The ERAP1 active site cannot productively access the N-terminus of antigenic peptide precursors stably bound onto MHC class I

Abstract

Processing of N-terminally elongated antigenic peptide precursors by Endoplasmic Reticulum Aminopeptidase 1 (ERAP1) is a key step in antigen presentation and the adaptive immune response. Although ERAP1 can efficiently process long peptides in solution, it has been proposed that it can also process peptides bound onto Major Histocompatibility Complex I molecules (MHCI). In a previous study, we suggested that the occasionally observed "ontο MHCI" trimming by ERAP1 is likely due to fast peptide dissociation followed by solution trimming, rather than direct action of ERAP1 onto the MHCI complex. However, other groups have proposed that ERAP1 can trim peptides covalently bound onto MHCI, which would preclude peptide dissociation. To explore this interaction, we constructed disulfide-linked MHCI-peptide complexes using HLA-B*08 and a 12mer kinetically labile peptide, or a 16mer carrying a phosphinic transition-state analogue N-terminus with high-affinity for ERAP1. Kinetic and biochemical analyses suggested that while both peptides could access the ERAP1 active site when free in solution, they were unable to do so when tethered in the MHCI binding groove. Our results suggest that MHCI binding protects, rather than promotes, antigenic peptide precursor trimming by ERAP1 and thus solution trimming is the more likely model of antigenic peptide processing.

Figure 1

Construction and validation of a disulfide-linked HLA-B*08/peptide complex. Panel (A), molecular model of the HLA-B*08(E76C)-ARAALRSRYWCI complex (peptide shown in green sticks, disulfide bond between the HLA residue Cys76 and peptide residue Cys11, is indicated). Panel (B), size-exclusion chromatogram after refolding of HLA-B*08(E76C) with the peptide ARAALRSRYWCI. Panel (C), SDS-PAGE of peaks A and B of chromatogram shown in Panel (B). Panel (D), MALDI-TOF–MS of purified complex in the presence of 100 mM DTT. Panel (E), MALDI-TOF–MS of purified complex in the absence of DTT. Panel (F), 1st derivative of thermal shift assay of purified complex (three repetitions shown). Panel (G), kinetic analysis of binding of SYPRO Οrange to purified complex HLA-B*08(E76C)-ARAALRSRYWCI (each data point is the average of three repetitions). Panel (H) Kinetic analysis of binding of SYPRO Οrange to non-covalent complex HLA-B*08 – ARAALRSRYWAI (each data point is the average of two repetitions).

Figure 2

Stability of HLA-B*08(E76C)-ARAALRSRYWCI versus peptide exchange and ERAP1 trimming. Panel (A) Νative-PAGE of HLA-B*08(E76C)-ARAALRSRYWCI mixed with excess ALRSRYWAI peptide: lane 1, complex incubated for 60 min at RT; lane 2, complex incubated with 100 μΜ ALRSRYWAI peptide for 60 min at RT; lane 3, complex without incubation; lane 4, HLA-B*08(E76C)-ALRSRYWCI complex; lane 5, HLA-B*08-ARAALRSRYWAI complex. Panel (B) lane 1, HLA-B*08(E76C)-ARAALRSRYWCI complex incubated at RT for 60 min; lane 2, HLA-B*08(E76C)-ARAALRSRYWCI complex + 100 μΜ ALRSRYWAI peptide, 60 min incubation; lane 3, HLA-B*08(E76C)-ARAALRSRYWCI + 100 μΜ DTT, 60 min incubation; lane 4, HLA-B*08(E76C)-ARAALRSRYWCI + 100 μΜ DTT + 100 μΜ ALRSRYWAI peptide; lane 5, HLA-B*08(E76C)-ALRSRYWCI; lane 6, HLA-B*08-ARAALRSRYWAI. Panel (C) Native PAGE of HLA-B*08(E76C)-ARAALRSRYWCI (lanes 1–4) or HLA-B*08-ARAALRSRYWAI (lanes 5–8) complexes incubated for 0–120 min with 20 nM ERAP1 (full-length gel is presented in Supplementary Fig. 1). Panel (D) MALDI-TOF–MS analysis of the HLA-B*08(E76C)-ARAALRSRYWCI complex incubated with ERAP1.

Figure 3

Construction of a disulfide-linked HLA-B*08/ phosphinic peptide complex. Panel (A) chemical structure of phosphinic pseudopeptide DG080. Panel (B) size-exclusion chromatogram of the refolding mixture of HLA-B*08(E76C) with DG080. Panel (C) SDS-PAGE analysis of peak A from the size-exclusion chromatogram shown in panel (B). Panel (D) MALDI-TOF–MS analysis of purified complex in the presence of 100 mM DTT. Panel (E) MALDI-TOF–MS analysis of purified complex in the absence of DTT. Panel (F) 1st derivative of thermal shift assay of purified complex (two representative experiments shown).

Figure 4

Characterization of the interaction between DG080 and the active site of ERAP1. Panel (A) effect of titration of free DG080 peptide on ERAP1 enzymatic activity (each point is the average of two measurements). Panel (B) time-dependent hydrolysis of fluorigenic ERAP1 substrate in the presence of 1 μΜ of either free DG080 or HLA-B*08 (E76C)-DG080 complex (two measurements per condition are shown). Panel (C) native-PAGE analysis of the stability of HLA-B*08(E76C)-DG080 complex: lane 1, complex incubated at 4 °C for 1 h; lane 2, complex incubated at 37 °C for 1 h; lane 3, complex incubated at 37 °C for 1 h in the presence of 20 nM ERAP1; lane 5, complex incubated at 37 °C for 1 h; lane 6, complex incubated at 37 °C for 1 h in the presence of 100 μM of peptide ALRSRYWAI; lane 7, HLA-B*08 (E76C)-ALRSRYWCI complex (full-length gel is presented in Supplementary Fig. 2). Panel (D) Native-PAGE of ERAP1 and HLA-B*08 (E76C)-DG080 complex: lane 1, ERAP1; lane 2, HLA-B*08 (E76C)-DG080 complex; lane 3, equimolar mixture of ERAP1 and HLA-B*08 (E76C)-DG080.

Figure 5

Length distribution of antigenic peptides that bind to HLA-B*08:01 and HLA-A*02:01. Data were extracted from the immune epitope database and analysis resource (www.iedb.org), 7231 total peptides for HLA-B*08:01 and 33,362 total peptides for HLA-A*02:01. Frequency distribution analysis was performed with Graphpad Prism 8.0™.