Somatically mutated Ig V_H3-21 genes characterize a new subset of chronic lymphocytic leukemia

Biased immunoglobulin variable heavy-chain (IgV_H) gene usage has been reported in different entities of B-cell lymphoma and in B-cell chronic lymphocytic leukemia (B-CLL).1-10 This nonrandom usage of individual V_H genes has been suggested to reflect that the immunoglobulin structure may play a role in leukemia/lymphoma development, possibly through chronic antigen stimulation.

Prior studies have revealed a preferential usage of the V_H1 family member, V_H1-69 in B-CLL (12%-21%) compared to normal peripheral blood B cells (1.6%).3,4,7,11,12 Elevated expression of the V_H1-69 gene has also been demonstrated as a feature of B-CLL with unmutated V_H genes.6,11Interestingly, the V_H1-69 rearrangements in B-CLL display distinctive molecular characteristics, such as skewed usage of certain D and J_H genes and a significantly longer average length of the complementarity determining region (CDR) 3, which is in contrast to normal V_H1-69 expressing peripheral blood B cells.4,7 13 Therefore, it has been proposed that these V_H1-69 rearrangements may provide a specific immunoglobulin structure that predisposes the respective B cells to neoplastic transformation, thus suggesting that yet unknown antigens could be involved in leukemogenesis of B-CLL.

We have performed V_H gene analysis in 119 B-CLL cases to study the V_H gene usage in subsets of mutated and unmutated B-CLL. In addition, we have correlated the V_H gene analysis to overall survival.

Tumor samples were collected from 119 patients with B-CLL, which had been identified from the archives of the Departments of Pathology and Clinical Immunology at the Uppsala University Hospital and from the Department of Pathology at the Umeå University Hospital between 1981 and 1998. Tumor material was obtained mainly from peripheral blood (64 cases) and bone marrow (44 cases), but in a few cases also from lymph nodes and spleen. According to the Royal Marsden scoring system, the tumor cells typically expressed CD5, CD19, and CD23 and showed a weak expression of immunoglobulin.14In 6 cases, CD23 expression was not analyzed.

There were 82 men and 37 women. The median age at diagnosis was 65 years and the median follow up 90 months. Survival data were obtained in 112 of 119 cases from the Swedish cancer registries in Uppsala and Umeå. Overall survival was calculated from date of diagnosis until death or last follow-up. Kaplan-Meier survival analysis and log-rank test were performed using the Statistica 5.5A software (Stat Soft, Tulsa, OK).

V_H gene family-specific polymerase chain reaction (PCR) amplification was performed as previously described.15 In the majority of samples clonal PCR products were sequenced directly using the BigDye Terminator Cycle Sequencing Reaction Kit (Perkin-Elmer, ABI, Foster City, CA), but in 13 cases cloning of the PCR products was performed as described elsewhere.15 The sequences were aligned to IgH sequences from the BLAST, V-BASE, and IMGT databases. V_H gene sequences deviating more than 2% from the corresponding germline gene were defined as mutated. The distribution ofreplacement mutations and silent mutations within the CDRs and framework regions was assessed according to the multinomial distribution model published by Lossos and colleagues to determine the extent of antigen selection in mutated V_H genes.16 The length of the CDR3 was calculated between codon 95 and 102.

A total of 134 IgH gene rearrangements from 119 cases were sequenced, including 15 cases with 2 gene rearrangements. Sixty-nine cases (58%) displayed unmutated V_H genes and 50 cases (42%) showed somatically hypermutated V_H genes. The median survival was significantly shorter for unmutated cases (71 months, n = 63) compared to mutated cases (124 months, n = 49;P < .001). Indeed, our finding is in line with previous studies and confirms the prognostic significance of V_H gene mutational status in B-CLL.11,17,18 Using the multinomial distribution model,16 evidence for antigen selection was indicated in 20 (39.2%) of 51 mutated rearrangements, 8 of which used the V_H3-21 gene.

Overall, the most frequently used V_H genes were V_H1-69 (n = 21; 15.7%) and V_H3-21 (n = 15; 11.2%). In accordance with prior studies, the V_H1-69 gene was exclusively used in unmutated B-CLL and demonstrated the highest frequency accounting for 25.3% of unmutated V_H genes (Table 1).3,4,7,11 A restricted J_H6 gene and D2-2 gene usage was found (75.0% and 23.8%, respectively) and the mean CDR3 length was longer than expected (18.7 codons) in the cases expressing V_H1-69.4,13 Our data strongly support the previously reported finding of a V_H1-69–using subset with specific molecular features in B-CLL.4 13 

The V_H3-21 gene was used mainly in the mutated (87%, n = 13) versus the unmutated (13%, n = 2) group. The restricted usage of V_H3-21 in mutated CLL cases is a novel finding because there are little previous frequency data available on V_H3-21 usage in normal B cells or B-CLL. Interestingly, the mutated V_H3-21 rearrangements displayed shorter average length of the CDR3 (8.3 codons) compared to the unmutated V_H1-69 rearrangements (18.7 codons) and the remaining of the IgH rearrangements (14.8 codons). Four of the mutated V_H3-21 cases (nos. 36, 62, 75, and 113) demonstrated identical or almost identical CDR3s (Table2), though having individual V_H gene mutation patterns (data not shown). In addition, all mutated V_H3-21 cases expressed clonal λ light chains, which is in contrast to the expected frequency in B-CLL. Preliminary sequence data of the Igλ gene rearrangements has revealed a preferential usage of 1 of the V_λ3 family genes (V2-14). The combined usage of the V_H3-21 gene and the V_λ3 gene is a new finding not previously described.

Most interestingly, the survival was significantly shorter for patients with mutated V_H3-21 compared to the remaining cases with mutated V_H genes (P = .018, Figure1). The median survival for mutated cases with V_H3-21 gene usage (n = 12) was 63 months, which was considerably shorter than the remainder of the mutated group excluding V_H3-21 (147 months, n = 37). They also showed an overall survival similar to the unmutated group (median survival, 63 versus 71 months). All together, these findings suggest that mutated V_H3-21 genes may constitute an additional entity of B-CLL, displaying distinctive genotypic/phenotypic features and a significantly poorer survival than mutated cases in general. However, larger numbers of cases expressing V_H3-21 have to be analyzed to further investigate the prognostic impact of V_H3-21 usage in mutated B-CLL.

The V_H3-21 gene has been implicated in the production of rheumatoid factors in rheumatoid arthritis.19 More than one third of cases (8 of 20) that showed signs of antigen selection used the V_H3-21 gene, which may indicate selective pressure for this particular V_H gene. Moreover, the short and in some cases identical CDR3s combined with a preferential V_λ3 gene usage may also suggest a unique binding of an antigen to the V_H3-21 encoded Ig heavy chains. It is therefore tempting to speculate on a possible role for antigens in the development of aggressive forms of B-CLL, considering our findings of poor prognostic outcome associated with the usage of the V_H1-69 gene as well as the V_H3-21 gene. Hypothetically, the biased usage in B-CLL could reflect that these rearrangements encode for specific epitopes to which unknown antigen(s) could bind and stimulate proliferation. However, future studies are necessary to clarify the possible role of specific antigens in the pathogenesis of B-CLL.

The authors are grateful to Anita Lindström and Inger Eriksson for skillful technical assistance and professor Dan Holmberg for scientific advice.