Expression of the retinoblastoma protein in low-grade B-cell lymphoma: relationship to cyclin D1.

The product of the retinoblastoma tumor-suppressor gene (pRB), a nuclear phosphoprotein that regulates transcription factors such as E2F, is involved in cell cycle control and differentiation. Its activity is regulated by phosphorylation; the underphosphorylated form inhibits transcription whereas the highly phosphorylated form is inactive. Cyclin D1 and its associated kinase (CDK 4/6) phosphorylate pRB in vitro, and therefore are thought to contribute to the regulation of pRB function. To examine the effect of cyclin D1 overexpression on pRB in primary tumor tissue, we studied pRB expression in low-grade B-cell neoplasms, with particular regard to mantle cell lymphoma, which is characterized by cyclin D1 (bcl-1) overexpression. pRB expression was studied by immunostaining with a well-characterized anti-pRB antibody; the phosphorylation status of pRB was examined by immunoblots; and the functional binding capacity of pRB was examined by in vitro binding to adenovirus E1A protein. We studied 3 reactive lymph nodes, 28 low grade B-cell lymphomas, 4 cases of hairy cell leukemia (HCL) and 3 plasmacytomas. Reactive lymph nodes showed intense pRB staining of germinal centers, with strongest (2+) staining in the large cells (centroblasts) of the proliferating (dark) zone and weak or no staining of small lymphocytes, including those of the mantle zone. In B-chronic lymphocytic leukemia (B-CLL) (4 cases), follicular lymphoma (3 cases) and mucosa-associated (MALT) lymphoma (3 cases) strong (2+) pRB staining was limited to centroblasts in reactive and neoplastic follicles and occasional proliferation centers, with only faint staining of small lymphoid cells. In contrast, 15 of 16 cases of mantle cell lymphoma showed strong (1-2+) staining of most cells; one blastoid mantle cell lymphoma showed only faint pRB staining. All cases of (HCL) and plasmacytoma showed strong pRB staining. Although most lymphomas with strong pRB expression were cyclin D1(+), three cyclin D1(+) cases showed only weak pRB expression (1 B-CLL, 1 blastoid mantle cell, 1 unclassifiable low grade B-cell lymphoma). Conversely, of the 4 pRB(+) HCLs and 3 pRB(+) plasmacytomas, only 1 of each was cyclin D1(+). pRB appeared to exist primarily in the underphosphorylated (fastest migrating) form on Western blot, despite the fact that cyclin D1 was complexed to CDK4, a form in which it normally phosphorylates pRB. In addition, pRB appeared to be unmutated, because it bound normally to the adenovirus E1A protein and showed nuclear localization by immunostaining. We conclude that most cases of mantle cell lymphoma, HCL, and plasmacytoma show high levels of pRB in contrast to follicle center lymphoma and small lymphocytic lymphoma; however, pRB expression does not appear to be consistently related to cyclin D1 overexpression. The pRB appears to be unmutated and underphosphorylated, and therefore should be in its active form. Our data from primary lymphoma tissue suggests that overexpression of cyclin D1, whereas tumorigenic, does not lead to pRB loss or hyperphosporylation. Thus, the mechanism by which cyclin D1 contributes to tumorigenesis and the significance of the restricted expression of pRB in low-grade lymphoid neoplasms remain to be determined.