Conserved Walker A Cysteines 431 and 1074 in Human P-Glycoprotein Are Accessible to Thiol-Specific Agents in the Apo and ADP-Vanadate Trapped Conformations

Abstract

P-glycoprotein (P-gp) is an ATP-binding cassette efflux transporter involved in the development of multidrug resistance in cancer cells. Although the mechanism of P-gp efflux has been extensively studied, aspects of its catalytic and transport cycle are still unclear. In this study, we used the conserved C431 and C1074 in the Walker A motif of nucleotide-binding domains (NBDs) as reporter sites to interrogate the interaction between the two NBDs during the catalytic cycle. Disulfide crosslinking of the C431 and C1074 residues in a cysless background can be observed in the presence of M14M and M17M crosslinkers, which have spacer arm lengths of 20 Å and 25 Å, respectively. However, crosslinking with both crosslinkers was prevented in the ADP-vanadate trapped (closed) conformation. Both C431 and C1074 alone or together (double mutant) in the apo and closed conformations were found to be accessible to fluorescein-5-maleimide (FM) and methanethiosulfonate derivatives of rhodamine and verapamil. In addition, C1074 showed 1.4-fold and 2-fold higher degrees of FM labeling compared to C431 in the apo and closed conformations, respectively, demonstrating that C1074 is more accessible than C431 in both conformations. In the presence of P-gp substrates, crosslinking with M17M is still observed, suggesting that binding of substrate in the transmembrane domains does not change the accessibility of the cysteines in the NBDs. In summary, the cysteines in the Walker A motifs of NBDs of human P-gp are differentially accessible to thiol-specific agents in the apo and closed conformations.

About P‐glycoprotein: a new drugable domain is emerging from structural data

“…Even mutated residues L175C/N820C, located at TMH3 and TMH9, respectively, could be cross‐linked after treatment with 1,1‐methanediyl bismethanethiosulfonate (M1M, 4 Å) in the absence of any drug substrate or ATP, suggesting again that in its native environment, both residues are already in close proximity . More recently, Loo et al demonstrated that mutant P‐gp I517C/I1050C was able to be cross‐linked with both short (M4M, 7.8 Å) and long (M17M, 22 Å) thiol‐reactive cross‐linkers; Wen et al calculated mean NBD distances from probability distributions obtained from double electron–electron resonance (DEER) experiments, locating them within a 30–40 Å range; and Sim et al showed that cysteine mutants C431 (NBD1) and C1074 (NBD2) in the Walker A motif of human P‐gp could be cross‐linked by disulfide cross‐linking agents such as M14M and M17M, allowing the authors to infer NBD distances between a 20 and 25 Å range (Table ) …”

“…The first ones to be identified were based on the efflux of two fluorescent dyes, Hoechst 33342 (H‐site) and Rhodamine‐123 (R‐site) . As these molecules achieve higher concentrations within the lipid bilayer, it was proposed that both dyes are able to accumulate within the membrane, to diffuse toward the transporter and permeate into the internal drug‐binding pocket where they interact on one or both sites . With the publication of the first murine crystallographic structure, an additional drug‐binding site was defined by two molecules cocrystallized within the drug‐binding pocket (QZ59 isomers; M‐site), involving residues facing the drug‐binding pocket that were experimentally proven to interact with verapamil .…”

“…However, even high‐throughput, third‐generation modulators acting at the nanomolar range (e.g., tariquidar, Figure ) failed to show any clinical advantage when co‐administered with an antitumor drug. In addition, cellular toxicity and alterations in the pharmacokinetics of antitumor drugs also further impaired any clinical usage for MDR reversers . Thus, the problem seems to probably reside elsewhere, namely in the polyspecificity of the drug‐binding pocket.…”

“…56 More recently, Loo et al 57 demonstrated that mutant P-gp I517C/I1050C was able to be crosslinked with both short (M4M, 7.8 Å) and long (M17M, 22 Å) thiol-reactive cross-linkers; Wen et al 58 calculated mean NBD distances from probability distributions obtained from double electronelectron resonance (DEER) experiments, locating them within a 30-40 Å range; and Sim et al showed that cysteine mutants C431 (NBD1) and C1074 (NBD2) in the Walker A motif of human P-gp could be cross-linked by disulfide crosslinking agents such as M14M and M17M, allowing the authors to infer NBD distances between a 20 and 25 Å range ( Table 1). 59 In addition, the fact that P-gp and the surrounding lipid bilayer work together as single entity is frequently neglected. 60 In the MDR phenotype, the formation of lipid rafts is often upregulated, and Pgp is almost entirely found in these detergentresistant membranes (DRMs) due to the presence of an extracellular oligosaccharide side chain and long transmembrane domains (Table 1).…”

“…In addition, cellular toxicity and alterations in the pharmacokinetics of antitumor drugs also further impaired any clinical usage for MDR reversers. 59 Thus, the problem seems to probably reside elsewhere, namely in the polyspecificity of the drug-binding pocket. From structural data, the high structural flexibility described for P-gp suggested the existence of multiple redundant drug-binding sites that recognize substrates through an induced-fit mechanism.…”