Abstract
The expressed human kappa light chain gene repertoire utilized by healthy individuals was studied by two different single-sided specific PCR techniques to avoid bias for certain V genes. A total of 103 rearranged kappa sequences from peripheral blood mononuclear cells from healthy individuals were cloned from cDNA and assigned to the Vkappa and Jkappa germ-line genes with the closest overall homology. The use of cDNA rather than genomic DNA focused the analysis on activated B cells rich in mRNA. Accordingly, the sequences represented the applied repertoire and almost all were somatically mutated. V genes from the Jkappa-proximal duplication unit of the kappa locus were almost exclusively used. A total of 65% of the sequences could be assigned to four or five genes: A27 (humkv325), L6 (Vg), L2 (humkv328), and A3 and/or A19. N additions and P nucleotides were quite common and found in 32% and 21% of the sequences, respectively. Extended CDR3s more than nine residues in length were found in 18% of the sequences, and in 71% of cases this was due to insertion of an extra proline residue. This proline was usually explained from the germ-line sequences involved. These results are in good agreement with those of previous repertoire studies using potentially V-gene-biased techniques. Thus, it is clear that restricted V-gene usage, common N and P additions, and extended CDR3 regions are normal features and not, as has been claimed, characteristics of pathological autoantibodies.