant and susceptible rice
cultivars have seven polymorphisms between them, located in the genomic region
that harbors the Pi21 gene. Two of these polymorphisms correspond to deletions and
were associated with the corresponding phenotype. Polymorphisms in this region,
between different cultivated rice accessions, allowed for the identification of 12
haplotypes and revealed the natural variation of QDR genes. Only one of this
haplotypes was associated with resistance (Fukuoka et al., 2009). Further studies
115
that include the intermediate phenotypes that are in between the lines or accessions
that have already been evaluated, will resolve the role of polymorphisms in QDR. It is
important to stress that these studies were conducted on contrasting lines,
representing extreme resistant and susceptible phenotypes, and it would be
interesting to evaluate the expression levels of QDR genes in individuals showing a
gradient of phenotypic response.
Polymorphisms can be present even in promoter sequences, as happens with the
ZmWAK gene; however, no association with function or phenotype has been studied
for these variations (Zuo et al., 2015). Additional studies are required to reveal if
there is an association between these polymorphisms and the quantitative response.
In addition, polymorphisms located in introns or in promoters can modify
transcription factor binding and splicing events, generating a particular quantity of
transcripts or differential timing and tissue specificity gene expression (Mackay et al.,
2009).
The studies presented here to exemplify allelic variation are important, not just
because they show that these polymorphisms represent different alleles, but because
the polymorphisms were found in QDR genes that were cloned and validated, and, in
this sense, these sequence differences could be the cause of the quantitative
response. The number and nature of polymorphisms found in a QDR gene or in its
genomic region could define the level of the phenotype. Huard-Chauveau et al.,
suggested that the quantitative disease response phenotype could be due the additive
effect or interaction of SNPs present in the identified haplotypes (Huard-Chauveau et
al., 2013). In a similar fashion, we propose that there is a highly resistant phenotype
that is associated with a haplotype of a QDR gene or genomic region and those
variations of this haplotype would lead to the quantitative characteristic of the
resistance. Additional SNPs could be in other genes or other genomic regions that
account for the resistance. Furthermore, It could be that susceptible alleles compete
with resistant alleles for the interaction with scaffold proteins of molecular signaling
complexes (Huard-Chauveau et al., 2013). Complementary studies with accessions
that represent the range of the response showing different levels of resistance and
the corresponding sequence polymorphisms will help to tell if the polymorphisms
present in QDR sequences are associated with the phenotype in order to shape or
discard this hypothesis.
Kinases and signaling
116
Kinases are essential components in plant biology and regulate different processes,
such as biotic stress (Afzal et al., 2008; Parniske, 2008), hormone signaling (Santner
and Estelle, 2009), growth (Hematy and Hofte, 2008), cell differentiation and other
physiological processes (De Smet et al., 2009). Serine-threonine kinases are
important signaling transduction components of PTI and ETI (Zipfel, 2014) and MAP
kinase cascades regulate downstream defense responses (Schwessinger and Zipfel,
2008). Additionally, pseudokinases are described as being important in signaling
network control (Huard-Chauveau et al., 2013). In this context, it is not unreasonable
to consider genes involved in the signaling pathway, including MAP kinases, as key
elements of QDR.
Several proteins that have been characterized as responsible for QDR have proved to
be kinases or pseudokinases. RKS1 from Arabidopsis (Huard-Chauveau et al., 2013) is
a typical kinase, ZmWAK protein from maize contains a kinase domain (Zuo et al.,
2015) and Yr36 from wheat has a kinase domain similar to Arabidopsis WAK-like
kinases (Fu et al., 2009). In addition, through eQTLs in barley, a gene was identified
that encodes for a “putative histidin-kinase” as an important component of the
resistance to Puccinia graminis f. sp. tritici (Druka et al., 2008). This biochemical
characteristic opens the door to possibilities for the role that these proteins may play
in QDR, as for example like transmitting molecular signals.
Miscellaneous
One interpretation of how QDR works at the cellular level considers the phenotypes
of complete resistance (with hypersensitive response) or susceptibility as the
extreme responses of PTI or ETI, while QDR is the product of a weak PTI or ETI
(Lopez, 2011; Kushalappa et al., 2016). In this way, the resistance, known as
qualitative, could also be polygenic and is achieved if all the components are present
and functioning correctly (Kushalappa et al., 2016). Therefore, the quantitative
counterpart may have missing components. The missing concept here not necessarily
corresponds to complete absence of a particular component, but to differential
quantities of resistance related metabolites, proteins coded by R genes, or PRRs,
which in turn are regulated by other genes (Kushalappa et al., 2016). Therefore, the
more defense-related components that are missing (or diminished), the more the
resistance is reduced. An alternative, but not excluding, hypothesis states that
differences in defense responses are the consequence of the sensitivity of the
components to input signals. It was hypothesized that resistant plants display robust
117
responses because they are insensitive to small changes in input signals (Tao et al.,
2003); therefore, the remaining range of responses of QDR could be more sensitive to
this change. QDR have been recently redefined due to the cloning of some of the
corresponding genes and it has been stated that the involved proteins do not belong
to a specific group, such as in the case of R genes, but may have several functions
(Navabi et al., 2005; Poland et al., 2009; Bryant et al., 2014; Roux et al., 2014a). Thus
new molecules, which previously have not been described as important during plant-
microbe interactions, could be responsible for the resistance (Roux et al., 2014b).
Nevertheless, it seems that R genes actually have roles in QDR, but with low
representation, as compared to other genes with different structures and functions
(Corwin et al., 2016). The new evidence suggests that the key QDR molecules are not
R proteins, but this does not mean that these proteins have to be excluded from the
model. They could also participate to a lesser extent.
Conclusions
In the present review, we would like to highlight the impact that QDR should have on
plant breeding, but that, unfortunately, is not happening. Although information
coming from functional gene studies is scare and relatively new, there are thousands
of studies on QTL identification. Consequently, there is a misbalance between the
published QTLs studies and the application of this information in field. This reflects
the bottleneck in the application of QDR and the lack of efforts made to validate these
QTLs. In addition, results coming from QTL studies could lead to false conclusions. A
genomic region could be identified as responsible for disease resistance because of a
statistic artifact. Furthermore, unless QTL detection is done in well-controlled
greenhouses or growth cambers, these experiments should have repetitions in
different growing cycles and different seasons or weather conditions to be sure that
the identified QTL is real and stable. Another flaw is found in QDR studies; the
identification of these QTL is frequently done with only one strain of the pathogen or
even when it is not known which strain is being evaluated because some studies are
done with natural inoculation. In this way, QTL validation should include multiple
strain inoculations. Plant-microbe is a two-way interaction, so the genetic
characteristics of the pathogen are necessary components that should be taken into
account. Frequently, QTL are involved in resistance to populations of pathogens;
however, no information about evolution or diversity of the pathogen is included or
old information is often used. Finally, the problem of subjectivity that adds error to
QTL studies will be removed with the arrival of precision phenotyping that, in the
end, will result in the reproducibility and reliability of these studies.
118
We gathered results and experiences from different pathosystems for QDR because
we wanted to highlight some components of plant defense that are known, but that
have not been integrated or incorporated to plant immunity models. Thereby, plant
immunity must be seen from a different point of view. We propose that plant defense
is not a simple and single-layered outcome, but instead is a synergistic process and
represents the sum of several protein interactions that are being taken at the same
time. Even hypotheses or mechanisms proposed by other authors, such as Poland et
al (2009), may occur simultaneously. Here, we present different molecular
explanations of how QDR work. It is important to note that all of these explanations
can arise together and they are not exclusive. We also suggest that it is necessary to
consider alternatives to imposed models. A particular and specific model could be
applied to one pathosystem, but not to others. Each plant and/or pathogen could
have its own characteristics within a set of shared components, determining the
validity of one model or other. Even though the zig-zag model has helped to
understand plant immunity, it just explains a part of it, which represents a monogenic
interaction in which QDR are not included. To achieve a better understanding of plant
immunity, a holistic approach should be considered that integrates the intricate array
of interactions between molecules and cells, determining the complexity of
phenotypic traits.
References
Afzal, A.J., Wood, A.J., and Lightfoot, D.A. 2008. Plant receptor-like serine threonine kinases: roles in signaling and plant defense. Mol Plant Microbe Interact 21:507-517.
Ansorge, W.J. 2009. Next-generation DNA sequencing techniques. New biotechnology 25:195-203.
Araus, J.L., and Cairns, J.E. 2014. Field high-throughput phenotyping: the new crop breeding frontier. Trends in plant science 19:52-61.
Arruda, M.P., Brown, P., Brown-Guedira, G., Krill, A.M., Thurber, C., Merrill, K.R., Foresman, B.J., and Kolb, F.L. 2016. Genome-Wide Association Mapping of Fusarium Head Blight Resistance in Wheat using Genotyping-by-Sequencing. The Plant Genome 9.
Baker, N.R. 2008. Chlorophyll fluorescence: a probe of photosynthesis in vivo. Annu. Rev. Plant Biol. 59:89-113.
Basu, P.S., Srivastava, M., Singh, P., Porwal, P., Kant, R., and Singh, J. 2015. High-precision phenotyping under controlled versus natural environments. Pages 27-40 in: Phenomics in Crop Plants: Trends, Options and Limitations, Springer.
119
Bauriegel, E., and Herppich, W.B. 2014. Hyperspectral and chlorophyll fluorescence imaging for early detection of plant diseases, with special reference to Fusarium spec. infections on wheat. Agriculture 4:32-57.
Benson, J.M., Poland, J.A., Benson, B.M., Stromberg, E.L., and Nelson, R.J. 2015. Resistance to gray leaf spot of maize: genetic architecture and mechanisms elucidated through nested association mapping and near-isogenic line analysis. PLoS Genet 11:e1005045.
Bonierbale, M.W., Plaisted, R.L., and Tanksley, S.D. 1988. RFLP maps based on a common set of clones reveal modes of chromosomal evolution in potato and tomato. Genetics 120:1095-1103.
Bozso, Z., Ott, P.G., Kaman-Toth, E., Bognar, G.F., Pogany, M., and Szatmari, A. 2016. Overlapping yet response-specific transcriptome alterations characterize the nature of tobacco-Pseudomonas syringae Interactions. Front Plant Sci 7:251.
Bozsó, Z., Maunoury, N., Szatmari, A., Mergaert, P., Ott, P.G., Zsíros, L.R., Szabó, E., Kondorosi, É., and Klement, Z. 2009. Transcriptome analysis of a bacterially induced basal and hypersensitive response of Medicago truncatula. Plant molecular biology 70:627-646.
Brachi, B., Morris, G.P., and Borevitz, J.O. 2011. Genome-wide association studies in plants: the missing heritability is in the field. Genome biology 12:1-8.
Brun, H., Chèvre, A.M., Fitt, B.D., Powers, S., Besnard, A.L., Ermel, M., Huteau, V., Marquer, B., Eber, F., and Renard, M. 2010. Quantitative resistance increases the durability of qualitative resistance to Leptosphaeria maculans in Brassica napus. New Phytologist 185:285-299.
Bryant, R.R.M., McGrann, G.R.D., Mitchell, A.R., Schoonbeek, H.-j., Boyd, L.A., Uauy, C., Dorling, S., and Ridout, C.J. 2014. A change in temperature modulates defence to yellow (stripe) rust in wheat line UC1041 independently of resistance gene Yr36. BMC plant biology 14:1-10.
Bustamam, M., Tabien, R., Suwarno, A., Abalos, M., Kadir, T., Ona, I., Bernardo, M., VeraCruz, C., and Leung, H. (2002). Asian rice biotechnology network: Improving popular cultivars through marker-assisted backcrossing by the NARES. In Poster presented at the international rice congress, pp. 16-20.
Cavanagh, C., Morell, M., Mackay, I., and Powell, W. 2008. From mutations to MAGIC: resources for gene discovery, validation and delivery in crop plants. Current opinion in plant biology 11:215-221.
Collard, B., Jahufer, M., Brouwer, J., and Pang, E. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142:169-196.
Cook, D.E., Mesarich, C.H., and Thomma, B.P. 2015. Understanding plant immunity as a surveillance system to detect invasion. Annual review of phytopathology.
120
Corwin, J.A., Copeland, D., Feusier, J., Subedy, A., Eshbaugh, R., Palmer, C., Maloof, J., and Kliebenstein, D.J. 2016. The quantitative basis of the Arabidopsis innate immune system to endemic pathogens depends on pathogen genetics. PLoS Genet 12:e1005789.
Cui, H., Tsuda, K., and Parker, J.E. 2015. Effector-triggered immunity: from pathogen perception to robust defense. Annual review of plant biology 66:487-511.
Chisholm, S.T., Coaker, G., Day, B., and Staskawicz, B.J. 2006. Host-microbe interactions: shaping the evolution of the plant immune response. Cell 124:803-814.
De Smet, I., Voss, U., Jurgens, G., and Beeckman, T. 2009. Receptor-like kinases shape the plant. Nat Cell Biol 11:1166-1173.
Dinesh-Kumar, S.P., and Baker, B.J. 2000. Alternatively spliced N resistance gene transcripts: their possible role in tobacco mosaic virus resistance. Proceedings of the National Academy of Sciences of the United States of America 97:1908-1913.
Druka, A., Potokina, E., Luo, Z., Bonar, N., Druka, I., Zhang, L., Marshall, D.F., Steffenson, B.J., Close, T.J., Wise, R.P., Kleinhofs, A., Williams, R.W., Kearsey, M.J., and Waugh, R. 2008. Exploiting regulatory variation to identify genes underlying quantitative resistance to the wheat stem rust pathogen Puccinia graminis f. sp. tritici in barley. Theoretical and Applied Genetics 117:261-272.
El-Soda, M., Malosetti, M., Zwaan, B.J., Koornneef, M., and Aarts, M.G. 2014. Genotypexenvironment interaction QTL mapping in plants: lessons from Arabidopsis. Trends in plant science 19:390-398.
Falconer, D.S., and Mackay, T.F.C. 1996. Introduction to Quantitative Genetics Prentice Hall, London, UK.
Flor, H.H. 1955. Host-parasite interaction in flax rust—Its genetics and other implications. Phytopathology 45:680–685.
Fu, D., Uauy, C., Distelfeld, A., Blechl, A., Epstein, L., Chen, X., Sela, H., Fahima, T., and Dubcovsky, J. 2009. A kinase-START gene confers temperature-dependent resistance to wheat stripe rust. Science (New York, N.Y 323:1357-1360).
Fukuoka, S., Saka, N., Koga, H., Ono, K., Shimizu, T., Ebana, K., Hayashi, N., Takahashi, A., Hirochika, H., Okuno, K., and Yano, M. 2009. Loss of function of a proline-containing protein confers durable disease resistance in rice. Science (New York, N.Y 325:998-1001).
Gautami, B., Foncéka, D., Pandey, M.K., Moretzsohn, M.C., Sujay, V., Qin, H., Hong, Y., Faye, I., Chen, X., and BhanuPrakash, A. 2012. An international reference consensus genetic map with 897 marker loci based on 11 mapping populations for tetraploid groundnut (Arachis hypogaea L.). PLoS ONE 7:e41213.
Giannakopoulou, A., Steele, J.F., Segretin, M.E., Bozkurt, T.O., Zhou, J., Robatzek, S., Banfield, M.J., Pais, M., and Kamoun, S. 2015. Tomato I2 immune receptor can be engineered to confer partial resistance to the oomycete Phytophthora infestans in
121
addition to the fungus Fusarium oxysporum. Mol Plant Microbe Interact 28:1316-1329.
Glazier, A.M., Nadeau, J.H., and Aitman, T.J. 2002. Finding genes that underlie complex traits. Science (New York, N.Y 298:2345-2349.
Griffiths, A.J. 2005. An introduction to genetic analysis. Macmillan.
Gutiérrez, L., Germán, S., Pereyra, S., Hayes, P.M., Pérez, C.A., Capettini, F., Locatelli, A., Berberian, N.M., Falconi, E.E., and Estrada, R. 2015. Multi-environment multi-QTL association mapping identifies disease resistance QTL in barley germplasm from Latin America. Theoretical and Applied Genetics 128:501-516.
Heffner, E.L., Sorrells, M.E., and Jannink, J.-L. 2009. Genomic selection for crop improvement. Crop Science 49:1-12.
Helentjaris, T., Weber, D., and Wright, S. 1988. Identification of the genomic locations of duplicate nucleotide sequences in maize by analysis of restriction fragment length polymorphisms. Genetics 118:353-363.
Hematy, K., and Hofte, H. 2008. Novel receptor kinases involved in growth regulation. Current opinion in plant biology 11:321-328.
Houterman, P.M., Ma, L., van Ooijen, G., de Vroomen, M.J., Cornelissen, B.J., Takken, F.L., and Rep, M. 2009. The effector protein Avr2 of the xylem-colonizing fungus Fusarium oxysporum activates the tomato resistance protein I-2 intracellularly. Plant J 58:970-978.
Huard-Chauveau, C., Perchepied, L., Debieu, M., Rivas, S., Kroj, T., Kars, I., Bergelson, J., Roux, F., and Roby, D. 2013. An atypical kinase under balancing selection bonfers broad-spectrum disease resistance in Arabidopsis. PLoS Genet 9:e1003766.
Iquira, E., Humira, S., and François, B. 2015. Association mapping of QTLs for sclerotinia stem rot resistance in a collection of soybean plant introductions using a genotyping by sequencing (GBS) approach. BMC plant biology 15:1.
Johnson, R. 1983. Genetic Background of Durable Resistance. Pages 5-26 in: Durable Resistance in Crops, F. Lamberti, J.M. Waller, and N.A. Graaff, eds. Springer New York, Boston, MA.
Jones, J.D., and Dangl, J.L. 2006. The plant immune system. Nature 444:323-329.
Kou, Y., and Wang, S. 2010. Broad-spectrum and durability: understanding of quantitative disease resistance. Current opinion in plant biology 13:181-185.
Krattinger, S.G., Lagudah, E.S., Spielmeyer, W., Singh, R.P., Huerta-Espino, J., McFadden, H., Bossolini, E., Selter, L.L., and Keller, B. 2009. A putative ABC transporter confers durable resistance to multiple fungal pathogens in wheat. Science (New York, N.Y 323:1360-1363).
Kushalappa, A.C., Yogendra, K.N., and Karre, S. 2016. Plant innate immune response: qualitative and quantitative resistance. Critical Reviews in Plant Sciences 35:38-55.
122
López, C. 2011. Descifrando las bases moleculares de la resistencia cuantitativa. Acta Biol Col 16.
Mackay, T.F.C., Stone, E.A., and Ayroles, J.F. 2009. The genetics of quantitative traits: challenges and prospects. Nature reviews 10:565-577.
Mahlein, A.-K., Steiner, U., Hillnhütter, C., Dehne, H.-W., and Oerke, E.-C. 2012. Hyperspectral imaging for small-scale analysis of symptoms caused by different sugar beet diseases. Plant methods 8:1.
Miklas, P.N., Kelly, J.D., Beebe, S.E., and Blair, M.W. 2006. Common bean breeding for resistance against biotic and abiotic stresses: from classical to MAS breeding. Euphytica 147:105-131.
Mittler, R., and Blumwald, E. 2010. Genetic engineering for modern agriculture: challenges and perspectives. Annual review of plant biology 61:443-462.
Mutka, A.M., and Bart, R.S. 2015. Image-based phenotyping of plant disease symptoms. Frontiers in Plant Science 5:734.
Mutka, A.M., Fentress, S.J., Sher, J.W., Berry, J.C., Pretz, C., Nusinow, D.A., and Bart, R. 2016. Quantitative, image-based phenotyping methods provide insight into spatial and temporal dimensions of plant disease. Plant physiology:pp. 00984.02016.
Narusaka, M., Shirasu, K., Noutoshi, Y., Kubo, Y., Shiraishi, T., Iwabuchi, M., and Narusaka, Y. 2009. RRS1 and RPS4 provide a dual Resistance-gene system against fungal and bacterial pathogens. Plant J 60:218-226.
Navabi, A., Singh, R.P., Huerta-Espino, J., and Tewari, J.P. 2005. Phenotypic association of adult-plant resistance to leaf and stripe rusts in wheat. Canadian Journal of Plant Pathology 27:396-403.
Navarro, L., Zipfel, C., Rowland, O., Keller, I., Robatzek, S., Boller, T., and Jones, J.D. 2004. The transcriptional innate immune response to flg22. Interplay and overlap with Avr gene-dependent defense responses and bacterial pathogenesis. Plant physiology 135:1113-1128.
Niks, R.E., Qi, X., and Marcel, T.C. 2015. Quantitative resistance to biotrophic filamentous plant pathogens: concepts, misconceptions, and mechanisms. Annual review of phytopathology 53:445-470.
Ntare, B., and Williams, J. 1998. Heritability and genotype x environment interaction for yield and components of a yield model in segregating population of groundnut under semi-arid conditions. Afr. Crop Sci. J 6:119-127.
Olukolu, B.A., Tracy, W.F., Wisser, R., De Vries, B., and Balint-Kurti, P.J. 2016. A Genome-Wide Association Study for Partial Resistance to Maize Common Rust. Phytopathology:11-15.
Parniske, M. 2008. Arbuscular mycorrhiza: the mother of plant root endosymbioses. Nat Rev Micro 6:763-775.
Phillips, R.L., and Vasil, I.K. 2013. DNA-based markers in plants. Springer Science & Business Media.
123
Poland, J.A., Balint-Kurti, P.J., Wisser, R.J., Pratt, R.C., and Nelson, R.J. 2009. Shades of gray: the world of quantitative disease resistance. Trends in plant science 14:21-29.
Rafalski, A. 2002. Applications of single nucleotide polymorphisms in crop genetics. Current opinion in plant biology 5:94-100.
Roux, F., Noël, L., Rivas, S., and Roby, D. 2014a. ZRK atypical kinases: emerging signaling components of plant immunity. New Phytologist 203:713-716.
Roux, F., Voisin, D., Badet, T., Balague, C., Barlet, X., Huard-Chauveau, C., Roby, D., and Raffaele, S. 2014b. Resistance to phytopathogens e tutti quanti: placing plant quantitative disease resistance on the map. Mol Plant Pathol 15:427-432.
Sadhu, M.J., Bloom, J.S., Day, L., and Kruglyak, L. 2016. CRISPR-directed mitotic recombination enables genetic mapping without crosses. Science (New York, N.Y 352:1113-1116.
Salvi, S., and Tuberosa, R. 2015. The crop QTLome comes of age. Curr Opin Biotechnol 32:179-185.
Santner, A., and Estelle, M. 2009. Recent advances and emerging trends in plant hormone signalling. Nature 459:1071-1078.
Sax, K. 1923. The association of size differences with seed-coat pattern and pigmentation in Phaseolus vulgaris. Genetics 8:552-560.
Schwessinger, B., and Zipfel, C. 2008. News from the frontline: recent insights into PAMP-triggered immunity in plants. Current opinion in plant biology 11:389-395.
Segretin, M.E., Pais, M., Franceschetti, M., Chaparro-Garcia, A., Bos, J.I., Banfield, M.J., and Kamoun, S. 2014. Single amino acid mutations in the potato immune receptor R3a expand response to Phytophthora effectors. Mol Plant Microbe Interact 27:624-637.
Sehgal, D. 2016. Advances in Molecular Breeding of Pearl Millet. Pages 397-419 in: Molecular Breeding for Sustainable Crop Improvement, Springer.
Soriano, J.M., and Royo, C. 2015. Dissecting the genetic architecture of leaf rust resistance in wheat by QTL meta-analysis. Phytopathology 105:1585-1593.
Soto, J.C., Ortiz, J.F., Perlaza-Jiménez, L., Vásquez, A.X., Lopez-Lavalle, L.A.B., Mathew, B., Léon, J., Bernal, A.J., Ballvora, A., and López, C.E. 2015. A genetic map of cassava (Manihot esculenta Crantz) with integrated physical mapping of immunity-related genes. BMC genomics 16:1.
St. Clair, D.A. 2010. Quantitative disease resistance and quantitative resistance loci in breeding. Annual review of phytopathology 48.
Stephens, A., Lombardi, M., Cogan, N.O., Forster, J.W., Hobson, K., Materne, M., and Kaur, S. 2014. Genetic marker discovery, intraspecific linkage map construction and quantitative trait locus analysis of Ascochyta blight resistance in chickpea (Cicer arietinum L.). Molecular Breeding 33:297-313.
124
Summers, R., and Brown, J. 2013. Constraints on breeding for disease resistance in commercially competitive wheat cultivars. Plant Pathology 62:115-121.
Tanksley, S.D., Medina-Filho, H., and Rick, C.M. 1982. Use of naturally-occurring enzyme variation to detect and map genes controlling quantitative traits in an interspecific backcross of tomato. Heredity 49:11-25.
Tao, Y., Xie, Z., Chen, W., Glazebrook, J., Chang, H.S., Han, B., Zhu, T., Zou, G., and Katagiri, F. 2003. Quantitative nature of Arabidopsis responses during compatible and incompatible interactions with the bacterial pathogen Pseudomonas syringae. Plant Cell 15:317-330.
Turuspekov, Y., Ormanbekova, D., Rsaliev, A., and Abugalieva, S. 2016. Genome-wide association study on stem rust resistance in Kazakh spring barley lines. BMC plant biology 16:13.
Venuti, S., Copetti, D., Foria, S., Falginella, L., Hoffmann, S., Bellin, D., Cindrić, P., Kozma, P., Scalabrin, S., and Morgante, M. 2013. Historical introgression of the downy mildew resistance gene Rpv12 from the Asian species Vitis amurensis into grapevine varieties. PLoS ONE 8:e61228.
Wang, Y., Xu, J., Deng, D., Ding, H., Bian, Y., Yin, Z., Wu, Y., Zhou, B., and Zhao, Y. 2016. A comprehensive meta-analysis of plant morphology, yield, stay-green, and virus disease resistance QTL in maize (Zea mays L.). Planta 243:459-471.
Xiao, S., Ellwood, S., Calis, O., Patrick, E., Li, T., Coleman, M., and Turner, J.G. 2001. Broad-spectrum mildew resistance in Arabidopsis thaliana mediated by RPW8. Science (New York, N.Y 291:118-120.
Xu, Y., and Crouch, J.H. 2008. Marker-assisted selection in plant breeding: from publications to practice. Crop Science 48:391-407.
Yang, J., Jiang, H., Yeh, C.T., Yu, J., Jeddeloh, J.A., Nettleton, D., and Schnable, P.S. 2015. Extreme‐phenotype genome‐wide association study (XP‐GWAS): a method for identifying trait‐associated variants by sequencing pools of individuals selected from a diversity panel. The Plant Journal 84:587-596.
Zhang, J., Yu, J., Pei, W., Li, X., Said, J., Song, M., and Sanogo, S. 2015. Genetic analysis of Verticillium wilt resistance in a backcross inbred line population and a meta-analysis of quantitative trait loci for disease resistance in cotton. BMC genomics 16:1.
Zhao, B.Y., Ardales, E., Brasset, E., Claflin, L.E., Leach, J.E., and Hulbert, S.H. 2004. The Rxo1/ Rba1 locus of maize controls resistance reactions to pathogenic and non-host bacteria. TAG. Theoretical and applied genetics 109:71-79.
Zhu, C., Gore, M., Buckler, E.S., and Yu, J. 2008. Status and prospects of association mapping in plants. The Plant Genome 1:5-20.
Zipfel, C. 2014. Plant pattern-recognition receptors. Trends in immunology 35:345-351.
125
Zuo, W., Chao, Q., Zhang, N., Ye, J., Tan, G., Li, B., Xing, Y., Zhang, B., Liu, H., Fengler, K.A., Zhao, J., Zhao, X., Chen, Y., Lai, J., Yan, J., and Xu, M. 2015. A maize wall-associated kinase confers quantitative resistance to head smut. Nature genetics 47:151-157.
126
CHAPTER 3
“I know I'm supposed to hate humans, but there's something about them. They don't
just survive, they discover, they create...I mean, just look at what they do with food”
-Ratatouille (Ratatouille, 2007)
127
A genetic map of cassava (Manihot esculenta Crantz) with integrated
physical mapping of immunity-related genes
Johana Carolina Soto Sedano1, Juan Felipe Ortiz1,5, Laura Perlaza-Jiménez2,6,, Andrea
Ximena Vásquez Chacón1 , Luis Augusto Becerra Lopez-Lavalle3, Boby Mathew4, Jens
Léon4, Adriana Jimena Bernal Giraldo 2, Agim Ballvora4, Camilo Ernesto López
Carrascal1
1 Manihot Biotec Laboratory, Biology department, Universidad Nacional de Colombia,
Bogotá, Colombia. 2 Laboratory of Mycology and Plant Pathology, Universidad de los Andes, Bogotá,
Colombia. 3 International Center for Tropical Agriculture (CIAT), Cali, Colombia. 4 INRES-Plant Breeding University of Bonn, Bonn, Germany. 5 Present address Department of Biological Sciences, Vanderbilt University,
Tennessee, USA. 6 Present address Max Planck Institute for Molecular Plant Physiology, Potsdam-
Golm, Germany
Published in BMC Genomics 2015. doi: 10.1186/s12864-015-1397-4.
Abstract
Cassava, Manihot esculenta Crantz, is one of the most important crops world-wide
representing the staple security for more than one billion of people. The development
of dense genetic and physical maps, as the basis for implementing genetic and
molecular approaches to accelerate the rate of genetic gains in breeding program
represents a significant challenge. The advent of novel molecular and bioinformatics
technologies makes it possible to generate and analyze thousands of DNA markers in
order to accomplish this task. A reference genome sequence for cassava has been
made recently available and community efforts are underway for improving its
quality. Cassava is threatened by several viral, bacterial and fungal pathogens, but the
mechanisms of defense are far from being understood. These genomic resources
could be useful for breeding resistance to these biotic stress factors. In addition, there
has been a lack of information about the number of genes related to immunity as well
as their distribution and genomic organization in the cassava genome. A new high
128
dense genetic map of cassava containing 2,141 SNPs generated by genotyping by
sequencing (GBS) approach has been constructed. Eighteen linkage groups were
resolved with an overall size of 2,571 cM and an average distance of 1.26 cM between
markers. More than half of mapped SNPs (57.4%) are located in coding DNA
sequences, 27% within introns, 10.4% within promoters and 5% within un-
translated regions (UTR). Physical mapping of scaffolds of cassava whole genome
sequence draft using the mapped markers as anchors resulted in the orientation of
687 scaffolds covering 45.6% of the genome. One hundred eighty nine new scaffolds
are anchored to the genetic cassava map leading to an extension of the present
cassava physical map with 30.7Mb. Comparative analysis based on anchor markers
showed strong co-linearity to previously reported cassava genetic and physical maps.
In silico based searching for conserved domains allowed the annotation of a repertory
of 1,061 cassava genes coding for immunity-related proteins (IRPs). Based on
physical map of the corresponding sequencing scaffolds, unambiguous genetic
localization was possible for 569 of them on the 18 linkage groups. The higher
density of genes coding for IRPs was found on chromosomes 10, 3, 7 and 18.This is
the first study reported so far of an integrated high density genetic map using SNPs
obtained from GBS analysis with integrated genetic and physical localization of newly
annotated immunity related genes in cassava. These data build a solid basis for future
studies to map and associate markers with single loci or quantitative trait loci for
agronomical important traits and molecular cloning of genes controlling these traits.
The enrichment of the physical map with novel scaffolds is in line with the efforts of
the cassava genome sequencing consortium. Considering these improvements, the
size of the genome sequence draft aligned to the genetic map is increased to 344Mb
corresponding to 64% of total cassava genome.
Keywords: linkage mapping, physical mapping, genotyping by sequencing, single
nucleotide polymorphisms, immunity-related genes
Introduction
The advent and progress made in the last two decades of DNA based molecular
markers has contributed to the generation of dense genetic maps (Davey et al., 2011;
Elshire et al., 2011; Takagi et al., 2013). New technologies like next generation
sequencing (NGS) have made possible the high throughput identification and
genotyping of thousands of molecular markers in a relatively short time and
potentially at a low cost (Nielsen et al., 2011). A fast cost-effective approach to next-
129
generation molecular marker discovery called genotyping by sequencing (GBS), has
been proposed to reduce the turnaround time significantly and increases the
availability of thousands of SNP (single nucleotide polymorphism) molecular
markers evenly distributed throughout the genome (Elshire et al., 2011; Poland et al.,
2012).
High-density genetic maps built using SNPs derived from the GBS approach have
been reported in important crop species such as barley (Poland et al., 2012; Liu et al.,
2014), wheat (Poland et al., 2012), rice (Spindel et al., 2013), raspberry (Ward et al.,
2013) and cotton (Gore et al., 2014). In non-model crops, new technologies as GBS
have not been widely used so far. However in cassava, one of the most highly dense
genetic maps was created using GBS-based SNPs, for mapping the resistance to
cassava mosaic geminiviruses (Rabbi et al., 2014b).
Cassava (Manihot esculenta Crantz) belongs to the Euphorbiaceae family, which
includes approximately 6,300 species (Wurdack et al., 2005). Botanically it is a
tropical perennial shrub whose origin center is the Amazon Basin (Olsen and Schaal,
1999). Cassava typically is a diploid species (2n=36) (Raji et al., 2009; Sakurai et al.,
2013) highly heterozygous and vegetative propagation through stakes in agriculture.
Cassava is important for food security in tropical regions of the world. It represents
an important source for calories for more than one billion of people (Ceballos et al.,
2010). The species tolerates drought and has been considered as a well adapted crop
facing climate change which could position it as one of the best alternatives for
providing food for the rapidly growing world population in future (FAO, 2008; Jarvis
et al., 2012; FAO, 2013).
Cassava is cultivated in more than 100 countries and its leaves and roots can be
consumed as food and feed (Taylor et al., 2012). The plant has also important
industrial uses, mainly for its low-cost starch which finds a diverse range of
applications (Ospina et al., 2002; FAO, 2008). For many decades the use of cassava
was limited to subsistence of farmers, but since several years is becoming
increasingly important for agro-processing industries mainly due to its biofuel
potential (Jansson et al., 2009). Despite the fact that cassava is one of the major crops
in the world, a decade ago this crop was listed as one of the least studied plant
species (Okogbenin and Fregene, 2003). The employment of modern molecular tools
will help to go deeper in the understanding of the genetic basis and even lead to the
identification and cloning of genes controlling agro-economic importance traits. Most
of the genes characterized so far in model and cultivated plants have been cloned
employing map based cloning approach (Bent, 1996; Pflieger et al., 2001; Jander et
al., 2002; Gebhardt et al., 2007). The application of this strategy requires the
130
development of high resolution genetic maps (Pflieger et al., 2001; Collard et al.,
2005). The lack of these maps has hampered so far the cloning of interesting genes in
cassava (Fregene et al., 1997; Mba et al., 2001; Okogbenin et al., 2006; Lopez et al.,
2007; Chen et al., 2010; Kunkeaw et al., 2010; Kunkeaw et al., 2011; Sraphet et al.,
2011; Whankaew et al., 2011; Rabbi et al., 2012).
While in genetic maps, markers, genes or loci are ordered based on recombination
frequencies at meiosis (Paterson, 1996), physical maps present ordered fragments of
cloned genomic DNA fragments and whose sizes and distances are given in base pairs
(bp). Genetic maps have considerable relevance for the construction of
comprehensive physical maps. Combining the relative location and order of genetic
markers on a map, with their location on scaffolds or contigs allows the assembly of
these fragments into a genome-wide physical map (Meyers et al., 2004).
The current draft of the cassava genome sequence (draft v4.1) is publicly available at
the JGI’s Phytozome v10 platform and it was obtained by a whole genome shotgun
(WGS) strategy (Green, 2001), using 454 Life Sciences technology. The cassava
genome assembled into 12,977 scaffolds span a total of 532.5 Mb (Prochnik et al.,
2012). However, based on nuclear DNA quantity, it has been estimated that the
cassava genome to be 772 Mb (Awoleye et al., 1994). Strategies based on
correlations between physical and genetic maps could serve as one valuable tool for
subsequent identification of genes involved in interesting traits (Moroldo et al., 2008;
Shulaev et al., 2011), for genome organization studies (Chen et al., 2002), assessment
of genetic diversity (Lu et al., 2011) and comparative genome analysis (Amarillo and
Bass, 2007).
One the main advantages of genetic and physical mapping is the possibility to
integrate traits of interest and the corresponding function of genes (Bakker et al.,
2011; Swamy et al., 2011; Whankaew et al., 2011). The availability of the functional
maps is of importance not only to better understand the evolution of plant species
through synteny but also for marker-assisted breeding programs.
Cassava, like other crops is affected by pests and diseases caused by bacteria, viruses,
fungi, phytoplasms and oomycetes (FAO, 2013). The molecular analysis of plant
pathogen interactions in several model plants and crops has allowed the
identification of two main branches in plant immunity depending on the receptor
molecules involved (Jones and Dangl, 2006). One branch is defined on the presence of
pattern recognition receptors (PRRs) that are able to detect microbe-associated
molecular patterns (MAMPs) (Gohre and Robatzek, 2008). The PRRs have conserved
domains as for example leucine rich repeats (LRR), LysM and kinases (Zipfel, 2014).
131
The MAMP-triggered immunity (MTI) is effective against non-adapted or non-host
pathogens. Some pathogens adapted to infect and colonize particular plants species,
suppressing the plant MTI by delivering effector proteins into the plant cytoplasm
(Buttner and He, 2009). However, plants evolved resistance (R) proteins, which
recognize specifically some of these effectors and trigger the second branch of
immunity named race specific, gene for gene resistance, or effector triggered
immunity (ETI) (Tsuda and Katagiri, 2010). The largest class of R proteins contains
NB-ARC (Nucleotide-binding domain shared by Apaf-1, R gene products, and CED-4)
and LRR domains which can be accompanied by the presence of a TIR
(Toll/interleukin-1 receptor) domain in their N-terminus. (Bent, 1996; Jones and
Jones, 1997; Zhang et al., 2014). Several studies have employed the presence of these
conserved domains to identify R genes in plant genomes to gain insight about their
genome organization and evolution (Jupe et al., 2012; Zhang et al., 2014). The
genome-wide identification of a set of classical defense-encoding sequences and their
localization in a genetic map will provide insights into the diversity of genes coding
for immunity-related proteins (IRPs) available in cassava and also can contribute to
accelerating the process of isolation and cloning of PRR and/or R genes.
In the present study a new genetic map of cassava is constructed based on a
population of 132 F1 full-sib progeny derived from a biparental cross and SNP
markers obtained using the GBS approach. Physical mapping of scaffolds from
cassava whole genome sequencing using the mapped markers as anchors is
presented. Furthermore we present a genome-comprehensive repertoire of cassava
IRPs based on the presence of conserved domains. Finally, more than five hundred of
genes encoding for IRPs were unambiguously localized on the sequencing scaffolds
and on the genetic map.
Materials and methods
Mapping population and DNA extraction
The mapping population consists of a full sib F1 segregating population of 132
individuals derived from single seeds of a cross between cultivars TMS30572 and
CIAT’s elite cultivar CM2177-2 (Fregene et al., 1997). Total genomic DNA was
extracted from young leaf tissue of 132 individuals of the F1 population and their
parents TMS30572 and CM2177-2, using the commercial kit QIAGEN DNeasy Plant
Mini Kit® (Hilden, Germany), following the manufacturer’s protocol and adjusting
132
the final concentration to 100ng/µL. To assess the quality of DNA and absence of
enzymatic inhibitors, a restriction digestion was performed using HindIII and
visualized on a 1% agarose gel.
Genotyping by sequencing (GBS) approach
GBS libraries were prepared and analyzed at the Institute for Genomic Diversity (IGD,
Cornell University, USA), according to Elshire et al. (Elshire et al., 2011). The partial
methylation sensitive ApeKI restriction enzyme that recognizes a five base pair
sequence (GCWGC) was used for digestion and a library was generated with 134
unique barcodes for progeny and parents. Two lanes of Illumina Hi-seq (Illumina,
Inc.) were used for the all samples.
The GBS analysis pipeline 3.0.139 version, an extension to the Java program TASSEL
(Bradbury et al., 2007), was used to call SNPs from the sequenced GBS libraries
(Elshire et al., 2011). The mean sequencing depth was 8 to 10 times. The alignment of
the resulting tags to the reference genome was performed using BWA Version 0.6.2-
r126 (Li and Durbin, 2009), checking that each SNP has a unique position within the
genome scaffolds with 89% of identity. The markers were delivered as Hapmap and
VCF (v0.1.10) (Variant Call Format) format files (Danecek et al., 2011).
Filtering of GBS data
From the complete set of markers an initial filtering was performed using SAS® 9.3
(Inc, 2011) (script, unpublished), to select those SNPs with Mendelian segregation for
1:1 if segregating only in one parent and 1:2:1 if segregating in both parents. Less
than 10% of distorted markers were allowed. Monomorphic homogeneous SNPs and
those with identical segregation were discarded. The segregation in the population,
corresponding to 132 individuals was analyzed for markers that exhibited
polymorphisms between TMS30572 and CM2177-2.
Linkage analysis and map construction
Both linkage analysis and map construction were performed with JoinMap 4.1, and
data were analyzed using the CP (outbreedering full-sib family) population type (Van
Ooijen, 2006). The X2 test was used to assess goodness-of-fit to the expected 1:1 or
1:2:1 segregation ratio for each marker. Linkage groups were established using a
133
grouping LOD (logarithm base 10 of odds) threshold upper than 3. Markers were
assigned to correct linkage groups using two-point grouping analysis and within each
group were mapped based on the strongest cross-link (SCL). The map was generated
using a recombination frequency below 0.50 and the “ripple” procedure was applied.
Recombination frequencies were converted to relative distances in centiMorgans
(cM) using Kosambi function (Kosambi, 1943). The graphical presentation of the
linkage groups was performed using R/qtl (Broman et al., 2003).
Comparative genetic map of cassava
The map developed in this study was compared to the other cassava reported maps.
For that the SNP markers located at the same position on scaffolds were used as
anchors. The genetic positions of these markers were compared and the co-linearity
of the maps was determined. The comparison revealed the number of newly mapped
scaffolds and their size was determined.
Physical mapping
All SNP markers obtained were physically localized in the scaffolds of the cassava
draft genome sequence (www.phytozome.com), based on minimum sequence
similarity of 89%. For that, the core sequence of the marker locus (64bp) was aligned
towards the available genome sequence information to order the position of the
markers on the scaffolds. The scaffolds were anchored and the corresponding
positions along the cassava chromosomes were defined by comparing the positions
of markers on the scaffolds and on the genetic map. The percentage of coverage was
calculated as sequence covered by all mapped scaffolds to the estimated total cassava
genome size. The graphical presentation of the physical map was done by using
Circos algorithm (Krzywinski et al., 2009).
Mapping of immunity-related proteins
The genes taken into account were those encoding for proteins containing any of the
following domains or domain-combination: LRR (Leucine-rich repeat), WRKY, LRR-
kinase, NB-ARC (Nucleotide Binding domain shared by Apaf-1 R gene products, and
CED-4)-LRR, TIR (Toll/interleukin-1 receptor)-NB-ARC-LRR, LysM (Lysin motif)-
134
kinase. All these domains or domain-combination correspond to essential part of the
most studied immunity-related protein encoding genes (van Ooijen et al., 2008;
Swiderski et al., 2009). Models for each domain were downloaded from
http://pfam.sanger.ac.uk (Finn et al., 2011). HMMscan was used with the
downloaded models to search the cassava proteome for proteins containing one or
more of the selected domains, using an e-value cutoff of 10. Proteins containing
several of the domains were identified collapsing the information of the position and
presence/absence of each domain. The genomic coordinates of each protein were
retrieved using BioMart tool from http://www.phytozome.net/cassava.
In order to detect orthologous clusters in Manihot esculenta, Arabidopsis thaliana,
Ricinus communis, and Populus trichocarpa the protein prediction using HMMER
(Finn et al., 2010) was performed. R. communis and P. trichicarpa are chosen as the
closest relatives of cassava and A. thaliana as model organism for which detailed
analysis of IRGs has been reported (Meyers et al., 2003). The Orthologous Cluster
Analysis was done using QuartetS (Yu et al., 2011). Two programs, Single Linkage
Cluster (SLC) and Markov Cluster Algorithm (MCL) were implemented to cluster
genes into orthologous clusters.
Using the obtained catalog of cassava IRPs, the annotated regions containing GBS-
markers were identified, to subsequently locate them on the map according to their
genome-scaffolds positions. IRP clusters were determined using scaffolds and map
positions. The definition of cluster was according to Meyers et al (Meyers et al., 2003)
and Jupe et al (Jupe et al., 2012). A maximum distance between two or more IRPs of
200 kb was allowed and less than eight non-IRPs between them.
Results
Genotyping by sequencing
To identify polymorphisms the parents and the progeny of the mapping population
were genotyped using the GBS approach. On average 2,920,870 reads were generated
for each of 134 samples and 2,173,235 tags were obtained in total. Considering that
the average length of each tag was 64 bp, the total amount of DNA sequence analyzed
was 139 million base pairs. To eliminate possible false positive SNPs, only tags
aligned to unique positions in the cassava reference genome were selected. After the
135
alignment to the cassava genome (Prochnik et al., 2012), 1,185,928 tags (54.6%)
were aligned to unique positions while 229,629 tags (10.6%) were aligned to
multiple positions and the remaining 757,678 tags (34.9 %) could not be aligned.
In total, 78,854 SNP markers were obtained which corresponds, on average, to one
SNP every 1,763 base pairs. They are distributed across 3,450 scaffolds from 12,977
constituting the current cassava genome sequence draft, corresponding to 87%
(463.2 Mb) of the genome. The distribution of tagged scaffolds, the number of SNPs
representing the scaffolds and the cumulative scaffold length in base pair across the
genome is shown in Additional file 1.
From the resulting set of 78,854 SNPs, 51.4% (40,561) of the total set of SNPs
correspond to transitions and 48.6% (38,293) to transversions, for a transition-
transversion ratio of 1.06. A meaningful number of SNPs, 62.6% (49,429), were
located in annotated cassava genome regions. Of these, 52.6% (26,030) were found
within annotated CDS (Coding DNA regions). For non-coding regions, 31.7% (15,708)
were found within introns, 10% (4,940) within promoters and 5.5% (2,751) within
UTRs (Additional file 2).
The gene ontology (GO) analysis was performed for 14,384 unique cassava genome
annotated sequences that contain at least one of the 49,429 annotated SNPs obtained
by GBS. On average, each annotated region contains three SNPs. In total for the three
groups, 2,682 unigenes (counts for gene product characteristics) were obtained
corresponding to the 49,429 annotated SNPs. The functional group with the highest
gene product counts was biological process with 58.2% (1,562 tags) followed by
molecular function 30.2% (811 tags) and cellular component 11.6% (309 tags)
(Additional file 3).
High-density genetic map construction
The obtained 78,854 SNPs were subjected to a series of selective criteria in order to
choose the useful SNPs for the purpose of genetic mapping. From the total set of
markers, 43,921 SNPs (55.6%) correspond to polymorphic markers in the two
parents, from which 25,968 (59.1%) correspond to genotypes derived from a cross
between heterozygous and homozygous parents. Monomorphic homogeneous (both
parents having the same allele) markers as well as those with missing data in more
than 10% of the population individuals were excluded. After the quality control
filters the number of useful and informative loci for mapping was reduced to 7,146.
136
More heterozygous markers were identified in the female parental than in the male.
Of the 7,146 markers, 2,528 (35.4%) were heterozygous only in the male parent
while 2,158 (30.2%) were heterozygous only in the female and 2,460 (34.4%) were
heterozygous for both parents. After the filtering of identical segregation and
distortion for linkage analysis and map construction 5,300 SNPs were taken into
account to be analyzed using Joinmap 4.1. From them, the software integrated,
unambiguously, 2,141 SNP markers onto the newly constructed genetic map. These
were distributed in 18 linkage groups, which corresponds to the number of haploid
cassava chromosomes (2n = 36; n = 18) (Raji et al., 2009; Sakurai et al., 2013). The
numbering was done according to previous studies (see below). The pairwise
recombination fractions and LOD scores obtained using R/qtl indicate strong linkage
for all pairs of markers on each of the 18 LGs (Additional file 4).
The number of SNPs in each linkage group ranged from 35 to 176, with an average of
118.9. The map spanned a total of 2,571 cM, with an average distance of 1.26 cM
between markers (Figure 3-1 and Table 3-1). The LG5 was the largest group, with a
total length of 208.5 cM, while the smallest was LG9, with 36.48 cM. The LG2 and LG8
were the groups with the highest marker density, with an interval of 0.7 cM, whereas
the LG17 was the least saturated group, with an interval of 2 cM. Longer intervals
were present in linkage groups 5, 4 and 14, with values of 20.7, 18 and 16.6 cM
respectively (Table 3-1 and additional file 5).
From the total of 2,141 mapped SNPs, 54.6% correspond to transitions and the
remaining 45.4% to transversions. 76.1% or 1,631 markers are located in annotated
regions, 57.4% (937) are within annotated CDS, 10.4% (170) within promoters, 27%
(442) within introns, and 5% (82) within UTRs regions (Figure 3-4). The total
number of annotated markers in the linkage groups varied from 28 annotated SNPs
for the LG9 to 139 SNPs for the LG1, with an average of 90.61 SNPs. The LG1 has the
highest number of SNPs positioned in CDS regions, followed by LG2.2, while LG9 has
the lowest number. For SNPs positioned within intronic regions, the linkage group
that has the highest number of counts corresponds to LG15, whereas LG9 again has
the lowest number. On the other hand, SNPs positioned in promoter regions, the LG2
shows the highest number of counts while LG9 does not have any. Finally, for SNPs
positioned within UTR regions, the LG2.2, LG10 and LG1 have the highest counts
(Figure 3-2).
137
Figure 3-1. Cassava genetic map containing 2,141 markers. The linkage groups
are named LG1 to LG18. On each linkage group, the black lines represent mapped
markers. Genetic distances are given in Kosambi map units in centi-Morgans and are
calculated using JoinMap 4.1 software (Van Ooijen, 2006).
Table 3-1. Genetic map data summary. The linkage groups, loci number, total
length per group, average distance between markers (density) and scaffolds for each
linkage group are shown.
Linkage
group
No. of
markers
Total
length(cM)
Density
Interval (cM)
Largest
interval (cM)
1 169 199.09 1.19 7.25
2 156 108.24 0.7 5.42
2.2 176 183.51 1.05 5.06
3 80 142.84 1.81 9.85
4 117 129.03 1.11 16.6
5 120 208.47 1.75 18.03
6 106 132.22 1.26 8.62
7 123 171.58 1.41 6.95
8 146 100.64 0.69 9.32
9 35 36.47 1.07 5.27
10 118 151.38 1.29 8.35
11 113 137.08 1.22 10.87
138
13 87 148.08 1.61 5.52
14 137 169.35 1.25 20.7
15 154 149.15 0.97 11.11
16 136 147.8 1.09 7.11
17 63 124.07 2 9.23
18 105 13,217 1.27 6.87
Total 2,141 2,571 1.26
Figure 3-2. Summary of mapped annotated SNPs. Linkage groups and the
corresponding annotated loci numbers. The positions of analyzed SNPs in the gene
structure are shown by different colors. Coding DNA Sequence CDS (), introns,
promoters or UTR (Un-translated Region).
139
Comparative genetic map of cassava
The map constructed here was compared to the previously reported genetic maps
(Rabbi et al., 2014a; Rabbi et al., 2014b). Only for the reported LG12 no homologous
linkage group could be identified. The rest of linkage groups show high co-linearity
when the markers are compared according to the corresponding scaffoldings they
tag. The identities of the scaffolds shared for each LG among the maps was in the
range between 52% (LG4) and 83% (LG13) with an overall average of 66%
throughout all the linkage groups (Figure 3-3 and additional file 6). In total 389
anchor markers between the maps were identified. The LG2.2 and LG14 contain the
highest anchor markers (34), while the LG17 with 8 markers was the lowest. On
average each LG have 21.6 anchor markers (Table 3-2 and Additional file 6). An
additional comparative analysis was done with the cassava map developed by Rabbi
et al. (2014) (Rabbi et al., 2014a). Eight anchor markers distributed in LG1, LG6,
LG14, LG16 and LG19 were identified (Additional file 6).
Figure 3-3. Anchor markers showing co-linearity between different cassava
genetic maps. Markers with the same genomic position (determined by the
corresponding scaffolds) are connected by lines. Comparison was carried out
employing the genetic map reported by Rabbi et al, 2014 (Rabbi et al., 2014b).
140
Figure 3-3. Continue. Anchor markers showing co-linearity between different
cassava genetic maps.
141
Figure 3-3. Continue. Anchor markers showing co-linearity between different
cassava genetic maps.
142
Table 3-2. Comparative analysis of cassava physical maps. Unique scaffolds in
the reported map version (A, Rabbi et al. 2014), in the map from the present study
(B), common scaffolds between them, new mapped scaffolds from this study (B)
anchored, their size in bp and the anchor markers per linkage group .
Linkage
group
Nr. of scaf.
(A)
Nr. of scaf.
(B)
Common
scaffolds
Nr. of new
scaffolds (B)
Size of new
scaffolds (bp)
Anchor
markers
1 65 45 33 8 885,261 23
2 50 43 25 18 3,391,767 17
2.2 58 43 30 11 1,428,130 34
3 56 35 25 10 1,066,798 15
4 56 33 21 10 813,846 17
5 60 48 28 17 4,888,465 21
6 72 42 28 14 1,766,509 21
7 66 34 25 8 3,061,711 23
8 79 53 37 15 2,262,879 23
9 19 10 10 0 0 16
10 54 37 24 11 1,608,816 23
11 41 32 13 16 2,783,942 11
13 60 32 25 6 1,124,354 16
14 91 46 39 6 329,758 34
15 50 36 25 11 1,026,082 31
16 64 46 31 12 1,996,766 25
17 71 28 22 5 890,465 8
18 60 44 32 11 1,396,841 31
total 1,072 687 473 189 30,722,390 389
Physical mapping of scaffolds in the genetic map
To orient the scaffolds of the cassava genome draft sequence into the genetic map,
the mapped markers were employed as anchors. A total of 687 unique scaffolds were
localized on the genetic map, representing 45.6% (242.6Mb) of the current cassava
reference genome. The linkage groups with the highest number of scaffolds were LG8
(53), LG5 (48), LG16 and LG14 with 46 each. LG9 and LG17 have the lowest numbers
of scaffolds with 10 and 28, respectively (Table 2). A total of 46% (316) of the
selected scaffolds were tagged by single-markers, 41% (282) were tagged by 2-5
SNPs and 13% (89) by more than five markers. Scaffold 1551 has the highest count of
markers with 45 SNPs in LG15. Only 3.4% (24) of the scaffolds were present in two
143
different linkage groups (Additional file 5). In this way, the previously reported map
(Rabbi et al., 2014a) could be enriched with 189 new scaffolds which were mapped in
this study. These scaffolds are disturbed on 17 LGs and the number varied between
six for LG13 or LG14 and 18 for LG2. Only for LG9 could not be anchored new
scaffolds. In total, the physical map of cassava is extended with 30.7Mb (Table 3-2),
which correspond to the sum of all new anchored scaffolds.
Figure 3-4. Repertoire of genes coding for immune related proteins (IRPs)
identified in the cassava genome. Numbers on right of bars show the number for
each class of immune related protein. Numbers in parenthesis show the mapped
IRPs. The branches of IRPs are indicated by the color code as shown on the upper
right side.
The relationship between physical and genetic distances in cassava genome was
determined. For that, three representative regions were selected from different areas
of the LG, one from the middle part and one for each of the distal parts. The scaffolds
analyzed contain at least three SNPs. The overall physical map anchored analyzed
comprises 32.1 Mb that corresponds to a genetic distance of 215 cM giving a mean
value of 603.2 kbp per 1 cM. However, this ratio varies strongly between the linkage
144
groups, from 76.8 to 2,429 kbp per 1 cM in LG13 and LG18, respectively. This
variability is calculated also inside of the linkage groups indicating uneven
recombination events. In LG11, 1cM can correspond to 0.1 or to 2,395kbp, whereas in
LG2 it ranges from 288.6 to 1,148kbp (Table 3-3).
Table 3-3. Relationships between genetic and physical maps, representative
for each linkage group and for the whole genome.
Linkage group
Physical
length
analyzed
(kbp)
Genetic
length
analyzed
(cM)
Mean value of
relationship of
genetic (1cM) to
physical (kbp) length
Range of relationship
of genetic (1cM) to
physical (kbp) length
1 2,550 26.4 169.1 95.2 – 269.3
2 2,048 5.3 751.6 288.6- 1,148
2.2 991 9.37 98.1 83.6 - 125
3 1,780 12.4 144.8 21.6 – 234.7
4 1,810 4.91 1,554 3.5 – 5,273.3
5 1,547 16 167.1 42.3 – 245.8
6 796 3.8 288 18.1 – 680.2
7 1,908 18.2 1,062 32.3 – 3,052.8
8 2,516 8.3 323 62.5 – 562.7
9 577 8.1 92.1 28.9 – 208.1
10 2,065 17.1 332 5.2 – 940.6
11 3,13 4.5 913 0.1 – 2,395
13 2,175 22.8 76.8 7.7 – 209.3
14 2,293 7.9 400 72.6 – 570.2
15 4,633 11 1,561 18.9 – 4,634.6
16 1,759 17.8 296 64.2 – 665.4
17 1,289 6.6 201 82.8 – 420.1
18 1,333 14.1 2,429 8.3 – 699.8
Genome-wide 32,1Mb 215 603.2 0.1 - 5,273
Repertoire of Immunity-related proteins
Employing a bioinformatics approach, the cassava proteome was investigated for
proteins containing the conserved domains present in PRRs and R proteins. A
repertoire of proteins with a complex pattern of combinations of these conserved
145
domains was obtained (Figure 3-4). In total 1,061 IRPs were identified (Additional
file 7). From them, 253 were classified as LRR-kinases based on the presence of
leucine-rich-repeat and kinase specific domains. These proteins, also known as
receptor-like kinases (RLKs), which contain an extracellular LRR and a cytoplasmatic
kinase domain are involved in MTI pathways. Seventeen putative proteins containing
only the LysM domain and eleven proteins containing both the LysM and kinase
domains were detected (Figure 3-4).
The cassava proteome contains 28 TIR-NB-ARC-LRR, 177 non-TIR-NB-ARC-LRR
putative proteins, and two with TIR-LRR domains. Proteins containing only the NB-
ARC domain or only the TIR domain were relatively well represented, with 29 and 14,
respectively. Proteins with an extracellular LRR domain are also known as receptor
like proteins (RLPs) can participate as immune receptors, while other RLPs
participate in plant development. The cassava proteome contains 425 of these RLPs
proteins. Although the WRKY domain separately is not present in any known R
protein, it is present in an important family of plant transcription factors related with
defense against pathogens. The cassava proteome has 105 WRKY proteins and none
of them contains additional conserved domains (Figure 3-4 and Additional file 7).
Genomic organization of immunity related annotated genes
In total, 554 scaffolds containing genes coding for IRPs were identified. Most of the
genes, 713 (67%) were localized in scaffolds containing two or more IRPs. However
349 genes (33%) were localized in scaffolds as single genes. The scaffolds containing
the highest number of annotated genes encoding for IRPs were 8265 with 13 (5 LRR,
4 LRR-kinase, 3 NB-ARC-LRR and 1 WRKY) and 05875 with 12 (4 LRR, 4 LRR-kinase,
2 NB-ARC-LRR, 1 LysM-kinase and 1 NB-ARC). Scaffold 8686 contains 11 genes all
from the LRR class. Three scaffolds contained ten genes: 6914 (4 NB-ARC-LRR, 3 LRR,
2 LRR-Kinase, 1 WRKY), 7520 (5 LRR, 3 LRR-kinase, 1 NB-ARC-LRR, and 1 WRKY)
and 10217 (6 NB-ARC-LRR and 4 LRR). Interestingly, from the 28 annotated genes
coding for putative TIR-NB-ARC-LRR proteins, 10 were grouped into only two
scaffolds, one containing six genes (scaffold 97) and the other one (scaffold 11897)
containing four of these genes. The six genes in scaffold 97 are located in a region of
just 77,359 bp, whereas the four genes in scaffold 11897 cover 116,966 bp. Scaffolds
3,921 and 11,106 also harbor a relatively high number of genes of the NB-ARC-LRR
class, with six genes each. The scaffolds containing genes coding for proteins with a
WRKY domain harbor only one or two of this class of genes and only a few have three
(Additional file 7).
146
The annotation of the immunity genes in the cassava genome was performed with an
Ortholog Cluster Analysis (sequence homology) (Figure 3-5). Arabidopsis thaliana,
Ricinus communis, and Populus trichocarpa were selected as related species and the
same pipeline employed to identify conserved domains in cassava was applied for
these species. From the 425 putative proteins of cassava classified as LRR proteins by
HMMscan, 189 have orthologs with LRR proteins from at least one of the other
species analyzed (Figure 3-5A). A cluster with 57 LRR family proteins was shared by
all the three species. Cassava shares 40 orthologous LRR proteins with P. trichocarpa,
26 with R. communis, and eight with A. thaliana (Figure 3-5A). The second biggest
group was the LRR-kinase family. Of the 253 proteins LRR-kinase proteins predicted
in cassava, 168 had an orthologous at least in one of the other plant species analyzed.
There were 68 orthologs of LRR-kinases shared by all species (Figure 3-5B). Of the
105 WRKY proteins from cassava, 66 have an ortholog in at least one of the other
plant species analyzed and 23 are in a cluster in all species (Figure 3-5C). In the case
of the NB-ARC family, all the 29 cassava predicted proteins had an ortholog in at least
one other plant species evaluated and one protein is shared by all of the species
(Figure 3-5D). Of the 177 proteins predicted in the non-TIR-NB-ARC-LRR family, 55
cassava proteins had an ortholog in at least one other analyzed species and six
proteins had orthologs in all the studied species (Figure 3-5E). Finally, less than 15
orthologs are found among the analyzed species for the predicted ORFs of each of the
following classes: LysM, LysM-kinase, TIR and TIR-NB-ARC-LRR (Figure 3-5F-I).
Mapping of immunity related proteins
Based on the cassava IRP repertoire (1,061 in total), those located on scaffolds
oriented in the physical map were selected. In total, 569 IRPs were mapped, 198 of
them (34.7%) belonging to LRR class, 1609 (28.1%) to the LRR-kinase, 88 (15.4%) to
NB-ARC-LRR, 80 (14%) to WRKY, 8 (1.4%) to NB-ARC, 13 (2.3%) to TIR-NB-ARC-
LRR, 8 (1.4%) to LysM, 6 (1.1%) to TIR and 9 (1.6%) to LysM-kinase (Figure 3-4,
Additional file 7).
147
Figure 3-5. Orthology clusters between of the predicted immunity-related
proteins in Manihot esculenta, Arabidopsis thaliana, Ricinus communis, Populus
trichocarpa. A. LRR. B. LRR-kinase. C. WRKY. D. NB-ARC. E. NB-ARC-LRR. F. LysM. G.
LysM-kinase. H. TIR. I. TIR-NB-ARC-LRR.
These 569 genes coding for IRPs were physically located in 226 scaffolds and
distributed in all the 18 linkage groups with an average of 31.6 per linkage group.
LG2.2, LG7 and LG8 had the highest counts with 45, 45 and 40 genes, respectively.
The linkage groups with the lowest counts were LG17 and LG9 with 16 genes each
(Additional file 7). In total, 128 clusters were identified, with 382 genes, counting for
almost 67% of the total mapped IRPs. Clusters were found in all 18 linkage groups.
The cluster with highest number had 11 IRPs (LRR) and was located in LG10,
followed by LG3, LG7 and LG18 with clusters of 9 IRPs each. Seventy clusters, on 17
LGs, except for LG13, have two IRPs each. These clusters had diverse combinations of
IRP classes (Figure 3-6).
148
Anchoring previous QTLs for disease resistance
We searched to localize loci or QTLs previously reported in our genetic or physical
map. The markers SSRY28 (CMD2), S5214_78931 and S5214_30911 have been
genetically associated with CMD resistance (Akano et al., 2002; Lokko et al., 2005;
Okogbenin et al., 2012; Rabbi et al., 2014a; Rabbi et al., 2014b). These markers were
anchored in the scaffold 5214 in LG16 (Figure 3-6 and Additional file 7) at the same
position as reported by Rabbi et al. 2014 (Rabbi et al., 2014b). In this study it was
possible to anchor the markers SSRNS158 and SSRNS169 previously associated with
CMD resistance (Okogbenin et al., 2007) in the scaffold 6906, while in the scaffolds
4,175 and 7,933, localized in the LG16, were anchored the markers SSRNS198 and
SSRY106 where a QTL for CMD resistance have been reported (Lokko et al., 2005;
Okogbenin et al., 2012). Interestingly, from these scaffolds, the 5,214 and 4,175, one
(from the LysM family) and five genes (two LRR-kinase, two NB-ARC-LRR and one
LysM) coding for IRPs are present (Additional file 7). A fine mapping and/or
association studies will allow if these candidate genes are directly related to CMD
resistance.
149
Figure 3-6. The cassava genetic and physical map enriched with IRPs and QTLs
for cassava disease resistance. The linkage groups are highlighted with different
colors and the markers in blue lines. In the inner part the black curves mark the
anchored scaffolds, their number and cumulative length in Mb per linkage group,
orientation based on map positions of markers. In red are shown the IRPs families,
their number per linkage group is shown in parenthesis. In green the reported loci
and QTLs for cassava mosaic virus resistance. The grey lines mark the link between
genetic and physical scaffold positions of marker clusters in the same scaffold.
Diagram was plotted using Circos software(Krzywinski et al., 2009).
150
Discussion
In this work a GBS approach was carried out to identify SNP derived markers in a
cassava population for genetic and physical mapping purposes. The 78,854 GBS-SNPs
obtained cover 87% (463.2 Mb) of the current cassava genome sequence. These
markers were distributed homogenously through 3,450 scaffolds of the genome
sequence draft. These scaffolds cover the majority of the cassava genome, although
they represent 16.5% of the total number of genome scaffolds. This due to just 487 of
almost 13,000 scaffolds covers half of the current cassava genome (Prochnik et al.,
2012). No SNPs were identified in small scaffolds representing the remaining 13% of
the cassava genome. Consequently, these data constitute the most representative
genotyping information for a cassava population until now, and can be relevant for
future applications where DNA fingerprint is pivotal.
The transition-transvertion ratio of the total of SNPs was 1.06. This figure is lower
when compared to previous cassava reports on genome-wide polymorphic discovery
(1.24) (Sakurai et al., 2013) and expressed sequence tags (EST) (1.27) (Ferguson et
al., 2012). Surprisingly, more than 60% of the SNP markers obtained were located
within annotated and coding regions. The enzyme ApeKI used for preparation of GBS
libraries is partially methylation sensitive (Elshire et al., 2011), and this leads to the
preferential restriction of coding sequences. Similar results were obtained in cattle
using the enzyme PstI, also a methylation sensitive enzyme (De Donato et al., 2013).
SNPs located more often in cassava CDS than in UTRs, which has also been reported
in a previous study based on genome-wide analysis (Sakurai et al., 2013). Those SNPs
located within a CDS can potentially modify the encoding amino acid chain, resulting
in proteins with new functions or introduction of a stop codon. These represent an
outstanding source of information to validate the function of genes (Wilson et al.,
2004; Kumar et al., 2014) and constitute a direct and effective way to conduct
phenotype association analysis.
On the other hand, the SNPs positioned in non-coding regions such as introns might
also play key roles in processes of alternative splicing and can be employed in
evolution and diversity studies (Yamanaka et al., 2004). Those SNPs residing in UTR
regions or promoters represent control points to regulate gene transcription and
translation. Interestingly, some non-coding regions have been reported as key in
151
regulating and controlling the expression of genes responsible for agronomical
important traits such as flowering time in maize (Salvi et al., 2007; Studer et al.,
2011) and loss of seed shattering in rice (Konishi et al., 2006). Therefore, in this
version on the cassava genetic map the description and putative function for the
sequences containing SNPs was not limited to coding regions, but to all annotated
sequences containing a marker.
The cassava population used in this study is derived from a cross between highly
contrasting parents for several phenotypic and phenological traits (Okogbenin and
Fregene, 2003; Okogbenin et al., 2008). This cross has been employed so far to
identify genomic regions involved in morphological traits (Fregene et al., 1997;
Okogbenin and Fregene, 2003) resistance to CMD (Cassava Mosaic Disease) (Akano
et al., 2002) and Cassava Bacterial Blight (Jorge et al., 2000; Jorge et al., 2001) . The
highly dense genetic map reported here could contribute to future research focused
on studies of allelic variation and the effect on different traits, as well QTL analysis
and marker-assisted breeding programs.
The linkage map we have constructed is the second most saturated map on cassava
reported so far (Rabbi et al., 2014b). However, although these two maps employed
GBS derived markers and the same restriction enzyme for library construction, the
total number of SNPs obtained was different. This could be due to library
preparation, technical issues, pipeline used for the SNP calling (Sonah et al., 2013),
the quality, quantity and concentration of the DNA sample, but also because of the
level of genetic diversity between the parents.
The map contained 2,141 SNP markers, distributed in homogenous manner in 18
linkage groups, with a density of 1.26 cM. Some regions of this map are sparsely
saturated, as has previously been reported for other species using SNPs obtained
from GBS (Ward et al., 2013; Liu et al., 2014; Rabbi et al., 2014b). This fact could be
explained by the scarcity or even lack of polymorphisms in these regions. However,
more than 93% of the map shows a high saturation and reduced interval lower than
3cM. It will be very useful establishing close relationships between markers and QTLs
(Falconer and Mackay, 1996; Davey et al., 2011), facilitating the subsequent
identification of genes involved in interesting traits.
Almost half (264.4Mb) of the current cassava genome draft sequence could be
anchored to the genetic map through 687 scaffolds. Comparative map analysis with
the reported cassava maps (Rabbi et al., 2014b) revealed high correlations between
linkage groups based on anchor markers. Moreover, the physical map of cassava was
extended with 30.7Mb by anchoring 189 new scaffolds. This will contribute to the
152
efforts of improve the current cassava genome sequence draft. It is expected that
SNPs belonging to the same scaffolds to be in clusters on the same linkage groups.
Nevertheless, cluster of markers from the same scaffold are disrupted by some
markers from other scaffolds. For instance in LG15, scaffold 1,551 was disrupted by
scaffold 3,241; in LG2.2, scaffold 2,895 was disrupted by scaffold 4,060. Similar
scenarios have also been reported (Sraphet et al., 2011; Whankaew et al., 2011; Rabbi
et al., 2012). On the other hand, it was found that 24 scaffolds are located at two
locations belonging to different linkage groups as already reported by Sraphet et al.
(Sraphet et al., 2011). The scaffolds 8,265 and 4,165 seem to harbor duplications,
because these two scaffolds are located in more than one LG in the cassava maps
(Sraphet et al., 2011; Rabbi et al., 2014b). Scaffold 8,254 is located in LG2.2, LG4 and
LG16 in the map constructed in this study as well as in that reported by Rabbi et al,
2014 (Rabbi et al., 2014a). Scaffold 4,165 is located in LG4 and LG9 in our study but
only in LG9 in Rabbi et al, 2014 (Rabbi et al., 2014a). It is common to assume that the
genomes of plants of the same species are similar, however, there is increasing
evidence for rearrangements, translocations, gains or losses of DNA segments and
copy number variations (CNV) usually found in all chromosomes among the genomes
of different genotypes of the same species (Swanson-Wagner et al., 2010; Zmienko et
al., 2014). This might be the case between the genotypes used for the draft genome
sequence and the parents used in this study and might explain the differences
observed between the genetic and physical map found. Undoubtedly, a consensus
genetic map for cassava could be helpful in this regard, as has been performed for
other species with high heterozygosity level such as grapevine and apple (Velasco et
al., 2007; Clark et al., 2014). Other explanations might be that some of the markers
identifying these scaffolds are not properly mapped or because of errors during
assembly of the reads, that are still present in the draft genome sequence.
The relationship between physical and genetic distances found is the range of
reported data for other plant species. The value of 603 kbp for 1 cM determined in
this study for cassava varies between 139 kbp in Arabidopsis to 510 in tomato or
2140 in maize
(http://www.ndsu.edu/pubweb/~mcclean/plsc731/analysis/analysis5.htm). This
information is useful when detailed genome structure analysis or gene cloning by
map-based cloning approaches will be undertaken in the future.
A high number of SNP-tagged genes were classified in different GO categories,
showing a wide variety of functions in the annotated regions containing markers.
This represents a meaningful source of genes/markers, which can be employed to
answer important biological questions and set up of further experiments to confirm
gene functions and links with phenotypes. GO analysis is a basis for construction of
153
functional maps for a particular group of genes of one of the functional categories,
such as responses to abiotic or biotic stress. Moreover, it allows the quick mapping of
gene families or even gene pathways for interesting traits.
Based on the presence of conserved domains in the PRR and R proteins, it was
possible to identify a large IRP repertoire in the cassava genome. In total 1,061 IRPs
were identified, although probably not all of them are involved in plant immunity.
The next challenge will be to identify the MAMP or effectors that are recognized by
these predicted proteins. The numbers of IRPs varies enormously between plant
species. For example, the quantity of NB-ARC-LRR, the largest class of R proteins,
ranges from 92 in Brassica rapa (Mun et al., 2009) and 150 in Arabidopsis thaliana
(Meyers et al., 2003) to 438 in potato (Jupe et al., 2012). The reasons for the number
variation of IRPs between different plant species have not been explained so far.
In other plant genomes, more than 40% of genes encoding for IRPs are clustered and
the cluster size can be highly variable (Meyers et al., 2003; Mun et al., 2009; Jupe et
al., 2012). In cassava we found a range from two to eleven members per cluster
whereas in Arabidopsis was from two to seven (Meyers et al., 2003), or two to
eighteen in potato (Jupe et al., 2012). As the physical map reported here represents
45.6% of the current cassava genome, it is expected that more IRPs and clusters of
them lie in the remaining genome regions that could not be analyzed. The 1,061 IRPs
were analyzing 532 Mb sequence information. This information will be important to
infer the evolutionary history of these important genes and better understand how
their genome organization has influenced on their structure dynamics and adaptation
to pathogen-derived selective forces.
In addition, in this study it was possible to anchor some markers with scaffolds
present in the LG16 with a region containing loci associated with CMD reported
previously. This example has shown the utility of how dense genetic and physical
map information in addition of phenotypic is an excellent way to accelerate the
cloning of agronomic interest trait genes or to develop markers useful in marker
assisted selection programs. With more phenotypic and QTL analysis the association
between the markers identified in this study and traits will increase.
Acknowledgments
We thank COLCIENCIAS for the financial support through grand 528-2011 and PhD
scholarship call 528. We would like to extend our gratitude to Alvaro Perez and Dr.
Teresa Mosquera from Universidad Nacional de Colombia, for their scientific support
154
and advices. Also to Wiebke Sannemann from INRES-Plant Breeding, Bonn
University, for her support with Circos software. Finally, to the Institute for Genomic
Diversity, Cornell University core facilities who conducted Illumina sequencing of the
GBS libraries.
References
Akano, O., Dixon, O., Mba, C., Barrera, E., and Fregene, M. 2002. Genetic mapping of a dominant gene conferring resistance to cassava mosaic disease. TAG. Theoretical and applied genetics 105:521-525.
Amarillo, F.I., and Bass, H.W. 2007. A transgenomic cytogenetic sorghum (Sorghum propinquum) bacterial artificial chromosome fluorescence in situ hybridization map of maize (Zea mays L.) pachytene chromosome 9, evidence for regions of genome hyperexpansion. Genetics 177:1509-1526.
Awoleye, F., Duren, M., Dolezel, J., and Novak, F.J. 1994. Nuclear DNA content and in vitro induced somatic polyploidization cassava (Manihot esculenta Crantz) breeding. Euphytica 76:195-202.
Bakker, E., Borm, T., Prins, P., Vossen, E., Uenk, G., Arens, M., Boer, J., Eck, H., Muskens, M., Vossen, J., Linden, G., Ham, R., Klein-Lankhorst, R., Visser, R., Smant, G., Bakker, J., and Goverse, A. 2011. A genome-wide genetic map of NB-LRR disease resistance loci in potato. Theoretical and Applied Genetics 123:493-508.
Bent, A.F. 1996. Plant Disease Resistance Genes: Function Meets Structure. Plant Cell 8:1757-1771.
Bradbury, P.J., Zhang, Z., Kroon, D.E., Casstevens, T.M., Ramdoss, Y., and Buckler, E.S. 2007. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics (Oxford, England) 23:2633-2635.
Broman, K.W., Wu, H., Sen, S., and Churchill, G.A. 2003. R/qtl: QTL mapping in experimental crosses. Bioinformatics (Oxford, England) 19:889-890.
Buttner, D., and He, S.Y. 2009. Type III Protein Secretion in Plant Pathogenic Bacteria. Plant physiology 150:1656-1664.
Ceballos, H., Okogbenin, E., Pérez, J.C., Becerra López-Lavalle, L.A., and Debouck, D. 2010. Cassava. Pages 53-96 in: Root and Tuber Crops, J.E. Bradshaw, ed. Springer New York.
Clark, M., Schmitz, C., Rosyara, U., Luby, J., and Bradeen, J. 2014. A consensus ‘Honeycrisp’ apple (Malus × domestica) genetic linkage map from three full-sib progeny populations. Tree Genetics & Genomes 10:627-639.
155
Collard, B., Jahufer, M., Brouwer, J., and Pang, E. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142:169-196.
Chen, M., Presting, G., Barbazuk, W.B., Goicoechea, J.L., Blackmon, B., Fang, G., Kim, H., Frisch, D., Yu, Y., Sun, S., Higingbottom, S., Phimphilai, J., Phimphilai, D., Thurmond, S., Gaudette, B., Li, P., Liu, J., Hatfield, J., Main, D., Farrar, K., Henderson, C., Barnett, L., Costa, R., Williams, B., Walser, S., Atkins, M., Hall, C., Budiman, M.A., Tomkins, J.P., Luo, M., Bancroft, I., Salse, J., Regad, F., Mohapatra, T., Singh, N.K., Tyagi, A.K., Soderlund, C., Dean, R.A., and Wing, R.A. 2002. An integrated physical and genetic map of the rice genome. Plant Cell 14:537-545.
Chen, X., Xia, Z., Fu, Y., Lu, C., and Wang, W. 2010. Constructing a genetic linkage map using an F1 population of non-inbred parents in cassava (Manihot esculenta Crantz). Plant Molecular Biology Reporter:1-8.
Danecek, P., Auton, A., Abecasis, G., Albers, C.A., Banks, E., DePristo, M.A., Handsaker, R.E., Lunter, G., Marth, G.T., Sherry, S.T., McVean, G., Durbin, R., and Group, G.P.A. 2011. The variant call format and VCFtools. Bioinformatics (Oxford, England) 27:2156-2158.
Davey, J.W., Hohenlohe, P.A., Etter, P.D., Boone, J.Q., Catchen, J.M., and Blaxter, M.L. 2011. Genome-wide genetic marker discovery and genotyping using next-generation sequencing. Nature reviews 12:499-510.
De Donato, M., Peters, S.O., Mitchell, S.E., Hussain, T., and Imumorin, I.G. 2013. Genotyping-by-Sequencing (GBS): A Novel, Efficient and Cost-Effective Genotyping Method for Cattle Using Next-Generation Sequencing. PLoS ONE 8:e62137.
Elshire, R.J., Glaubitz, J.C., Sun, Q., Poland, J.A., Kawamoto, K., Buckler, E.S., and Mitchell, S.E. 2011. A Robust, Simple Genotyping-by-Sequencing (GBS) Approach for High Diversity Species. PLoS ONE 6:e19379.
Falconer, D.S., and Mackay, T.F.C. 1996. Introduction to Quantitative Genetics Prentice Hall, London, UK.
FAO. 2008. Oficina de prensa. Yuca para la seguridad alimentaria y energética (Rome).
FAO. 2013. Save and grow: Cassava. A guide to sustainable production intensification. Food and Agriculture Organization of the United Nations, Rome.
Ferguson, M.E., Hearne, S.J., Close, T.J., Wanamaker, S., Moskal, W.A., Town, C.D., de Young, J., Marri, P.R., Rabbi, I.Y., and de Villiers, E.P. 2012. Identification, validation and high-throughput genotyping of transcribed gene SNPs in cassava. TAG. Theoretical and applied genetics 124:685-695.
Finn, R.D., Clements, J., and Eddy, S.R. 2011. HMMER web server: interactive sequence similarity searching. Nucleic acids research 39:W29-37.
Finn, R.D., Mistry, J., Tate, J., Coggill, P., Heger, A., Pollington, J.E., Gavin, O.L., Gunasekaran, P., Ceric, G., Forslund, K., Holm, L., Sonnhammer, E.L.L., Eddy, S.R., and
156
Bateman, A. 2010. The Pfam protein families database. Nucleic acids research 38:D211-D222.
Fregene, M., Angel, F., Gomez, R., Rodriguez, F., Chavarriaga, P., Roca, W., Tohme, J., and Bonierbale, M. 1997. A molecular genetic map of cassava (Manihot esculenta Crantz). TAG Theoretical and Applied Genetics 95:431-441.
Gebhardt, C., Li, L., Pajerowska-Mukthar, K., Achenbach, U., Sattarzadeh, A., Bormann, C., Ilarionova, E., and Ballvora, A. 2007. Candidate Gene Approach to Identify Genes Underlying Quantitative Traits and Develop Diagnostic Markers in Potato. Crop Science 47.
Gohre, V., and Robatzek, S. 2008. Breaking the Barriers: Microbial Effector Molecules Subvert Plant Immunity. Annual review of phytopathology 46.
Gore, M.A., Fang, D.D., Poland, J.A., Zhang, J., Percy, R.G., Cantrell, R.G., Thyssen, G., and Lipka, A.E. 2014. Linkage Map Construction and Quantitative Trait Locus Analysis of Agronomic and Fiber Quality Traits in Cotton. Plant Gen. 7.
Green, E.D. 2001. Strategies for the systematic sequencing of complex genomes. Nature reviews 2:573-583.
Inc, S.A.S.I. (2011). SAS/STAT Software.
Jander, G., Norris, S.R., Rounsley, S.D., Bush, D.F., Levin, I.M., and Last, R.L. 2002. Arabidopsis map-based cloning in the post-genome era. Plant physiology 129:440-450.
Jansson, C., Westerbergh, A., Zhang, J., Hu, X., and Sun, C. 2009. Cassava, a potential biofuel crop in China. Applied Energy 86:95-99.
Jarvis, A., Ramirez-Villegas, J., Campo, B.V.H., and Navarro-Racines, C. 2012. Is cassava the answer to African climate change adaptation? Tropical Plant Biology 5:9-29.
Jones, D.A., and Jones, J.D.G. 1997. The Role of Leucine-Rich Repeat Proteins in Plant Defences. Pages 89-167 in: Advances in Botanical Research, I.C.T. J.H. Andrews and J.A. Callow, eds. Academic Press.
Jones, J.D., and Dangl, J.L. 2006. The plant immune system. Nature 444:323-329.
Jorge, V., Fregene, M.A., Duque, M.C., Bonierbale, M.W., Tohme, J., and Verdier, V. 2000. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). TAG Theoretical and Applied Genetics 101:865-872.
Jorge, V., Fregene, M., Vélez, C.M., Duque, M.C., Tohme, J., and Verdier, V. 2001. QTL analysis of field resistance to Xanthomonas axonopodis pv. manihotis in cassava. Theoretical and Applied Genetics 102:564-571.
Jupe, F., Pritchard, L., Etherington, G., MacKenzie, K., Cock, P., Wright, F., Sharma, S.K., Bolser, D., Bryan, G., Jones, J., and Hein, I. 2012. Identification and localisation of the NB-LRR gene family within the potato genome. BMC genomics 13:75.
157
Konishi, S., Izawa, T., Lin, S.Y., Ebana, K., Fukuta, Y., Sasaki, T., and Yano, M. 2006. An SNP caused loss of seed shattering during rice domestication. Science (New York, N.Y 312:1392-1396.
Kosambi, D.D. 1943. The estimation of map distances from recombination values. Annals of Eugenics 12:172-175.
Krzywinski, M., Schein, J., Birol, I., Connors, J., Gascoyne, R., Horsman, D., Jones, S.J., and Marra, M.A. 2009. Circos: an information aesthetic for comparative genomics. Genome research 19:1639-1645.
Kumar, B., Abdel-Ghani, A.H., Pace, J., Reyes-Matamoros, J., Hochholdinger, F., and Lübberstedt, T. 2014. Association analysis of single nucleotide polymorphisms in candidate genes with root traits in maize (Zea mays L.) seedlings. Plant Science 224:9-19.
Kunkeaw, S., Tangphatsornruang, S., Smith, D.R., and Triwitayakorn, K. 2010. Genetic linkage map of cassava (Manihot esculenta Crantz) based on AFLP and SSR markers. Plant Breeding 129:112-115.
Kunkeaw, S., Yoocha, T., Sraphet, S., Boonchanawiwat, A., Boonseng, O., Lightfoot, D., Triwitayakorn, K., and Tangphatsornruang, S. 2011. Construction of a genetic linkage map using simple sequence repeat markers from expressed sequence tags for cassava (Manihot esculenta Crantz). Molecular Breeding 27:67-75.
Li, H., and Durbin, R. 2009. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics (Oxford, England) 25:1754-1760.
Liu, H., Bayer, M., Druka, A., Russell, J., Hackett, C., Poland, J., Ramsay, L., Hedley, P., and Waugh, R. 2014. An evaluation of genotyping by sequencing (GBS) to map the Breviaristatum-e (ari-e) locus in cultivated barley. BMC genomics 15:1-11.
Lokko, Y., Danquah, E., Offei, S., Dixon, A., and Gedil, M. 2005. Molecular markers associated with a new source of resistance to the cassava mosaic disease. African Journal of Biotechnology 4.
Lopez, C.E., Quesada-Ocampo, L.M., Bohorquez, A., Duque, M.C., Vargas, J., Tohme, J., and Verdier, V. 2007. Mapping EST-derived SSRs and ESTs involved in resistance to bacterial blight in Manihot esculenta. Genome / National Research Council Canada = Genome / Conseil national de recherches Canada 50:1078-1088.
Lu, Y., Shah, T., Hao, Z., Taba, S., Zhang, S., Gao, S., Liu, J., Cao, M., Wang, J., Prakash, A.B., Rong, T., and Xu, Y. 2011. Comparative SNP and Haplotype Analysis Reveals a Higher Genetic Diversity and Rapider LD Decay in Tropical than Temperate Germplasm in Maize. PLoS ONE 6:e24861.
Mba, R.E.C., Stephenson, P., Edwards, K., Melzer, S., Nkumbira, J., Gullberg, U., Apel, K., Gale, M., Tohme, J., and Fregene, M. 2001. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: towards an SSR-based molecular genetic map of cassava. TAG Theoretical and Applied Genetics 102:21-31.
158
Meyers, B.C., Scalabrin, S., and Morgante, M. 2004. Mapping and sequencing complex genomes: let's get physical! Nature reviews 5:578-588.
Meyers, B.C., Kozik, A., Griego, A., Kuang, H., and Michelmore, R.W. 2003. Genome-wide analysis of NBS-LRR-encoding genes in Arabidopsis. Plant Cell 15:809-834.
Moroldo, M., Paillard, S., Marconi, R., Fabrice, L., Canaguier, A., Cruaud, C., De Berardinis, V., Guichard, C., Brunaud, V., Le Clainche, I., Scalabrin, S., Testolin, R., Di Gaspero, G., Morgante, M., and Adam-Blondon, A.-F. 2008. A physical map of the heterozygous grapevine 'Cabernet Sauvignon' allows mapping candidate genes for disease resistance. BMC plant biology 8:66.
Mun, J.H., Yu, H.J., Park, S., and Park, B.S. 2009. Genome-wide identification of NBS-encoding resistance genes in Brassica rapa. Mol Genet Genomics 282:617-631.
Nielsen, R., Paul, J.S., Albrechtsen, A., and Song, Y.S. 2011. Genotype and SNP calling from next-generation sequencing data. Nature reviews 12:443-451.
Okogbenin, E., and Fregene, M. 2003. Genetic mapping of QTLs affecting productivity and plant architecture in a full-sib cross from non-inbred parents in Cassava (Manihot esculenta Crantz). TAG. Theoretical and applied genetics 107:1452-1462.
Okogbenin, E., Marin, J., and Fregene, M. 2006. An SSR-based molecular genetic map of cassava. Euphytica 147:433-440.
Okogbenin, E., Marin, J., and Fregene, M. 2008. QTL analysis for early yield in a pseudo F2 population of cassava. African Journal of Biotechnology 7:131-138.
Okogbenin, E., Egesi, C., Mba, C., Espinosa, E., Santos, L.G., Ospina, C., Marín, J., Barrera, E., Gutiérrez, J., Ekanayake, I., Iglesias, C., Fregene, M.A., and Porto, M.C.M. 2007. Marker-assisted introgression of resistance to cassava mosaic disease into latin american germplasm for the genetic improvement of cassava in Africa. Crop Science 47:1895-1904.
Okogbenin, E., Egesi, C.N., Olasanmi, B., Ogundapo, O., Kahya, S., Hurtado, P., Marin, J., Akinbo, O., Mba, C., Gomez, H., de Vicente, C., Baiyeri, S., Uguru, M., Ewa, F., and Fregene, M. 2012. Molecular Marker Analysis and Validation of Resistance to Cassava Mosaic Disease in Elite Cassava Genotypes in Nigeria. Crop Sci. 52:2576-2586.
Olsen, K.M., and Schaal, B.A. 1999. Evidence on the origin of cassava: phylogeography of Manihot esculenta. Proceedings of the National Academy of Sciences of the United States of America 96:5586-5591.
Ospina, P.B., Ceballos, H., Alvarez, E., Bellotti, A.C., Calvert, L.A., Arias V, B., Cadavid, L.F., Pineda L, B., Llano, G.A., and Cuervo, M.I. 2002. La yuca en el Tercer Milenio. Sistemas modernos de producción, procesamiento, utilización y comercialización.
Paterson, A.H. 1996. Genome Mapping in Plants. Elsevier Science Publishing Co Inc, San Diego.
Pflieger, S., Lefebvre, V., and Causse, M. 2001. The candidate gene approach in plant genetics: a review. Molecular Breeding 7:275-291.
159
Poland, J.A., Brown, P.J., Sorrells, M.E., and Jannink, J.L. 2012. Development of high-density genetic maps for barley and wheat using a novel two-enzyme genotyping-by-sequencing approach. PLoS ONE 7:28.
Prochnik, S., Marri, P., Desany, B., Rabinowicz, P., Kodira, C., Mohiuddin, M., Rodriguez, F., Fauquet, C., Tohme, J., Harkins, T., Rokhsar, D., and Rounsley, S. 2012. The Cassava Genome: Current Progress, Future Directions. Tropical Plant Biology 5:88-94.
Rabbi, I., Hamblin, M., Gedil, M., Kulakow, P., Ferguson, M., Ikpan, A.S., Ly, D., and Jannink, J.-L. 2014a. Genetic Mapping Using Genotyping-by-Sequencing in the Clonally Propagated Cassava. Crop Sci. 54:1384-1396.
Rabbi, I.Y., Kulembeka, H.P., Masumba, E., Marri, P.R., and Ferguson, M. 2012. An EST-derived SNP and SSR genetic linkage map of cassava (Manihot esculenta Crantz). Theoretical and Applied Genetics 125:329-342.
Rabbi, I.Y., Hamblin, M.T., Kumar, P.L., Gedil, M.A., Ikpan, A.S., Jannink, J.L., and Kulakow, P.A. 2014b. High-resolution mapping of resistance to cassava mosaic geminiviruses in cassava using genotyping-by-sequencing and its implications for breeding. Virus Res 186:87-96.
Raji, A.A., Anderson, J.V., Kolade, O.A., Ugwu, C.D., Dixon, A.G., and Ingelbrecht, I.L. 2009. Gene-based microsatellites for cassava (Manihot esculenta Crantz): prevalence, polymorphisms, and cross-taxa utility. BMC plant biology 9:118.
Sakurai, T., Mochida, K., Yoshida, T., Akiyama, K., Ishitani, M., Seki, M., and Shinozaki, K. 2013. Genome-Wide Discovery and Information Resource Development of DNA Polymorphisms in Cassava. PLoS ONE 8:e74056.
Salvi, S., Sponza, G., Morgante, M., Tomes, D., Niu, X., Fengler, K.A., Meeley, R., Ananiev, E.V., Svitashev, S., Bruggemann, E., Li, B., Hainey, C.F., Radovic, S., Zaina, G., Rafalski, J.A., Tingey, S.V., Miao, G.H., Phillips, R.L., and Tuberosa, R. 2007. Conserved noncoding genomic sequences associated with a flowering-time quantitative trait locus in maize. Proceedings of the National Academy of Sciences of the United States of America 104:11376-11381.
Shulaev, V., Sargent, D.J., Crowhurst, R.N., Mockler, T.C., Folkerts, O., Delcher, A.L., Jaiswal, P., Mockaitis, K., Liston, A., Mane, S.P., Burns, P., Davis, T.M., Slovin, J.P., Bassil, N., Hellens, R.P., Evans, C., Harkins, T., Kodira, C., Desany, B., Crasta, O.R., Jensen, R.V., Allan, A.C., Michael, T.P., Setubal, J.C., Celton, J.M., Rees, D.J., Williams, K.P., Holt, S.H., Ruiz Rojas, J.J., Chatterjee, M., Liu, B., Silva, H., Meisel, L., Adato, A., Filichkin, S.A., Troggio, M., Viola, R., Ashman, T.L., Wang, H., Dharmawardhana, P., Elser, J., Raja, R., Priest, H.D., Bryant, D.W., Jr., Fox, S.E., Givan, S.A., Wilhelm, L.J., Naithani, S., Christoffels, A., Salama, D.Y., Carter, J., Lopez Girona, E., Zdepski, A., Wang, W., Kerstetter, R.A., Schwab, W., Korban, S.S., Davik, J., Monfort, A., Denoyes-Rothan, B., Arus, P., Mittler, R., Flinn, B., Aharoni, A., Bennetzen, J.L., Salzberg, S.L., Dickerman, A.W., Velasco, R., Borodovsky, M., Veilleux, R.E., and Folta, K.M. 2011. The genome of woodland strawberry (Fragaria vesca). Nature genetics 43:109-116.
Sonah, H., Bastien, M., Iquira, E., Tardivel, A., Légaré, G., Boyle, B., Normandeau, É., Laroche, J., Larose, S., Jean, M., and Belzile, F. 2013. An Improved Genotyping by
160
Sequencing (GBS) Approach Offering Increased Versatility and Efficiency of SNP Discovery and Genotyping. PLoS ONE 8:e54603.
Spindel, J., Wright, M., Chen, C., Cobb, J., Gage, J., Harrington, S., Lorieux, M., Ahmadi, N., and McCouch, S. 2013. Bridging the genotyping gap: using genotyping by sequencing (GBS) to add high-density SNP markers and new value to traditional bi-parental mapping and breeding populations. TAG. Theoretical and applied genetics 126:2699-2716.
Sraphet, S., Boonchanawiwat, A., Thanyasiriwat, T., Boonseng, O., Tabata, S., Sasamoto, S., Shirasawa, K., Isobe, S., Lightfoot, D.A., Tangphatsornruang, S., and Triwitayakorn, K. 2011. SSR and EST-SSR-based genetic linkage map of cassava (Manihot esculenta Crantz). Theoretical and Applied Genetics 122:1161-1170.
Studer, A., Zhao, Q., Ross-Ibarra, J., and Doebley, J. 2011. Identification of a functional transposon insertion in the maize domestication gene tb1. Nature genetics 43:1160-1163.
Swamy, B.M., Vikram, P., Dixit, S., Ahmed, H., and Kumar, A. 2011. Meta-analysis of grain yield QTL identified during agricultural drought in grasses showed consensus. BMC genomics 12:319.
Swanson-Wagner, R.A., Eichten, S.R., Kumari, S., Tiffin, P., Stein, J.C., Ware, D., and Springer, N.M. 2010. Pervasive gene content variation and copy number variation in maize and its undomesticated progenitor. Genome research 20:1689-1699.
Swiderski, M.R., Birker, D., and Jones, J.D. 2009. The TIR domain of TIR-NB-LRR resistance proteins is a signaling domain involved in cell death induction. Mol Plant Microbe Interact 22:157-165.
Takagi, H., Abe, A., Yoshida, K., Kosugi, S., Natsume, S., Mitsuoka, C., Uemura, A., Utsushi, H., Tamiru, M., Takuno, S., Innan, H., Cano, L.M., Kamoun, S., and Terauchi, R. 2013. QTL-seq: rapid mapping of quantitative trait loci in rice by whole genome resequencing of DNA from two bulked populations. Plant J 74:174-183.
Taylor, N.J., Halsey, M., Gaitan-Solis, E., Anderson, P., Gichuki, S., Miano, D., Bua, A., Alicai, T., and Fauquet, C.M. 2012. The VIRCA Project: virus resistant cassava for Africa. GM Crops Food 3:93-103.
Tsuda, K., and Katagiri, F. 2010. Comparing signaling mechanisms engaged in pattern-triggered and effector-triggered immunity. Current opinion in plant biology 13:459-465.
van Ooijen, G., Mayr, G., Kasiem, M.M., Albrecht, M., Cornelissen, B.J., and Takken, F.L. 2008. Structure-function analysis of the NB-ARC domain of plant disease resistance proteins. Journal of experimental botany 59:1383-1397.
Van Ooijen, J.W. (2006). JoinMap ® 4, Software for the calculation of genetic linkage maps in experimental populations. In K. B.V., ed (Wageningen, Netherlands. ).
Velasco, R., Zharkikh, A., Troggio, M., Cartwright, D.A., Cestaro, A., Pruss, D., Pindo, M., FitzGerald, L.M., Vezzulli, S., Reid, J., Malacarne, G., Iliev, D., Coppola, G., Wardell, B.,
161
Micheletti, D., Macalma, T., Facci, M., Mitchell, J.T., Perazzolli, M., Eldredge, G., Gatto, P., Oyzerski, R., Moretto, M., Gutin, N., Stefanini, M., Chen, Y., Segala, C., Davenport, C., Demattè, L., Mraz, A., Battilana, J., Stormo, K., Costa, F., Tao, Q., Si-Ammour, A., Harkins, T., Lackey, A., Perbost, C., Taillon, B., Stella, A., Solovyev, V., Fawcett, J.A., Sterck, L., Vandepoele, K., Grando, S.M., Toppo, S., Moser, C., Lanchbury, J., Bogden, R., Skolnick, M., Sgaramella, V., Bhatnagar, S.K., Fontana, P., Gutin, A., Van de Peer, Y., Salamini, F., and Viola, R. 2007. A High Quality Draft Consensus Sequence of the Genome of a Heterozygous Grapevine Variety. PLoS ONE 2:e1326.
Ward, J., Bhangoo, J., Fernández-Fernández, F., Moore, P., Swanson, J.D., Viola, R., Velasco, R., Bassil, N., Weber, C., and Sargent, D. 2013. Saturated linkage map construction in Rubus idaeus using genotyping by sequencing and genome-independent imputation. BMC genomics 14:1-14.
Whankaew, S., Poopear, S., Kanjanawattanawong, S., Tangphatsornruang, S., Boonseng, O., Lightfoot, D., and Triwitayakorn, K. 2011. A genome scan for quantitative trait loci affecting cyanogenic potential of cassava root in an outbred population. BMC genomics 12:266.
Wilson, L.M., Whitt, S.R., Ibanez, A.M., Rocheford, T.R., Goodman, M.M., and Buckler, E.S.t. 2004. Dissection of maize kernel composition and starch production by candidate gene association. Plant Cell 16:2719-2733.
Wurdack, K.J., Hoffmann, P., and Chase, M.W. 2005. Molecular phylogenetic analysis of uniovulate Euphorbiaceae (Euphorbiaceae sensu stricto) using plastid RBCL and TRNL-F DNA sequences. American Journal of Botany 92:1397-1420.
Yamanaka, S., Nakamura, I., Watanabe, K.N., and Sato, Y. 2004. Identification of SNPs in the waxy gene among glutinous rice cultivars and their evolutionary significance during the domestication process of rice. TAG. Theoretical and applied genetics 108:1200-1204.
Yu, C., Zavaljevski, N., Desai, V., and Reifman, J. 2011. QuartetS: a fast and accurate algorithm for large-scale orthology detection. Nucleic acids research 39:e88.
Zhang, R., Murat, F., Pont, C., Langin, T., and Salse, J. 2014. Paleo-evolutionary plasticity of plant disease resistance genes. BMC genomics 15:1-17.
Zipfel, C. 2014. Plant pattern-recognition receptors. Trends in immunology 35:345-351.
Zmienko, A., Samelak, A., Kozlowski, P., and Figlerowicz, M. 2014. Copy number polymorphism in plant genomes. TAG. Theoretical and applied genetics 127:1-18.
Supplementary data
The SNP data set supporting the results of this article is available in the SNiPlay
repository, http://sniplay.cirad.fr/cgi-bin/public_data.cgi.
162
The Cassava draft genome sequence used in this research is available at:
http://phytozome.jgi.doe.gov/pz/portal.html#!bulk?org=Org_Mesculenta.
The additional files are available at:
https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-015-1397-
4#MOESM1
163
CHAPTER 4
“We keep moving forward, opening new doors, and doing new things, because we're
curious and curiosity keeps leading us down new paths”
-Walt Disney
164
Novel genetic factors involved in cassava bacterial blight resistance
detected through QTL analysis
Johana Carolina Soto Sedano1, Fabio Andrés Gómez Cano4, Rubén Eduardo Mora
Moreno1, Boby Mathew2, Jens Léon2, Adriana Jimena Bernal, Agim Ballvora2, Camilo
Ernesto López Carrascal1 1 Manihot Biotec Laboratory, Biology department, Universidad Nacional de Colombia,
Bogotá, Colombia. 2 INRES-Plant Breeding University of Bonn, Bonn, Germany.
Abstract
Cassava, Manihot esculenta Crantz, is one of the most important crops world-wide
representing the staple security for more than 1 billion people. Cassava’s production
is constantly threatened by several diseases, one of the most important is cassava
bacterial blight caused by Xanthomonas axonopodis pv. manihotis (Xam). A high
dense genetic map developed with Single Nucleotide Polymorphisms (SNPs) through
a Genotyping by Sequencing (GBS) approach was used to QTL (Quantitative Trait
Loci) detection for cassava bacterial blight (CBB) resistance. As a mapping population
a F1 highly segregant of 117 full sibs was used and tested for resistance to two Xam
strains (Xam318 and Xam681) at two locations in Colombia: La Vega, Cundinamarca
and Arauca. The evaluation was conducted in two years during rainy and dry season.
A third evaluation was carried out under greenhouse conditions. The phenotypic
evaluation of the response to Xam revealed continuous variation. Based on composite
interval mapping analysis, 16 strains-specific QTLs were detected, explaining
between 11.7 and 22.1% of phenotypic variance of resistance to Xam. From these
QTLs, nine show stability between the two seasons evaluated. A genotype by
environment analysis was performed, in order to evaluate responses to Xam
performance of genotypes under CBB incidence. QTL by environment interaction was
detected for ten QTLs and broad sense heritability of the resistance showed values of
23% and 53% for Xam318 and Xam681 respectively. In total 147 genes were found in
the QTLs intervals, thirteen genes correspond to genes coding for resistance related
domains, LRR, LRR-Kinase, NB-ARC-LRR and WRKY. Four genes co-localizing with
three QTLs exhibited differentially expression during infection of parental TMS30572
(resistance background) with Xam681.
165
Keywords: Xanthomonas axonopodis p.v manihotis, genotyping by sequencing, QTLs,
high dense genetic map, RNA-seq.
Introduction
Cassava, Manihot esculenta Crantz, is a starchy root crop and one of the most
important crops over the world due to its importance for food security in tropical
regions. This crop represents an important source for calories for about one billion
people (Ceballos et al., 2010). Cassava tolerates drought and it has been considered
as one of the best alternatives for providing food for the world population in the
context of climatic change (FAO, 2013). The major bacterial vascular disease affecting
this crop is Cassava Bacterial Blight (CBB), caused by Xanthomonas axonopodis pv.
manihotis (Xam). This disease has a very high destructive power causing losses
between 12 and 100% in affected areas (Lozano, 1986; López and Bernal, 2012). Xam
was described among the top 10 most important plant pathogenic bacteria
(Mansfield et al., 2012). CBB has been reported in all regions where cassava is grown
(Taylor et al., 2012; López y Bernal, 2012; FAO, 2013), and has been identified in 56
countries distributed in Asia, Africa, Oceania and North, Central and South America
(http://www.cabi.org/). Additionally the number of reports in countries where the
disease was not previously identified is increasing and it has been described the
bacterial movement worldwide (Bart et al., 2012). Recent studies have shown that
Colombian Xam populations remain highly dynamic and exhibit a high genetic
diversity (Trujillo et al., 2014). The analysis of 65 Xam genomes revealed that this
pathogen harbors between 14 to 22 effector genes, from which nine are conserved in
all the strains (Bart et al., 2012).
The best alternative and most efficient strategy to control CBB is to take advantage of
natural plant genetic resistance and planting resistant cultivars. Plants have
developed strategies to defend themselves against pathogens. Unfortunately these
mechanisms have been mainly studied in model plants and knowledge generated in
cassava is relatively scarce. Histology and cytochemistry studies of the resistance
mechanisms in cassava during Xam infection showed callose deposits as a barrier in
cortical parenchyma cells and phloem, contributing to block bacterial multiplication
(Kpémoua, 1996; Sandino et al., 2105). Other mechanisms like cell wall fortification,
lignification and suberization associated with callose deposition and production of
flavonoids and polysaccharides were also observed during cassava response being
faster and stronger in resistant cultivars compared to susceptible ones (Kpémoua et
al., 1996). On the other side, in the last years important efforts have been conducted
166
to identify molecular determinants of the CBB resistance, including the amplification
of genes coding for proteins containing NBS and TIR domains through PCR (López et
al., 2003) or by bioinformatics approach on the cassava genome (Lozano et al. 2015;
Soto et al., 2015). Several gene expression studies have identified genes induced and
repressed in response to Xam in both susceptible and resistant cultivars (Santaella et
al., 2004; López et al., 2005; Muñoz et al., 2014). The identification of non-coding
microRNAs (miRNAs) (Pérez et al., 2012) and trans-acting small interfering RNAs (ta-
siRNAs) induced and repressed in during Xam infection (Perez et al., 2012; Quintero
et al., 2013) have been recently described.
Resistance to CBB has been described as quantitative, with polygenic inheritance and
additive (Hahn et al., 1974; Jorge et al., 2000, 2001) contrary as it occurs with
resistance to cassava mosaic disease (CMD) (Hahn, et al., 1980; Lokko, 2004; Rabbi et
al., 2014). Several quantitative trait loci (QTL) for resistance to CBB have been
described for cassava employing the full-sib population derived from the cross
TMS30572 x CM1477-2. Eight QTLs, explaining between 7.2% and 18.2% of the
variance were detected in field conditions under high disease pressure and over two
consecutive crop cycles (Jorge et al, 2001). On the other hand, twelve QTLs were
identified under controlled conditions to five Xam strains (Jorge et al. 2000). These
QTLs explained 9% to 27% of the phenotypic variance (Jorge et al. 2000). Two new
QTLs were identified to the Xam strains CIO151 and CIO121 explaining 62% and 21%
resistance, respectively (López et al., 2007). Moreover, Wydra et al. (2004) reported
nine QTL explaining from 16% to 33% of the phenotypic variance to four African Xam
strains.
The environment plays an important role in the phenotypic response of traits
governed quantitatively (Weinig and Schmitt, 2004; Anderson et al., 2014). In
consequence, the detection and stability of the QTLs between environments is an
important aspect to consider in the study of genetic determinants of complex traits
(Anderson et al., 2013; Mitchell-Olds, 2013; El-Soda et al., 2014). In QTL studies there
is a QTL by environment interaction (Q x E) when a QTL has a different effect under a
trait in different environments, even if it has a significant effect in one environment
but not in another (El-soda et al., 2014). This situation is even more complex when is
considered the plant-pathogen interaction. According to the classic quantitative
genetic model the phenotype is the result of the genetic composition of the plant (G),
the environment (E) and the interaction between them. However, G x E interaction
can be outspread in the case of the plant-pathogen interactions where another
genotype (corresponding to the pathogen organism) is included. In this case an
equation of this form will result in G x G x E (Jorgensen et al., 2012). In this context
the detection of QTL and the dissection of their allelic composition are necessary to
167
better understand the genetic basis of these interactions and the phenotypic
responses to specific environmental conditions (El-Soda et al., 2014).
In literature can be found a wide range of studies focused on plant resistance QTLs
and many others concentrated to characterize and quantify transcriptomes as well as
to understand the mechanisms of variation of gene expression during plant pathogen
interactions (Verhage et al., 2010; Lodha and Basak, 2012; Schenk et al., 2012).
However, studies where these two approaches come together for the analysis of gene
expression of loci governing the quantitative resistance in plants are scarce. For
example, in rice has been elucidated the contribution of the quantitative resistance to
blast and sheath blight through an analysis of gene expression of a cluster of genes
belonging to a Germin-Like Protein gene family. These genes correspond to a
previously reported QTL (Manosalve et al., 2009). Recently, in Arabidopsis thaliana a
strategy combining genome wide association (GWA) with the large public gene
expression data it was possible to identify co-expression components involved in
quantitative resistance to distinct isolates of the pathogen Botrytis cinerea (Corwin et
al., 2016). Besides these two examples, large-scale gene expression analysis in
repertories of resistance candidate genes for plant disease has not been described so
far.
The molecular bases and mechanisms of quantitative resistance are not known in
detail. However, recent reports, including the cloning of genes associated to QTLs,
have provided some important clues. The proteins involved in pathogen recognition
can be similar in both qualitative and quantitative resistance (Gebhardt, 2001; López
et al., 2003; Ramalingam et al., 2003; Calenge and. Durel, 2006; Poland et al., 2009;
Kou and Wang, 2010; Roux et al, 2014; Corwin et al., 2016). In addition, genes
involved in quantitative resistance can correspond to those coding for proteins
related to signal transduction (Fukuoka et al., 2009; Corwin et al., 2016) and/or with
antimicrobial activity (Ramalingam et al., 2003; Liu et al., 2006; Guimaraes and Stotz,
2004; Van Kan, 2006).
Here we report 16 novel QTLs for CBB resistance, specifically for two Xam strains,
detected in field evaluations during rainy and dry seasons in two Colombian
localities, as well under greenhouse conditions. Furthermore an analysis of G x E and
QTL x E interaction (Q x E), allowed establishing the effect of the environment over
genotypes and QTLs. A repertory of genes located within the QTL intervals were
identified and their transcription profiles during Xam infection was studied through
RNA-seq.
168
Materials and methods
Plant material
The mapping population consisted in a cross between the Nigerian cultivar
TMS30572 and CIAT’s elite cultivar CM2177-2 provided a full sib F1 segregating
population of 117 individuals (Fregene, 1997). This population has been extensively
used in mapping studies (Fregene et al., 1997; Jorge et al., 2000; Jorge et al., 2001;
Mba et al., 2001; López et al., 2007), and recently for high a dense genetic map
construction integrated with the physical map of the cassava genome (Soto et al.,
2015). To produce multiple clonal plants of each individual, parental and F1
individuals were grown and vegetative propagated from stem-cuttings in CBB-free
fields conditions, at La Vega Cundinamarca and Universidad Nacional de Colombia, at
Arauca, Arauca.
Bacterial strains and phenotypic evaluation of CBB resistance
For long-term storage, bacteria strains were kept in 60% glycerol at −80 °C, and
streaked on LPGA medium [yeast extract (5 g/Lt), peptone (5 g/Lt), glucose (5 g/Lt),
bacto-agar (15 g/Lt)] at 28°C for 12 hours before use as inoculum. A single colony of
Xam was grown in liquid culture at 28°C with shaking at 230 rpm for 24 hours. Cells
were harvested by centrifugation at 3,000 g and re-suspended in 10mM MgCl2.
For the evaluation of CBB disease response, plants were grown on two different
locations: La Vega Cundinamarca, Latitude 0,5°,00´44.188´´N, Longitude
74°21´31.005´´N, which belongs to the Andina region and Arauca, Arauca, Latitude 7°
1' 22.32"N, Longitude 70° 44' 42.50"W, in the Orinoquía region. Plants of six-weeks
old were inoculated on July 2013 and December 2014, corresponding to rainy season
and dry season respectively. Besides, an evaluation was done under greenhouse
controlled conditions, 30/20+/-2°C day/night temperature, 12h of photoperiod and
70% of relative humidity. The inoculation was conducted by puncturing the stem of
each plant between the second and third true leaf. The bacterial suspension was
placed using a tip filled with 10 ul of the inoculum (1x106UFC/mL). As a mock one
plant was inoculated with 10 mM of MgCl2. The disease severity was scored at 7, 14,
21 and 30 days after inoculation, using a rating from 0 to 5 according to the
symptoms scale proposed by Verdier et al (1994), where 0= no symptoms, 1=
169
necrosis at the inoculation point, 2= stem exudates, 3= one or two wilted leaves, 4=
more than three wilted leaves and 5= plant death. Disease progress in time was
calculated for each replicate through the area under disease progress curve (AUDPC)
(Shaner and Finney, 1977; Jeger and Viljanen-Rollinson, 2001). Once the phenotyping
evaluation had finished, the inoculated material and the substrate were burned with
the purpose of ensure the CBB-free fields condition. In order to determine resistance
and susceptibility to the strain in parental and F1 genotypes, the AUDPC value was
taking into account as well as the criteria previously described (Trujillo et al., 2014;
Restrepo et al., 2000; Jorge et al., 2000). A statistical t test was performed to establish
contrasting responses of resistance and susceptibility to each Xam strain evaluated.
Broad-sense heritability (H2) of the response to Xam strains was determined through
calculating variance components using lme4 package (Bates et al., 2015) in R
software (Ripley, 2001). The measure of the variance components of genotype,
genotype by environment, genotype by year and experimental error, were measure
according to Holland (2006). The value of heterobeltiosis (better-parent heterosis),
as heterosis estimated over the resistant parent (Jinks and Jones, 1958), was
measured in order to establish the percentage of progeny that exhibited higher levels
of resistance than the resistant parental. In order to evaluate performance (response
to bacteria) of cassava genotypes under CBB incidence, a genotype x environment
analysis was performed through a GGE-biplot analysis of multi-environment trials
(MET) independently by bacteria strain and year of evaluation (rainy and dry
seasons), based on the model for two principal components (Yan et al., 2000; Yan
2001, 2002; Yan and Tinker, 2006). Data from all locations, including greenhouse,
were combined to construct the GGE-biplots in order to compare also the behavior of
the genotypes even under controlled environment. The GGE-biplot criteria were
centered by two (centered G + GE), no scaling and singular value partitioning (SVP) of
row metric preserving. GGE-biplot analysis was performed using the R package
GGEBiplotGUI (Frutos and Galindo, 2012).
Statistical analysis
For each location, year and strain, five biological replications per genotype, parental
and mock were disposed according to a randomized complete design. A Log
transformation for the AUDPC values was done according to previous CBB
evaluations (Restrepo et al., 2000). The distribution of frequency of the AUDPC data
was evaluated for normal distribution by the Shapiro-Wilk test, analysis of variance
(ANOVA) and correlation of environment phenotyping data were also performed
170
through Pearson test. The AUDPC scores were averaged per experiment and used for
further QTL analysis. All statistical analyses were performed using R software
(Ripley, 2001).
QTL mapping
The QTL mapping analysis was performed by Composite Interval Mapping (CIM)
using the Haley–Knott regression model of R/qtl V1.37-11 (Broman, 2015). Three
markers as covariant in a window size of 10 cM were used. A high dense genetic map
of cassava containing 2,141 SNPs was employed. This map has eighteen linkage
groups with an overall size of 2,571 cM and an average distance of 1.26 cM between
markers (Soto et al., 2015). In addition, a set of 2,236 GBS-SNP markers with
unknown genetic position but with physical position known were join to the map as
linkage group nineteen. A LOD threshold was calculated from 1,000 permutation
tests. However a LOD score higher than 3 was also chosen to declare the presence of
a QTL. The LOD peak of a significant QTL was considered as the QTL location on the
linkage group in the map. The QTL interval was determined through a LOD decrease
of 1.5 from the LOD peak position. The phenotypic variation (R2) explained by each
QTL was determined through the function calc.Rsq in R. In order to determine
possible environment effects under QTLs, an analysis using the significant additive
phenotypic effects of QTLs was performed through the software QTL IciMapping
(Meng et al., 2015). Candidate genes were located using the knowledge of physical
positions of the SNP-based genetic map. To define possible clusters, a maximum
distance between two or more genes of 200 kb and less than eight different
annotated genes between them was allowed, according to previous criteria applied in
other genomes (Meyers et al., 2003; Jupe et al., 2012). Moreover, BLAST (Basic Local
Alignment Search Tool) and gene ontology analysis were conducted to the current
cassava genome (v6.1) implemented at the JGI’s Phytozome platform for genes
identified within the QTLs intervals.
Gene expression of genes co-localized with QTLs in resistant parental
background during Xam681 infection
Plant inoculation, RNA extraction and sequencing
TMS30572 plants of six-weeks old in 500gr substrate pots were grown under
greenhouse conditions and maintained at 30/20+/-2°C day/night temperature, 12h
171
of photoperiod and 70% of relative humidity before and during the inoculation
process. Preparation of the inoculum of strain Xam681, and inoculation procedure
was performed as previously described for the evaluation of CBB resistance in F1
population. Four cms of stem area around the inoculation point (two cm above and
two below the inoculation point) were collected at 1, 3 and 5 days post inoculation
(dpi) flash introduced into RNAlater® solution and stored at −80°C until RNA
extraction. Additionally tissue of not inoculated plants, grown under the same
conditions, was taken as a control of the inoculation puncturing effect. Three plants
as biological replicates per time were used.
Total RNA was isolated separately from approximately 100 mg of stem tissue from
each plant using Trizol protocol (Invitrogen). mRNA was isolated from total RNA
using oligo (dT) magnetic beads. First-strand cDNA synthesis was performed using
reverse transcriptase (Invitrogen) and second-strand cDNA was synthesized using
RNase H and DNA polymerase I (Invitrogen). Size selection for the paired-end
libraries was 101bp. The integrity and quality of samples was evaluated with the
bioanalyzer Agilent 2100 (Agilent Technologies). RNA sequencing was performed
through Illumina HiSeqTM 2000.
Data analysis and gene expression level
RNAseq data and cassava transcriptome were generated by Gomez et al 2016
(manuscript in preparation). Sequence quality and filtered was evaluated by FastQC
(V0.11.2) (Andrews, 2010) and FASTX-trimmer from FASTX-Toolkit (V0.0.13.2)
(Gordon and Hannon, 2010), respectively. Clean reads were mapped to the current
cassava genome V6.1 (phytozome.com) using RSUBREAD, mapping option uniquely
(Shi and Shi, 2013). The count per millions account (CPM) were calculated using
FeatureCount (Liao et al., 2014). The CPM was normalized by trimmed mean of M-
values normalization method (TMM) and the differential expression profiles were
generated using the R package edgeR (Robinson et al., 2010) using the QL F-test
(p<0.05). The differentially expressed genes were identified using Log2Fold-change
and pairwise comparisons of treatments.
Results
Bacterial strains and phenotypic evaluation of CBB resistance
In order to identify Xam strains that generate highly contrasting responses in
parents, an initial evaluation was performed under greenhouse conditions. Seven
172
Xam strains were chosen due to the fact that they belong to different haplotypes and
represent the genetic diversity of Xam in Colombia (Trujillo et al., 2014). These
experiments were conducted by duplicated under controlled, greenhouse conditions.
Five of seven strains inoculated (Xam226, Xam571, Xam645, Xam306 and CIO151)
shown to be virulent in both parents and the symptoms/response were similar, with
AUDPC values not showing significant differences. The AUDPC average values for the
female parent (TMS30572) against these strains were 1.50, 1.68, 1.34, 1.46 and 1.47,
while for the male parent (CM2177-2) the AUDPC average values were of 1.65, 1.72,
1.55, 1.69 and 1.64, respectively (Figure 4-1). Considering the threshold establish by
Jorge et al (2000) and Restrepo et al (2000), both parents were susceptible to these
five strains. On the other hand, the parents exhibited a highly contrasting phenotypic
response when were inoculated with the strains Xam318 and Xam681 (Figure 4-1).
The parent TMS30572 shown to be resistant with an AUDPC average value of 1,19 for
Xam318 and 1.49 for Xam681, while the parental CM2177-2 was susceptible with a
AUDPC values of 1,71 and 1,88 for Xam318 and Xam681, respectively (ɑ= 0.05).
Symptoms in parental CM2177-2 against Xam318 and Xam681 appeared from 14
days post inoculation (dpi), while in the parental TMS30572 symptoms started to be
evident around 21 dpi. Moreover, during the evaluation time this parental not
exceeded note 3 in the scale of symptoms. For strain Xam318 the AUDPC average
values, based in two experiments with five biological replicates each, range from 1.49
to 1.84 for parental CM2177-2 while for the parental TMS30572 ranged from 1.22 to
1.55. On the other side, for strain Xam681, the AUDPC average values range from 1.62
to 1.79 for parental CM2177-2 and for the parental TMS30572 ranged from 0.81 to
1.41. The contrasting responses induced in the parental by strains Xam318 and
Xam681 were confirmed in a third experiment under natural conditions in La Vega.
Based on these results, strains Xam318 and Xam681 were selected for further
inoculations and phenotyping evaluation in the biparental mapping population.
173
Figure 4-1. Evaluation of parental responses to different bacterial strains.
Phenotypic evaluation of the parents TMS30572 and CM2177-2 inoculated with
seven strains of Xam was determined through the measured of the area under
disease progress curve (AUDPC). Significant differences as contrasting responses of
resistance and susceptibility, T-test ɑ= 0.05*.
Phenotypic evaluation of CBB resistance in mapping population
Once established the strains for which the parental shown a contrasting response, the
F1 population and parents were evaluated on two different locations (Arauca and La
Vega) and during a two-year period. The first one corresponds to Arauca during a
rainy season in 2013. In this period the maximum and minimal temperatures were
31°C and 22°C, respectively, relative humidity of 88%, with a mean precipitation of
301 mm. The second evaluation conducted at the same location was done in 2014
during the dry season, with maximum and minimal temperatures of 31°C and 22°C,
relative humidity of 73%, with a mean precipitation of 18.7 mm (IDEAM.
www.ideam.gov.co). On the other hand, at La Vega, the maximum and minimal
temperatures during the evaluation time in the rainy season were 20°C and 9°C,
1,19
1,50
1,68
1,34
1,49 1,46
1,47
1,71 1,65
1,72
1,55
1,88
1,69 1,64
0,00
0,20
0,40
0,60
0,80
1,00
1,20
1,40
1,60
1,80
2,00
Xam318 Xam226 Xam571 Xam645 Xam681 Xam306 CIO151
Log
AU
DP
C
TMS30572 CM2177-2
*
*
174
respectively, relative humidity of 79%, with a mean precipitation of 106 mm. The dry
season showed maximum and minimal temperatures of 22°C and 10°C, relative
humidity of 70%, with a mean precipitation of 30 mm (IDEAM. www.ideam.gov.co).
In addition, the evaluation was conducted in two years under greenhouse conditions,
30/20+/-2°C day/night temperature, 12h of photoperiod and 70% of relative
humidity. The corresponding codes for the different conditions, localities, strain and
seasons, are presented in Table 4-1.
Table 4-1. Codes for the localities, strain and seasons where the inoculation and
phenotyping was conducted.
Location Season Xam
strain Code
Arauca (Arauca)
Rainy Xam318
AR318-R
Dry AR318-D
Rainy Xam681
AR681-R
Dry AR681-D
La Vega (Cundinamarca)
Rainy Xam318
LV318-R
Dry LV318-D
Rainy Xam681
LV681-R
Dry LV681-D
Greenhouse (controlled
conditions)
2013 Xam318
G318-2013
2014 G318-2014
2013 Xam681
G681-2013
2014 G681-2014
All AUDPC values for the F1 genotypes for each location, Xam strain and season
showed a continuous and normal distribution, except for LV681-D (supplementary
figure 4-1). As was expected, both parents exhibited AUDPC values in both extremes
of the distribution curve. The genotype TMS30572 was considered as resistant with
AUDPC values ranged from 1.24 for LV318-D, to 1.39 for AR681-R. On the other hand,
the genotype CM2177-2 was considered as susceptible, exhibiting high AUDPC values
which ranged from 1.72 for AR318-R to 1.92 for LV681-R. The number of genotypes
evaluated, the AUDPC by parental and the distribution of AUDPC values in the
progeny by location condition is shown in Table 4-2. The widest range of AUDPC
distribution in the genotypes evaluated was found at AR318-D with a range from 1.23
to 1.78 and G681-2013 with a range of 1.08 to 1.89. While the minor range of AUDPC
distribution was for AR318-R from 1.23 to 1.78 (Table 4-1).
175
In all the locations, for both Xam strains and seasons tested, the number of resistant
genotypes were higher that the susceptible ones (Table 4-2). Moreover, under dry
season conditions for both Arauca and La Vega against Xam318 and Xam681, were
found more resistant genotypes than those found under rainy season. In the
evaluations performed under dry season against Xam318, it was found that 84% and
75% of the progeny was resistant at Arauca and La Vega, respectively, while under
rainy season the percentages were 77% and 72%. A resistant phenotype was
observed for 72% and 70% of the progeny at Arauca and La Vega, respectively under
dry season against Xam681. These values dropped to 65% and 59% during the rainy
season (supplementary Table 4-1).
In order to detect a differential behavior of genotypes against the pathogen, the
phenotypic plasticity was also determined. For the strain Xam318, 48% of the
genotypes showed a distinct phenotype in at least one location and/or season with
respect to the others, while 38% of the genotypes presented this behavior for
Xam681 (supplementary Table 4-1). For instance genotypes g5, g40, g53 and g116
were susceptible in AR318-R but were resistant under the other conditions tested.
On the other hand, genotypes g15, g51 and g97, which showed a resistant response to
Xam681 in La Vega during rainy and dry seasons, were susceptible under the rainy
season in Arauca.
The mapping population exhibited a transgressive segregation for resistance to Xam
strains. Transgressive segregants with higher resistance or susceptibility than the
parents were identified in the two locations, against the two Xam strains and also
under rainy and dry seasons (Table 4-2). For all conditions, the total of resistant
transgressives was higher than susceptible ones. At LV318-D and LV681-D were
obtained the highest number of transgressives, with seven and six respectively. On
the other hand, the highest number of resistant transgressives was identified at
AR318-R and under greenhouse conditions against Xam681 for both years of
evaluation (G681-2013; G681-2014) with 25, 18 and 18 individuals respectively
(Table 4-2). The genotype g79 exhibited a resistant transgressive phenotype in six of
the twelve conditions evaluated (Arauca, for both Xam strains, and also during rainy
and dry seasons). Besides, the greenhouse evaluation against Xam681 showed lower
AUDPC values in relation to the resistant parent. However, the genotype g29 was
identified as a susceptible transgressive for six of the twelve conditions: AR681-R,
AR681-D, LV681-D, G318-2013, G681-2013 and G681-2014 (supplementary Table 4-
2).
Pairwise Pearson correlation between AUDPC values measured at two locations and
greenhouse, two seasons and for the two Xam strains employed was highly
176
significant (P <0.05) with correlation coefficients ranging between 0.62 for pairwise
AR318-R and AR318-D to 0.99 for G681-2013 and G681-2014 (Table 4-3).
Broad sense heritability of the resistance to Xam strains showed values of 23% and
53% for Xam318 and Xam681 respectively. Meanwhile regarding to the heterosis
exhibited by the progeny for the resistance to Xam strains in cassava, the highest
values of heterobeltiosis were those corresponding to the evaluations conducted
under greenhouse conditions. For Xam681 was 21.11% and 20.39% in 2013 and
2014, respectively. For Xam318 was 16.62% for the two years of evaluation. The
lowest value of heterobeltiosis corresponds to AR318-R (Table 4-4).
The analysis of variance showed significant differences (p<0.001) for genotype (g),
environment (location, Xam strain, season) and genotype x environment among the
F1 cassava genotypes tested (supplementary Table 4-3). The GGE-Biplot analysis
showed that the AUDPC evaluations of the genotypes were able to discriminate
between environments. Moreover, the different environments presented large
vectors, meaning that there is information for discriminate between genotypes. In
addition, the behavior of genotypes differed between environments and these fell
into two sectors of the graphic, except for Xam318 rainy season, which fell in three
sectors. Taken together these results indicate there is not just one "best genotype" for
all environments. However, it was possible to distinguish two genotypes as the
“extremes” for a resistant and susceptible behavior for almost all environments. The
two principal components (PC1=genotypes and PC2=Environments) explained
77.83%, 83.96%, 81.12% and 83.97% of the total variance caused by G + GE for
cassava resistance to Xam318 during rainy season, Xam318 during dry season,
Xam681 in rainy season and Xam681 in dry season respectively.
For Xam318 rainy season it was possible to distinguish g29 and g131 as extreme
genotypes as susceptible and resistant respectively (Figure 4-2a). For Xam318 in the
evaluation during the dry season, the extreme susceptible genotype was g124.
However, for Xam318 during the same dry season, the extreme resistant genotype
was g131, the same as for Xam318 during the rainy season (Figure 4-2b). On the
other hand, the evaluation conducted under the rainy season employing Xam681
allowed to distinguish g79 and g29 as extreme genotypes for resistance and
susceptibility respectively (Figure 4-2c). The genotype g93 was identified as the
extreme resistant for the evaluation using Xam681 in the dry season. On the other
hand under these conditions the extreme susceptible genotype was g29, which was
also identified as the extreme susceptible for Xam681 during the rainy season and for
Xam318 in the rainy season. In addition g29 was also the highest susceptible
transgressive (Figure 4-2d). For evaluations against Xam681 under dry season, all
177
genotypes tested in both, Arauca and La Vega locations, seems to have similar
responses to this particular Xam strain (Figure 4-2d).
Table 4-2. Distribution of AUDPC values in the mapping population. Number of
genotypes evaluated, AUDPC values showed by parental, range of AUDPC obtained in
the phenotype distribution, resistant and susceptible genotypes in the population
and number of resistant (R) and susceptible (S) transgressive segregants by
environment, Xam strain and season. AR= Arauca; LV= La Vega; G= greenhouse; R
rainy season; D= dry season.
Location Genotypes
evaluated
AUDPC
CM2177-2
AUDPC
TMS30572
AUDPC
range
R.
genotypes
S.
genotypes
R.
Transg.
S.
Transg.
AR318-R 103 1.72 1.39 1.23 - 1.78 79 24 25 4
AR318-D 100 1.76 1.26 0.97 - 1.78 84 16 11 1
AR681-R 104 1.83 1.35 1.28 - 1.94 68 36 6 2
AR681-D 100 1.8 1.32 1.16 - 1.94 72 28 8 3
LV318-R 93 1.85 1.21 1.16 - 1.82 67 26 3 0
LV318-D 106 1.73 1.24 1.04 - 1.82 80 26 15 7
LV681-R 93 1.92 1.39 1.19 - 1.97 55 38 6 2
LV681-D 106 1.88 1.36 1.21 - 1.97 74 32 12 6
G318-
2013 117 1.85 1.28 1.09 - 1.88 78 39 15 1
G318-
2014 109 1.87 1.28 1.10 - 1.86 73 36 16 0
G681-
2013 117 1.87 1.25 1.08 - 1.89 74 43 18 1
G681-
2014 112 1.87 1.25 1.10 - 1.90 73 39 18 2
Table 4-3. Pairwise Pearson correlation coefficients between AUDPC values.
Correlation measured at two environments and greenhouse in two seasons and for
the two Xam strains employed. AR= Arauca; LV= La Vega; G= greenhouse; R= rainy
season; D= dry season P value = 0.05.
Location AR318-R AR681-R LV318-R LV681-R G318-2013 G681-2013
AR318-D 0.62
AR681-D 0.06 0.84
LV318-D 0.03 0.08 0.81
LV681-D 0.15 0.35 0,05 0.89
G318-2014 0.07 0.15 0,04 0.05 0.91
G681-2014 -0.03 0.34 0,04 0.22 0.34 0.99
178
Table 4-4. Better-parent heterosis. Percentage of heterosis over the better parent
values for the quantitative genetic trait of resistance to the two Xam strains evaluated
in cassava under the environments AR= Arauca; LV= La Vega; G= greenhouse; R=
rainy season; D= dry season
AR-R AR-D LV-R LV-D G-2013 G-2014
Xam681 14.36% 14.60% 13.46% 14.15% 21.11% 20.39%
Xam318 7.43% 14.24% 14.60% 16.59% 16.62% 16.62%
QTL mapping
In total 16 QTLs were detected which were distributed in 12 of the 19 linkage groups.
The linkage groups 19 and 5 had the higher number of QTLs with four and two QTLs
respectively (Table 4-4). The phenotypic variance of resistance to Xam explained for
these QTLs ranged from 11.7 to 22.1%. In particular, for the evaluation against
Xam318 were detected six QTLs explaining from 11.8% to 18.8% of the phenotypic
variance. Ten QTLs were associated to resistance to Xam681 and explained from
11.7% to 22.1% of the phenotypic variance. From these, six QTLs were detected
under the Arauca conditions, while in La Vega and greenhouse conditions were
detected four and six QTLs respectively.
Five QTLs showed high significance based on LOD threshold obtained by permutation
test. The QTL detected in the evaluation at La Vega during the rainy season
(QLV681RD-6) explaining 22.1% of the resistance to Xam681, showed the highest
LOD (Table 4-4).
In average the interval length of the QTLs was 3.1 cM. The QTL with the lowest
interval length was QAR681R-17 with 1 cM, while the QTL with the highest interval
length was QAR318R-5, with 7.5 cM (Table 4-4). Nine of the 16 QTLs were stable
between seasons for the same location (Figure 4-2). From these, five correspond to
QTLs detected under greenhouse conditions, three were detected from the evaluation
conducted at La Vega and one detected under Arauca conditions. Five of the stable
QTLs were associated to the resistance to Xam681 and the other four to Xam318
179
(Table 4-5). There were not QTLs detected for all environments and/or seasons with
significant LOD score.
Figure 4-2. Which Won Where/What graphic of GGE-Biplot analysis. Axis1=PC1=
Genotypes, Axis2=PC2= Environments. A. GGE-Biplot analysis for phenotype
evaluation against Xam318 under rainy season. B. GGE-Biplot analysis for phenotype
evaluation against Xam318 under dry season. C. GGE-Biplot analysis for phenotype
evaluation against Xam681 during the rainy season. D. GGE-Biplot analysis for
phenotype evaluation against Xam681 during the dry season. For all e1=Arauca, e2=
La Vega, e3=Greenhouse.
-40 -20 0 20
-30
-20
-10
01
02
03
0
Which Won Where/What
AXIS1 47.58 %
AX
IS2
30
.25
%
g1
g2
g3
g4
g5
g6
g7
g8
g9
g10
g11
g12
g14
g15
g16
g18
g20 g21
g22
g23
g24
g25
g26
g29
g30
g31
g32
g33
g35g36
g37g38
g39
g40
g41
g42g45
g46
g47g51g52
g53g55
g56
g57
g61
g62
g63
g64
g66
g67
g68
g69g70
g71
g74
g75
g76
g77
g78
g79
g80
g81g82
g84
g85
g86
g88
g89g91
g92
g93
g95
g96
g97
g99
g100
g101
g102
g103
g104
g105
g107g108
g109
g111
g112g114
g115
g116
g118
g120
g121
g122g123
g124
g125
g126
g127
g128
g129
g131
g132
g133
g135g138
g139
g140
g142
g143
g144
g145
g146
g147
g148g149
g150
e1
e2
e3
-40 -20 0 20 40
-30
-20
-10
01
02
03
0
Which Won Where/What
AXIS1 47.07 %
AX
IS2
35
.99
%
g1
g2
g3
g4
g5
g6
g7
g8
g9
g10
g11g12 g14
g15
g16
g18
g20g21
g22g23
g24g25
g26
g29 g30g31
g32
g33
g35
g36
g37g38
g39
g40g41
g42
g45
g46
g47
g51
g52
g53
g55
g56g57
g61
g62
g63
g64
g66g67
g68g69
g70
g71
g74
g75
g76
g77
g78
g79
g80
g81
g82
g84
g85
g86
g88
g89
g91
g92
g93
g95
g96g97
g99
g100
g101
g102
g103
g104g105
g107
g108
g109g111
g112
g114
g115
g116
g118
g120
g121
g122
g123
g124
g125
g126
g127
g128
g129
g131g132
g133
g135
g138
g139
g140
g142
g143
g144
g145
g146
g147
g148
g149
g150
e1
e2
e3
-30 -20 -10 0 10 20 30 40
-30
-20
-10
01
02
0
Which Won Where/What
AXIS1 56.45 %
AX
IS2
24
.67
%
g1
g2
g3 g4
g5
g6
g7g8
g9
g10
g11
g12
g14
g15g16
g18g20
g21
g22
g23
g24
g25
g26
g29
g30
g31
g32
g33
g35
g36
g37
g38 g39
g40
g41g42
g45
g46
g47
g51
g52
g53
g55
g56
g57
g61
g62
g63
g64g66
g67
g68g69g70
g71
g74
g75
g76
g77
g78
g79g80
g81 g82
g84
g85
g86
g88
g89
g91
g92
g93
g95
g96g97
g99
g100
g101
g102g103
g104
g105
g107
g108
g109
g111
g112
g114
g115
g116
g118
g120
g121g122
g123
g124
g125
g126
g127
g128
g129
g131
g132
g133
g135
g138g139
g140
g142
g143
g144
g145
g146
g147
g148g149
g150
e1
e2
e3
-40 -20 0 20
-40
-30
-20
-10
01
02
03
0
Which Won Where/What
AXIS1 53.34 %
AX
IS2
30
.63
%
g1
g2
g3g4
g5
g6
g7g8
g9
g10
g11
g12g14g15
g16
g18
g20
g21
g22
g23g24
g25
g26
g29
g30
g31
g32
g33
g35
g36
g37
g38
g39g40
g41g42
g45
g46
g47
g51
g52
g53
g55
g56
g57
g61
g62 g63
g64g66
g67
g68
g69
g70
g71
g74
g75
g76
g77
g78g79
g80
g81
g82
g84
g85
g86
g88g89
g91
g92
g93
g95 g96
g97
g99
g100
g101
g102 g103
g104
g105
g107
g108
g109
g111g112
g114
g115
g116
g118
g120
g121
g122
g123
g124
g125
g126g127
g128
g129
g131
g132
g133
g135g138
g139 g140
g142
g143g144
g145
g146
g147
g148
g149
g150
e1e2
e3
a
.
b
c d
180
Table 4-5. Summary of QTLs ligados to CBB resistance. QTL name, location, Xam
strain and linkage group), LOD threshold, R2= percentage of phenotypic variance
explicated by the QTL, peak marker, position of the peak marker in the genetic map
given in cM. QTL marker interval, interval length in cM and Kb and number of genes
within the intervals is shown. Underlined stable QTL between seasons for the same
location. In bold QTL highly significant based on LOD obtained by permutation test. *
Unknown genetic position. **. ND= Not determinated.
QTL
name
LOD
score R2
Peak
Marker
Pos.
cM
Marker
Interval
Interval
length cM
Interval
length Kb
Genes in
interval
QAR318R-5 3.3 13.7% MB_36197 78.6 MB_11835/MB_25647 7.5 ND ND
QAR318RD-3 3.0 12.2% MB_49048 79.0 MB_43944/MB_49052 3.3 44.8 3
QAR681R-17 3.4 13.9% MB_3100 107.1 MB_57042/MB_57031 1,0 25.7 9
QAR681R-19 3.2 13.1% MB_76581 * MB_75272/MB_76594 * 6.7 2
QAR681D-14 4.3 18.1% MB_9956 90.6 MB_62539/MB_10857 4.1 1.5 2
QAR681D-19 3.1 13.1% MB_48318 * MB_48251/MB_48264 * 331.3 19
QLV318RD-19 3.8 17.3% MB_38006 * MB_37785/MB_38030 * 80.8 8
QLV681RD-6 5.0 22.1% MB_50599 31.0 MB_36352/MB_63835 2.2 1.2 3
QLV681R-7 3.4 15.4% MB_78306 81.3 MB_78255/MB_78336 3.6 303.9 40
QLV681RD-4 3.4 13.8% MB_74425 30.8 MB_74354/MB_74453 5.2 275.7 34
QGH318-8 3.5 13.6% MB_62295 38.8 MB_10205/MB_10214 1.5 25.4 6
QGH318-13 3.2 12.7% MB_0787 76.0 MB_69697/MB_29269 3.3 1.1 1
QGH318-19 4.9 18.8% MB_49863 * MB_49851/MB_49878 * 58.8 8
QGH681-5 3.4 12.7% MB_0645 115.3 MB_56217/MB_0643 1.8 1.1 2
QGH681-10 3.2 11.7% MB_55582 107.0 MB_55587/MB_55581 1.9 5.4 2
QGH681-2.2 4.2 15.8% MB_23160 93.5 MB_23143/MB_26945 2.3 110.1 8
In order to identify the environment effect on the QTLs, a Q x E interaction analysis
based on the additive phenotypic effect (APE) was conducted. In this case, QTLs
detected under greenhouse conditions were not taken into account due to this
condition was not considered as an environment. Significant Q x E interaction was
established for ten QTLs (Figure 4-2). Four QTLs (QAR318R-5, QAR681R-17,
QAR681R-19 and QLV318RD-19) exhibit positive APE, which were correlated with
high AUDPC values or susceptibility, while five QTLs (QAR318RD-3, QAR681D-14,
QAR681D-19, QLV681R-7 and QLV681RD-4) shown a negative APE, which were
correlated with resistance or low AUDPC values. The QTLs QAR318R-5, QAR681R-17,
QAR681R-19, QAR681D-14, QAR681D-19 and QLV681R-7 are conditionally neutral,
due to they were detected in specific environments. On the other hand, the QTLs
QAR318RD-3, QLV318RD-19, QLV681RD-6 and QLV681RD-4 are stable QTLs, showing
181
different effect levels in dry and rainy season. Despite that the QTL QLV681RD-6 was
detected during rainy and dry seasons, the APE was detected only under dry
conditions (Figure 4-3).
The genomic regions corresponding to each QTL interval were searched for coding
genes. Several genes were identified in all regions covering the QTLs intervals, except
for the QTL QAR318R-5. In total 147 genes were present in the regions containing
QTLs (supplementary Table 4-4). The QTLs physical intervals comprised in total
1275.5 Kb corresponding to a genetic distance of 37.7 cM with a mean value of 33.8
kb per 1 cM. In average it was found one gene each 8.67 Kb. However, this ratio
varies between the QTLs from 0.4 in QLV681RD-6 to 17.4 genes per Kb in QAR681D-
19. QTL regions QLV681R-7 and QAR681D-19 which shown the highest intervals
lengths (303.9Kb and 331.3Kb, respectively) contained also the highest number of
genes with 40 and 19 respectively. However the gene density (GD) was very
different: one gene every 7.5Kb for QLV681R-7 and one gene every 17.4 Kb for
QAR681D-19. In this case both QTLs have almost the same interval but the difference
in number of genes is relatively high. On the other hand QTLs with lowest interval
length, as for example QGH318-13 and QGH681-5, both with 1.1 kb each, have only
one and two genes, respectively.
From the 147 genes co-localizing with the QTLs, 89 genes (60.5%) had an annotation
in PANTHER and PFAM, while 30.5% (45 genes) were annotated in the three data
bases. 29 genes (19.7%) had no annotation described so far in the current cassava
genome, based on PFAM (Finn et al., 2010), PANTHER (Thomas et al., 2003) or
EuKaryotic Orthologous Groups (KOG) (Koonin et al., 2004). Protein kinase domain
was the most represented with thirteen counts (8.84%). A gene cluster of nine
kinases was identified in QLV681D-4. This cluster comprises a region of 140kb in
length, a GD of one gene every 15.5 kb. Also, within this cluster it was found a NB-ARC
and an LRR-Kinase encoded immunity-related gene (IRG). One genes coding for a
kinase was identified in QAR681R-19 and QGH681-2.2 while in the QLV681R-7 two
kinase were found (supplementary Table 4-3).
182
Figure 4-3. QTL x environment interaction based on additive phenotypic effects
(APE). QTLs exhibit significant Q x E interaction based on additive phenotypic effects
(APE). Positive APE was correlated with negative effect under the response to Xam
while negative APE with positive effect under resistance.
A cluster of six NB-ARC was found in QAR681D-19. This cluster comprises a region of
226 kb, with a GD of one gene every 37.6kb. On the other hand, two IRGs co-localized
with QLV318RD-19, a gene coding for a WRKY DNA -binding domain protein and a
gene coding for an NB-ARC-LRR, which were 1.4Mb apart. The QTLs QAR681R-19,
QGH318-8 and QAR681R-17 co-localized with a gene coding for LRR-Kinase, a protein
with a site for AvrRpt-cleavage (Cleavage site for pathogenic type III effector
avirulence factor Avr), and a Defensin like protein respectively (Figure 4-4). The
QTLs QLV681RD-6, QGH318-19, QAR681D-14 and QGH681-2.2, which are those QTLs
with higher percentage of phenotypic variance explained, co-localized with different
-0,06
-0,05
-0,04
-0,03
-0,02
-0,01
0
0,01
0,02
0,03
0,04
0,05
0,06
0,07
0,08
AR318-R AR318-D
AR681-R AR681-D
LV318-R LV318-D
LV681-R LV681-D
QA
R3
18
R-5
QA
R3
18
RD
-3
Ad
dit
ive
ph
eno
typ
ic e
ffec
t
QA
R6
81
R-1
7
QA
R6
81
R-1
9
QA
R6
81
D-1
4
QA
R6
81
D-1
9
QL
V3
18
RD
-19
QL
V6
81
RD
-6
QL
V6
81
R-7
QL
V6
81
RD
-4
183
annotated genes, however no IRG were found in these regions (supplementary Table
4-4).
Gene expression of genes co-localizing with QTLs in resistant parental against
Xam681
The expression profiles for 109 (74.1%) of the 147 genes identified in the QTLs were
obtained from transcriptomic data. The RNAseq was obtained from libraries
constructed using mRNA extracted at 1, 3 and 5 dpi. From the 147 genes only four
showed significant differential expression (p-value<0.01) when were compared to
the mock, in at least one the post inoculation times. Three genes were down
regulated and one up-regulated (Figure 4-4). Two of the down-regulated genes,
Manes.06G091400 and Manes.06G091700, co-localized with QLV681R-7 and
corresponds to a chlorophyll A-B binding protein and to a non-annotated gene,
respectively. The third down-regulated gene, Manes.07G107000, corresponds to a
protein tyrosine kinase and co-localized with QLV681D-4. This gene belongs to the
kinase cluster present in this QTL (Figure 4-4). On the other hand, the up regulated
gene Manes.10G091100 co-localized with QAR681D-19. This gene has no annotation
described so far.
Figure 4-4. Gene expression of genes that co-localizes with QTLs in resistant
parental against Xam681. Log2 of the fold-change of the relative mRNA expression
levels of the two down-regulated genes, Manes.06G091400 and Manes.06G091700 at
1dpi, and the down-regulated gene, Manes.07G107000 and up regulated gene
Manes.10G091100 at 5dpi. * p-value<0.01.
-0,89
-2,09
-0,46
0,4
0,01
-0,8
-0,35
0,93
-0,17
0,28
-0,84
1,54
-2,5
-2
-1,5
-1
-0,5
0
0,5
1
1,5
2
log
2 o
f f
old
ch
an
ge
1dpi 3dpi 5dp1
*
*
*
*
Manes.06G091400
Manes.06G091700
Manes.07G107000
Manes.10G091100
184
Discussion
In this study we evaluated the phenotypic response of 117 genotypes of a F1
mapping population against two Xam strains in two different environments and
under rainy and dry seasons, which allowed us to identify 16 strain-specific QTLs
associated with resistance to CBB. In these QTLs were found 147 genes, from which
four were differentially expressed in the parental TMS30572 after infection with
Xam681.
From the seven Xam strains evaluated, two showed a contrasting phenotype between
the parents which was reproducible in all field experiments, showing that is a stable
contrasting response. These two strains, Xam318 and Xam681, were then selected to
inoculate the F1 progeny and to evaluate their response. Strains Xam318 and
Xam681 were isolated from plant samples with typical symptoms of CBB disease in
fields located at Ciénaga de Oro (N 08.889°, W 075.569°) a typical Caribbean savanna
region and Palmitos (N 09.450°, W 075,160°) located on mountains, respectively.
These strains are part of a set of Xam strains evaluated for virulence in nine cassava
accessions. Both showed a high virulent behavior, causing disease in six and eight of
nine accessions tested, respectively (Trujillo et al., 2014). Moreover, these strains
have been reported as part of prevalent haplotypes described in Colombian Xam
populations and to be the product of migratory processes between regions in
Colombia (Trujillo et al., 2014). Even though several studies on Xam populations have
been carried out, it is still needed to examine the population structure of Xam in other
Colombian geographic regions like the Orinoquía and Andina, regions where the
current phenotypic evaluation was conducted. The Xam populations in these regions
remained unexplored since more than a decade (Restrepo et al., 1999, 2000). In these
two particular regions although cassava is cultivated, the crop area is relatively
limited. It will be important for example to establish if the two strains employed in
this study are present in these two regions or if other strains belonging to the same
haplotypes exist. This will allow directing the breeding programs based on these
QTLs toward the resistance to these Xam strains for these particular Colombian
regions.
The broad sense heritability measured in this study was 23% and 53% for resistance
to Xam318 and Xam681 respectively. Broad sense heritability for CBB resistance has
been previously reported with values ranged from 10% to 69% (Hahn et al., 1998,
Jorge et al., 2000; Fregene et al., 2001; Ly et al., 2013). These values of heritability
185
highlight the important effect of the environmental condition on CBB response. In
particular the humidity has a clear influence on the phenotypic response in some of
the F1 individuals. Thus, for example, the number of susceptible individuals was
higher during the rainy season compared with the dry season in both localities (68
vs. 64 for LV and 60 vs. 44 for Arauca) suggesting a positive effect of the humidity in
the phenotypic response of the genotypes tested. This can be related with the fact
that several studies have shown a positive (favorable) effect of humidity not only on
the speed of symptoms, but also on the growing of Xam (Banito et al., 2001; Wydra
and Verdier, 2002; Restrepo et al., 2004).
Even though the environmental wet conditions seem to favor the disease, also their
changes can generate different effects under a particular genotype. This can be
perceived as a change in the plant phenotype, which is known as phenotypic
plasticity. In all the conditions evaluated were found genotypes that exhibited a
resistant behavior under some environmental conditions but susceptible in others.
Thus is the case for example of the genotypes g29, g30 and g116, which showed a
resistant phenotype under the dry season but shown to be susceptible under the
rainy one. Other example is the g51 and g97 which were resistant to Xam681 at La
Vega during rainy and dry seasons, but susceptible under the rainy season at Arauca.
The phenotypic plasticity has been widely described in model and non-model crops
for several traits, including resistance to plant pathogens (Agrawal, 1999; Dicke and
Hilker, 2003). The individuals showing phenotypic plasticity identified in this study
can be used in local breeding programs as superior genotypes adapted to specific
environmental conditions as an approach exploiting adaptive plasticity (Nicotra et al.,
2010).
The identification of resistant transgressive segregants represents an important
source of resistance for breeding programs. Several genotypes showing resistant
transgressive phenotype were found in this work. The highest values of better-parent
heterosis were those corresponding to the evaluations conducted under controlled
conditions. A particular case was the genotype g79, which was categorized as a
resistant transgressive phenotype for almost all the conditions evaluated. Other
interesting case is g131, which was detected by the GGBiplot analysis. The genotype
g131 has an extreme resistance genotype in rainy and dry season against Xam318.
These two genotypes become an important source of resistance for its future use in
recombination strategies in a cassava breeding program. Transgressive segregation
of resistance has been previously described in cassava against Xam strains (Jorge et
al., 2000, 2001) as well as for other important traits for this crop (Akinwale et al.,
2010; Whankaew et al., 2011; Thanyasiriwat et al., 2013; Njenga et al., 2014). The fact
of identifying some resistant transgressive genotypes suggests the presence of
186
additive and dominant genes playing a pivotal role in CBB resistance. This type of
heterosis has been explained by the presence of blocks of dominant genes from both
parents (Bingham, 1998; Jorge et al., 2000), variations in the chromosome number,
chromosome rearrangements (Rieseberg et al., 1999), and even DNA methylation,
epigenetics and silencing by small RNAs (Shivaprasad et al., 2012).
Sixteen QTLs distributed in eight chromosomes were successfully identified in
cassava to Xam. Six were associated to resistance to strain Xam318 and ten to
Xam681, indicating these are strain specific QTLs. These loci correspond to QTLs
which explains the phenotypic variance of resistance to Xam up to 10%. In order to
not discard QTLs with relative small effect two criteria were used to declare them as
significant. The first one employed a permutation test and the second one consider a
LOD >3. Even if the QTLs explaining up 10% can be considered as minor QTLs, and
the identification of QTLs for plant resistance has focused principally on those
explaining more than 20% of phenotypic variance or large effect (Roux et al., 2014),
QTLs with relative small effect as those identified here, can also have a relevant
importance not only in breeding programs, but also to the understanding the
molecular mechanisms involved in plant disease resistance.
Previously, twelve QTLs were identified to five Colombian Xam strains by Jorge et al
(2000), while Wydra et al (2004) reported nine QTLs to four Xam strains. None of
these QTLs were identified in the present study. Several reasons can explain this fact.
First of all, the environments where the evaluations were conducted were completely
different. As mentioned before the environmental conditions play an important role
in the response to Xam infection and in consequence certainly influence the QTL
detection. Second, several of the previously identified QTLs were strain-specific. As
the present study employed two strains different to the previous studies it is
expected not to find the same QTLs. Actually all the QTLs identified in this study were
strain-specific. Detection of strain-specific QTLs has already been reported for
quantitative resistance in the past for several crops such as rice (Li et al., 1999),
tomato (Wang et al., 2000; Carmeille et al., 2006), melon (Perchepied et al., 2005) and
apple (Calenge et al., 2004), showing that is common phenomenon. Taken together,
these results suggest a strong strain x cultivar x environment interaction. In order to
prove this hypothesis it will be important to consider carrying out phenotypic
evaluations in multi-environments with the same Xam strains as well as to expand
the panel (repertoire) of strains belonging to the same and different haplotypes.
The Q x E interaction was established for all QTLs except for those detected under
greenhouse conditions. Remarkable, the majority of QTLs detected under rainy
season showed a positive APE, while those detected under dry conditions exhibit a
187
negative one. This suggests a positive contribution in the response to Xam of these
QTLs under dry conditions in the population evaluated. This evidences once again the
important role of the environment and especially the humidity conditions in favoring
CBB disease. The detection of QTLs with a significant APE in one environment but not
in another, (unstable QTLs) or conditionally neutral QTLs, are evidence of the pivotal
role of the environment conditions in the instability of the detection of QTLs between
environments. Conditionally neutral resistance QTLs have been reported in some
crops like rice (Li et al., 2007), wheat (Ramburan et al., 2004) and apple (Calenge and
Durel, 2006). In cassava in spite that environmental unstable QTLs have been
reported (Jorge et al., 2001), Q x E interaction for CBB has not been described so far.
Although, stable detected QTLs are the best ones for breeding programs, due to its
genetic component plays a major role, the knowledge behind a conditional neutral
QTL also has a place in plant improvement. These resistance loci indicate the great
influence on the phenotype of external conditions, such as environmental factors.
Thus, they can be exploited in local programs that present particular environmental
conditions and adapted pathogens.
All the QTLs identified covering a region corresponding to 1,275.5 Kb. In this region
were found 147 genes. In the past, studies on QTLs worked with low saturated
genetic maps developed mainly with anonymous markers. In consequence they
generate QTLs with large intervals with narrow knowledge about the number and the
nature of the genes present in these regions. This fact made the identification of
genes responsible of the variation of the most interesting traits in crops time-
consuming. Nowadays, with the advent of next generation sequencing, strategies
such as genotyping by sequencing (GBS), it is possible to obtain thousands of markers
with known physical positions in the genome. This expedited the development of
high dense genetic maps containing non-anonymous markers. In spite of some
associations between candidates genes with QTL has been reported (Faris et al.,
1999; Ramalingam et al., 2003), the isolation of genes present in QTLs regions are
scarce. In this study through the use of a high dense genetic map obtained by GBS we
were able to identify 147 putative genes that co-localized with QTLs in short interval
lengths (3.1 cM in average).
The corresponding functional gene annotation was obtained for 118 of the genes,
from which thirteen showed annotations directly related to plant immunity. These
genes co-localized with five QTLs. In the past, other studies have shown the co-
localizing of genes coding for typical resistance genes with QTLs (Ramalingam et al.,
2003; St. Clair, 2010; López, 2011). Thus in spite that resistance to CBB has been
considered as a quantitative trait and that in cassava-Xam the Avr-R interaction has
not been demonstrated, the presence of genes coding for typical R genes within the
188
QTLs intervals strengthen the idea of an overlapping between qualitative and
quantitative resistance.
In the whole repertoire of genes present in the QTL intervals, the kinase was the most
represented group of proteins. A cluster of nine kinases, one of them being a LRR-
kinase, co-localized with QLV681D-4. Despite the kinase family is one of the most
widely distributed protein families in plant genomes, find a co-localization of
members of this family with resistance QTLs not seems to be due by chance. Several
proteins of this family have been involved in plant resistance in model plants (Huard-
Chauveau et al., 2013) as well as in some of the most important economical crops
such as barley (Druka et al., 2008), wheat (Fu et al., 2009) and maize (Zuo et al.,
2015). The kinases can be an important element in quantitative disease resistance,
either as a receptor or as part of the signaling pathway. Clusters of kinase related
proteins have been reported as an import part of defense responses in Arabidopsis
and in other crops (Roux et al., 2014). In tomato for example, the Pseudomonas
resistance gene (Pto) is a Ser/Thr kinase (Oh and Martin, 2011) that belongs to a
kinase cluster. The large group of kinases has been exploited to be used as “decoys”
for the plant immunity (Van der Hoorna and Kamoun, 2008). One of the kinases
(Manes.07G107000) identified in the QTL region (QLV681RD-4) and explaining
13.8% of the resistance to Xam, was down-regulated at 5dpi. The down regulation-of
protein kinases have been previously reported as a negative regulator of plant
immunity (Petersen et al., 2000). This is the case of the Mitogen-activated protein
kinase MPK4 from Arabidopsis, which when is down regulated, result in the
activation of the plant pathogen-mediated defense responses (Gao et al., 2008; Zhang
et al., 2012).
Despite the fact that the majority of the genes (121 genes) present in the QTL regions
do not correspond to IRGs and only four were differentially expressed they should
not be disposal as interesting genes. Conversely, it has been demonstrated that the
classical immunity-related proteins could be part of a small fraction of the total genes
associated with quantitative resistance (Corwin et al., 2016). Also, evidence from the
cloning of some genes responsible for plant quantitative resistance, shows that the
corresponding proteins do not belong to any specific group of immunity related
proteins or lack of the classical domains present in R proteins. These genes are
involved in different functions and/or process (Poland et al., 2009; Bryant et al.,
2014; Roux et al., 2014). The repertoire of genes co-localizing with the QTLs reported
here, represents a first step in the dissection of the biological mechanisms that
govern CBB resistance and a new sources of genes to be validated through different
approaches. With the advent of gene editing methodologies it will be possible to
know the function of these particular genes in CBB resistance (Sander and Joung,
189
2014). Moreover, this gene repertoire and the SNP markers associated to them
become a source of data directly related to CBB resistance. They can be used in plant
breeding strategies focused to develop cassava materials resistant to CBB adapted to
the regions here evaluated.
Acknowledgments
We thank COLCIENCIAS for the financial support through grand 521-2011 and PhD
scholarship call 528. We would like to extend our gratitude to the International
Center for Tropical Agriculture (CIAT) for enabling the plant material, to the staff of
the Universidad de Colombia, Orinoquía, especially to Mr. Alexis Landaeta. Special
thanks to Mr. Lisímaco López† for their hospitality, kindness and for the support
during the field experiments.
References
Agrawal, A. A. 1999. Induced plant defense: evolution of induction and adaptive phenotypic plasticity. Inducible plant defenses against Pathog. Herbiv. Biochem. Ecol. Agric. Am. Phytopathol. Soc. Press. St. Paul, MN:251–268
Akinwale, M. G., Aladesanwa, R. D., Akinyele, B. O., Dixon, A. G. O., and Odiyi, A. C. 2010. Inheritance of-carotene in cassava (Manihot esculenta crantza). Int. J. Genet. Mol. Biol. 2:198–201
Anderson, J. T., Lee, C. R., Rushworth, C. A., Colautti, R. I., and Mitchell-Olds, T. 2013. Genetic trade-offs and conditional neutrality contribute to local adaptation. in: Molecular Ecology. 22:699-708
Anderson, J., Wagner, M., Rushworth, C., Prasad, K., and Mitchell-Olds, T. 2014. The evolution of quantitative traits in complex environments. Heredity (Edinb). 11233:4–12
Andrews, S., and others. 2010. FastQC: A quality control tool for high throughput sequence data. Ref. Source.
Banito, A., Kpémoua, K. E., Wydra, K., and Rudolph, K. 2001. Bacterial blight of cassava in Togo: its importance, the virulence of the pathogen and the resistance of varieties. Pages 259–264 in: Plant Pathogenic Bacteria, Springer.
Bart, R., Cohn, M., Kassen, a., McCallum, E. J., Shybut, M., Petriello, a., Krasileva, K., Dahlbeck, D., Medina, C., Alicai, T., Kumar, L., Moreira, L. M., Neto, J. R., Verdier, V., Santana, M. a., Kositcharoenkul, N., Vanderschuren, H., Gruissem, W., Bernal, a., and Staskawicz, B. J. 2012. PNAS Plus: High-throughput genomic sequencing of cassava
190
bacterial blight strains identifies conserved effectors to target for durable resistance. Proc. Natl. Acad. Sci. 109:E1972–E1979
Bates, D., Maechler, M., Bolker, B., and Walker, S. 2015. lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1--7. 2014. Inst. Stat. Math. WU website. http//CRAN. R-project. org/package= lme4. Accessed March. 18
Bingham, E. T. 1998. Role of chromosome blocks in heterosis and estimates of dominance and overdominance. Concepts Breed. heterosis Crop plants. :71–87
Broman, K. W. 2015. R/qtlcharts: interactive graphics for quantitative trait locus mapping. Genetics. 199:359–361
Bryant, R. R. M., McGrann, G. R. D., Mitchell, A. R., Schoonbeek, H., Boyd, L. A., Uauy, C., Dorling, S., and Ridout, C. J. 2014. A change in temperature modulates defence to yellow (stripe) rust in wheat line UC1041 independently of resistance gene Yr36. BMC Plant Biol. 14:1
Calenge, F., and Durel, C. E. 2006. Both stable and unstable QTLs for resistance to powdery mildew are detected in apple after four years of field assessments. Mol. Breed. 17:329-339.
Calenge, F., Faure, A., Goerre, M., Gebhardt, C., de Weg, W. E., Parisi, L., and Durel, C.-E. 2004. Quantitative trait loci (QTL) analysis reveals both broad-spectrum and isolate-specific QTL for scab resistance in an apple progeny challenged with eight isolates of Venturia inaequalis. Phytopathology. 94:370–379
Carmeille, A., Caranta, C., Dintinger, J., Prior, P., Luisetti, J., and Besse, P. 2006. Identification of QTLs for Ralstonia solanacearum race 3-phylotype II resistance in tomato. Theor. Appl. Genet. 113:110–121
Ceballos, H., Okogbenin, E., Pérez, J. C., López-Valle, L. A. B., and Debouck, D. 2010. Cassava. Pages 53–96 in: Root and tuber crops, Springer.
St. Clair, D. A. 2010. Quantitative disease resistance and quantitative resistance loci in breeding. Annu. Rev. Phytopathol. 48:247–268
Corwin, J. A., Copeland, D., Feusier, J., Subedy, A., Eshbaugh, R., Palmer, C., Maloof, J., and Kliebenstein, D. J. 2016. The Quantitative Basis of the Arabidopsis Innate Immune System to Endemic Pathogens Depends on Pathogen Genetics. PLoS Genet.
Dicke, M., and Hilker, M. 2003. Induced plant defences: from molecular biology to evolutionary ecology. Basic Appl. Ecol. 4:3–14
Druka, A., Potokina, E., Luo, Z., Bonar, N., Druka, I., Zhang, L., Marshall, D. F., Steffenson, B. J., Close, T. J., Wise, R. P., and others. 2008. Exploiting regulatory variation to identify genes underlying quantitative resistance to the wheat stem rust pathogen Puccinia graminis f. sp. tritici in barley. Theor. Appl. Genet. 117:261–272
El-Soda, M., Malosetti, M., Zwaan, B. J., Koornneef, M., and Aarts, M. G. M. 2014. Genotype x environment interaction QTL mapping in plants: Lessons from Arabidopsis. Trends Plant Sci. 19:390-398.
191
FAO. 2013. Save and grow: Cassava. A guide to sustainable production intensification. Food and Agriculture Organization of the United Nations, Rome.
Faris, J. D., Li, W. L., Liu, D. J., Chen, P. D., and Gill, B. S. 1999. Candidate gene analysis of quantitative disease resistance in wheat. Theor. Appl. Genet. 98:219–225
Fregene, M., Angel, F., Gomez, R., Rodriguez, F., Chavarriaga, P., Roca, W., Tohme, J., and Bonierbale, M. 1997. A molecular genetic map of cassava ( Manihot esculenta Crantz). TAG Theor. Appl. Genet. 95:431–441
Frutos-Bernal, E., and Galindo, M. P. 2012. GGEBiplotGUI: Interactive GGE Biplots in R [Programa informatico]. Salamanca, España Dep. Estadistica, Univ. Salamanca. http//cran. r-project. org/web/packages/GGEBiplotGUI/index. html.
Fu, D., Uauy, C., Distelfeld, A., Blechl, A., Epstein, L., Chen, X., Sela, H., Fahima, T., and Dubcovsky, J. 2009. A kinase-START gene confers temperature-dependent resistance to wheat stripe rust. Science (80). 323:1357–1360
Fukuoka, S., Saka, N., Koga, H., Ono, K., Shimizu, T., Ebana, K., Hayashi, N., Takahashi, A., Hirochika, H., Okuno, K., and others. 2009. Loss of function of a proline-containing protein confers durable disease resistance in rice. Science (80). 325:998–1001
Gao, M., Liu, J., Bi, D., Zhang, Z., Cheng, F., Chen, S., and Zhang, Y. 2008. MEKK1, MKK1/MKK2 and MPK4 function together in a mitogen-activated protein kinase cascade to regulate innate immunity in plants. Cell Res. 18:1190–1198
Gebhardt, C., and Valkonen, J. P. T. 2001. Organization of genes controlling disease resistance in the potato genome. Annu. Rev. Phytopathol. 39:79–102
Gordon, A., and Hannon, G. J. 2010. Fastx-toolkit. FASTQ/A short-reads preprocessing tools (unpublished) http://hannonlab. cshl. edu/fastx_toolkit.
Guimaraes, R. L., and Stotz, H. U. 2004. Oxalate production by Sclerotinia sclerotiorum deregulates guard cells during infection. Plant Physiol. 136:3703–3711
Hahn, S. K., Howland, A. K., and Okoli, C. A. 1974. Breeding for resistance to cassava bacterial blight at IITA. in: Okpala, EU; Glaser, HJ (eds.). Workshop on Cassava Bacterial Blight in Nigeria (1, 1974, Umudike, Nigeria). Proceedings.,
Hahn, S. K., Howland, A. K., and Terry, E. R. 1980. Correlated resistance of cassava to mosaic and bacterial blight diseases. Euphytica. 29:305–311
Holland, J. B. 2006. Estimating genotypic correlations and their standard errors using multivariate restricted maximum likelihood estimation with SAS Proc MIXED. Crop Sci. 46:642–654
van der Hoorn, R. A. L., and Kamoun, S. 2008. From guard to decoy: a new model for perception of plant pathogen effectors. Plant Cell. 20:2009–2017
Huard-Chauveau, C., Perchepied, L., Debieu, M., Rivas, S., Kroj, T., Kars, I., Bergelson, J., Roux, F., and Roby, D. 2013. An atypical kinase under balancing selection confers broad-spectrum disease resistance in Arabidopsis. PLoS Genet. 9:e1003766
192
Jeger, M. J., and Viljanen-Rollinson, S. L. H. 2001. The use of the area under the disease-progress curve (AUDPC) to assess quantitative disease resistance in crop cultivars. Theor. Appl. Genet. 102:32–40
Jinks, J. L., and Jones, R. M. 1958. Estimation of the components of heterosis. Genetics. 43:223
Jorge, V., Fregene, M. a., Duque, M. C., Bonierbale, M. W., Tohme, J., and Verdier, V. 2000a. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). TAG Theor. Appl. Genet. 101:865–872
Jorge, V., Fregene, M. A., Duque, M. C., Bonierbale, M. W., Tohme, J., and Verdier, V. 2000b. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). Theor. Appl. Genet. 101:865–872
Jorge, V., Fregene, M., Vélez, C. M., Duque, M. C., Tohme, J., and Verdier, V. 2001. QTL analysis of field resistance to Xanthomonas axonopodis pv. manihotis in cassava. Theor. Appl. Genet. 102:564–571
Jorgensen, T. H. 2012. The effect of environmental heterogeneity on RPW8-mediated resistance to powdery mildews in Arabidopsis thaliana. Ann. Bot. 109:833–842
Jupe, F., Pritchard, L., Etherington, G. J., MacKenzie, K., Cock, P. J. A., Wright, F., Sharma, S. K., Bolser, D., Bryan, G. J., Jones, J. D. G., and others. 2012. Identification and localisation of the NB-LRR gene family within the potato genome. BMC Genomics. 13:1
Kou, Y., and Wang, S. 2010. Broad-spectrum and durability: understanding of quantitative disease resistance. Curr. Opin. Plant Biol. 13:181–185
Kpémoua, K., Boher, B., Nicole, M., Calatayud, P., and Geiger, J.-P. 1996. Cytochemistry of defense responses in cassava infected by Xanthomonas campestris pv. manihotis. Can. J. Microbiol. 42:1131–1143
Li, Y. B., Wu, C. J., Jiang, G. H., Wang, L. Q., and He, Y. Q. 2007. Dynamic analyses of rice blast resistance for the assessment of genetic and environmental effects. Plant Breed. 126:541–547
Li, Z.-K., Luo, L. J., Mei, H. W., Paterson, A. H., Zhao, X. H., Zhong, D. B., Wang, Y. P., Yu, X. Q., Zhu, L., Tabien, R., and others. 1999. A “defeated” rice resistance gene acts as a QTL against a virulent strain of Xanthomonas oryzae pv. oryzae. Mol. Gen. Genet. MGG. 261:58–63
Liao, Y., Smyth, G. K., and Shi, W. 2014. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 30:923–930
Liu, X., Huang, B., Lin, J., Fei, J., Chen, Z., Pang, Y., Sun, X., and Tang, K. 2006. A novel pathogenesis-related protein (SsPR10) from Solanum surattense with ribonucleolytic and antimicrobial activity is stress-and pathogen-inducible. J. Plant Physiol. 163:546–556
193
Lodha, T. D., and Basak, J. 2012. Plant-pathogen interactions: what microarray tells about it? Mol. Biotechnol. 50:87–97
Lokko, Y., Gedil, M., and Dixon, A. 2004. QTLs associated with resistance to the cassava mosaic disease. Proc 4th Intl Crop Sci Congr.
Van Loon, L. C., Rep, M., and Pieterse, C. M. J. 2006. Significance of inducible defense-related proteins in infected plants. Annu. Rev. Phytopathol. 44:135–162
López, C. 2011. Descifrando las bases moleculares de la resistencia cuantitativa. Acta Biol{ó}gica Colomb. 16:3
López, C. E., and Bernal, A. J. 2012. Cassava Bacterial Blight: Using Genomics for the Elucidation and Management of an Old Problem. Trop. Plant Biol. 5:117–126
López, C. E., Quesada-Ocampo, L. M., Bohorquez, A., Duque, M. C., Vargas, J., Tohme, J., and Verdier, V. 2007. Mapping EST-derived SSRs and ESTs involved in resistance to bacterial blight in Manihot esculenta. Genome. 50:1078–1088
López, C. E., Zuluaga, A. P., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2003. Isolation of resistance gene candidates (RGCs) and characterization of an RGC cluster in cassava. Mol. Genet. genomics. 269:658–671
López, C., Soto, M., Restrepo, S., Piégu, B., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2005. Gene expression profile in response to Xanthomonas axonopodis pv. manihotis infection in cassava using a cDNA microarray. Plant Mol. Biol. 57:393–410
Lozano, J. 1986. Cassava Bacterial Blight: a Manageable Disease. Plant Dis. 70:1089–1093
Lozano, R., Hamblin, M. T., Prochnik, S., and Jannink, J.-L. 2015. Identification and distribution of the NBS-LRR gene family in the Cassava genome. BMC Genomics. 16:1
Manosalva, P. M., Davidson, R. M., Liu, B., Zhu, X., Hulbert, S. H., Leung, H., and Leach, J. E. 2009. A germin-like protein gene family functions as a complex quantitative trait locus conferring broad-spectrum disease resistance in rice. Plant Physiol. 149:286–296
Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., Dow, M., Verdier, V., Beer, S. V., Machado, M. a., Toth, I., Salmond, G., and Foster, G. D. 2012. Top 10 plant pathogenic bacteria in molecular plant pathology. Mol. Plant Pathol. 13:614–629
Mba, R. E. C., Stephenson, P., Edwards, K., Melzer, S., Nkumbira, J., Gullberg, U., Apel, K., Gale, M., Tohme, J., and Fregene, M. 2001. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: Towards an SSR-based molecular genetic map of cassava. Theor. Appl. Genet. 102:21–31
Meng, L., Li, H., Zhang, L., and Wang, J. 2015. QTL IciMapping: Integrated software for genetic linkage map construction and quantitative trait locus mapping in biparental populations. Crop J. 3:269–283
Meyers, B. C., Kozik, A., Griego, A., Kuang, H., and Michelmore, R. W. 2003. Genome-Wide Analysis of NBS-LRR–Encoding Genes in Arabidopsis. Plant Cell. 15:809–834
194
Mitchell-Olds, T. 2013. Selection on QTL and complex traits in complex environments. Mol. Ecol. 22:3427-3429.
Muñoz-Bodnar, A., Perez-Quintero, A. L., Gomez-Cano, F., Gil, J., Michelmore, R., Bernal, A., Szurek, B., and Lopez, C. 2014. RNAseq analysis of cassava reveals similar plant responses upon infection with pathogenic and non-pathogenic strains of Xanthomonas axonopodis pv. manihotis. Plant Cell Rep. 33:1901–1912
Nicotra, A. B., Atkin, O. K., Bonser, S. P., Davidson, A. M., Finnegan, E. J., Mathesius, U., Poot, P., Purugganan, M. D., Richards, C. L., Valladares, F., and others. 2010. Plant phenotypic plasticity in a changing climate. Trends Plant Sci. 15:684–692
Njenga, P., Edema, R., and Kamau, J. 2014. Combining ability for beta-carotene and important quantitative traits in a cassava F1 population.
http://dx.doi.org/10.5897/JPBCS12.069
Oh, C.-S., and Martin, G. B. 2011. Effector-triggered immunity mediated by the Pto kinase. Trends Plant Sci. 16:132–140
Perchepied, L., Dogimont, C., and Pitrat, M. 2005. Strain-specific and recessive QTLs involved in the control of partial resistance to Fusarium oxysporum f. sp. melonis race 1.2 in a recombinant inbred line population of melon. Theor. Appl. Genet. 111:65–74
Pérez-Quintero, Á. L., Quintero, A., Urrego, O., Vanegas, P., and López, C. 2012. Bioinformatic identification of cassava miRNAs differentially expressed in response to infection by Xanthomonas axonopodis pv. manihotis. BMC Plant Biol. 12:1
Petersen, M., Brodersen, P., Naested, H., Andreasson, E., Lindhart, U., Johansen, B., Nielsen, H. B., Lacy, M., Austin, M. J., Parker, J. E., and others. 2000. Arabidopsis MAP kinase 4 negatively regulates systemic acquired resistance. Cell. 103:1111–1120
Poland, J. A., Balint-Kurti, P. J., Wisser, R. J., Pratt, R. C., and Nelson, R. J. 2009. Shades of gray: the world of quantitative disease resistance. Trends Plant Sci. 14:21-29.
Quintero, A., Pérez-Quintero, A. L., and López, C. 2013. Identification of ta-siRNAs and cis-nat-siRNAs in cassava and their roles in response to cassava bacterial blight. Genomics. Proteomics Bioinformatics. 11:172–181
Rabbi, I. Y., Hamblin, M. T., Kumar, P. L., Gedil, M. A., Ikpan, A. S., Jannink, J.-L., and Kulakow, P. A. 2014. High-resolution mapping of resistance to cassava mosaic geminiviruses in cassava using genotyping-by-sequencing and its implications for breeding. Virus Res. 186:87–96
Ramalingam, J., Cruz, C. M. V., Kukreja, K., Chittoor, J. M., Wu, J.-L., Lee, S. W., Baraoidan, M., George, M. L., Cohen, M. B., Hulbert, S. H., Leach, J. E., and Leung, H. 2003. Candidate Defense Genes from Rice, Barley, and Maize and Their Association with Qualitative and Quantitative Resistance in Rice. Mol. Plant-Microbe Interact. MPMI. 16
195
Ramburan, V. P., Pretorius, Z. A., Louw, J. H., Boyd, L. A., Smith, P. H., Boshoff, W. H. P., and Prins, •. R. 2004. A genetic analysis of adult plant resistance to stripe rust in the wheat cultivar Kariega. Theor Appl Genet. 108:1426–1433
Restrepo, S. 1999. Etude de la structure des populations de Xanthomonas axonopodis pv. manihotis en Colombie. (Doctoral dissertation).
Restrepo, S., Duque, M. C., and Verdier, V. 2000. Characterization of pathotypes among isolates of Xanthomonas axonopodis pv. manihotis in Colombia. Plant Pathol. 49:680–687
Restrepo, S., Velez, C. M., Duque, M. C., and Verdier, V. 2004. Genetic structure and population dynamics of Xanthomonas axonopodis pv. manihotis in Colombia from 1995 to 1999. Appl. Environ. Microbiol. 70:255–261
Rieseberg, L. H., Archer, M. A., and Wayne, R. K. 1999. Transgressive segregation, adaptation and speciation. Heredity (Edinb). 83:372
Ripley, B. D. 2001. The R project in statistical computing. MSOR Connect. Newsl. LTSN Maths, Stats OR Netw. 1:23–25
Robinson, M. D., McCarthy, D. J., and Smyth, G. K. 2010. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 26:139–140
Roux, F., Voisin, D., Badet, T., Balagu??, C., Barlet, X., Huard-Chauveau, C., Roby, D., and Raffaele, S. 2014. Resistance to phytopathogens e tutti quanti: Placing plant quantitative disease resistance on the map. Mol. Plant Pathol. 15:427-432.
Sander, J. D., and Joung, J. K. 2014. CRISPR-Cas systems for genome editing, regulation and targeting. Nat. Biotechnol. 32:347–355
Sandino, T., López-Kleine, L., López, C., and Marqu’\inez, X. 2015. Characterization of the morphological response of susceptible and resistant varieties of cassava (Manihot esculenta Crantz) to vascular bacterial blight caused by Xanthomonas axonopodis pv manihotis. Summa Phytopathol. 41:94–100
Santaella, M., Suárez, E., López, C., González, C., Mosquera, G., Restrepo, S., Tohme, J., Badillo, A., and Verdier, V. 2004. Identification of genes in cassava that are differentially expressed during infection with Xanthomonas axonopodis pv. manihotis. Mol. Plant Pathol. 5:549–558
Schenk, P. M., Carvalhais, L. C., and Kazan, K. 2012. Unraveling plant-microbe interactions: can multi-species transcriptomics help? Trends Biotechnol. 30:177–184
Shaner, G., and Finney, R. E. 1977. The effect of nitrogen fertilization on the expression of slow-mildewing resistance in Knox wheat. Phytopathology. 67:1051–1056
Shi, W., and Shi, M. W. 2013. Package “Rsubread.”
Shivaprasad, P. V, Dunn, R. M., Santos, B. A. C. M., Bassett, A., and Baulcombe, D. C. 2012. Extraordinary transgressive phenotypes of hybrid tomato are influenced by epigenetics and small silencing RNAs. EMBO J. 31:257–266
196
Soto, J. C., Ortiz, J. F., Perlaza-Jiménez, L., Vásquez, A. X., Lopez-Lavalle, L. A. B., Mathew, B., Léon, J., Bernal, A. J., Ballvora, A., and López, C. E. 2015. A genetic map of cassava (Manihot esculenta Crantz) with integrated physical mapping of immunity-related genes. BMC Genomics. 16:190
Taylor, N. J., Fauquet, C. M., and Tohme, J. 2012. Overview of Cassava Special Issue. Trop. Plant Biol. 5:1–3
Thanyasiriwat, T., Sraphet, S., Whankaew, S., Boonseng, O., Bao, J., Lightfoot, D. A., Tangphatsornruang, S., and Triwitayakorn, K. 2014. Quantitative trait loci and candidate genes associated with starch pasting viscosity characteristics in cassava (Manihot esculenta Crantz). Plant Biol. 16:197–207
Trujillo, C. A., Ochoa, J. C., Mideros, M. F., Restrepo, S., López, C., and Bernal, A. 2014a. A Complex Population Structure of the Cassava Pathogen Xanthomonas axonopodis pv. manihotis in Recent Years in the Caribbean Region of Colombia. Microb. Ecol. 68:155–167
Trujillo, C. A., Ochoa, J. C., Mideros, M. F., Restrepo, S., López, C., and Bernal, A. 2014b. A Complex Population Structure of the Cassava Pathogen Xanthomonas axonopodis pv. manihotis in Recent Years in the Caribbean Region of Colombia. Microb. Ecol. 68: 155-167
Verdier, V., Boher, B., Maraite, H., and Geiger, J.-P. 1994. Pathological and molecular characterization of Xanthomonas campestris strains causing diseases of cassava (Manihot esculenta). Appl. Environ. Microbiol. 60:4478–4486
Verhage, A., van Wees, S. C. M., and Pieterse, C. M. J. 2010. Plant immunity: it’s the hormones talking, but what do they say? Plant Physiol. 154:536–540
Wang, J.-F., Olivier, J., Thoquet, P., Mangin, B., Sauviac, L., and Grimsley, N. H. 2000. Resistance of tomato line Hawaii7996 to Ralstonia solanacearum Pss4 in Taiwan is controlled mainly by a major strain-specific locus. Mol. plant-microbe Interact. 13:6–13
Weinig, C., and Schmitt, J. 2004. Environmental effects on the expression of quantitative trait loci and implications for phenotypic evolution. Bioscience. 54:627–635
Whankaew, S., Poopear, S., Kanjanawattanawong, S., Tangphatsornruang, S., Boonseng, O., Lightfoot, D. A., and Triwitayakorn, K. 2011. A genome scan for quantitative trait loci affecting cyanogenic potential of cassava root in an outbred population. BMC Genomics. 12:1
Wydra, K., and Verdier, V. 2002. Occurrence of cassava diseases in relation to environmental, agronomic and plant characteristics. Agric. Ecosyst. Environ. 93:211–226
Wydra, K., Zinsou, V., Jorge, V., and Verdier, V. 2004. Identification of Pathotypes of Xanthomonas axonopodis pv. manihotis in Africa and Detection of Quantitative Trait Loci and Markers for Resistance to Bacterial Blight of Cassava. Phytopathology. 94:1084–1093
197
Yan, W., Hunt, L. A., Sheng, Q., and Szlavnics, Z. 2000. Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Sci. 40:597–605
Yan, W., and Tinker, N. A. 2006. Biplot analysis of multi-environment trial data: Principles and applications. Can. J. Plant Sci. 86:623–645
Zhang, M., Pan, J., Kong, X., Zhou, Y., Liu, Y., Sun, L., and Li, D. 2012. ZmMKK3, a novel maize group B mitogen-activated protein kinase kinase gene, mediates osmotic stress and ABA signal responses. J. Plant Physiol. 169:1501–1510
Zuo, W., Chao, Q., Zhang, N., Ye, J., Tan, G., Li, B., Xing, Y., Zhang, B., Liu, H., Fengler, K. A., and others. 2015. A maize wall-associated kinase confers quantitative resistance to head smut. Nat. Genet. 47:151–157
Supplementary data
Supplementary Figure 4-1. Distribution of AUDPC values for the F1 mapping
population for each environment, Xam strain and season. ar= Arauca; lv= La Vega. In
figure it is show the AUDPC value for the parents: TMS= parental TMS30572; CM=
parental CM2177-2. Shapiro-W Test ɑ= 0,05, **AUDPC normal distributed.
198
Supplementary Table 4-1. Phenotypic responses to Xam318 and Xam681
during multi-environment evaluation. K= genotype, R= resistant, S= susceptible.
(year (2013= rainy season, 2014= dry season; Location (ara= Arauca, lv= La Vega,
gh= Greenhouse; Xam strain (318 or 681). File available at:
https://docs.google.com/spreadsheets/d/1jly7-0nviwh7kmt6eugK9j8P-X9YBpGdN-
tF-WNphxY/pubhtml
Supplementary Table 4-2. Transgressive segregants of the phenotypic
responds to Xam318 and Xam681 during multi-environment evaluation. K=
genotype, R= resistant, S= susceptible. (year (2013= rainy season, 2014= dry season;
Location (ara= Arauca, lv= La Vega, gh= Greenhouse; Xam strain (318 or 681). File
available at:
https://docs.google.com/spreadsheets/d/1GPUk7KQZUULS_WRSgRHr1QH625y6cp
dBOIXvmpTpzyU/pubhtml
199
Supplementary Table 4-3. Analysis of variance for genotype (g), environment
(location, Xam strain, season) and genotype x environment among the F1 mapping
population. (p<0.001)
Df Sum Sq Mean Sq F value Pr(>F)
Rep 4 1.81 0.45 14.56 8.52e-12 ***
Environment (E) 7 9.40 1.34 43.13 <2.2e-16 ***
Genotype (G) 116 34.5 0.29 9.56 <2.2e-16 ***
GxE 681 62.5 0.09 2.95 <2.2e-16 ***
Error 3200 99.6 0.03
Supplementary Table 4-3. Repertoire of candidate defense-related genes
identified in QTLs intervals. CBB resistance QTLs with its positions in the cassava
genetic map and in the cassava genome v4.1 and v.6.1. Candidate defense-related
genes identified in QTLs intervals with its functional annotation. File available at:
https://docs.google.com/spreadsheets/d/1uiDF9OKWSYkyoF7gvmFZA5P9adfcu2ko
0d8Ui0socBs/pubhtml
200
CHAPTER 5
"Just because I cannot see it, doesn't mean I can't believe it!"
-Jack Skellington, A Nightmare Before Christmas, 1993
201
QTL identification for cassava bacterial blight resistance under
natural infection conditions.
Johana Carolina Soto Sedano1, Rubén Eduardo Mora Moreno1, Fernando Calle2,
Camilo Ernesto López Carrascal1
1 Manihot Biotec Laboratory, Biology department, Universidad Nacional de Colombia,
Bogotá, Colombia. 2 Unidad de Mejoramiento y Genética de yuca. Centro Internacional de Agricultura
Tropical. CIAT, Palmira, Colombia
Submitted to Acta biológica Colombiana. June 2016
Abstract
Cassava, Manihot esculenta Crantz, represents the main food source for more than
one billion people. Cassava’s production is affected by several diseases, one of the
most serious is cassava bacterial blight (CBB) caused by Xanthomonas axonopodis pv.
manihotis (Xam). A quantitative trait loci (QTL) analysis for CBB resistance was
performed under natural infection conditions, using a mapping population of 99 full-
sibs genotypes highly segregant and a SNP-based high dense genetic map. The
phenotypic evaluation was carried out in Puerto López, Meta, Colombia, during the
rainy season in 2015. Both resistant and susceptible transgresive segregants were
detected in the mapping population. Through a non-parametric interval mapping
analysis, two QTLs were detected, explaining 10.9 and 12.6% of phenotypic variance
of resistance to field CBB. After a bioinformatics exploration four genes were
identified in the QTLs intervals. This work represents a contribution to the
elucidation of the molecular bases of quantitative cassava resistance to Xam.
Keywords: Xanthomonas axonopodis pv. manihotis, cassava, QTL, resistance,
molecular marker, SNPs
202
Introduction
Cassava, Manihot esculenta Crantz, is a cross-pollinated species, belong to the
Euphorbiaceae family, is a perennial shrub and its origin is the Amazon Basin (Olsen
and Schaal, 1999). The cassava diploid genome is 2n = 36 and has sexual
reproduction but for agro-economical purposes farmers use vegetative propagation
(Carvalho and War, 2002; Raji et al., 2009). Cassava is one of the most important
crops worldwide. It is the third most important source of calories in the tropics, after
rice and maize. Cassava has been considered essential in protecting food security,
especially for developing countries in Africa, Asia and Latin America (FAO, 2008).
Due to the high adaptability to drought and acid and poor soils, cassava has been
considered as an excellent alternative for an eventual world food crisis (FAO, 2008,
2013).
Brazil, Thailand, Indonesia, Angola and Ghana are the countries with the largest
cassava planted area. Colombia was ranked fifteenth in world cassava production and
third in Latin America after Brazil and Paraguay (Aguilera, 2012). In Colombia,
Departments such as Bolívar, Córdoba, Sucre, Magdalena and Meta are those with the
largest cassava planted area and production. The total production in these
Departments was more than 500 thousand tons in 2014 (http://agronet.gov.co).
Cassava, as any other crop, is affected by several diseases produced by virus, fungus,
oomycetes and bacteria (FAO, 2013). The most important bacterial disease affecting
cassava is Cassava Bacterial Blight (CBB), which is caused by the vascular and foliar
pathogen Xanthomonas axonopodis pv. manihotis (Xam). Recently, Xam was ranked as
one of the top 10 most important bacterial phytopathogens (Mansfield et al., 2012).
CBB is a devastating disease, generating significant losses, which can reach between
12 and 100% in infected fields (Lozano, 1986; López and Bernal, 2012). In Colombia,
the Xam populations are highly dynamic and diverse (Restrepo et al., 2004; Trujillo et
al., 2014).
Conventional breeding strategies have been used to address CBB but with limited
success. The most efficient strategy to manage CBB is planting resistant cultivars.
However, the knowledge of the molecular mechanisms which governs the resistance
in cassava is scarce. Nevertheless, histology and cytochemistry studies of the
resistance to CBB shows the callose depositions (Kpémoua, 1996; Sandino et al.,
2105), cell wall fortification, lignification and suberization associated with callose
203
deposition and production of flavonoids and polysaccharides as important
mechanisms of resistant cultivars in response to Xam (Kpémoua et al., 1996). Also,
different molecular approaches have conducted to identify resistance and defense
genes (López et al., 2003; López et al., 2005).
Resistance to CBB is a quantitative trait, with polygenic and additive inheritance
(Hahn et al., 1974; Jorge et al., 2000, 2001). A number of quantitative trait loci (QTL)
for resistance to CBB, with major and minor effects as well as stable and unstable
have been detected. In 2000, Jorge and coworkers reported twelve QTLs explaining
9% to 27% of the phenotypic variance. These QTLs were detected under greenhouse
conditions to five Xam strains (CIO-84, CIO-1, CIO-136, CIO-295 and ORST X-27).
Eight novel QTLs, explain between 7.2% and 18.2% of the resistance, were identified
under field conditions of natural disease pressure and during two consecutive crop
cycles (Jorge et al., 2001). Nine QTLs explaining from 16% to 33% of the phenotypic
variance to four African Xam strains were also reported (Wydra et al., 2004). More
recently, two new QTLs explaining 62% and 21% of the CBB resistance were
identified to the Xam strains CIO151 and CIO121 (López et al., 2007).
Undoubtedly the environmental conditions plays a key role in traits governed
quantitatively and even more in plant pathogen interactions (Weinig and Schmitt,
2004; Anderson et al., 2014). In fact, several studies had shown that the environment
conditions are a key factor in the cassava - Xam interaction. In particular, the
humidity favors the dispersion and proliferation of the bacteria and favors the
disease (Banito et al., 2000, 2008; Wydra and Verdier, 2002; Restrepo et al., 2004).
Thus, is essential to perform field evaluations with high disease pressure in order to
detect genetic determinants involved in CBB resistance under natural conditions
where cassava grows.
Here we report two novel QTLs for CBB field resistance, based on the evaluation of a
biparental population of 99 F1 segregating full sib progeny. These QTLs were
detected during a rainy season in 2015 in Meta, Colombia. A bioinformatics research
for genes present in the QTLs regions was carried out finding some candidate genes.
204
Materials and methods
Mapping population and field design experiment
The mapping population is a full sib F1 segregating population of 99 individuals
obtained by a cross between the Nigerian cultivar TMS30572 and CIAT’s elite cultivar
CM2177-2 (Fregene, 1997). This population has been used for several mapping
studies (Fregene et al., 1997; Jorge et al., 2000; Jorge et al., 2001; Mba et al., 2001;
Lopez et al., 2007), and for the construction of a high-density cassava genetic map
(Soto et al., 2015). Each genotype was grown from stakes at the research institute “La
Libertad” Corpoica, located in Puerto López, Meta, Colombia (4°03'40.3"N
73°27'22.5"W). This region belongs to ecozone 2 (ECZ2): a lowland tropical region in
the Colombian eastern plains (Restrepo et al., 1999; Jorge et al., 2001; Trujillo et al.,
2014). Ten plants of each parent and each genotype were planted with a density of
1m2, in an area of 1.9 ha under a complete random design. The phenotyping
evaluation was performed during July 2015 corresponding to rainy season
(www.ideam.gov.co).
Field evaluation of the response to CBB
Under a natural pressure of Xam, the disease severity was scored in ten plants by
genotype and parental at 10 months after planting, using a rating from 0 to 5, using
the symptoms scale described by Jorge et al (2001). Symptom 1=no symptoms;
2=angular leaf spots; 3=wilting of leaves; 4=dieback of one or several apices;
5=dieback of whole plan. The average of the symptoms at the observation time was
calculated for each genotype and taken as a disease index (DI). The DI of each
genotype was used for QTL analysis. The transgressive segregants were also
evaluated in the mapping population. The distribution of frequency of the DI was
tested for normal distribution by the Shapiro-Wilk test. An analysis of variance
(ANOVA) and its non-parametric test (Kruskal-Wallis) was also performed.
QTL mapping
Interval Mapping (IM) analysis with the “np” model was used for QTL detection
through R/qtl V1.37-11 (Broman, 2015). The high dense genetic map of cassava (Soto
205
et al., 2015) was employed plus a set of 2,236 GBS-SNP markers with unknown
genetic position but with known physical position in the current cassava genome
v6.1. To declare the presence of a QTL a LOD score equal or higher than 2.5 was used
as criteria. The QTL interval was established by a LOD decrease of 0.5 from the
marker peak position. Phenotypic variation explained by each QTL was determined
with calc.Rsq in R. Physical positions of the genes identified within the QTLs intervals
were established based on the SNP-based genetic map. The gene annotation was
consulted in the JGI’s Phytozome platform.
Results
Field evaluation of the response to CBB
The F1 population was planted in August, 2014. The plants were cultivated according
to the agronomical practices employed by the farmers. No control to diseases was
conducted. During the evaluation period of the response to natural disease pressure
of Xam in Meta, Colombia, in July 2015, the maximum and minimal temperatures
were 32°C and 22°C, respectively, 87% of relative humidity and a mean precipitation
of 400 mm (www.ideam.gov.co). Insects or other diseases did not attack the plants,
which grew as expected.
At the end of the productive cycle (10 months after plantation) and before collecting
the roots, the plants were scored for the presence of symptoms. Five genotypes (5%)
were symptomless, while 94 genotypes (95%) exhibit at least one of the typical
symptoms related to CBB, being the angular leaf spots the most common. Taking into
account these observations it is possible to conclude that CBB was present in the field
and in consequence it was possible to evaluate the differential responses between the
genotypes. The plants were categorized according to the presence of symptoms using
a field scale previously established (Jorge et al., 2001) and a disease index (DI) was
calculated for each genotype. The DI in the mapping population did not exhibit a
normal distribution, (P<0.05) (Figure 5-1). However, the Kruskal-Wallis test showed
significant differences (p<0.05) for the DI values obtained for the genotypes tested,
sugesting that the response to CBB is genotype-dependent (Supplementary Table 5-
1). Also both parents exhibited DIs statistically different (significant P<0.05),
indicating contrasting responses to CBB (Supplementary Table 5-2). The resistant
parental TMS30572 had a DI value of 0.2 while the DI for the susceptible parent was
0.6. The DI in the mapping population ranged from 0 to 2, with a mean of 0.75 and a
206
standard deviation of 0.45. Most IDs values were found near the average of the
sample (34 genotypes with ID=1) and very few values near the upper (4 genotypes
with ID >1.8) and lower extremes (5 genotypes with ID =0) (Fig. 1). From the 99
genotypes evaluated, 38 have equal or lower IDs values than 0.2 (ID value of the
resistant parent TMS30572), these were considered as resistant genotypes. On the
other hand, 61 genotypes exhibited equal or higher IDs values than 0.6 (ID value for
the susceptible parent CM772-14) and those were considered as susceptible.
Transgressive segregants with DIs higher or lower than the parents were identified
in the mapping population. The total trasngressive segregants were 8 for DIs lower
than 0.2 (ID value of the parent TMS30572) and 60 higher than 0.6 (ID value of the
parent CM2177-2). The genotypes g52 and g135 were the extreme genotypes for
susceptibility with ID values of 1.83 and 2, respectively. For resistance, the extreme
genotypes were g23, g89, g92, g93 and g104, which did not exhibit any symptom
related to CBB (Supplementary Table 5-2).
QTL mapping
Due to the ID values in the mapping population did not exhibit a normal distribution,
a non-parametric QTL interval mapping approach using the model “np” of R/qtl, was
applied. Based on the phenotyping evaluation of the response to CBB in the F1
population and the previous cassava genetic map developed (Soto et al., 2015) a QTL
analysis was carried out using IM. This analysis allowed the identification of two
QTLs explaining CBB field resistance. These QTLs were located in linkage groups 4
and 8 with LOD 2.5 and 2.9 respectively (Figure 5-2). The QTL in the linkage group 4
was named as QLB-4 and explains the 12.6% of the field resistance to CBB. It covers
an interval length of 2.4 cM. The interval flanking markers of QLB-4 were MB_21980
and MB_25367. The physical distance from the peak (MB_21974) to the flanking
marker MB_21980 was 317bp. While the distance from the peak marker to the
flanking marker MB_24367 could not be established due to these two markers belong
to different scaffolds in the cassava genome v4.1.
207
Figure 5-1. Histogram of the distribution of the disease index values obtained
in the field evaluation of the response to CBB. The X-axis represents the classes of
the distribution of ID values for the 99 genotypes evaluated. The Y-axis shows the
frequency of genotypes in each category. ID values of the parents are shown by the
arrows.
Figure 5-2. QTL detection for field resistance to CBB in linkage groups 4 and 8
by non-parametric interval mapping. The Y-axis indicates the LOD values and the
X-axis the linkage group with the corresponding molecular markers. The QTL peak is
shown with the SNP peak marker. The red line indicates the LOD 2.5 threshold.
208
The second QTL was located on the linkage group 8, explained 10.9% of CBB
resistance and was named as QLB-8. The QLB-8 covers an interval of 1.8 cM, whit a
peak marker matches to the SNP MB_8500 and interval flanking markers MB_2801
and MB_14991 (Table 5-1). The physical distance from the peak marker to the
flanking markers could not be established due to the two flanking markers belonging
to different scaffolds in the cassava genome v4.1.
For each QTL interval, the corresponding genomic regions were searched for the
presence of coding regions based on the new cassava genome version v6.1. The
positions of the SNPs markers MB_21974 (peak marker) and MB_21980 (flanking
marker) of QLB-4 match with the gene Manes.07G062100. According to PFAM the
annotation, this gene coded for a protein related to the vacuolar-sorting receptor 3.
The position of the other flanking marker of QLB-4 (MB_25367) matches with the
gene Manes.07G053100 which following the PFAM annotation corresponds to a serine
protease carboxypeptidase. While in the interval of QLB-8 two genes were detected:
Manes.S010100a and Manes.03G002800, which code for a C2HC zinc finger-containing
protein and for a core-2/i-branching beta-1,6-n-acetylglucosaminyltransferase
protein, respectively.
Table 5-1. Summary of QTLs detected for field resistance to CBB. The QTL name
(Q=QTL; LB=La Libertad; number of linkage group), LOD score, percentage of
phenotypic variance explicated (R2), peak marker and its position in the genetic map,
QTL interval, interval length in cM and number of genes within the intervals. ND= Not
determinate.
QTL
name
LOD
score R2
Peak
SNP Marker
Pos.
cM
QTL
Interval
Interval
length cM
Genes in
interval
QLB-4 2,5 12,6% MB_21974 98.34 MB_25367- MB_21980 2.4 2
QLB-8 2.9 10,9% MB_8500 7.21 MB_2801- MB_14991 1.8 2
Discussion
The present study evaluated the phenotypic response of 99 full-sib segregant
genotypes to CBB in field during the rainy season at Meta (Eastern plains), one of the
most productive areas of cassava in Colombia (http://agronet.gov.co). An adequate
209
high disease pressure was observed during the field evaluation. Differences in the
severity of the disease between parental genotypes, as well as differences within the
individuals of the mapping population, could be detected. Based on the phenotype
data obtained it was possible the identification of two QTLs explaining 10.9 and
12.6% of cassava resistance to Xam. In each of these QTLs regions were found two
coding genes, representing novel candidate genes for CBB resistance.
The Colombian Eastern plains belong to the ECZ2 which is characterized by savannas
of acid soils, with mean temperature of 26.1°C and mean precipitation of 400 mm per
month (Jorge et al., 2001; Ospina et al., 2002; Restrepo et al., 2004). This ECZ2 has
been described as an area with one of the highest incidence of CBB in Colombia
(CIAT, 1975; Jorge et al., 2001). Several studies have shown that Xam populations
present ecozone-differentiation as well as pathogenic specialization to the local
adapted cassava material (Restrepo and Verdier, 1997; Restrepo, 1999). Thus, the
QTLs reported here could be useful for further breeding strategies whose interest
will be developing new CBB-resistant cassava varieties highly adapted to this ECZ. It
will be important to carry out studies on the pathogen in this area in order to dissect
the current status of the presence of different Xam strains and its dynamics. The last
information available on Xam in this particular ECZ was obtained almost two decades
ago (Restrepo, 1999).
A higher number of susceptible genotypes (61.6%) compared with the susceptible
ones were identified in the mapping population. Due to the phenotypic evaluation
was performed during a rainy year, it is expected that the high humidity had
contributed to this finding. This is consistent with several studies showing that high
humidity favored the development and speed of symptoms as well Xam growing and
dispersion (Leu, 1978; Banito et al., 2000, 2001; Wydra and Verdier, 2002; Restrepo
et al., 2004).
Both, resistant and susceptible transgressive segregants were identified in the
phenotypic evaluation of the mapping population. Ten resistant transgressive were
identified. This type of segregation has been described for several crops (Whankaew
et al., 2011; Akinwale et al., 2010; Njenga et al., 2014; Thanyasiriwat et al., 2013), as
well for cassava to CBB resistance in the same mapping population used in this study
(Jorge et al., 2000, 2001). The finding of these segregants suggests the action of
additive and dominant genes for CBB resistance in the TMS30571 x CM2177-2 cross.
The transgresive genotypes and especially the extreme resistant individuals g23, g89,
g92, g93 and g104, became an important source of CBB resistance to be employed in
different cassava breeding programs.
210
In this study only two QTLs were identified. A previous QTL detection study for CBB
conducted also at ECZ2 (Jorge et al., 2001) revealed the presence of eight QTLs from
which six were stable. On greenhouse and controlled conditions with particular Xam
strains more than twenty QTLs have been identified (Jorge et al., 2000; Wydra et al.,
2004; Lopez et al., 2007). Several of these QTLs were strain-specific QTLs. In order to
evaluate the stability of the QTLs reported here, it will be necessary to perform
additional field evaluations during different years and dry seasons. Another
important aspect will be to determinate the Xam strain (s) present on the infected
plants to know if these QTLs are strain specific.
The two QTLs identified in this study cover a genetic region of 4.2 cM, 2.4 cM for QLB-
4 and 1.8 cM for QLB-8, respectively. The QTLs cover a short interval length given the
high marker density exhibited for this genetic map (Soto et al., 2015). In addition, this
genetic map was anchored to the cassava genome which allowed the identification of
the genes present in the QTLs intervals. Even though some associations of candidate
genes with QTL have been reported (Faris et al., 1999; Ramalingam et al., 2003; Liu et
al., 2004), these types of studies are scarce. Here it was possible to identify four
candidate genes in a relative short interval length. The presence of only two genes in
each of these QTLs will facilitate the number of genes to be functionally validated.
The functional annotation of the four genes present within the QTL intervals are not
directly related to known plant immunity related genes. Other studies have reported
the presence of genes coding for proteins related in plant immunity process as
pathogen perception or in signal pathways (Ramalingam et al., 2003; St. Clair, 2010;
López, 2011). However, some studies have established that the typical immunity-
related genes are only a small part of the whole genes related to plant resistance
(Corwin et al., 2016). Recently, several genes have been cloned from QTLs and none
of them correspond to classical R genes, but have different functions not directly
related with pathogen recognition or defense (Poland et al., 2009; Bryant et al., 2014;
Roux et al., 2014). Thus, the four genes here detected become new genetic factors
that may be playing an important role in CBB resistance. The functional validation of
these genes should be addressed in order to deepen the understanding of the cassava
response to Xam.
Acknowledgments
We want to thank Universidad Nacional de Colombia and COLCIENCIAS for the PhD
scholarship call 528 of the author Johana Soto. We would like to extend our gratitude
211
to the “Unidad de Mejoramiento y Genética”- CIAT International Center for Tropical
Agriculture (CIAT) for enabling the plant material and to the staff of Corpoica La
libertad for the support during the field experiments.
References
Aguilera, M. 2012. La yuca en el Caribe colombiano: De cultivo ancestral a agroindustrial. Banco de la República de Colombia
Akinwale, M. G., Aladesanwa, R. D., Akinyele, B. O., Dixon, A. G. O., and Odiyi, A. C. 2010. Inheritance of-carotene in cassava (Manihot esculenta crantza). Int. J. Genet. Mol. Biol. 2:198–201
Anderson, J. T., Wagner, M. R., Rushworth, C. A., Prasad, K., and Mitchell-Olds, T. 2014. The evolution of quantitative traits in complex environments. Heredity (Edinb). 112:4–12
Banito, A., Kpemoua, K. E., and Wydra, K. 2008. Expression of resistance and tolerance of cassava genotypes to bacterial blight determined by genotype x environment interactions. J. Plant Dis. Prot. 115:152–161
Banito, A., Kpémoua, K. E., Wydra, K., and Rudolph, K. 2001. Bacterial blight of cassava in Togo: its importance, the virulence of the pathogen and the resistance of varieties. Pages 259–264 in: Plant Pathogenic Bacteria, Springer.
Broman, K. W. 2015. R/qtlcharts: interactive graphics for quantitative trait locus mapping. Genetics. 199:359–361
Bryant, R. R. M., McGrann, G. R. D., Mitchell, A. R., Schoonbeek, H., Boyd, L. A., Uauy, C., Dorling, S., and Ridout, C. J. 2014. A change in temperature modulates defence to yellow (stripe) rust in wheat line UC1041 independently of resistance gene Yr36. BMC Plant Biol. 14:1
De Carvalho, R., and Guerra, M. 2002. Cytogenetics of Manihot esculenta Crantz (cassava) and eight related species. Hereditas. 136:159–168
St. Clair, D. A. 2010. Quantitative disease resistance and quantitative resistance loci in breeding. Annu. Rev. Phytopathol. 48:247–268
CIAT (Centro International de Agricultura Tropical).Annual report. 1975. Cali, Colombia.
Corwin, J. A., Copeland, D., Feusier, J., Subedy, A., Eshbaugh, R., Palmer, C., Maloof, J., and Kliebenstein, D. J. 2016. The quantitative basis of the Arabidopsis innate immune system to endemic pathogens depends on pathogen genetics. PLoS Genet. 12:e1005789
212
Faris, J. D., Li, W. L., Liu, D. J., Chen, P. D., and Gill, B. S. 1999. Candidate gene analysis of quantitative disease resistance in wheat. Theor. Appl. Genet. 98:219–225
FAO. Oficina de prensa. Yuca para la seguridad alimentaria y energética. 2008 http://www.fao.org/newsroom/es/news/2008/1000899/index.html.
FAO. 2013. Panorama de la seguridad alimentaria y nutricional de América Latina y el Caribe. Santiago: Oficina Regional para América Latina y el Caribe de FAO. Food Agric. Organ. United Nations, Rome.
Fregene, M., Angel, F., Gomez, R., Rodriguez, F., Chavarriaga, P., Roca, W., Tohme, J., and Bonierbale, M. 1997. A molecular genetic map of cassava (Manihot esculenta Crantz). TAG Theor. Appl. Genet. 95:431–441
Hahn, S. K., Howland, A. K., and Okoli, C. A. 1974. Breeding for resistance to cassava bacterial blight at IITA. in: Okpala, EU; Glaser, HJ (eds.). Workshop on Cassava Bacterial Blight in Nigeria (1, 1974, Umudike, Nigeria). Proceedings.
Jorge, V., Fregene, M. a., Duque, M. C., Bonierbale, M. W., Tohme, J., and Verdier, V. 2000. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). TAG Theor. Appl. Genet. 101:865–872
Jorge, V., Fregene, M., Vélez, C. M., Duque, M. C., Tohme, J., and Verdier, V. 2001. QTL analysis of field resistance to Xanthomonas axonopodis pv. manihotis in cassava. Theor. Appl. Genet. 102:564–571
Kpémoua, K., Boher, B., Nicole, M., Calatayud, P., and Geiger, J.-P. 1996. Cytochemistry of defense responses in cassava infected by Xanthomonas campestris pv. manihotis. Can. J. Microbiol. 42:1131–1143
Lopez, C. 2011. Descifrando las bases moleculares de la resistencia cuantitativa. Acta Biológica Colomb. 16:3
Lopez, C. E., Quesada-Ocampo, L. M., Bohorquez, A., Duque, M. C., Vargas, J., Tohme, J., and Verdier, V. 2007. Mapping EST-derived SSRs and ESTs involved in resistance to bacterial blight in Manihot esculenta. Genome. 50:1078–1088
Lopez, C. E., Zuluaga, A. P., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2003. Isolation of resistance gene candidates (RGCs) and characterization of an RGC cluster in cassava. Mol. Genet. genomics. 269:658–671
Lopez, C., Soto, M., Restrepo, S., Piégu, B., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2005. Gene expression profile in response to Xanthomonas axonopodis pv. manihotis infection in cassava using a cDNA microarray. Plant Mol. Biol. 57:393–410
Lozano, J. 1986. Cassava Bacterial Blight: a Manageable Disease. Plant Dis. 70:1089–1093
Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., Dow, M., Verdier, V., Beer, S. V., Machado, M. a., Toth, I., Salmond, G., and Foster, G. D. 2012.
213
Top 10 plant pathogenic bacteria in molecular plant pathology. Mol. Plant Pathol. 13:614–629
Mba, R. E. C., Stephenson, P., Edwards, K., Melzer, S., Nkumbira, J., Gullberg, U., Apel, K., Gale, M., Tohme, J., and Fregene, M. 2001. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: Towards an SSR-based molecular genetic map of cassava. Theor. Appl. Genet. 102:21–31
Njenga, P., Edema, R., and Kamau, J. 2014. Combining ability for beta-carotene and important quantitative traits in a cassava F1 population.
Olsen, K. M., and Schaal, B. a. 1999. Evidence on the origin of cassava: phylogeography of Manihot esculenta. Proc. Natl. Acad. Sci. U. S. A. 96:5586–5591
Ospina, B., and Ceballos, H. 2002. La yuca en el tercer Milenio: Sistemas Modernos de producción, procesamiento, utilización y comercialización. CIAT.
Poland, J. A., Balint-Kurti, P. J., Wisser, R. J., Pratt, R. C., and Nelson, R. J. 2009. Shades of gray: the world of quantitative disease resistance. Trends Plant Sci. 14:21–29
Raji, A. A. J., Anderson, J. V, Kolade, O. A., Ugwu, C. D., Dixon, A. G. O., and Ingelbrecht, I. L. 2009. Gene-based microsatellites for cassava (Manihot esculenta Crantz): prevalence, polymorphisms, and cross-taxa utility. BMC Plant Biol. 9:1
Ramalingam, J., Vera Cruz, C. M., Kukreja, K., Chittoor, J. M., Wu, J.-L., Lee, S. W., Baraoidan, M., George, M. L., Cohen, M. B., Hulbert, S. H., and others. 2003. Candidate defense genes from rice, barley, and maize and their association with qualitative and quantitative resistance in rice. Mol. Plant-Microbe Interact. 16:14–24
Restrepo, S. 1999. Etude de la structure des populations de Xanthomonas axonopodis pv. manihotis en Colombie.
Restrepo, S., Velez, C. M., Duque, M. C., and Verdier, V. 2004. Genetic structure and population dynamics of Xanthomonas axonopodis pv. manihotis in Colombia from 1995 to 1999. Appl. Environ. Microbiol. 70:255–261
Restrepo, S., and Verdier, V. 1997. Geographical Differentiation of the Population of Xanthomonas axonopodis pv. manihotis in Colombia. Appl. Environ. Microbiol. 63:4427–4434
Roux, F., Voisin, D., Badet, T., Balagué, C., Barlet, X., Huard-Chauveau, C., Roby, D., and Raffaele, S. 2014. Resistance to phytopathogens e tutti quanti: placing plant quantitative disease resistance on the map. Mol. Plant Pathol. 15:427–432
Sandino, T., López-Kleine, L., López, C., and Marquinez, X. 2015. Characterization of the morphological response of susceptible and resistant varieties of cassava (Manihot esculenta Crantz) to vascular bacterial blight caused by Xanthomonas axonopodis pv manihotis. Summa Phytopathol. 41:94–100
Soto, J. C., Ortiz, J. F., Perlaza-Jiménez, L., Vásquez, A. X., Lopez-Lavalle, L. A. B., Mathew, B., Léon, J., Bernal, A. J., Ballvora, A., and López, C. E. 2015. A genetic map
214
of cassava (Manihot esculenta Crantz) with integrated physical mapping of immunity-related genes. BMC Genomics. 16:190
Thanyasiriwat, T., Sraphet, S., Whankaew, S., Boonseng, O., Bao, J., Lightfoot, D. A., Tangphatsornruang, S., and Triwitayakorn, K. 2014. Quantitative trait loci and candidate genes associated with starch pasting viscosity characteristics in cassava (Manihot esculenta Crantz). Plant Biol. 16:197–207
Trujillo, C. A., Ochoa, J. C., Mideros, M. F., Restrepo, S., López, C., and Bernal, A. 2014. A Complex Population Structure of the Cassava Pathogen Xanthomonas axonopodis pv. manihotis in Recent Years in the Caribbean Region of Colombia. Microb. Ecol.
Weinig, C., and Schmitt, J. 2004. Environmental effects on the expression of quantitative trait loci and implications for phenotypic evolution. Bioscience. 54:627–635
Whankaew, S., Poopear, S., Kanjanawattanawong, S., Tangphatsornruang, S., Boonseng, O., Lightfoot, D. A., and Triwitayakorn, K. 2011. A genome scan for quantitative trait loci affecting cyanogenic potential of cassava root in an outbred population. BMC Genomics. 12:1
Wydra, K., and Verdier, V. 2002. Occurrence of cassava diseases in relation to environmental , agronomic and plant characteristics. Agric. Ecosyst. Environ. 93:211–226
Wydra, K., Zinsou, V., Jorge, V., and Verdier, V. 2004. Identification of Pathotypes of Xanthomonas axonopodis pv. manihotis in Africa and Detection of Quantitative Trait Loci and Markers for Resistance to Bacterial Blight of Cassava. Phytopathology. 94:1084–1093
Supplementary data
Supplementary Table 5-1. CBB disease index values of the mapping population
evaluated under natural conditions of infection. Disease index obtained from 99
genotypes during field evaluation (Meta, Colombia). The additional file is available at:
https://drive.google.com/file/d/0B_L_gXNSRAr1MWloYUFBRVYtUjA/view?usp=shar
ing
Supplementary Table 5-2. T test of the response (disease index) of mapping
population. Parents exhibited indicating statistically different contrasting responses
P<0.05. The additional file is available at:
https://docs.google.com/spreadsheets/d/1LJzkdfLvWiih801gW3wE8byrNbcZwaeZt
PDNAOdjuzo/pubhtml
215
216
General discusion
Improving CBB resistance: a general diagnostic
In the last decade, according to the FAO cassava has emerged as one of the most
important crops, after rice, wheat and maize. The demand of this crop has increased
dramatically, not only for the millions of people in developing countries who base
their diet on this crop, but also for its use in the industry (FAO, 2015). However the
production, and thus the supply of this product, is under threat as consequence of
devastating diseases such as CBB. Notwithstanding its importance, cassava is a crop
that has been scarcely studied, including the molecular mechanisms related with CBB
resistance.
A century ago CBB was first described as one of the most devastating diseases of
cassava (Lozano, 1986). Xam, its causative agent, has been considered a quarantine
pathogen. However, new reports of CBB have emerged over time, being one of the
most recent those from Burkina-Faso (Wonni et al., 2015). CBB has also been
described as a disease highly influenced by the environment (Banito et al., 2001).
Xam populations have been described as highly diverse (Restrepo et al., 2004;
Trujillo et al., 2014). This scenario makes the study of this disease a challenge,
requiring the isolation of several genes conferring resistance to the large repertoire
of Xam strains reported.
Undoubtedly there is a need for detailed knowledge of the pathogen populations, the
effect of the environment on the resistance response and the inherent genetic factors
governing resistance. All this information should be integrated in order to generate
novel and innovative strategies to improve CBB resistance through complementary
traditional and molecular breeding approaches.
For years, new cassava varieties with different levels of CBB resistance have been
obtained through conventional breeding in different countries (Russell, 2013).
However, these materials are not necessarily adapted to different regions. Thus it will
be necessary to implement resistance evaluations in multi-environments into the
breeding schemes. On the other hand, the knowledge of Xam populations, diversity
and dynamics has been rarely taking into account within the cassava breeding
programs. As a consequence there is a lack of knowledge concerning the level of
resistance of the new materials developed to the spectrum of Xam strains, making
impossible to predict the level and durability of resistance in these materials.
217
In this thesis, through a combination of approaches including mapping, QTL and
transcriptomic, analysis we sought out to get a closer to the genetic factors governing
CBB resistance. A special attention was given to the effect of the environmental
conditions on CBB responses to two Xam strains evaluated on the F1 population used
to construct the cassava genetic map and on the QTL detection. Once the QTLs will be
validated, they could be successfully integrated into cassava breeding programs
which aim to develop adapted cassava varieties resistant to CBB
The bottlenecks: molecular markers in the past, phenotypic data nowadays
Different approaches have been developed for the detection of the genetic factors
governing a quantitative trait. The most popular is the QTL mapping and also more
recently the GWAS (Korte and Farlow, 2013). The idea behind these approaches is to
understand the dynamics between allele variation and how it influences the
phenotype. Thus, both genotype and phenotype data are the bedrock of these sort of
analysis.
In the QTL linkage mapping approach, solid statistical analysis such as interval
mapping and composite interval mapping allow the association between molecular
markers localized on the genetic map with the phenotypic data obtained for a
particular trait in the same population (Collard et al., 2005). However in the past the
big challenge was to obtain high volumes of genotypic (molecular markers) data.
Thus there was always a particular interest in the generation of high number of
molecular markers which could be employed for the construction of genetic maps.
The first generation of molecular markers included RFLPs, RAPDs, AFLPs and SSRs
among others. With high effort (time-consuming, low high-throughput possibilities)
hundreds of these markers were obtained for a population and positioned on genetic
maps. In addition these markers were not widely distributed in the genome, were
anonymous and come from unpractical techniques in terms of cost and time.
However, they were used successfully for the development of the first genetic maps
and for QTL detection purposes in the most important economical species (McCouch
et al., 1988; Graner et al., 1991; Reinisch et al., 1994).
The limitations of the above-mentioned molecular markers were overcome with the
generation of the new and massive sequencing technologies (Ansorge, 2009) which
revolutionized the ability not only of obtain complete genome sequences but also to
detect variations in DNA through high-throughput genotyping. Nowadays it is
218
possible to obtain thousands of SNPs markers with a precise position on the genome
in a relatively short period of time and on large size of population (Elshire et al.,
2011). This growth in the capacity to obtain genotypic data raised a big challenge on
the ability to store and to analyze high volume of information. The demand of
specialized programs used for genetic mapping and QTL detection as well as
algorithms able to perform genetic analysis of these volumes of data tends to grow.
Nevertheless, new algorithms and statistical models are in constant development to
improve the accurate and speed of genetic analysis based on DNA high-throughput
data.
The high-throughput genotyping also consolidated a new generation of more
accurate genetic maps. Today we have the possibility to integrate large set of
molecular markers for the development of saturated genetic maps. This translates
into a greater genetic resolution, short gaps between non-anonymous molecular
makers, more efficient computing processes and a diminution in price. In fact, in this
thesis a genotyping by sequencing approach was applied to the mapping population.
In less than a year we were able to generate more than 78,000 SNPs for 150
segregating individuals, covering around 87% of the current cassava genome. Based
on this genotypic data we made a big jump concerning the construction of cassava
genetic maps producing one of the most saturated maps reported so far. The cassava
genetic map here reported increased the number of markers from a few hundred of
anonymous markers with big gaps between them (Fregene et al., 1997; Jorge et al.,
2000; Mba et al., 2001) to 2,141 SNPs markers with known positions in the genome
and an average distance of 1.26 cM between markers.
Moreover, with the availability of the cassava genome sequence, we were able to
assemble the genetic map into a genome-wide physical map, as a first attempt to give
an order to the sequence of the cassava genome v.4.1, which at that time was
represented by thousands of scaffolds. Also, this map increased the last version of the
cassava map (Rabbi et al., 2014) in 30.7Mb, leading 64% of the genome sequence
draft v.4.1 anchored to a genetic map. Thus, the high-throughput genotyping used in
this work allowed us to develop a genetic map that represents a great contribution in
different ways for the cassava scientific community. This map can be exploited to
identify specific genotypes through the presence, absence and inheritance of
particular molecular markers associated with a particular trait within populations.
This map can also be employed as a reference in future studies of the cassava genome
evolution, contributing to detect chromosome rearrangements, intraspecific genome
duplications and search for syntenic regions between species. In addition this map
can be exploited for further comparative genomic studies with other cassava
populations. One of the most important practical contributions of the genetic map
219
developed during this work is the fact that it constitutes a fundamental element for
gene cloning by positional mapping in cassava.
Positional gene cloning requires the molecular markers delimiting the QTL are
separated for reduced intervals. One of the ways to reduce this interval is through a
fine mapping approach (Lynch et al., 1998). Nevertheless, since this approach
requires the production of populations with thousands of recombinant individuals, it
is not an expeditious strategy to be followed in cassava given it has a low seed
production and low germination rate (Hahn et al. 1973) (Fregene et al., 2001).
Another way to reduce the QTL intervals is through the use of high dense genetic
maps. In the development of this kind of maps, the greater the size of the mapping
population, the greater number of recombinants that can be detected, and thus the
map resolution will increase. Here we were able to integrate a good number of
molecular markers into the map, however this number was affected by our mapping
population size. As a consequence a number of markers not were integrated to the
map. Nevertheless the resolution achieved in this version of the cassava genetic map
was enough to detect 18 strain-specific QTLs with low sizes intervals ranged from 1
to 7 cM; in which relatively few number of CBB candidate defense-related genes
(CDRGs) were found, ranged from 1 to 40 for a single QTL interval. In the 18 QTLs
which represents a total map region of 38 cM were identified a repertoire of 151
CDRGs. Moreover, this dense genetic map has already been used for the dissection of
other cassava traits such as the cassava vegetal architecture (Mora et al., 2016).
Despite the easiness in obtain genotypic data useful to construct high dense genetic
maps there is a contrast with the rate in the generation of phenotypic information.
This unbalance can affect the accurate interpretation of how a change in the
phenotype is the result of the variation in loci and its subsequent identification; since
this interpretation requires both genotype and phenotype data, as high and precise as
possible. To achieve significant levels of precision in the QTL detection it is necessary
the assessment of large number of plants and its biological replicates. This
constitutes a big challenge if we consider the areas needed for these evaluations and
the time required. Here for example for QTL linkage mappping, the defense responses
were evaluated for more than 100 cassava materials with its respective replicates
and mocks, against two Xam strains, under multi-environments and two seasons,
required the generation and visual inspection of more than 9,000 plants at five
different time points. In the same way, for AM the defense responses have to be
evaluated in panels of at least 100 diverse accessions also with its respective
replicates and mocks. Each accession could probably have its own rate growth and
requirements that have to be taken into account for the phenotyping evaluation. All
220
this phenotyping process implied different types of efforts, including field areas, big
budget, agricultural supplies, human work and a lot of time.
CBB resistance is a quantitative trait highly influenced by the environment (Banito et
al., 2001). In fact, in this study were detected 18 strain-specific QTLs, from which half
of them were stable between the rainy and dry seasons. Also it was described a
genotype and QTL by environment interaction. These results highlight and confirm
the important effect of the environmental conditions on the response to Xam
infection and thus in the QTL detection. Therefore, under this scenario, the evaluation
of the resistance to CBB becomes a bigger challenge, because it requires evaluations
in multi-environment and multi-Xam strains assessments. One alternative to
overcome this challenge is to perform the CBB resistance phenotyping evaluation in
cassava seedlings using growth chambers simulating different environmental
conditions and with different Xam strains. The uses of the growth chambers for
resistance phenotyping have been successfully applied in other pathosystems
(Dannon et al., 2004; Hao et al., 2015). Several advantages and constraints can be
considered with this system. The detection of potentially resistant genotypes will be
easier and faster and obtaining lights about the behavior of a genotype can be
conducted in less time than would be required if the assessment take place under
field conditions. The phenotyping can be achieved through monitoring of symptoms
and measuring the bacterial growth. However, even if it is possible to obtain
phenotypic data it will be important to consider that the growth chamber conditions
will not be never exactly the same when the genotype is under natural conditions.
Despite that this system can simulate parameters such as temperature, humidity,
photoperiod, light intensity, etc, it cannot take into account other natural
environment factors playing a role in the plant defense responses, such as wind
speed, precipitations, and interactions with soil microbiota.
The application of approaches such as MAS and GS which aim to reduce the
phenotyping steps and give more strength to the genotype predictive value (Heffner
et al., 2009) are alternative to avoid the big efforts to phenotyping. To achieve this,
the best molecular markers associated with the trait are selected to screen directly a
target population. Thus it is to expect that the genotypes carrying the selected
molecular markers contain also the trait of interest. This assumption is that the
molecular markers are genetically linked to the QTLs that govern the trait. For this
strategy a DNA extraction of the materials to be evaluated is followed by the
detection of the molecular markers by a PCR to determinate the presence/absence of
the corresponding marker. On the other hand, GS can estimate the phenotype from a
genomic perspective based on a prediction equation, which is obtained from the
whole-genome DNA polymorphisms, and previous phenotypic data collected from
221
training populations. Notwithstanding, the power of this approach resides in an
initial high quality phenotyping of several training populations under different
environments. Only when it is possible to get this information the predictive model
will be enough accurate to define a real association between genomic regions and the
trait. Therefore once again, the high challenge and the current bottleneck is the
development of efficient methods for phenotyping, but in this case only for training
populations.
An alternative strategy is the use of high-throughput phenotyping technologies
(Araus and Cairns, 2014). The use of complex image machines, the use of geographic
information systems (GIS), screenings coupled to microscopes are some of the
technologies, which have been proposed. Several high-throughput phenotyping
technologies have been applied successfully for the detection of pathogens and
symptoms caused during infection in several crops (Mahlein et al., 2012a, 2012b,
2013; Fahlgren et al., 2015). However, this is not the case for the diseases affecting
cassava such as CBB. In consequence before to implement these technologies it is
necessary to standardize the procedures for this pathosystem. Some parameters can
be explored, for example the reduction on the photosynthetic ratio which can be
measure in a high-throughput manner (Fahlgren et al., 2015), the diminution in foliar
area or the detection of necrotic areas (Stewart and McDonald, 2014). The access to
these technologies continues to be limited because some of them are based on
sophisticated machines (Cabrera et al., 2012). It will be important to note that not all
the laboratories and breeding programs can have access to these technologies. This
point is even more complex for cassava, which is a crop cultivated mainly for small
farmers and where the research is, in many of the cases, conducted in developing
countries. Thus, it will require greater efforts to develop novel, accurate, fast and
affordable disease diagnostic methods for CBB.
Identifying and validating the most promising CBB-CDRG
The search for genetic factors involved in CBB resistance has been of great interest
for the cassava scientific community since long ago, arguably almost from the time
when the disease was described. Different efforts have been achieved to identify the
genomic regions that govern the resistance of this devastating disease (Jorge et al.,
2000b, 2001; Lopez et al., 2003). However, the identification of these factors has
proven to be a big challenge, basically because as any other quantitative trait, the
governing loci have usually small effect and is highly conditioned by the environment.
222
The main aim of this thesis was to search for these factors through the construction,
in a first time, of a dense genetic map built with thousands of SNPs markers obtained
by GBS. In a second time, the response evaluation to CBB in the segregating mapping
population was performed in multi-environments (two localities during rainy and
dry seasons and under greenhouse conditions). The response was followed in the F1
segregating population after inoculation with two Xam strains and also under natural
pressure of the pathogen. All this work allowed us to identify novel QTLs and, more
important, a repertoire of 151 CDRGs with a putative role during the cassava defense
response to CBB. From these genes four showed gene differential expression in the
resistant parent during Xam681 infection. The detected QTLs explained between 10.9
to 22.1% of the phenotypic variance of resistance to Xam.
Selecting the best candidates
Since the QTLs detected are the result of linkage mapping, it is very probable that not
all of the CDRGs that reside in the QTL interval are involved in the resistance to CBB.
Therefore it is necessary to “reduce the interval” where these genes are located and
determine which of them are the most promising. Different alternatives can be
applied for the selection of CDRGs. The first one is to select directly those genes that
co-localized with QTLs which satisfies one or more of the following criteria. i) Explain
the higher percentage of phenotypic variance. It is to expect that these loci represent
responsible gene alleles, which in a greater manner confer the defense response to
Xam. Thus the genes located in these QTLs could be involved in pathogen recognition
or playing a pivotal role in an immunity pathway. ii) Genes localized on stable QTLs.
The effect in conferring defense response to Xam of these genes is strong and is not
strongly affected by the environment. In consequence these genes have the potential
to be used in breeding programs and MAS strategies with multi-environment
purposes. iii) Genes coding for IRP. Despite it has been reported that several proteins
lacking of the classical domains present in R proteins can have an important roles in
the plant defense response, a selection of genes coding for proteins containing NB or
LRR domains and co-localizing with a resistance QTL is a straight alternative. iv)
CDRGs showing a differential expression during Xam infection and displaying a
contrast in gene expression (induced in one case and repressed in other) between
resistant and susceptible genotypes can be privileged. The differential gene
expression of these candidates can be taken as a clue of its role in the cassava defense
response.
223
Confirming the CDRGs
Once established the best CDRGs, the next step is to confirm its role in CBB
resistance. The most expedites strategies to accomplish this goal could are: the
application of AM, the search for polymorphisms in a panel of cassava accessions to
associate them with the phenotype and the analysis of the gene expression in several
resistant/susceptible backgrounds against different Xam strains.
The use of AM allows detecting a direct association between the CBB resistance trait
and particular SNPs markers (possibly corresponding to CDRGs). One advantage of
AM is its possibility to increase the level of resolution. The AM analysis is based on
the exploration of the historical recombination events at population level allowing
the exploration of a large number of recombination events taking place through all
the genome and during large periods of time (Zhu et al., 2008). In contrast, the
linkage mapping analysis has a limited detection resolution because it depends on
the number of recombinants present in the biparental population studied. Thus, the
genetic map developed, and therefore the QTL linkage mapping analysis applied, in
this study had a limited detection resolution, exemplified for the QTL intervals of
more than 2 cM. To increase the resolution of the map and QTL detection by the
employment of the same genetic map it will be necessary to increase the number of
segregating individuals and thus cover a greater number of recombination events. As
was mentioned in the previous section, this represents a difficult task considering the
cassava reproductive biology. The AM approach offers a promising alternative. The
strategy to follow will be performing a high-throughput genotyping of a diverse panel
of cassava accessions, including the parentals TMS30572 and CM21772 employed to
construct the cassava genetic reported in this work. This will offer information
concerning to the polymorphisms present in the genes localized in the QTLs reported
in this study. At the same time, it will be imperious to conduct an evaluation of the
phenotypic response to Xam infection for these cassava accessions. In this way it will
be possible to determine the level of significance in the association of the allelic
variability of the SNPs, contained in the CDRGs or in other genomic regions, with the
variability of the cassava phenotypic response to Xam. The information generated in
this study will guide the primary sources of genes to direct the search of
polymorphisms, which will be the first task to be accomplished. In a second time, but
not in an exclusive manner, other genes or sequences (promoters, for instance) can
be screened to identify associations. In this schema the AM and QTL mapping are
considered as complementary approaches and one can guide or direct the sampling
area to particular genomic regions (Mammadov et al., 2011). The limiting factor to
accomplish the ambitious goal of detecting association between DNA sequences and a
224
complex trait is the generation of phenotypic data. It was reiteratively stressed in the
previous section. The challenge now is to develop new tools to evaluate in a rapid and
accurate way the cassava resistance to Xam. This aspect takes more relevance in the
actual context where the genome of several cassava accessions and related species is
available (Bredeson et al., 2016). The exploitation of this information can be only
accomplished with phenotypic data. For the particular case of CBB resistance it will
be important not only to develop phenotypic analysis for a restraint number of Xam
strains or in a particular locality, but should be extended to the more representative
and stable Xam strains in different Colombian regions. All this information will help
us to determine the G x E interactions and to detect eventual strain-specific marker-
trait associations.
The gene expression analysis of the CDRGs in a resistant (female parent) background
allowed us to identify four genes showing differential expression during Xam681
infection. The gene expression analysis has allowed for decades to identify and
quantified new and known transcripts related to plant defense for many diseases of
economically important crops (Matsumura et al., 2003; Soto and López, 2014).
Accordingly, it will be interesting to evaluate the gene expression profiles of the four
CDRGs differentially expressed as well as the complete CDRGs repertoire in a
susceptible background (male parent) and against different Xam strains. These
profiles will help to establish more accurately the dynamic of expression of these
genes on a resistance and susceptible host; as well as provide information regarding
to strain-specific gene responses.
Functional validation
Once identified the most relevant CDRG within the QTL intervals, several possibilities
for its functional gene validation can be implemented. The loss-of-function and gain-
of function strategies can be envisaged. The list includes gene down expression
through interfering RNA (RNAi), virus-induced gene silencing (VIGS), employment of
artificial microRNAs and genome edition using the new CRISPR/Cas9 system
(Hamilton and Baulcombe, 1999; He and Hannon, 2004; Burch-Smith et al., 2004; Ran
et al., 2013). In cassava there are not mutant collections, which will offer a rapid and
efficient opportunity to study the function of particular genes. These mutants exist
for Arabidopsis and rice and have been largely employed (Anderson et al., 1992;
Hirochika et al., 2004). The VIGS system to knock down genes is broadly used in plant
species of the Solanaceae family (Liu et al., 2002; Ryu et al., 2004; Unver and Budak,
225
2009) and in fewer cases for other plant species (Liu et al., 2002b; Robertson, 2004).
In cassava a VIGS system has been reported but its use has been very restraint
(Fofana et al., 2004). The remaining alternatives of loss-of-function (RNAi, use of
artificial microRNAs and genome edition), as well the gain-of-function are all of them
dependent of a transformation system. Some consideration should be taking into
account before to follow one of these approaches.
Genetic transformation has been achieved in different cassava cultivars but with a
very low efficiency (Liu et al., 2011). Currently the cultivar employed for
transformation is cv. 60444 (TMS60444 or NGA11 in CIAT germoplasm collection),
which has become the model given the high efficiency in transformation and
regeneration (Sayre et al., 2011). The cv. 60444 is a West African cultivar not broadly
employed for farmers. Concerning to CBB resistance, it has been established that this
genotype is susceptible to most of the Xam strains, including Xam318 and Xam681
(Trujillo et al., 2014). The use of the loss-of-function approach by genetic
transformation in 60444 will produce more susceptible plants if a CDRG is effectively
involved in CBB resistance. In this case, once obtained these highly susceptible plants
could be difficult to obtain and/or to study. Therefore, the alternative is employing a
gain-of function approach through overexpression, as it has already been
implemented in our laboratory (Diaz et al., 2016, unpublished results). It is expected
that the overexpression of a particular CDRG in a susceptible cultivar as cv.60444
show a reduction or a delayed onset of CBB symptoms after Xam infection. For this
strategy the principle would be to clone the coding DNA sequence of the CDRG in a
binary vector under a strong constitutive promoter, for example the cauliflower
mosaic virus (35S). Then this construction could be used for Agrobacterium
tumefaciens-mediated genetic transformation of friable embryogenic callus (FEC) of
cv. 60444. After a molecular characterization of several lines, showing only a unique
insertion event and an accurate level of expression of the transgene, these can be
employed to study the effect on CBB resistance. If the candidate gene is involved in
resistance it is expected a reduction of symptoms on the cassava plants after Xam
inoculation. Classically and for the case of qualitative resistance it is achieved through
the observation of an Hypersensitive Response (HR). However in cassava this HR has
never been observed and taking into account the quantitative nature of CBB
resistance it is not a real possibility. Alternatively, and for the particular case of this
disease resistance, it will be possible to measure the effect as a decreasing or
retarding on the development of symptoms. However, given that the CDRGs
correspond to QTLs, the value of resistance expected to be transfer into the candidate
gene-overexpressed plants, should be in the order of the percentage of the
phenotypic variance explained by the QTL. In this study for example it was possible
to identify QTL with a phenotypic variance explaining 22.1% and 18.8%, (QLV681RD-
226
6 and QGH318-19 respectively). This implies that the CDRG-transgene may confer
limited quantitative resistance to the disease. The question raise here is how to
quantify the effect in reduction or retard of CBB symptoms for genomic regions
contributing in these percentages to CBB resistance. Some alternatives can be
considered. In a first time, the AUDPC is an excellent way to follow the speed and
intensity in the development of symptoms. A reduction on the AUDPC will represent
a good hint of the function of the candidate gene. This strategy implies to produce
adult plants in a relative high number representing a time-consuming task. The
production of stem cuttings from transgenic plants can take more than two or three
years. In this sense it will be important to test the possibility to conduct an evaluation
of “in vitro” plants and develop a new AUDPC scale based on symptoms observed on
these types of plants. In consequence, it will be imperative to validate the in vitro
system to study CBB resistance. However, considering the important effect of the
environment on the CBB response, the in vitro system should be considered as an
initial step. It would be convenient that all phenotypic evaluation of the regenerated
transgenic plants should be conducted under multi-environmental conditions
(locality and strains), including those conditions under which the QTLs that contains
the CDRGs were identified. In this way it will be possible to evaluate the environment
influence on the effectiveness of the transgene to reduce the susceptibility in the
transgenic plant. Other possibility is to quantify the bacterial growth in plants after
inoculation. This strategy offers the possibility to have a quantitative measure and it
will be relatively easy to identify a quantitative reduction, even a soft one, in bacterial
growth in the transgenic plants if the candidate gene is involved in CBB resistance.
There are several efforts using the bacterial growth in cassava plants as a measure of
the level of resistance/susceptibility as well the virulence/aggressiveness of Xam
(Cohn et al., 2015; Muñoz et al., 2015).
CBB-CDRGs into conventional and molecular breeding
As any other complex trait, the introduction of a quantitative allele conferring
resistance to CBB is not an easy task. In the case of qualitative disease resistance, the
introduction of a single resistance gene into a susceptible cultivar represents an
expedited scenario. In contrast, given the quantitative resistance is governed by
multiple genes, the level of resistance achieved through different breeding strategies
is dependent on the number and effect of the loci introduced into the susceptible
cultivar. Based on this, some efforts have been accomplished in order to introduce
several plant resistance QTLs by conventional breeding. In these cases even without
the cloning of the corresponding genes the genomic regions corresponding to these
227
loci have been introgressed into elite varieties of some of the most important
economical crops such as barley (Toojinda et al., 1998; van Berloo et al., 2001),
tomato (Robert et al., 2001) and rice (Ahmadi et al., 2001) within others. The
backcrossing and the development of near-isogenic lines (NILs) supported by MAS
approach for the identification of the desirable QTL alleles are part of the recurrent
breeding scheme followed for this purpose (Tanksley and Nelson, 1996). The cases of
the introgression of QTLs have allowed increasing the resistance compared to the
susceptible background. The level of resistance achieved into the introgressed lines
varies with the environment conditions and according to the number of introgressed
QTLs. For example in rice, the introgression of four QTLs into an elite cultivar confers
different levels of resistance depending on the combination of QTLs alleles present in
the introgressed lines (Fukuoka et al., 2014). More recently, several studies have
reported the cloning of genes from QTLs (Zuo et al., 2015; Zheng et al., 2016). This
represents a big advance in the sense it will contribute to develop in a more expedite
way new varieties containing a single gene to increase the resistance level for a group
of variants of a particular pathogen species. In this context, this works provided the
first steps towards the identification of CDRG, which in a near future can be employed
into breeding schemes to be introgressed into cassava elite cultivars.
Since the QTLs described here were strain-specifics and none of them were detected
in all the environments studied, the candidate genes, once validated, could be useful
for breeding purposes but taking into account some considerations. First, the
breeding program including these genes should be focused on obtaining adapted
materials to the regions where the QTLs were identified, either Arauca or La Vega.
The aim to extrapolate this information towards breeding programs targeting other
regions should be considered with caution. To accomplish this it will be imperative to
know in a first time if the strains evaluated in this study (Xam318 and Xam681) are
present in the target region. At least it will be important to know if in the Xam
population in the region present strains belonging to the same haplotypes of Xam318
and Xam681. In a second time it will be important to consider the agro-ecological
conditions present in the zone where the breeding program will be conducted.
Considering the high influence of environmental factors on the CBB response, the
data generated in this study will be only valid for other zones showing similar
conditions and even high caution should be taken before to extrapolate the
information. Once the big challenge of obtaining phenotypic data will be
accomplished, it will be expected to have new and complementary information
concerning to the CBB response in multi environments. In this way, in a near future,
the breeding programs will be directed to zones with overlapping characteristic
concerning the environmental conditions. In this situation we will know if the genes
228
underlying the QTLs detected here can be employed for different or similar
environments.
The validated CDRGs can also have a potential use in the MAS approach. Based on the
genome sequence, and knowing the position of SNP markers it is possible to identify
genes or molecular markers associated to the trait, even knowing that not necessarily
they correspond to the functional gene. In consequence cassava accessions, hybrids
and wild-relative species carrying those molecular markers will also contain the
genomic region for CBB resistance conferred by the “unknown” CDRG allele. This
strategy can increase the accuracy in the selection of resistant cassava materials and
in turn decreasing the time necessary for it. Once this correlation is established, it is
possible to identify at early stages of the plant the presence of the DNA markers and
predict the behavior in response to a particular Xam strain before to conduct the
phenotyping. In practical breeding, the MAS strategy has been successfully applied
for the detection of molecular markers associated to genes conferring resistance to
CMD in cassava (Okogbenin et al., 2007, 2012). Concerning to disease resistance, it
has been done for economically important crops such as wheat (Kuchel et al., 2007;
http://maswheat.ucdavis.edu/), barley (Miedaner and Korzun, 2012) and common
bean (Miklas et al., 2006). However it is important to highlight that the MAS requires
the generation of thousands of markers in a first step as well as very confident
phenotypic data. Only when there is a good combination of these two types of
information it is possible to conduct judicious selections. The problem turns more
complex for quantitative traits. Besides quantitative disease resistance, in some cases
it has been reported MAS application for complex traits such as resistance to abiotic
stress (Miklas et al., 2006; Patto et al., 2006), quality (Dubcovsky, 2004), yield (Stuber
et al., 1999), etc. However, for these quantitative traits is expected that the genotype
selection based on MAS would have a limit in the number of described loci and its
effect under the trait. The application of MAS for cassava qualitative traits such as
resistance to CMD was expediently achieved, for quantitative traits it requires the
validation of loci with small and large effects. Also the evaluation of these effects in
different cassava backgrounds and under a range of environment conditions would
be essential.
In our study no epistatic effects between the QTLs were described. However it is
possible that some interactions may occur between these QTLs and other loci in
different genetic backgrounds. In the last years the MAS-based approach “mapping as
you go” (MAYG) has been developed aiming to solve some problems associated with
the underestimation of the power of the described QTLs (Podlich et al., 2004). MAYG
takes into account the fact that the effect of quantitative alleles can change by the
influence of the genetic background, by epistatic interactions or by environmental
229
effects. In this strategy the genotype and phenotype is evaluated for elite materials
within a breeding program. Then, this data is used in a QTL analysis for the re-
estimation of those detected previously. This will ensure that the initially identified
QTLs persist for the elite materials tested. Also ensures the detection of possible new
QTLs. Considering the MAYG approach and its application for the particular case of
CBB resistance, the idea will be to re-estimate the effect of the validated QTLs in each
cycle of MAS during a cassava breeding program. If a change in the QTL-allele (CDRG
allele) effect is detected, it could be due to the genotypic background or by the
environment. In consequence, the novel markers linked to the new detected QTLs can
be used for the next selection cycle of MAS. If any change in the CDRG allele effect is
detected, it means that neither the genotype nor the environment is affecting the
detected QTL. Thus the markers linked to the QTL could be applied directly in MAS.
Despite that the MAYG offers an option for the incorporation of the QTLs and the
CDRGs identified in this thesis into breeding programs, this approach requires new
genotyping and phenotyping experiments and new QTL mapping in each MAS cycle.
As it has been stressed before, it demands an important human effort, costs, large
infrastructure and is time-consuming. This will be only practical if new phenotypic
evaluation of CBB resistance is developed.
An alternative option to overcome the bottleneck of phenotyping is avoid it. In recent
years the approach of GS has offered new ways to incorporate molecular and some
phenotypic data on breeding programs. GS is a prediction-based breeding strategy
which applies mathematical modeling to predict the plant phenotype in early states
using whole genome molecular markers as predictors of breeding values, avoiding
the necessity to conduct field evaluations in adult plants. A predictive model is
created with high and accurate volumes of genotypic and phenotypic data of the
traits of interest obtained from “training populations”. Afterwards the model
generates genomic estimated breeding values (GEBV) for the evaluated lines. These
GEBVs represent the phenotypic predicted value of the line; in other words, the
GEBVs contain the information of how this line will perform in the field. Currently
some efforts around GS are underway for accelerated breeding cycles in cassava (de
Oliveira et al., 2012; nextgencassava.org). This approach seems to be promising for
cassava since the crop breeding cycle can takes several years and it is not easy to
evaluate the phenotype of thousands of adult cassava plants.
The GS approach seems to be an interesting choice to incorporate the SNP markers of
those validated QTLs for further CBB resistance breeding purposes. In simulated data
for GS-Bayesian model analysis, several results have shown that the accuracy in GS
prediction improves with the inclusion of QTLs into the analysis (Zhong et al., 2009;
230
Jannink et al., 2010). Therefore for CBB the strategy could consist in include into the
“training population” the most resistant individuals from the mapping population
evaluated here; which carry the resistant QTL alleles. Thus the GS genotyping dataset
will contain the SNPs markers linked to the validated QTLs generated during this
thesis. Even knowing the percentage explained by the QTLs identified in this work it
will not easy to establish the level of CBB resistance that can be predicted through GS.
The idea behind this approach is to strengthen the selection of cassava materials
resistant to CBB and thus accelerate the breeding cycle. Additionally, GS opens the
opportunity to identify QTL with small effect, which could have escaped from the QTL
mapping analysis performed here.
References
Lozano, J. 1986. Cassava Bacterial Blight: a Manageable Disease. Plant Dis. 70:1089–1093
Ahmadi, N., Albar, L., Pressoir, G., Pinel, A., Fargette, D., and Ghesquière, A. 2001. Genetic basis and mapping of the resistance to Rice yellow mottle virus. III. Analysis of QTL efficiency in introgressed progenies confirmed the hypothesis of complementary epistasis between two resistance QTLs. Theor. Appl. Genet. 103:1084–1092
Anderson, M., Mulligan, B., Koncz, C., Chua, N. H., Schell, J., and others. 1992. Arabidopsis mutant collection. Methods Arab. Res. :419–437
Ansorge, W. J. 2009. Next-generation DNA sequencing techniques. N. Biotechnol. 25:195–203
Araus, J. L., and Cairns, J. E. 2014. Field high-throughput phenotyping: the new crop breeding frontier. Trends Plant Sci. 19:52–61
Banito, A., Kpémoua, K. E., Wydra, K., and Rudolph, K. 2001. Bacterial blight of cassava in Togo: its importance, the virulence of the pathogen and the resistance of varieties. Pages 259–264 in: Plant Pathogenic Bacteria, Springer.
van Berloo, R., Aalbers, H., Werkman, A., and Niks, R. E. 2001. Resistance QTL confirmed through development of QTL-NILs for barley leaf rust resistance. Mol. Breed. 8:187–195
Bredeson, J. V, Lyons, J. B., Prochnik, S. E., Wu, G. A., Ha, C. M., Edsinger-Gonzales, E., Grimwood, J., Schmutz, J., Rabbi, I. Y., Egesi, C., and others. 2016. Sequencing wild and cultivated cassava and related species reveals extensive interspecific hybridization and genetic diversity. Nat. Biotechnol. 34:562–570
Burch-Smith, T. M., Anderson, J. C., Martin, G. B., and Dinesh-Kumar, S. P. 2004. Applications and advantages of virus-induced gene silencing for gene function studies in plants. Plant J. 39:734–746
231
Cabrera-Bosquet, L., Crossa, J., von Zitzewitz, J., Serret, M. D., and Luis Araus, J. 2012. High-throughput Phenotyping and Genomic Selection: The Frontiers of Crop Breeding ConvergeF. J. Integr. Plant Biol. 54:312–320
Cohn, M., Shybut, M., Dahlbeck, D., and Staskawicz, B. 2015. Assays to Assess Virulence of Xanthomonas axonopodis pv. manihotis on Cassava. Bio-Protocol. 5:e1522
Collard, B. C. Y., Jahufer, M. Z. Z., Brouwer, J. B., and Pang, E. C. K. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica. 142:169-196
Dannon, E. A., and Wydra, K. 2004. Interaction between silicon amendment, bacterial wilt development and phenotype of Ralstonia solanacearum in tomato genotypes. Physiol. Mol. Plant Pathol. 64:233–243
Dubcovsky, J. 2004. Marker-assisted selection in public breeding programs: the wheat experience. Crop Sci. 44
Elshire, R. J., Glaubitz, J. C., Sun, Q., Poland, J. A., Kawamoto, K., Buckler, E. S., and Mitchell, S. E. 2011. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One. 6:e19379
Fahlgren, N., Gehan, M. A., and Baxter, I. 2015. Lights, camera, action: high-throughput plant phenotyping is ready for a close-up. Curr. Opin. Plant Biol. 24:93–99
FAO. 2015. Food Outlook, Biannual report on global food markets. Food Agric. Organ. United Nations, Rome.
Fofana, I. B. F., Sangare, A., Collier, R., Taylor, C., and Fauquet, C. M. 2004. A geminivirus-induced gene silencing system for gene function validation in cassava. Plant Mol. Biol. 56:613–624
Fregene, M., Angel, F., Gómez, R., Rodriguez, F., Chavarriaga, P., Roca, W., Tohme, J., and Bonierbale, M. 1997. A molecular genetic map of cassava (Manihot esculenta Crantz). Theor. Appl. Genet. 95:431–441
Fregene, M., Okogbenin, E., Mba, C., Angel, F., Suarez, M. C., Janneth, G., Chavarriaga, P., Roca, W., Bonierbale, M., and Tohme, J. 2001. Genome mapping in cassava improvement: Challenges, achievements and opportunities. in: Euphytica,
Fukuoka, S., Yamamoto, S.-I., Mizobuchi, R., Yamanouchi, U., Ono, K., Kitazawa, N., Yasuda, N., Fujita, Y., Nguyen, T. T. T., Koizumi, S., and others. 2014. Multiple functional polymorphisms in a single disease resistance gene in rice enhance durable resistance to blast. Sci. Rep. 4
Graner, A., Jahoor, A., Schondelmaier, J., Siedler, H., Pillen, K., Fischbeck, G., Wenzel, G., and Herrmann, R. G. 1991. Construction of an RFLP map of barley. Theor. Appl. Genet. 83:250–256
Hahn, S. K., Howland, A. K., and Terry, E. R. 1973. Cassava Breeding at IITA. IITA,[sd].
Hamilton, A. J., and Baulcombe, D. C. 1999. A species of small antisense RNA in posttranscriptional gene silencing in plants. Science (80). 286:950–952
232
Hao, Y., Parks, R., Cowger, C., Chen, Z., Wang, Y., Bland, D., Murphy, J. P., Guedira, M., Brown-Guedira, G., and Johnson, J. 2015. Molecular characterization of a new powdery mildew resistance gene Pm54 in soft red winter wheat. Theor. Appl. Genet. 128:465–476
He, L., and Hannon, G. J. 2004. MicroRNAs: small RNAs with a big role in gene regulation. Nat. Rev. Genet. 5:522–531
Heffner, E. L., Sorrells, M. E., and Jannink, J.-L. 2009. Genomic selection for crop improvement. Crop Sci. 49:1–12
Hirochika, H., Guiderdoni, E., An, G., Hsing, Y., Eun, M. Y., Han, C., Upadhyaya, N., Ramachandran, S., Zhang, Q., Pereira, A., and others. 2004. Rice mutant resources for gene discovery. Plant Mol. Biol. 54:325–334
Jannink, J.-L., Lorenz, A. J., and Iwata, H. 2010. Genomic selection in plant breeding: from theory to practice. Brief. Funct. Genomics. 9:166–177
Jorge, V., Fregene, M. A., Duque, M. C., Bonierbale, M. W., Tohme, J., and Erdier, - V. 2000a. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). Theor Appl Genet. 101:865–872
Jorge, V., Fregene, M. A., Duque, M. C., Bonierbale, M. W., Tohme, J., and Verdier, V. 2000b. Genetic mapping of resistance to bacterial blight disease in cassava (Manihot esculenta Crantz). Theor. Appl. Genet. 101:865–872
Jorge, V., Fregene, M., Vélez, C. M., Duque, M. C., Tohme, J., and Verdier, V. 2001. QTL analysis of field resistance to Xanthomonas axonopodis pv. manihotis in cassava. Theor. Appl. Genet. 102:564–571
Korte, A., and Farlow, A. 2013. The advantages and limitations of trait analysis with GWAS: a review.
Kuchel, H., Fox, R., Reinheimer, J., Mosionek, L., Willey, N., Bariana, H., and Jefferies, S. 2007. The successful application of a marker-assisted wheat breeding strategy. Mol. Breed. 20:295–308
Liu, J., Zheng, Q., Ma, Q., Gadidasu, K. K., and Zhang, P. 2011. Cassava Genetic Transformation and its Application in BreedingF. J. Integr. Plant Biol. 53:552–569
Liu, Y., Schiff, M., and Dinesh-Kumar, S. P. 2002a. Virus-induced gene silencing in tomato. Plant J. 31:777–786
Liu, Y., Schiff, M., Marathe, R., and Dinesh-Kumar, S. P. 2002b. Tobacco Rar1, EDS1 and NPR1/NIM1 like genes are required for N-mediated resistance to tobacco mosaic virus. Plant J. 30:415–429
Lopez, C. E., Zuluaga, A. P., Cooke, R., Delseny, M., Tohme, J., and Verdier, V. 2003. Isolation of resistance gene candidates (RGCs) and characterization of an RGC cluster in cassava. Mol. Genet. genomics. 269:658–671
Lynch, M., Walsh, B., and others. 1998. Genetics and analysis of quantitative traits. Sinauer Sunderland, MA.
233
Mahlein, A.-K., Oerke, E.-C., Steiner, U., and Dehne, H.-W. 2012a. Recent advances in sensing plant diseases for precision crop protection. Eur. J. Plant Pathol. 133:197–209
Mahlein, A.-K., Rumpf, T., Welke, P., Dehne, H.-W., Plümer, L., Steiner, U., and Oerke, E.-C. 2013. Development of spectral indices for detecting and identifying plant diseases. Remote Sens. Environ. 128:21–30
Mahlein, A.-K., Steiner, U., Hillnhütter, C., Dehne, H.-W., and Oerke, E.-C. 2012b. Hyperspectral imaging for small-scale analysis of symptoms caused by different sugar beet diseases. Plant Methods. 8:1
Mammadov, J., Sun, X., Gao, Y., Ochsenfeld, C., Bakker, E., Ren, R., Flora, J., Wang, X., Kumpatla, S., Meyer, D., Thompson, S., and Agrosciences, D. 2011. Combining powers of linkage and association mapping for precise dissection of QTL controlling resistance to gray leaf spot disease in maize (Zea mays L.). BMC Genomics. 16:1
Matsumura, H., Reich, S., Ito, A., Saitoh, H., Kamoun, S., Winter, P., Kahl, G., Reuter, M., Krüger, D. H., and Terauchi, R. 2003. Gene expression analysis of plant host--pathogen interactions by SuperSAGE. Proc. Natl. Acad. Sci. 100:15718–15723
Mba, R. E. C., Stephenson, P., Edwards, K., Melzer, S., Nkumbira, J., Gullberg, U., Apel, K., Gale, M., Tohme, J., and Fregene, M. 2001. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: Towards an SSR-based molecular genetic map of cassava. Theor. Appl. Genet. 102:21-31
McCouch, S. R., Kochert, G., Yu, Z. H., Wang, Z. Y., Khush, G. S., Coffman, W. R., and Tanksley, S. D. 1988. Molecular mapping of rice chromosomes. Theor. Appl. Genet. 76:815–829
Miedaner, T., and Korzun, V. 2012. Marker-assisted selection for disease resistance in wheat and barley breeding. Phytopathology. 102:560–566
Miklas, P. N., Kelly, J. D., Beebe, S. E., and Blair, M. W. 2006. Common bean breeding for resistance against biotic and abiotic stresses: from classical to MAS breeding. Euphytica. 147:105–131
Mora, R. E., Soto, J. C., and LÓPEZ, C. 2016. Identification of QTLs Associated to Plant Architecture in Cassava (Manihot esculenta). Acta Biológica Colomb. 21:99–109
Muñoz, A., Gomez, C., Mariel, L., Bernal, A., Szurek, B., and López, C. E. 2015. Comparing inoculation methods to evaluate the growth of Xanthomonas axonopodis pv. manihotis ON CASSAVA PLANTS. Acta Biológica Colomb. 20:47–55
Okogbenin, E., Egesi, C. N., Olasanmi, B., Ogundapo, O., Kahya, S., Hurtado, P., Marin, J., Akinbo, O., Mba, C., Gomez, H., and others. 2012. Molecular marker analysis and validation of resistance to cassava mosaic disease in elite cassava genotypes in Nigeria. Crop Sci. 52:2576–2586
Okogbenin, E., Porto, M. C. M., Egesi, C., Mba, C., Espinosa, E., Santos, L. G., Ospina, C., Marin, J., Barrera, E., Gutiérrez, J., and others. 2007. Marker-assisted introgression of resistance to cassava mosaic disease into Latin American germplasm for the genetic improvement of cassava in Africa. Crop Sci. 47:1895–1904
234
de Oliveira, E. J., de Resende, M. D. V., da Silva Santos, V., Ferreira, C. F., Oliveira, G. A. F., da Silva, M. S., de Oliveira, L. A., and Aguilar-Vildoso, C. I. 2012. Genome-wide selection in cassava. Euphytica. 187:263–276
Patto, M. C. V., Skiba, B., Pang, E. C. K., Ochatt, S. J., Lambein, F., and Rubiales, D. 2006. Lathyrus improvement for resistance against biotic and abiotic stresses: from classical breeding to marker assisted selection. Euphytica. 147:133–147
Podlich, D. W., Winkler, C. R., and Cooper, M. 2004. Mapping as you go. Crop Sci. 44:1560–1571
Rabbi, I. Y., Hamblin, M. T., Kumar, P. L., Gedil, M. A., Ikpan, A. S., Jannink, J. L., and Kulakow, P. A. 2014. High-resolution mapping of resistance to cassava mosaic geminiviruses in cassava using genotyping-by-sequencing and its implications for breeding. Virus Res. 186:87-96
Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. A., and Zhang, F. 2013. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8:2281–2308
Reinisch, A. J., Dong, J.-M., Brubaker, C. L., Stelly, D. M., Wendel, J. F., and Paterson, A. H. 1994. A detailed RFLP map of cotton, Gossypium hirsutum x Gossypium barbadense: chromosome organization and evolution in a disomic polyploid genome. Genetics. 138:829–847
Restrepo, S., Velez, C. M., Duque, M. C., and Verdier, V. 2004. Genetic structure and population dynamics of Xanthomonas axonopodis pv. manihotis in Colombia from 1995 to 1999. Appl. Environ. Microbiol. 70:255–261
Robert, V. J. M., West, M. A. L., Inai, S., Caines, A., Arntzen, L., Smith, J. K., and Clair, D. A. S. 2001. Marker-assisted introgression of blackmold resistance QTL alleles from wild Lycopersicon cheesmanii to cultivated tomato (L. esculentum) and evaluation of QTL phenotypic effects. Mol. Breed. 8:217–233
Robertson, D. 2004. VIGS vectors for gene silencing: many targets, many tools. Annu. Rev. Plant Biol. 55:495–519
Russell, G. E. 2013. Progress in Plant Breeding—1. Elsevier.
Ryu, C.-M., Anand, A., Kang, L., and Mysore, K. S. 2004. Agrodrench: a novel and effective agroinoculation method for virus-induced gene silencing in roots and diverse Solanaceous species. Plant J. 40:322–331
Sayre, R., Beeching, J. R., Cahoon, E. B., Egesi, C., Fauquet, C., Fellman, J., Fregene, M., Gruissem, W., Mallowa, S., Manary, M., and others. 2011. The BioCassava plus program: biofortification of cassava for sub-Saharan Africa. Annu. Rev. Plant Biol. 62:251–272
Soto Sedano, J. C., and López Carrascal, C. E. 2014. RNA-seq: herramienta transcriptómica útil para el estudio de interacciones planta-patógeno. Fitosanidad. 16:101–113
235
Stewart, E. L., and McDonald, B. A. 2014. Measuring quantitative virulence in the wheat pathogen Zymoseptoria tritici using high-throughput automated image analysis. Phytopathology. 104:985–992
Stuber, C. W., Polacco, M., and Senior, M. L. 1999. Synergy of empirical breeding, marker-assisted selection, and genomics to increase crop yield potential. Crop Sci. 39:1571–1583
Tanksley, S. D., and Nelson, J. C. 1996. Advanced backcross QTL analysis: a method for the simultaneous discovery and transfer of valuable QTLs from unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 92:191–203
Toojinda, T., Baird, E., Booth, A., Broers, L., Hayes, P., Powell, W., Thomas, W., Vivar, H., and Young, G. 1998. Introgression of quantitative trait loci (QTLs) determining stripe rust resistance in barley: an example of marker-assisted line development. Theor. Appl. Genet. 96:123–131
Trujillo, C. A., Ochoa, J. C., Mideros, M. F., Restrepo, S., López, C., and Bernal, A. 2014. A Complex Population Structure of the Cassava Pathogen Xanthomonas axonopodis pv. manihotis in Recent Years in the Caribbean Region of Colombia. Microb. Ecol.
Unver, T., and Budak, H. 2009. Virus-induced gene silencing, a post transcriptional gene silencing method. Int. J. Plant Genomics. 2009
Wonni, I., Ouedraogo, L., Dao, S., Tekete, C., Koita, O., Taghouti, G., Portier, P., Szurek, B., and Verdier, V. 2015. First report of Cassava Bacterial Blight caused by Xanthomonas axonopodis pv. manihotis in Burkina Faso. Plant Dis.
Zheng, W., Wang, Y., Wang, L., Ma, Z., Zhao, J., Wang, P., Zhang, L., Liu, Z., and Lu, X. 2016. Genetic mapping and molecular marker development for Pi65 (t), a novel broad-spectrum resistance gene to rice blast using next-generation sequencing. Theor. Appl. Genet. 129:1035–1044
Zhong, S., Dekkers, J. C. M., Fernando, R. L., and Jannink, J.-L. 2009. Factors affecting accuracy from genomic selection in populations derived from multiple inbred lines: a barley case study. Genetics. 182:355–364
Zhu, C., Gore, M., Buckler, E. S., and Yu, J. 2008. Status and Prospects of Association Mapping in Plants. Plant Genome J. 1:5
Zuo, W., Chao, Q., Zhang, N., Ye, J., Tan, G., Li, B., Xing, Y., Zhang, B., Liu, H., Fengler, K. A., and others. 2015. A maize wall-associated kinase confers quantitative resistance to head smut. Nat. Genet. 47:151–157
236
General conclusions and perspectives
CBB is a serious disease that can be devastating for the cassava crop in all regions
where cassava is cultivated. Given the importance of cassava for developing
countries, its protection should be a priority. The final aim of this work was to
contribute to identify and deepen the understanding of the natural genetic factors
involved in CBB resistance. This information will guide the breeding programs to
confer resistance to the oncoming improved cassava materials.
In this thesis it was built one of the densest genetic maps reported so far in cassava.
This map has 2,141 SNPs obtained from GBS which integrates genetic and physical
localization of newly annotated IRG in cassava. In this map it was possible to anchor
almost half of the cassava genome v.4.1 sequence draft. Also this map was enriched
with 189 new scaffolds that increase the previous version of the cassava map in
30.7Mb. Nearly 344 Mb or 64% of the genome sequence draft is now anchored to the
genetic map. The cassava GBS-derived data and the genetic map will allow in the
future to map and associate markers with QTLs for particular traits and molecular
cloning of genes controlling them. Also, these data would assist future efforts in
closing the gaps between the scaffolds in the genome sequence and for the
construction of a consensus genetic map. The cassava IRG repertoire, as well as their
genetic and physical map position, accompanied with the SNP information will be a
reference for future genetic analysis and candidate gene approaches to identify genes
related to resistance to diverse biotic stresses.
The phenotypic evaluation of the F1 population to two pathogenic Xam strains was
conducted under multi-environments and during rainy and dry seasons. In addition
the population was evaluated under natural high pressure of the disease. An
evaluation under greenhouse conditions was also included. This represents a high
effort to collect as much information as possible to obtain an excellent set of
phenotypic data. This information was employed for a QTL mapping analysis. As a
result, 18 strains-specific QTLs were detected, explaining between 10,9 and 22.1% of
phenotypic variance of resistance to Xam. Nine of these QTLs show stability between
the rainy and dry seasons and a QTL by environment interaction was detected for ten
QTLs. Moreover, a genotype by environment analysis was accomplished, identifying
several resistant transgressive segregants which represent an important source of
resistant individuals for future breeding programs.
Within the QTLs intervals it was possible to identify 151 CDRGs, from which thirteen
correspond to genes coding for domains present in immunity proteins. Through an
237
RNAseq analysis four CDRGs were differentially expressed during Xam681 infection
in the resistant parental TMS30572. The repertoire of CDRGs co-localizing with the
QTLs reported here, represents a first step in the dissection of the biological
mechanisms that govern CBB resistance and constitutes a novel source of defense-
related genes to be validated. Moreover, in the near future once identified the most
promissory genes within the CDRGs repertoire and once they are functional
validated, these CDRGs can be used in plant breeding strategies focused to develop
cassava materials resistant to CBB adapted to the regions here evaluated.
238
Publications and presentations
Publications
Vasquez, A.X., Soto, J. C., & López, C. E. (Submitted to Molecular Plant-Microbe
Interactions Journal. September 2016). Unraveling the molecules hidden in the gray
shadows. Review.
Soto, J. C., Mora, R. E., Calle, F., & López, C. E. (Submitted to Acta Biologica
Coolombiana. June 2016). QTL identification for cassava bacterial blight resistance
under natural infection conditions.
Mora, R. E., Soto, J. C., & López, C. E. (2016). Identification of QTLs associated to plant
architecture in cassava (Manihot esculenta). Acta Biológica Colombiana, 21(1), 99-
109.
Soto, J. C., Ortiz, J. F., Perlaza-Jiménez, L., Vásquez, A. X., Lopez-Lavalle, L. A. B.,
Mathew, B.... and López, C. E. (2015). A genetic map of cassava (Manihot esculenta
Crantz) with integrated physical mapping of immunity-related genes. BMC
genomics, 16(1), 1.
Soto, J. C., and López, C. E. (2012). RNA-seq as a transcriptome useful tool for the
studies of plant pathogen interactions. Fitosanidad, 16(2), 101-113.
Oral presentations in scientific events
Soto, J. C and López C. E. Mapeando la resistencia a la bacteriosis vascular de la yuca,
Manihot esculenta a través de genotipificación por secuenciación. XIII Congreso
nacional de fitomejoramiento y producción de cultivos. Corpoica. Tibaitatá
(Cundinamarca), Colombia. Noviembre 6-8 de 2013.
Soto, J. C and López C. E. Genotipificación por secuenciación y mapeo de alta
resolución en yuca, para detección de QTLs de resistencia a la bacteriosis vascular.
XXX Congreso ASCOLFI (Asociación Colombiana de Fitopatología y ciencias afines).
Pereira (Risaralda), Colombia. Septiembre 18-20 de 2013.
239
Poster presentations in scientific events
Soto, J. C, Mathew, B., Léon, J., Bernal, A., Ballvora, A., And López, C. E. Molecular and
genetic analysis of cassava bacterial blight (Xanthomonas axonopodis pv. manihotis)
resistance in Manihot esculenta. GPZ (German Society for Plant Breeding) 12th
Conference: Genetic Variation in Plant Breeding. Kiel, Germany. September 24-27-
2014. Outstanding scientific poster award.