Immunoglobulin Light Chain Repertoire in Chronic Lymphocytic Leukemia (CLL): Recognition of Subsets with “CLL-Specific” CDR3 Regions and Associations with Heavy Chains.

We analyzed immunoglobulin light chain (IgLC) repertoire in a series of 253 typical, unselected CLL cases and compared CLL IgLC sequences to GenBank IgLC sequences from normal, autoreactive and neoplastic cells. The present series included 165 κ- and 88 λ-CLL cases. Twenty-three functional IGKV genes were used in IGKV-J rearrangements in κ-CLL; the most frequent genes were: 3-20/A27 (25 cases), 1-39-1D-39/O2-O12 (19 cases), 4-1/B3 (16 cases), 1-5/L12 (15 cases), 2-30/A17 (13 cases) and 1-8/L9 (10 cases). There were 55/165 unmutated sequences (33%), 44/165 sequences (27%) with 97-99,6% homology to germline and 66/165 sequences (40%) with less than 97% homology. KCDR3 region length ranged from 6–11 (median, 9) aminoacids (aa). N nucleotides (median 3, range 1–12) were detected in 85/165 rearrangements (51.5%). IGKJ3-5 gene usage was observed in 51/165 rearrangements (30%); interestingly, IGKJ3-5 genes were used in 7/8 IGKV3-11 and 4/5 IGKV1-9 rearrangements. Subsets with homologous and “CLL-specific” KCDR3 regions were identified: IGKV2-30, 5 mutated sequences with identical KCDR3 (MQGTYWPYT), 3/5 associated with IGHV4-34 utilizing heavy chains with a similar HCDR3 of 20 aa; IGKV1-39/1D-39, 3 unmutated sequences with identical KCDR3 (QQSYSTTPLT), all associated with IGHV4-39 utilizing heavy chains with a similar HCDR3 of 19 aa; IGKV1-5, 4 unmutated sequences with identical KCDR3 (QQYNSYPWT), 2/4 associated with unmutated IGHV4-39 utilizing heavy chains with a HCDR3 of unequal length. Twenty-six functional IGLV genes were used in IGLV-J rearrangements in λ-CLL; the most frequent genes were: IGLV2-8/1-2 (14 cases), 3-21/2-14 (13 cases), 2-14/1-4 and 1-44/1-16 (7 cases each). There were 24/88 unmutated sequences (27%), 33/88 sequences (37,5%) with 97-99,6% homology to germline and 31/88 sequences (35%) with less than 97% homology. LCDR3 region length ranged from 8-13 aa (median, 11). N nucleotides (median 3, range 1-15) were detected in 42/88 rearrangements (47.7%). The IGLJ1 gene was used in 18/88 rearrangements (20%); all other rearrangements used the IGLJ3*01/*02 genes. Subsets with homologous and “CLL-specific” LCDR3 regions were identified: IGLV1-44, 2 sequences with very similar LCDR3 (AAWDDSLNGP/QV), both associated with IGHV4-b utilizing heavy chains with a similar HCDR3 of 11 aa; IGLV3-21, 7 sequences all with identical LCDR3 (QVWDSGSDHPWV), 3/7 associated with IGHV3-21 utilizing heavy chains with a similar HCDR3 of 9 aa. These results document that IgLC repertoire in CLL is biased by both intrinsic molecular processes as well as selection after LC expression. Genes that have been reported to be overexpressed in the normal and autoimmune disorders were also found to be overrepresented in the CLL repertoire, often with “CLL-specific” molecular features. Finally, the existence of subgroups with homologous CDR3 regions associated with similar heavy chains provides further evidence for the role of antigen selection in CLL pathogenesis.