Standardization of a single-cell assay for sensitive detection of multidrug resistance protein expression in normal and malignant cells in archival clinical samples.

Multidrug resistance protein (MRP), like P170, confers multidrug resistance, but its clinical relevance is uncertain, whereas P170 is an accepted cause of chemotherapy failure for which ongoing reversal trials are being conducted. Because such trials have been only modestly successful, we must investigate alternative drug resistance mechanisms such as MRP, which is poorly blocked by P170 inhibitors. The significance of MRP has remained undefined because MRP mRNA is difficult to assay in archival material, does not necessarily reflect MRP levels, and is widely expressed in normal or hematopoietic cells within tumors and bone marrow. Because conventional immunoblot or immunocytochemistry may not be sensitive enough to detect low or heterogeneous MRP expression in clinical samples, we elected to score MRP in single tumor cells by modifying our P170 assays that have proven valuable for correlating P170 expression with the outcome of pediatric cancer chemotherapy. We enhanced the signal-to-noise ratio with several peroxidase-tagged secondary antibody layers and staining refinements, standardizing the assay with MRP-negative and MRP-positive but P170-negative transfected or drug-selected controls in which MRP was quantified by immunoblot. We confirmed sensitivity by staining a very low MRP-expressing revertant line and "mixed" samples containing small numbers of positive cells; we confirmed specificity by applying two antibodies directed against separate MRP epitopes. We examined neuroblastoma, osteosarcoma, rhabdomyosarcoma, and retinoblastoma samples, identifying MRP-positive malignant cells, which were distinguishable from MRP-positive normal cells. This assay may be valuable for early diagnosis of low but potentially important MRP expression, which would allow timely application of alternative therapy, perhaps with MRP-specific blockers.