Deregulation of CCND1 or BCL2 in the Immature B-Cell Stage Determines the Resulting Lymphoma Histology

Abstract 416

In B-cell lymphomas, several chromosomal translocations are associated with specific histological subtypes. CCND1/IGH is detected in more than 90 % of the cases with mantle cell lymphoma (MCL), and BCL2/IGH is characteristically observed in follicular lymphoma and diffuse large B-cell lymphoma (DLBCL) of germinal center (GC) B-cell origin. Although these strong correlations are clinically recognized, their biological mechanism is not clearly explained so far. According to the results of translocation breakpoint mapping, both CCND1 and BCL2 translocations are considered to be generated by an error during physiological VDJ rearrangement of the IGH gene in the precursor-B cell stage. We hypothesized that the occurrence of these translocations in the immature B-cell stage, probably as an initial genetic event, should have a special impact on the determination of resulting lymphoma histology.

We generated a mouse model mimicking human lymphoma with CCND1 or BCL2 translocation by lentivirally introducing these genes into Tp53^+/− B6 mouse bone marrow cells and transplanting them to lethally irradiated wild-type B6 mice. In this model, CCND1 or BCL2 is expressed from immature to mature B cell stage under the control of the CD19 promoter, and subsequently mutations accumulate in the background of Tp53 haploinsufficiency. Both mice developed B-cell lymphomas several months after the transplantation, but their tumors showed some different features. CCND1-Tp53^+/− mice developed B220^lowCD5⁺CD23⁻ tumors that expand into the B-cell area with sparing the GC, whereas BCL2-Tp53^+/− mice preferentially developed B220^highCD5⁻CD23⁺ tumors that tend to localize in the GC. Additionally, somatic hypermutation (SHM) analysis of the IGH gene of the tumor cells revealed obviously higher mutation frequency in the BCL2-Tp53^+/− mice than in the CCND1-Tp53^+/- mice (p<0.004). These results indicate that the primary gene deregulation of CCND1 and BCL2 determines the upcoming lymphoma of MCL-type and GC B-cell lymphoma-type, respectively. MCL has been postulated to be derived from naïve pre-GC B cells, with few SHM introduced in the immunoglobulin variable region. But the MCL-like tumor cells generated in CCND1-Tp53^+/−mice seem to be originated from B-1a B cells, a distinct B cell population that does not enter GC and has few SHM in nature.

There has been a debate whether B-1 cell population exists in humans, but it is recently proposed that the phenotype of human B-1 cells is CD20⁺CD27⁺CD43⁺CD70⁻ by testing sort-purified B cell fractions for fundamental B-1 cell functions based on mouse studies (J Exp Med 2011;208:67–80). Interestingly, flow cytometric analysis of the human MCL tumor cells has shown that they mostly express this provisional B-1 cell phenotype, supporting the idea that human MCL is also derived from B-1 cells.

Our mouse model precisely reproduces the link between CCND1 and BCL2 translocations and the resulting lymphoma subtypes in humans. It is assumed that these translocations trigger the cell expansion of different B-cell subgroups, which consequently leads to the development of lymphoma of distinct histology. Our findings provide new insights into the mechanism of lymphoma subtype determination.