Combined effects of ankylosing spondylitis-associated ERAP1 polymorphisms outside the catalytic and peptide-binding sites on the processing of natural HLA-B27 ligands

Abstract

ERAP1 polymorphism involving residues 528 and 575/725 is associated with ankylosing spondylitis among HLA-B27-positive individuals. We used four recombinant variants to address the combined effects of the K528R and D575N polymorphism on the processing of HLA-B27 ligands. The hydrolysis of a fluorogenic substrate, Arg-528/Asp-575 < Lys-528/Asp-575 < Arg-528/Asn-575 < Lys-528/Asn-575, indicated that the relative activity of variants carrying Arg-528 or Lys-528 depends on residue 575. Asp-575 conferred lower activity than Asn-575, but the difference depended on residue 528. The same hierarchy was observed with synthetic precursors of HLA-B27 ligands, but the effects were peptide-dependent. Sometimes the epitope yields were variant-specific at all times. For other peptides, concomitant generation and destruction led to similar epitope amounts with all the variants at long, but not at short, digestion times. The generation/destruction balance of two related HLA-B27 ligands was analyzed in vitro and in live cells. Their relative yields at long digestion times were comparable with those from HLA-B27-positive cells, suggesting that ERAP1 was a major determinant of the abundance of these peptides in vivo. The hydrolysis of fluorogenic and peptide substrates by an HLA-B27 ligand or a shorter peptide, respectively, was increasingly inhibited as a function of ERAP1 activity, indicating that residues 528 and 575 affect substrate inhibition of ERAP1 trimming. The significant and complex effects of co-occurring ERAP1 polymorphisms on multiple HLA-B27 ligands, and their potential to alter the immunological and pathogenetic features of HLA-B27 as a function of the ERAP1 context, explain the epistatic association of both molecules in ankylosing spondylitis.

Keywords: Aminopeptidase; Antigen Processing; Arthritis; Major Histocompatibility Complex (MHC); Pathogenesis; Peptides.

FIGURE 1.

Hydrolytic activity of ERAP1 variants toward L-AMC. The indicated ERAP1 variants were incubated with the fluorogenic substrate at an E/S ratio of 1:30 (w/w), and the fluorescence intensity of the generated fluorophore was measured as a function of time. The data are means of 10–13 experiments.

FIGURE 2.

Digestion rate of synthetic peptide precursors of HLA-B27 ligands by ERAP1 variants. A, combined digestion (%) of peptide species longer than the natural ligand, excluding the original substrate. It was calculated as: 100% − the combined yield of the corresponding species; B, yield of the natural ligand; C, yield of peptides shorter than the natural ligand, resulting from destructive cleavages. ♦, Arg-528/Asp-575; ■, Lys-528/Asp-575; ▴, Arg-528/Asn-575; ●, Lys-528/Asn-575. Yields are relative to the total amount of peptide, which was estimated as the added intensity of the ion peaks corresponding to each peptide species in the MALDI-TOF MS spectrum of the digestion mixture. The data are mean ± S.D. of 3–5 experiments. Note that for any given substrate and time point, the value in A is essentially accounted for by the added yield of the corresponding ligand and its digestion product(s) in B and C, respectively.

FIGURE 3.

Comparison of the digestion of a synthetic peptide substrate analyzed by MALDI-TOF MS and HPLC. A, peptide 7 was digested with Arg-528/Asp-575 at the indicated times in parallel experiments, and the digestion mixtures were analyzed by MALDI-TOF and HPLC, respectively. Conventions are as in Fig. 2. The left graph shows the combined digestion of peptide species longer than the natural ligand, excluding peptide 7 itself. Digestion was calculated as 100% − the combined yield of the corresponding species. The middle and right graphs show the % yield of the natural ligand and of shorter peptide(s), respectively. B, same substrate was digested with Lys-528/Asn-575 and analyzed as in A. The data in each panel are from a single experiment, but it was carried out separately for samples to be analyzed by either MS or HPLC.

FIGURE 4.

Identification of RRYQKSTELLIR and RRYQKSTEL from C1R-B*27:05 cells. A, MALDI-TOF MS spectra (left panel) from the HPLC fractions 150 and 121, respectively, corresponding to the maximum of the elution profile of each peptide during the fractionation of the B*27:05-bound peptide pool. Only the relevant ion peaks are shown. The sequence of both ligands was determined by MALDI-TOF/TOF MS/MS (right panel); B, MALDI-TOF/TOF MS/MS spectra of the corresponding synthetic peptides.

FIGURE 5.

Digestion of two related HLA-B27 ligands of distinct length by ERAP1 variants. The indicated peptides were digested at an E/S ratio 1:10 (w/w) at various times with the following ERAP1 variants: Arg-528/Asp-575 (♦); Lys-528/Asp-575 (■); Arg-528/Asn-575 (▴); and Lys-528/Asn-575 (●).The data are means ± S.D. of three experiments.

FIGURE 6.

Influence of ERAP1 polymorphism on the inhibition of L-AMC hydrolysis by a natural HLA-B27 ligand. A, L-AMC was incubated at 100 μm concentration, with the indicated ERAP1 variants at an E/S ratio of 1:30 (w/w), in the presence of 100, 75, or 50 μm of RRYQKSTEL for 1 h. The fluorescence intensity of the processed fluorogenic substrate in the absence (black) or in the presence of peptide (white) is indicated. The figures above the histograms indicate the % inhibition of L-AMC hydrolysis. B, L-AMC was incubated for 1 h at 100 μm concentration, with the indicated ERAP1 variants, at an E/S ratio of 1:30 (w/w), in the absence (black) or in the presence (white) of 100 μm of the RYQKSTEL octamer, resulting from the digestion of the nonamer used in A. Figures above the histogram bars of the most active variants indicate the % activation of L-AMC hydrolysis. C, percent yield of RYQKSTEL after incubation of 100 μm RRYQKSTEL with the indicated ERAP1 variants in exactly the same conditions as in A. The data are means ± S.D. of three experiments.

FIGURE 7.

Influence of ERAP1 polymorphism on the inhibition of peptide trimming by a short peptide. Peptide 6 (LYSELARYGKSPYLY) was incubated with the indicated ERAP1 variants in the presence of various amounts of RYQKSTEL for 4 h at an E/S ratio of 1:10. In each panel, the % yield of the precursor substrate, relative to the total digestion mixture, was plotted for the indicated substrate/inhibitor ratios. The figures above the histograms indicate the % inhibition of substrate hydrolysis. The data are means ± S.D. of three experiments.