Sequencing of the human IG light chain loci from a hydatidiform mole BAC library reveals locus-specific signatures of genetic diversity

Corey T Watson; Karyn Meltz Steinberg; Tina A Graves-Lindsay; Rene L Warren; Maika Malig; Jacqueline E Schein; Richard K Wilson; Rob Holt; Evan Eichler; Felix Breden

We present data for the first haplotypes of the human IGL and IGK loci from the same haploid genome, representing only the second full-length references constructed for these regions to date. From the CH17 clones, 12 novel alleles were identified in the two loci, including four IGLV alleles, seven IGKV alleles, and one IGLC allele. Two recent assessments of IGK allelic variation --one of a public dataset of 435 expressed sequences 33 , and a second of deepsequenced antibody repertoires from four individuals 34 --concluded that, unlike IGH, IGK allelic datasets are likely to be mostly complete, as only two putative novel alleles were identified from these analyses 33 . IGHV and IGKV repertoire sequencing in a single individual also supports . CC-BY-NC 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/006866 doi: bioRxiv preprint these observations, finding that nearly 25% of characterized alleles in IGHV were novel, compared to 0% in IGKV 35 . However, the fact that we identified 11 novel light chain V gene alleles from a single haploid genome implies that additional efforts to identify unreported alleles in IGL and IGK are warranted. Importantly, as noted previously in IGHV 7 , SNPs associated with novel alleles identified in IGL and IGK were present in the 1KG dataset, further supporting the notion that the 1KG dataset could serve as a useful resource for future investigations of IG gene diversity and the identification of novel polymorphisms. However, given the prevalence of segmental duplication in the IG loci, it will undoubtedly be essential to consider the impact of paralogous sequence variants (PSVs) when assessing 1KG SNP data in these regions, as complex and duplicated sequence structure are known to confound SNP characterization 36, 37 .