Fig. 5: K/R polymorphism at position 528 modulates a conformationally dependent interdomain electrostatic interaction and regulates binding of small-molecule ERAP1 modulators that induce closing.
a Lys or Arg at position 528 makes a long-distance electrostatic interaction with Glu913, in addition to interactions with Asn 414 and His 448. Open and closed conformations for proteins with Lys528 or Arg528 are shown. The closed Arg528 conformation is modeled based on 6M8P, all others are from crystal structures as indicated. Closest interatomic distances between Lys/Arg528 and Glu913 are 7.8 Å (closed) and 13.5 Å (open), respectively. Lys528 stabilizes the closed ERAP1 conformation more than does Arg528. Pairwise electrostatic potentials between residues 528 and 913 shown in kT units for each structure, calculated using DelPhiForce. b DG013 exhibits greater potency for Lys528 ERAP1 than Arg528 ERAP1. Leucine-AMC hydrolysis rate was measured and normalized to control condition without DG013. Data were fit to a sigmoidal curve with constrained top and bottom = 100 and 0% activity respectively. Significance calculated using two-tailed ANOVA (*P < 0.0001). c Compound 3 exhibits greater potency for Lys528 ERAP1 than Arg528 ERAP1. L-AMC hydrolysis rate was measured and normalized to control condition without compound 3. Data were fit to a sigmoidal curve with constrained bottom = 100% activity. Significance calculated using two-tailed ANOVA (***P < 0.0001). d Compound 2 does not show ERAP1 variant-specific potency. Leucine-AMC hydrolysis rate was measured and normalized to control condition without compound 2. Data were fit as in (b). For binding studies with all the three small molecules (b–d), two independent experiments were performed with triplicate (n = 3) samples in each experiment and variation between the replicates are represented by error bars. For both Lys528 (allele II) and Arg528 (allele III) variants, ERAP1 conformation remains open in the absence of ligand and in presence of compound 2, as shown by Rg (e) or SAXS/WAXS model fitting (f). Both Arg/Lys528 variants of ERAP1 remain closed in the presence of compound 3 and DG013, as shown in the same analysis in (e) and (f). A single independent experiment (n = 3) was performed for Rg analysis on Lys528 and Arg528 variants of ERAP1 and error of linear fit for each independent experiment is shown as error bars (e). Source data are provided as a Source Data file.
