Relationship Between Post-Translational Modification of CD20 Protein and the Responsiveness to Rituximab Treatment

Rituximab is a murine/human chimeric-anti-CD20 monoclonal antibody that has become a key molecular-targeting drug for CD20-positive B-cell lymphomas. Although combination chemotherapy with rituximab has provided remarkably favorable results for CD20-positive B-cell lymphoma patients, acquired resistance to rituximab has become a considerable problem. Several mechanisms of resistance have been predicted, but the clinical significance of those mechanisms has remained unclear. Previously, at the last ASH Meeting, we showed that down-regulation of CD20 protein expression after using rituximab-containing chemotherapy is one of the critical mechanisms of rituximab resistance, and some epigenetic mechanisms, in part, were related to the aberrant down-regulation of MS4A1 (CD20) gene transcription. On the other hand, we have also found that some patients show resistance to rituximab even in the presence of CD20 protein expression. With this backgrounds in mind, we here investigated he relationship between CD20 protein expression condition and the responsiveness to treatment with rituximab. First, using B-cell lymphoma cell lines and primary lymphoma cells obtained from B-cell lymphoma patients, the CD20 protein expression condition was confirmed by immuno-histochemistry, flow cytometry (FCM), and immunoblotting (IB). In IB analysis, two different sizes of CD20 protein, upper (~37 kDa) and lower (~35 kDa), were confirmed. In vitro phosphatase assay suggested that the upper band is a phosphorylated band of the wild-type CD20 protein. The intensity of those two bands was measured, and the upper/lower (U/L) ratio was calculated. Interestingly, the U/L ratio in diffuse large B-cell lymphoma (DLBCL) cells was significantly lower (range; 0.1 to 0.3) than that of grade 1 and 2 follicular lymphoma (FL) and B-chronic lymphocytic leukemia (B-CLL) cells (0.3 to 0.9 and 0.5 to 0.9, respectively). Next, we tried to analyze the rituximab-inducing complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC) activities using in vitro chrome-releasing assay. Generally, the CDC/ADCC activity in the CD20(−) cells was significantly lower than that of CD20(+) rituximab-sensitive cells. In contrast, the CDC/ADCC activity in CD20(+) cells varied considerably, with the result that the protein expression level measured by FCM and IB was not always proportional to rituximab sensitivity. In one FL patient showing clinically rituximab resistance in spite of the CD20(+) phenotype, CDC/ADCC activity in vitro was also significantly lower than that of rituximab-sensitive control cells, and the U/L ratio was remarkably high (~1.7) compared to CD20(+) DLBCL and FL cells obtained from rituximab-sensitive patents. Our findings suggest that some kind of B-cell lymphoma, e.g. DLBCL, can be differently classified by CD20 protein expression pattern according to the U/L ratio, and the responsiveness to rituximab may not be determined only by the whole CD20 protein expression level. Furthermore, our data may also suggest a possibility that the responsiveness to rituximab is, in part, modulated by post-translational modification of CD20 protein in vivo.