Defining the mutation sites in chickpea nodulation mutants PM233 and PM405

Like most legumes, chickpeas form specialized organs called root nodules. These nodules allow for a symbiotic relationship with rhizobium bacteria. The rhizobia provide fixed atmospheric nitrogen to the plant in a usable form. It is of both basic and practical interest to understand the host plant genetics of legume root nodulation. Chickpea lines PM233 and PM405, which harbor the mutationally identified nodulation genes rn1 and rn4, respectively, both display nodulation-deficient phenotypes. Previous investigators identified the rn1 mutation with the chickpea homolog of Medicago truncatula nodulation gene NSP2, but were unable to define the mutant rn1 allele. We used Illumina and Nanopore sequencing reads to attempt to identify and characterize candidate mutation sites responsible for the PM233 and PM405 phenotypes.

We aligned Illumina reads to the available desi chickpea reference genome, and did a de novo contig assembly of Nanopore reads. In mutant PM233, the Nanopore contigs allowed us to identify the breakpoints of a ~ 35 kb deleted region containing the CaNSP2 gene, the Medicago truncatula homolog of which is involved in nodulation. In mutant PM405, we performed variant calling in read alignments and identified 10 candidate mutations. Genotyping of a segregating progeny population narrowed that pool down to a single candidate gene which displayed homology to M. truncatula nodulation gene NIN.

We have characterized the nodulation mutation sites in chickpea mutants PM233 and PM405. In mutant PM233, the rn1 mutation was shown to be due to deletion of the entire CaNSP2 nodulation gene, while in mutant PM405 the rn4 mutation was due to a single base deletion resulting in a frameshift mutation between the predicted RWP-RK and PB1 domains of the NIN nodulation gene. Critical to characterization of the rn1 allele was the generation of Nanopore contigs for mutant PM233 and its wild type parent ICC 640, without which the deletional boundaries could not be defined. Our results suggest that efforts of prior investigators were hampered by genomic misassemblies in the CaNSP2 region of both the desi and kabuli reference genomes.