1- Evolution of resistance clusters (Co-2, B4 and Co-x)
To attain our aims, we are combining structural and comparative genomics approaches, high-throughput sequencing of small RNA, classical genetic analysis, and pathogenicity tests with 3 pathogens (the fungus Colletotrichum lindemuthianum –agent of anthracnose-, the bacteria Pseudomonas syringae pv phaseolicola –agent of Halo Blight-, the virus BPMV (Bean pod mottle virus)).
We are also developing FISH (Fluorescence In Situ Hybridization) analyses in collaboration with Prof. A. Pedrosa-Harand (Brazil).
In the genome of common bean, large disease resistance gene (R) clusters are most often localized to the end of linkage groups (LG). This is true for the Co-x, Co-2 and B4 R gene clusters, localized at one end of LG-B1, LG-B11 and LG-B4, respectively (Geffroy et al. 1998, 2000, 2008, 2009). Sequencing of the common bean genome confirmed that large NL clusters are often located at the ends of the chromosomes (Figure 1 ; Schmutz et al 2014).
Figure 1 : Physical location of NL sequences in the common bean genome (G19833) (Schmutz et al. 2014)
Sequencing of B4 and Co-2 clusters revealed that they are very large clusters each containing more than 40 NBS-LRR genes (David et al. 2008, 2009, 2010 ; Innes et al. 2008).
The origin of these NBS-LRR sequences was investigated through phylogenetic analysis which revealed that B4 and Co-2 CNL genes are closely related while belonging to nonhomologous chromosomes 4 and 11.
Our results suggest that B4 NBS-LRR are derived from the Co-2 cluster via an ectopic recombination event. FISH analyses revealed that both R clusters are located in a similar genomic environment, adjacent to terminal knobs (heterochromatic blocks) and associated with a 528-bp subtelomeric satellite repeat, referred to as khipu.
For both R gene clusters, khipu is interspersed between NBS-LRR sequences and is also present in the adjacent knobs. Khipu is specific to the Phaseolus genus and present on most P. vulgaris chromosomal termini, indicating the existence of frequent ectopic recombination events between P. vulgaris subtelomeric regions.
Our results show that subtelomeres correspond to favorable niches for R gene proliferation. In addition, we have also identified a centromeric satellite repeat presenting also a 528bp size unit (Richard et al 2013).
We have generated about ≈1Mb corresponding to B4 and Co-2 resistance clusters in two cultivated genotypes (G19833 : Andean ; BAT93 : Meso-American). Comparison between these two genotypes will allow us to study the evolution of these resistance clusters on a short time scale.
Genome-wide analysis of the khipu repeated sequence, of NL sequences and of their potential links is currently underway.
Map-based cloning of R-BPMV, an R gene against BPMV, and of Co-x (Richard et al 2014), an anthracnose atypical R gene (i.e. not a NL encoding gene), are currently underway.
Figure 2 : FISH experiments on pachytene chromosomes using CNL or khipu sequences as probes.
2- Functional validation of candidate genes for genes of interest in legumes using VIGS (Virus-Induced Gene Silencing)
VIGS is a reverse-genetic tool that allows rapid and transient analysis of gene function in crops such as legumes that are difficult and time-consuming to transform genetically.
VIGS exploits the natural plant anti-viral defense response i.e. post-transcriptional gene silencing (PTGS) that targets viral RNAs for sequence-specific degradation.
This technology implies the generation of recombinant viral vectors carrying a fragment of sequence which is homologous to an endogenous gene of interest. After infection with this recombinant virus, transcripts of the gene of interest will become targets for degradation by the PTGS mechanism, thus knocking down expression of the target gene.
● VIGS in common bean (Phaseolus vulgaris)
Previous studies have shown that BPMV (Bean pod mottle virus), a widely used VIGS vector in soybean (Glycine max), is also suitable for gene silencing in Phaseolus vulgaris cv. Black Valentine (Zhang et al. 2010 ; Diaz-Camino et al. 2011).
The success of VIGS relies on the ability of viral vectors to infect the genotype of interest. In contrast to soybean for which almost all commercial cultivars are susceptible to BPMV infection, only few common bean genotypes (including Black Valentine) are susceptible to BPMV. With the recent availability of the whole genome sequence of Phaseolus vulgaris (Schmutz et al. 2014), VIGS technology opens new pastures in common bean functional genomics to assess gene functions involved in traits of agronomic importance (agronomic, disease resistance, abiotic stress tolerance).
We have developed a high-throughput protocol for VIGS in Phaseolus vulgaris using BPMV-based vectors (Pflieger et al. 2014).
The susceptibility of common bean genotypes of interest was tested using BPMV-Green Fluorescent protein (GFP) vector (Figure 3). In susceptible genotypes, the efficiency of VIGS was subsequently tested using BPMV-Phytoene desaturase (BPMV-PDS) vector (Figure 4).
● VIGS in pea (Pisum sativum)
In the context of the PeaMUST project (Pea MUlti-STress adaptation and biological regulations for yield improvement and stability ; project ANR-11-BTBR-0002), we are developing a high-throughput VIGS tool for rapid validation of candidate genes involved in agronomic traits (disease resistances, frost tolerance…) in pea.
We use the BPMV VIGS vector developed in Glycine max and Phaseolus vulgaris. The effectiveness of VIGS in aerial parts (leaves) is tested using BPMV-PDS vector (Figure 5) and in roots using a BPMV-KOR1 vector targeting the endogene KORRIGAN-1 previously shown to be a good reporter gene for VIGS in roots (Constantin et al. 2004) (Figure 6).