Thiol-reactive drug substrates of human P-glycoprotein label the same sites to activate ATPase activity in membranes or dodecyl maltoside detergent micelles.

P-glycoprotein (P-gp, ABCB1) is an ABC drug pump that is clinically important because it is involved in multidrug resistance. Many studies have used purified P-gp in detergent (n-dodecyl-β-D-maltoside; DM) micelles to map the locations of the drug-binding sites. A potential problem is that DM could be a substrate and affect binding of drugs to P-gp. To test whether DM was a substrate of P-gp, we used an assay involving drug-rescue of the immature 150 kDa misprocessed P-gp mutant (L1260A) to show that DM is not substrate. By contrast, the detergents Triton X-100 or NP-35 were substrates because they rescued the L1260A P-gp mutant such that the major product was the mature 170 kDa protein. Cross-linking of mutant A80C/R741C in membranes can only be inhibited by the P-gp substrate tariquidar. We show that cross-linking A80C/R741C mutant was also inhibited by tariquidar in the presence of excess DM. This result suggests that the presence of DM did not affect the tariquidar-binding site. Similarly, the presence of DM did not alter the locations of other drug-binding sites since the thiol reactive forms of the substrates verapamil or rhodamine labeled the same sites in transmembrane segments 5 (I306C for verapamil) and 6 (F343C for rhodamine) whether P-gp was in native membranes or in detergent micelles. These results suggest that the presence of DM does not alter the locations of the P-gp drug-binding sites and that the detergent purified protein is suitable for mapping their locations using biochemical or structural assays.