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

Abstract

Germline variation at immunoglobulin (IG) loci is critical for pathogen-mediated immunity, but establishing complete haplotype sequences in these regions has been problematic because of complex sequence architecture and diploid source DNA. We sequenced BAC clones from the effectively haploid human hydatidiform mole cell line, CHM1htert, across the light chain IG loci, kappa (IGK) and lambda (IGL), creating single haplotype representations of these regions. The IGL haplotype generated here is 1.25 Mb of contiguous sequence, including four novel IGLV alleles, one novel IGLC allele, and an 11.9-kb insertion. The CH17 IGK haplotype consists of two 644 kb proximal and 466 kb distal contigs separated by a large gap of unknown size; these assemblies added 49 kb of unique sequence extending into this gap. Our analysis also resulted in the characterization of seven novel IGKV alleles and a 16.7-kb region exhibiting signatures of interlocus sequence exchange between distal and proximal IGKV gene clusters. Genetic diversity in IGK/IGL was compared with that of the IG heavy chain (IGH) locus within the same haploid genome, revealing threefold (IGK) and sixfold (IGL) higher diversity in the IGH locus, potentially associated with increased levels of segmental duplication and the telomeric location of IGH.

Figure 1

IGL V, J and C gene comparison between CH17 and the NG_0000002.1 assembly. The upper panel shows the tiling path of sequenced CH17 BAC clones, with SD annotations depicted below. In the lower panel, functional and ORF IGL genes annotated in NG_0000002.1 and the CH17 haplotype are depicted by filled boxes, with corresponding locus and allele names located above and below the respective representation. Genes that are annotated as ORFs are denoted by green (V genes) or grey (J and C genes) boxes with a magenta outline. Shared genes and alleles between NG_0000002.1 and the CH17 haplotype are indicated by filled green or grey boxes. IGLV alleles that, in the CH17 haplotype, are different from those in NG_0000002.1 are indicated by boxes with other colors (red for non-synonymous and blue for synonymous allelic differences). IGLJ and IGLC allelic differences are denoted by the labelling of allele names on the CH17 haplotype. A filled purple circle denotes a novel allele in the CH17 haplotype with a polymorphism resulting in a non-sense mutation (pseudogene). All novel allele names are shown in bold. The 11-kb insertion in the CH17 IGL path is indicated.

Figure 2

IGK V, J and C gene comparison between CH17 and the NG_00034.1/NG000833.1 assemblies. The upper panel shows the tiling path of sequenced CH17 BAC clones, with SD annotations depicted below. The previously sequenced BAC clone, CH17-158B1, is noted by an asterisk. In the lower panel, functional or ORF IGK genes annotated in the NG_000834.1 (proximal) and NG_000833.1 (distal) assemblies and CH17 are depicted by filled boxes, with corresponding locus and allele names located above and below the respective representation. Genes that are annotated as ORFs are denoted by green (V genes) or grey (J and C genes) boxes with a magenta outline. IGKV genes and alleles that are shared between NG_000834.1/NG_000833.1 assemblies and CH17 are indicated by filled green boxes. IGKV alleles that, in the CH17 haplotype, are different from those in the NG_000833.2 (distal) and NG_000834.1 (proximal) assemblies are indicated by boxes with other colors (red for non-synonymous and blue for synonymous). Filled blue circles denote loci at which proximal or distal alleles were observed at the alternate locus (e.g., proximal allele at distal locus); and orange circles denote allelic differences between the haplotypes with respect to V-RSs. The 21-bp indel polymorphism upstream of IGKV1-8 is indicated with a red arrow. The novel sequence extending into the gap in the CH17 IGK path is shown in red. Horizontal black arrows indicate IGKV genes in opposite orientation in the locus, according to the IGK IMGT locus representation (IMGT Repertoire, http://www.imgt.org).

Figure 3

Detection of putative sequence exchange event between IGK proximal and distal gene clusters. (a) Pair-wise alignments between proximal and distal SDs in the CH17 haplotype and NG_000834.1 (proximal) and NG_000833.1 (distal) assemblies (Abbreviations: Ka, NG_000834.1/NG_000833.1 assemblies; prox, proximal; dist, distal). The region where Ka-prox and Ka-dist show stronger similarity than CH17-dist and Ka-prox highlights a potential region of sequence exchange between the two NG_000834.1 and NG_000833.1 assemblies (red box). Coordinates (GRCh37) for the proximal cluster are shown on the X axis. (b) Top panel shows a four-way sequence alignment of a 22.5-kb region from the proximal and distal units from within the red box in (a). Blue tick marks indicate bp SNP differences between the sequences. Upward pointing red arrows indicate boundaries of regions where the NG_000833.1 distal sequence aligns with a higher sequence similarity to the NG_000834.1/NG_000833.1 assemblies and CH17 proximal sequences than to the CH17 distal sequence, indicative of exchange between proximal and distal regions of the NG_000834.1/NG_000833.1 assemblies (sequence similarities: Ka-dist/CH17-dist=98.7%; Ka-dist/Ka-prox=99.7%). A DSS recombination analysis (McGuire and Wright; see Materials and Methods) using the same four-way sequence alignment is shown in the bottom panel. The two peaks with the strongest DSS values (downward pointing red arrows) correspond to the predicted breakpoints shown in the top panel based on sequence similarity values. The dotted line across the chart indicates the significance threshold based on the null distribution of DSS values calculated assuming no recombination.

Figure 4

Representation of SNPs identified in the CH17 IG V gene regions in public SNP databases. The number of SNPs identified in IGKV, IGLV and IGHV gene regions based on alignments of CH17 to assemblies generated by Kawasaki et al.^, and Matsuda et al. (excluding gaps; NG_000834.1, IGK proximal; NG_000833.1. IGK distal; NG_000002.1, IGL; NG_001019.5 IGH). The fraction of SNPs represented in dbSNP135 (top) and 1KG (bottom) databases (db) is shown, and the total number of novel SNPs not found in either database is indicated at the bottom of the left panel.