Binding of CLL B-Cell Receptors to Viable and Apoptotic Human Cells Offers Insight into the Role of Autoantigens in Leukemic Transformation.

IgVH gene mutation status correlates with clinical course in chronic lymphocytic leukemia (CLL), with the unmutated group usually having a worse outcome. Because the binding domains of B-cell receptors (BCRs) among patients exhibit considerable genetic and structural similarity, it has been suggested that a limited set of antigens drive clonal expansion in CLL. We reasoned that characterizing antigenic reactivities of CLL BCRs would provide insight into the development of the disease. Hence we cloned and expressed BCRs from 28 CLL patients (18 unmutated and 10 mutated) as monoclonal antibodies (mAbs). We tested these by immunofluorescence for reactivity with the human epithelial cell line HEp-2, often used to identify nuclear autoantigens in autoimmune disorders such as systemic lupus erythematosus (SLE). Furthermore since apoptotic cells are a source of autoantigens in autoimmune conditions, we also tested the mAbs for reactivity with apoptotic and viable B and T cells, from RAMOS (B) and Jurkat (T) cell lines and from normal subjects and CLL patients. All 28 mAbs showed binding to intracellular structures of HEp-2 cells, reminiscent of SLE. Unlike SLE, the binding was predominantly cytoplasmic. Only one mAb, expressing VH4-59, bound HEp-2 cell nuclei. When tested for reactivity with B and T cell surfaces by indirect immunofluorescence and flow cytometry, only 13 (10 unmutated and 3 mutated) of the 28 mAbs bound at least one lymphoid cell type. Among the cell-reactive mAbs, 8 bound apoptotic cells without discriminating between B and T cells. All these expressed unmutated IgVH genes; 5 were VH1 family members (3 exhibited VH1-69-D3-3-JH6 with different HCDR3s and L chain rearrangements, 1 expressed VH1-03, and 1 expressed VH1-02) and the last 3 belonged to a stereotypic set comprising VH4-39-D6-13-JH5 + Vk1D-39-Jk1. Notably, 5 mAbs, expressing VH4-34, reacted only with viable cells. Two of these used unmutated 4-34 genes and bound both B and T cells; altering their companion VLJL rearrangements eliminated virtually all cell binding. The other 3 mAbs, members of the stereotypic set with a mutated VH4-34-D5-18-JH6 rearrangement, bound only viable B cells, predominantly of the naïve subset. Substitution of their L chains led to a decrease in B cell binding, although to a lesser degree than for the unmutated VH4-34 mAbs. The one mAb reacting with HEp-2 nuclei failed to react with apoptotic or viable B and T cells. Of note, no VH3-expressing mAbs reacted with cell surface targets. In sum, even though all 28 CLL mAbs reacted with intracellular targets, only ~45% (n=13) reacted with lymphoid cell surfaces. Of those, 8 bound only apoptotic cells and 5 bound viable cells. All mAbs binding viable cells expressed VH4-34, with mutated mAbs binding solely B cells, and unmutated binding both. Interestingly, although the 8 mAbs binding apoptotic B or T cells were all unmutated, another 8 unmutated, but HEp-2-reactive, mAbs failed to bind any surface targets. Thus unmutated polyreactive BCRs exhibit a degree of specificity in antigen binding conferred by VH family and gene use and their companion VHDJH and VLJL. Finally, these data highlight a role for apoptosis as a source of already existing or newly created autoantigens in the promotion and evolution of CLL.