Human Macrophages Phagocytose Rituximab Opsonised Leukemic Cells Via CD16, CD32 and CD64 but Do Not Mediate ADCC.

Rituximab (Mabthera®) is a chimeric monoclonal IgG1 antibody with therapeutic activity in non-Hodgkin B lymphomas (B-NHL) and B-Chronic Lymphocytic Leukemia (B-CLL). We have recently obtained evidence, using a bulky lymphoma xenograft model in nude mice, that both complement and macrophages are required for the therapeutic activity of rituximab. In order to further investigate the tumor cell killing potential of macrophages and its modulation by different factors, including complement, we have set up in vitro experiments with purified macrophage populations. Human macrophages were obtained from purified peripheral blood monocytes cultured for 4 days in presence of 20% FCS and 20 ng/ml M-CSF. FACS analysis confirmed the phenotype of these cells including CD11b and FcγRs expression (CD16, CD32, CD64). Phagocytosis assays were then carried out with CLL cell as targets in presence or absence of increasing concentrations of rituximab. Phagocytosis was evaluated by counting under an inverted microscope the stained cytospin preparations. From 9.8% to 60.8% of macrophages engulfed at least one tumor target cell in a series of 24 experiments (mean 29.7%± 18.3%). Control irrelevant IgG1k monoclonal antibodies (anti-erbB2 trastuzumab and anti-EGFR cetuximab) did not mediate phagocytosis, and rituximab did not lead to ingestion of CD20 negative cells, demonstrating the specificity of the assay. Phagocytosis was already maximal at around 0.1 μg/ml rituximab concentration. In contrast complement activation required Mab concentration of at least 1 μg/ml. Thus phagocytosis, like ADCC, is active at about 10 fold lower MAb concentrations than complement triggering. Levels of CD20 expression on targets did not significantly affect phagocytosis. The role of different FcγRs was also investigated by addition 5 μg/ml blocking antibodies to CD16, CD32 and CD64. All 3 blocking Mabs reduced significantly phagocytosis (by 45%, 42% and 40% respectively with respect to control). Inhibition increased to 64% in presence of all 3 antibodies. Since previous data had suggested a role of the Val/Phe polymorphism at position 158 of CD16A in the clinical response of lymphoma patients to rituximab as well as in NK-mediated ADCC, we investigated whether this polymorphism also affected phagocytosis. No significant differences in dose response curves were observed using macrophages from either Val-Val or Phe-Phe homozygotes. Perhaps surprisingly, concomitant complement activation induced by addition of human serum did not increase phagocytosis. Whether human macrophages can also mediate antibody dependent cellular cytotoxicity (ADCC) was also studied. CLL or BJAB cells were labeled with Calcein-AM and ADCC measured as released fluorescence after 4 hours at 37°C. Macrophages were unable to mediate ADCC in presence of rituximab even following treatment with IFNγ (100 U/ml) for 48 hours. We conclude that macrophages efficiently mediate phagocytosis but not ADCC in presence of low concentrations of rituximab.