Final Report Summary - GATFUN (Genetic basis of host adaptative traits in the plant pathogenic fungus Mycosphaerella fijiensis)
Project objectives
Elucidating the genetic basis of adaptive traits is a central issue in evolutionary biology. Plant pathogenic fungi are interesting organisms to address this question since they are able to adapt rapidly to their environment. Through changes in their aggressiveness, they are notably able to erode quantitative host resistances, thereby threatening their durability. This project aims to elucidate the genetic basis of aggressiveness traits involved in the adaptation of a plant pathogenic fungus to this type of resistance. The study model is the fungus Mycosphaerella fijiensis, responsible for a recent and devastating banana pandemic, Black Leaf Streak Disease (BLSD). This model displays biological characteristics which, when combined with available methods and resources, make it possible to take QTL mapping and population genomics approaches in a complementary manner. These methods have seldom been used in fungi. The power of these approaches is greatly increased by new-generation sequencing techniques. To achieve the objectives we developed two complementary approaches, one corresponding to the search for QTLs by genetic mapping, the other the search for loci selected by the host, taking a population genomics approach. During the outgoing phase, as a training, Jean Carlier joined Stephen Wright's laboratory (University of Toronto, Canada), which is one of the most proficient laboratories in genetics and population genomics on plants using new-generation sequencing techniques.
Results
QTL mapping and host selection effect. A classic QTL mapping approach has been conducted with the analysis of progenies derived from a laboratory cross between 2 isolates displaying differences in aggressiveness traits. Aggressiveness traits of the progenies were also evaluated by inoculation under controlled conditions. A genetic map was constructed with about 370 genetic makers. This map will be used soon to identify QTLs related to aggressiveness traits. To detected host selection effect on pathogen populations, a second approach has been conducted with the evaluation of aggressiveness traits of about 100 isolates from Cuba on 4 varieties of various quantitative resistant levels. We did not observed difference between populations. However, variability in aggressiveness traits was detected within the whole sample. This variability could be further exploited to identify QTL through an association genetic approach. On this way, the genome of about 130 individual was sequenced and we are currently running bioinformatic pipelines to call SNPs. This studies was not planned in the project, but we decided to run it based on the promising results from phenotyping and it will be probably achieve few months after the official end of the project.
Population genomic. The population genomics approach was involved comparing natural populations of M. fijiensis living on different hosts. Some samples taken from populations living for 5 to 10 years on susceptible and resistant hosts in Cuba and the Dominican Republic were compared using a large number of SNP (single nucleotide polymorphism) markers distributed over the genome. These SNPs were defined by re-sequencing the whole genome using pools of individuals for each of the populations.
About 700000 SNPs were detected from the pools coming from the Latin America-Caribbean region. Population tree was inferred from the 10000 most polymorphic SNPs. This analysis showed that population in each country stemmed probably from a single introduction with a bottleneck. However, most populations were grouped per location and no significant gene flow between countries was detected.
Descriptive genetic analysis suggests also existence of selection in some DNA regions between population pairs coming from susceptible and resistant hosts. However their existence could not be tested using methods based on genetic models. Because the populations analysed stemmed from a recent colonizing event, these failures might results from presence of incomplete sweeps or from demographic history. The presence of different putative selected regions among population pairs might results either from the effect of different environmental factors or of host resistances with independent paths of adaptation. Phenotyping on the different varieties and sequencing of individuals from a sub-sample are being conducted to strengthen this study
About 800000 SNPs were detected among the 62 isolates representing a global sample. A phylogenetic tree was constructed to infer the population structure. As expected we detected more diversity in the disease center of origin (Asia) and trace of serial bottleneck in the other regions. We are now using this data to compare among native and recently introduced populations the genome-wide levels of negative and positive selections.
Impact
The time with Stephen Wright (University of Toronto, Canada) enabled Jean Carlier to take advantage of his extensive experience in plant genetics and population genomics using next generation sequencing techniques and to transfer those skills to plant pathogenic fungi. On his return to his research unit (JRU BGPI, Montpellier, France), these new skills have enabled Jean Carlier to strengthen the scientific project of his own research group and to become a leader in his scientific field. This project has also broadened his network for greater international visibility. It has then contributed to the excellence of ERA, enhancing skills in research on plant pathogenic fungi and on banana in Europe. Lastly, the results of this project will serve to define durable strategies for controlling BLSD of banana, a plant grown in the EU outermost regions and important for world food security, and more generally for reduced pesticide use in crops.
Elucidating the genetic basis of adaptive traits is a central issue in evolutionary biology. Plant pathogenic fungi are interesting organisms to address this question since they are able to adapt rapidly to their environment. Through changes in their aggressiveness, they are notably able to erode quantitative host resistances, thereby threatening their durability. This project aims to elucidate the genetic basis of aggressiveness traits involved in the adaptation of a plant pathogenic fungus to this type of resistance. The study model is the fungus Mycosphaerella fijiensis, responsible for a recent and devastating banana pandemic, Black Leaf Streak Disease (BLSD). This model displays biological characteristics which, when combined with available methods and resources, make it possible to take QTL mapping and population genomics approaches in a complementary manner. These methods have seldom been used in fungi. The power of these approaches is greatly increased by new-generation sequencing techniques. To achieve the objectives we developed two complementary approaches, one corresponding to the search for QTLs by genetic mapping, the other the search for loci selected by the host, taking a population genomics approach. During the outgoing phase, as a training, Jean Carlier joined Stephen Wright's laboratory (University of Toronto, Canada), which is one of the most proficient laboratories in genetics and population genomics on plants using new-generation sequencing techniques.
Results
QTL mapping and host selection effect. A classic QTL mapping approach has been conducted with the analysis of progenies derived from a laboratory cross between 2 isolates displaying differences in aggressiveness traits. Aggressiveness traits of the progenies were also evaluated by inoculation under controlled conditions. A genetic map was constructed with about 370 genetic makers. This map will be used soon to identify QTLs related to aggressiveness traits. To detected host selection effect on pathogen populations, a second approach has been conducted with the evaluation of aggressiveness traits of about 100 isolates from Cuba on 4 varieties of various quantitative resistant levels. We did not observed difference between populations. However, variability in aggressiveness traits was detected within the whole sample. This variability could be further exploited to identify QTL through an association genetic approach. On this way, the genome of about 130 individual was sequenced and we are currently running bioinformatic pipelines to call SNPs. This studies was not planned in the project, but we decided to run it based on the promising results from phenotyping and it will be probably achieve few months after the official end of the project.
Population genomic. The population genomics approach was involved comparing natural populations of M. fijiensis living on different hosts. Some samples taken from populations living for 5 to 10 years on susceptible and resistant hosts in Cuba and the Dominican Republic were compared using a large number of SNP (single nucleotide polymorphism) markers distributed over the genome. These SNPs were defined by re-sequencing the whole genome using pools of individuals for each of the populations.
About 700000 SNPs were detected from the pools coming from the Latin America-Caribbean region. Population tree was inferred from the 10000 most polymorphic SNPs. This analysis showed that population in each country stemmed probably from a single introduction with a bottleneck. However, most populations were grouped per location and no significant gene flow between countries was detected.
Descriptive genetic analysis suggests also existence of selection in some DNA regions between population pairs coming from susceptible and resistant hosts. However their existence could not be tested using methods based on genetic models. Because the populations analysed stemmed from a recent colonizing event, these failures might results from presence of incomplete sweeps or from demographic history. The presence of different putative selected regions among population pairs might results either from the effect of different environmental factors or of host resistances with independent paths of adaptation. Phenotyping on the different varieties and sequencing of individuals from a sub-sample are being conducted to strengthen this study
About 800000 SNPs were detected among the 62 isolates representing a global sample. A phylogenetic tree was constructed to infer the population structure. As expected we detected more diversity in the disease center of origin (Asia) and trace of serial bottleneck in the other regions. We are now using this data to compare among native and recently introduced populations the genome-wide levels of negative and positive selections.
Impact
The time with Stephen Wright (University of Toronto, Canada) enabled Jean Carlier to take advantage of his extensive experience in plant genetics and population genomics using next generation sequencing techniques and to transfer those skills to plant pathogenic fungi. On his return to his research unit (JRU BGPI, Montpellier, France), these new skills have enabled Jean Carlier to strengthen the scientific project of his own research group and to become a leader in his scientific field. This project has also broadened his network for greater international visibility. It has then contributed to the excellence of ERA, enhancing skills in research on plant pathogenic fungi and on banana in Europe. Lastly, the results of this project will serve to define durable strategies for controlling BLSD of banana, a plant grown in the EU outermost regions and important for world food security, and more generally for reduced pesticide use in crops.