Fig 2: ATP13A1 deficiency affects MR1 levels at or immediately after protein synthesis.A and B, the THP-1 parent clone, the ATP13A1 KO clone 2, and the ATP13A1-sufficient clone 15 were incubated with MG-132, thapsigargin (TG), or NMS-873 for 4 h. Protein lysates were separated by SDS-PAGE and immuno-blotted for HERPUD1 (A, green) or MR1 (B, green) with GAPDH as the loading control (A and B, red). M denotes protein marker and molecular weights of visible marker bands are indicated in kDa to the left of the blot. Quantification of the blots is shown to the right. Dashed lines indicate 100% compared with DMSO. Western blots are representative of two separate experiments for MG-132 and TG (repeat not quantified). C and D, THP-1 parent clone, ATP13A1 KO clone 2, and ATP13A1-sufficient clone 15 overexpressing HA-tagged MR1 were pulsed with radioactively labeled cysteine and methionine for 1 min and then chased in the presence of cold Cys/Met and 100 µM cycloheximide for 0, 2, or 4 min before lysis and immunoprecipitation with anti-HA antibody 12CA5 (C) and then anti-MHC I antibody W6/32 (D). Immunoprecipitated proteins were separated by SDS-PAGE, transferred onto PVDF membranes, and visualized by autoradiography. Cycloheximide was included during the pulse in the control sample (CHX pulse). Molecular weight indications were inferred based on protein marker on the PVDF membranes. ATP13A1 KO clones are highlighted in red. Quantification in C and D shows pooled data from two independent experiments with median and range. E, quantification at t = 0 min for the two experiments shown in C and data from an experiment with steady-state analysis only. Each symbol represents one experiment and median is shown for each group. Statistical significance of the difference was calculated using a paired t test. F, quantification at t = 0 min for the two experiments shown in D. ns = p > 0.05. Red symbols represent ATP13A1 KO cells, black symbols represent ATP13A1 sufficient cells. MHC, major histocompatibility complex.

Fig 3: DB28 competes with 5-OP-RU for MR1 binding. (A and B) THP1-MR1 cells were pulsed with 20 μg/mL DB28 for 2 h before addition of the indicated concentrations of Ac-6-FP (A) or 5-OP-RU (B). Cells were stained after 18 h. (C) THP1-MR1 cells were pulsed with 20 μg/mL DB28 2 h before, concurrently, or 2 h after addition of 5-OP-RU (5 μg/mL). The compounds were kept in culture for 7 h or were washed after 2 h, and cells were chased for 5 h before staining with anti-MR1 (26.5) antibody. Geo MFI ± SD of technical duplicates are plotted in each graph. Data (A–C) are representative of three experimental replicates. (D) THP1-MR1 cells were pulsed with 20 μg/mL DB28 and MG or 5-OP-RU in the presence (hatched bars) or absence (black bars) of 10 μM cycloheximide (CHX) to block protein synthesis. Plotted is the percentage of MR1 down-regulation for each experimental condition (n = 4 experimental replicates for MG, 3 for 5-OP-RU; multiple t test, **P ≤ 0.005). (E) MR1-GFP expression was induced in BEAS2B cells with doxycycline; after 16 h, cells were pulsed with DB28 (20 μg/mL) for 8 h in the presence or absence of doxycycline. Geo MFI ± SD of surface MR1 expression (Left) or percentage of MR1 down-regulation (Right) are plotted. n = 5 experimental replicates with doxycycline, 3 without (t test, n.s).

Fig 4: DB28 down-regulates MR1 from the cell surface of primary cells. CD2-depleted PBMC were incubated overnight with MG (50 µM), 5-OP-RU (1 µg/mL) with or without DB28 (20 µg/mL), and MR1 expression on the surface of gated live B cells (A and B) or monocytes (C and D) was determined by flow cytometry. Cumulative data of geometrical mean fluorescence intensity of four different blood donors (A and C); representative FACS histograms (B and D). The background staining of isotype control is shown in black.

Fig 5: Ternary structures of AF-7-TCR-MR1 inhibitor complexes. (A) Overlay of the TCR-MR1 ternary structures of 5-OP-RU (PDB ID: 6PUC), DB28 and NV18.1 compounds. MR1 and ß2m are colored gray and pale cyan, respectively. TCR chains and the ligands are colored as follows: 5-OP-RU, green; DB28, orange; and NV18.1, wheat. (B) The top view of MR1 cleft with the CDRa, CDRß loops is shown as cartoon and colored as in A. (C and D) Omit maps contoured at 3s (Fo - Fc map; green mesh) of DB28 (C) and NV18.1 (D) after stimulated annealing refinements in Phenix crystallography package. (E) Superposition of DB28, NV18.1, and 5-OP-RU metabolites within the binding cleft. (F and G) Detailed interaction of DB28 (F) and NV18.1 (G) within the A'-pocket of MR1 binding cleft. Intermolecular and intramolecular H-bonds are represented as black and deep-teal dashes, respectively. (H) Superposition of all pocket residues of DB28, NV18.1, and 5-OP-RU compounds highlighting small conformational differences of MR1 residues between various structures. The residues of MR1-DB28 and NV18.1 are colored gray, whereas the interacting residues of MR1-5-OP-RU are colored light blue. (I–K) Interaction between CDR3a loop and CDR3ß loop with MR1 presenting DB28 (I) and NV18.1 (J) and 5-OP-RU (K), respectively. The CDR3a loop is colored blue in all panels, where the CDR3ß loop is colored light pink in DB28 and NV18.1 structures, and lemon in 5-OP-RU complex.