Restoration of CD20 mRNA and protein expression by treatment with the DNA methyltransferase inhibitor 5-Aza. (A) Primary B-cell lymphoma cells, which showed a CD20 protein-negative phenotype from UPN 3, were incubated with or without 5-Aza. Total RNA was prepared, and semiquantitative RT-PCR was performed. Restoration of CD20 mRNA expression after treatment with 5-Aza was observed in lane 3. As a positive control, tumor cells obtained at the initial diagnosis of that same patient were used in lane 1. (B) The CD20 protein-negative B-cell lymphoma cell line RRBL1,³²  which was derived from UPN 5 in Table 2, was incubated in culture medium with or without 5-Aza. After preparation of total RNA and whole-cell lysates from these cells, semiquantitative RT-PCR and immunoblotting (IB) were performed. Up-regulation of CD20 mRNA and protein expression was observed as shown in lane 2. (C) Quantitative RT-PCR was performed using the same mRNA as in (B). We observed an up-regulation of more than 10-fold in CD20 mRNA after treatment with 5-Aza (column 2). (D) RRBL1 cells were treated with 5-Aza under the same conditions as in (B), and FCM analysis using anti-CD20 antibody was performed. After treatment with 5-Aza, 30.7% of RRBL1 cells showed a CD20-positive phenotype. Positive cells are shown with black lines, and the percentage of positive cells is also shown. (E) In vitro ADCC analysis using the ⁵¹Cr-release assay. Cells from the CD20-positive B-cell lymphoma/leukemia cell lines Daudi and RRBL1 treated with or without 5-Aza were used for this assay. In Daudi cells (○), but not in RRBL1 cells (□), cytotoxic activity was observed in the presence of rituximab in a dose-dependent manner. Partial restoration of rituximab sensitivity in RRBL1 cells was observed after treatment with 5-Aza (▲). Error bars indicate plus or minus 1 standard deviation.