Abstract

We conducted a parallel analysis of somatic mutations in paired IG heavy (HC) and light chain (LC) chain genes in a series of 253 (165 κ- and 88 λ-LC expressing) chronic lymphocytic leukemia (CLL) cases. IGHV/IGKV/IGLV sequences were 100% homologous to germline in 32/33/27% of cases; had 97–99.7% homology in 50/27/37,5% of cases; less than 97% homology in 18/40/35% of cases. IGHV sequences were generally more mutated than the corresponding IGKV/IGKL sequences (average mutation load for IGHV/IGKV/IGLV genes: 94.4/96.4/96.8); 18/253 cases (7.1%) had more mutated LCs than HCs. The most frequent genes among 100% homologous sequences were IGHV4-39 and IGHV1-69 (10/16 and 9/15 cases, respectively) and IGKV1-39/1D-39 (9/19 cases); among IGLV genes, 3/13 IGLV3-21 sequences were unmutated and 9/13 had >98% homology. The most frequent genes in the “<97%-homology” subgroup were IGHV3-7 and IGHV4-34 (18/19 and 23/27 cases, respectively), IGKV3-20 (12/25 cases) and IGLV2-8 (8/14 cases). Analysis after the multinomial distribution model disclosed significant evidence for positive selection by classical T-dependent antigen from both chains in 43 cases, from HCs only in 53 cases and from LCs only in 16 cases; 27 HC+LC mutated cases (7/27 with <97% homology) had no evidence for antigen selection from either HCs or LCs. Mutations followed a similar pattern, with elevated replacement/silent (R/S) ratios in CDRs (mean, 2.8, 2.6 and 3.7 for IGHV, IGKV and IGLV, respectively) and decreased R/S in FRs (mean, 1.4, 1.5 and 1.6 for IGHV, IGKV and IGLV, respectively). R mutations in FRs usually led to substitution by aminoacids of similar charge and size. The somatic hypermutation process generally spread diversity to the periphery of the CDR1 region. Mutational “hotspots” differed not only between HCs and LCs but among subgroups (eg, IGHV1 vs. IGHV3 vs. IGHV4: thus, S-92 in IMGT-FR3 was mutated in 52/21/5% of IGHV4/3/1 sequences; IGKV1/3 vs. IGKV2 subgroup genes; IGLV1 vs. IGLV3 subgroup genes) or even individual genes, e.g. IGHV1-8, IGHV3-23, IGHV4-34, IGKV2-30, IGLV3-21; in this context, a subgroup of three mutated IGHV4-34 sequences was identified, all in sIgG+ cases, with similar HCDR3 regions and associated with IGKV2-30 κ-LCs with identical KCDR3; furthermore, in the case of IGLV3-21, an already described association with IGHV3-21 HCs was again noted in our series (three cases). Importantly, some IMGT positions were universal “hotspots” (eg. 32 in IMGT-CDR1) or “coldspots” (both invariant C at IMGT positions 23/104 in all sequences; IMGT positions 41, 43, 74, 75 and 102 for all IGHV subgroups; 16, 41, 76, 87, 89 and 98 for all IGKV subgroups; 41, 43, 50, 70, 76, 83, 89, 98, 99 and 102 for all IGLV subgroups). Finally, IGLV sequences showed a lower incidence of CDR1 and FR2/3 mutations compared to IGKV sequences. In conclusion, a complementary imprint of antigen selection in CLL might be witnessed either by IGHV, IGKV or IGLV rearranged sequences. These results provide further evidence for the role of antigen selection in the pathogenesis of at least a proportion of CLL cases.

Author notes

Corresponding author