TAPBPR bridges UDP-glucose:glycoprotein glucosyltransferase 1 onto MHC class I to provide quality control in the antigen presentation pathway
- University of Cambridge, United Kingdom
- University of Southampton, United Kingdom
- Institute of Medical Sciences, University of Aberdeen, United Kingdom
- Eberhard Karls University Tübingen, Germany
Figures
Figure 1
TAPBPR associates with UDP-glucose:glycoprotein glucosyltransferase 1
(A and B) TAPBPR or (C) tapasin were immunoprecipitated using PeTe4 or Pasta1, respectively, from (A) IFN-γ-treated HeLaM, TAPBPR KO, tapasin KO, and double tapasin KO/TAPBPR KO HeLaM cells or (B …
Figure 2 with 1 supplement
C94 in TAPBPR is not involved in an intramolecular disulphide bond
(A) Amino acid sequences of human TAPBPR (NP_060479.3) and tapasin (AAC20076.1) were aligned using ClustalW. Cysteine residues are marked in yellow boxes, with the cysteine in tapasin that interacts …
Figure 2—figure supplement 1
Transduction efficiency of the cysteine-mutant panel into HeLaM cells
(A) Bar graph showing the percentages of GFP-expressing cells for the various TAPBPR cysteine mutants transduced into HeLaM cells. (B) Bar graph showing mean fluorescence intensity for GFP …
Figure 3 with 2 supplements
TAPBPR binds to UDP-glucose:glycoprotein glucosyltransferase 1 in a C94-dependent manner, but not via a disulphide bond and bridges UDP-glucose:glycoprotein glucosyltransferase 1 to MHC class I molecules
(A–C) TAPBPR or (A) HC10-reactive MHC class I were isolated by immunoprecipitation from (A and B) HeLaM-TAPBPRKO cells (HeLaMKO) and HeLaMKO cells reconstituted with either TAPBPRWT, TAPBPRC94A, or …
Figure 3—figure supplement 1
Cysteine residues are conserved in TAPBPR across different species
Amino acid sequences of TAPBPR in human (Homo, NP_060479.3), mouse (Mus, NP_663366.2), rat (Rattus, NP_001100092.1), chicken (Gallus, NP_001026543.1), rainbow trout (Oncorhynchus, NP_001118026.1), …
Figure 3—figure supplement 2
Residues in the helix next to C94 in TAPBPR influence UDP-glucose:glycoprotein glucosyltransferase 1 association
(A) Location of a predicted helix next to residues C94 in our FFAS model for TAPBPR. (B) TAPBPR was immunoprecipitated from HeLaM and HeLaM cells expressing TAPBPRWT, TAPBPRUBS1 (I83K, E87K), TAPBPRU…
Figure 4 with 1 supplement
TAPBPRC94A still functions as a peptide editor in vitro
(A) Dissociation of the fluorescent peptide FLPSDC*FPSV from HLA-A*02:01 in the absence or presence of TAPBPRWT or TAPBPRC94A. 500 nM HLA-A02:01 molecules (refolded with UV-labile KILGFVFjV peptide, …
Figure 4—figure supplement 1
Comparison of the ability of FLPSDCFPSV or NLVPMVATV to inhibit binding of FLPSDC*FPSV to peptide-receptive HLA-A*02:01fos molecules
150 nM HLA-A02:01fos molecules (refolded with UV-labile KILGFVFjV peptide) were mixed with 1.5 µM human β2m and exposed to 366-nm UV light at 4°C for 20 min, and then incubated with 14 nM …
Figure 5 with 4 supplements
UDP-glucose:glycoprotein glucosyltransferase 1 bound to TAPBPR influences peptide selection on MHC class I molecules
(A) Cytofluorimetric analysis of W6/32-, HLA-A68-, and 4E-reactive MHC class I molecules on HeLaMKOcells (black line), HeLaMKOcells reconstituted with TAPBPRWT (blue line) or TAPBPRC94Acells (green …
-
Figure 5—source data 1
MHC class I peptide elution from IFN-γ treated HeLaMKOTAPBPRWT and HeLaMKOTAPBPRC94A
The total peptides list shows all of the peptides that were eluted from MHC class I molecules expressed in IFN-γ treated HeLaMKO (HeLa 7.9), HeLaMKOTAPBPRWT and HeLaMKOTAPBPRC94A cells using W6/32. The HLA-A peptide subgroup 1 list shows all of the peptides that were predicted to be HLA-A binders using NetMHC, comparing IFN-γ treated HeLaMKOTAPBPRWT and HeLaMKOTAPBPRC94A cells. The HLA-B peptide subgroup 1 list shows all of the peptides that were predicted to be HLA-B binders using NetMHC, comparing IFN-γ treated HeLaMKOTAPBPRWT and HeLaMKOTAPBPRC94A cells. The HLA-A peptide subgroup 2 list shows all of the peptides that were predicted to be HLA-A binders using NetMHC, comparing IFN-γ treated HeLaMKOTAPBPRWT with HeLaMKOTAPBPRC94A cells. All A-peptides that were identified in IFN-γ treated HeLaMKO (HeLa 7.9) cells were excluded from this analysis. The HLA-B peptide subgroup 2 list shows all of the peptides that were predicted to be HLA-B binders using NetMHC, comparing IFN-γ treated HeLaMKOTAPBPRWT with HeLaMKOTAPBPRC94A cells. All B-peptides that were identified in IFN-γ treated HeLaMKO (HeLa 7.9) cells were excluded from this analysis.
- https://doi.org/10.7554/eLife.23049.013
Figure 5—figure supplement 1
Surface expression of HLA-A68 upon IFN-γ treatment
Cytofluorimetric analysis of HLA-A68 on IFN-γ-treated HeLaMKO and IFN-γ-treated HeLaKOcells reconstituted with TAPBPRWT or TAPBPRC94A. The bar graph summarises the data generated from three …
Figure 5—figure supplement 2
WebLogo depictions of the peptide sequences of 9-mers isolated from TAPBPRWT- and TAPBPRC94A-expressing cells WT: wild-type.
https://doi.org/10.7554/eLife.23049.015
Figure 5—figure supplement 3
Statistical analysis of peptides isolated from cells expressing TAPBPRWT and TAPBPRC94A
(A) The reproducibility of peptide identifications in three out of three technical liquid chromatography tandem mass spectrometry runs is 55.21% for the TAPBPRWT and 51.51% for the TAPBPRC94A …
Figure 5—figure supplement 4
Associations between TAPBPR, MHC class I, and UDP-glucose:glycoprotein glucosyltransferase 1 in the HeLaM cell lines used to test peptide receptivity of HLA-A68
(A) TAPBPR and (B) HC10 immunoprecipitations from independently produced HeLaM, HeLaM-TAPBPRWT, or HeLaM-TAPBPRC94A cells (i.e. not in the TAPBPRKO background) are included to demonstrate that the …
Figure 6 with 2 supplements
The TAPBPR:UDP-glucose:glycoprotein glucosyltransferase 1 complex promotes reglucosylation of MHC class I molecules, enhancing their association with the peptide-loading complex
(A) Lysates were prepared from IFN-γ-treated HeLaM, HeLaMKO, HeLaMKOTAPBPRWT, and HeLaMKOTAPBPRC94A cells in 1% digitonin. After preclear, pulldowns were performed with GST/6xHis-tagged exogenous WT …
Figure 6—figure supplement 1
MHC class I molecules associated with GST-CRT
Scatter dot plots show (A) the fold change in the total GST-CRT-reactive MHC class I molecules relative to the IFN-γ-treated HeLaMKOcells and (B) the ratio of HLA-A68 to HLA-B15 associated with …
Figure 6—figure supplement 2
– Densitometry on the MHC class I molecules bound to tapasin and TAPBPR
Representative Cy5 total protein images of (A) tapasin and (B) TAPBPR immunoprecipitations obtained using the Amersham WB system used to generate the data displayed in Figure 6C and D. (C) …
Figure 7
Working model of the TAPBPR:UDP-glucose:glycoprotein glucosyltransferase 1 complex in the MHC class I presentation pathway.
In addition to (1) the loading and editing of peptides (shown in blue) via tapasin in the peptide-rich milieu of the peptide-loading complex (PLC), an environment that favours MHC class I molecules …
Tables
Table 1
Selected proteins identified in IgG-sepharose pulldowns on ZZ-TAPBPR
Affinity chromatography with IgG-sepharose was performed on HeLaM cells expressing a protein-A-tagged TAPBPR molecule (ZZ-TAPBPR) or HeLaM cells transduced with an empty vector (control). Immunoprecipitates were analysed by in gel tryptic digest followed by liquid chromatography-tandem mass spectrometry and data were processed using Scaffold. Identified proteins are shown with their exclusive unique peptide count, percentage coverage, and exclusive unique spectrum count as determined by Scaffold. Rank denotes the position when data are sorted by exclusive unique peptide count with all proteins present in the control removed. Pep: exclusive unique peptide count; Cov: percentage coverage; Count: exclusive unique spectrum count.
| Protein | Gene name | Control | ZZ-TAPBPR | Rank | ||
|---|---|---|---|---|---|---|
| Pep (Cov) | Count | Pep (Cov) | Count | |||
| Tapasin-related protein | TAPBPL | – | – | 8 (16) | 11 | 2 |
| HLA class 1, A-68 | HLA-A | – | – | 14 (35) | 21 | 1 |
| β-2-microglobulin | β2M | – | – | 1 (8.4) | 1 | 95 |
| UDP-glucose:glycoprotein glucosyltransferase 1 | UGGT1 | – | – | 10 (7.3) | 10 | 3 |
Table 2
Selected proteins identified in TAPBPR co-immunoprecipitates
TAPBPR was immunuoprecipitated using PeTe4 from IFN-γ-treated HeLaM-TAPBPRKO(HeLaMKO) cells reconstituted with either TAPBPRWT or TAPBPRC94A. Immunoprecipitates were analysed by in gel tryptic digest followed by liquid chromatography-tandem mass spectrometry and data were processed using Scaffold. Identified proteins are shown with their exclusive unique peptide count, total percentage coverage, and exclusive unique spectrum count as determined by Scaffold. Pep: exclusive unique peptide count; Cov: percentage coverage; Count: exclusive unique spectrum count.
| Protein | Gene name | TAPBPRWT | TAPBPRC94A | ||
|---|---|---|---|---|---|
| Pep (Cov) | Count | Pep (Cov) | Count | ||
| Tapasin-related protein | TAPBPL | 32 (43) | 54 | 29 (47) | 50 |
| HLA class 1, A-68 | HLA-A | 50 (64) | 88 | 41 (59) | 70 |
| β-2-microglobulin | β2M | 4 (46) | 7 | 1 (8.4) | 1 |
| UDP-glucose:glycoprotein glucosyltransferase 1 | UGGT1 | 19 (11) | 25 | – | – |
Table 3
Primer sequences used for the mutation of individual cysteine residues to alanine in TAPBPR
| Name | Primers used for site-directed mutagenesis | Predicted TAPBPR domain |
|---|---|---|
| C18A | 5'-CAGTGGACGTGGTCCTAGACGCTTTCCTGGTGAAGGACGGTG-3' 5'-CACCGTCCTTCACCAGGAAAGCGTCTAGGACCACGTCCACGT-3' | Unique N-terminal |
| C94A | 5'-GAGGCCTTGCTCCATGCTGACGCCAGTGGGAAGGAGGTGACCTG-3' 5'-CAGGTCACCTCCTTCCCACTGGCGTCAGCATGGAGCAAGGCCTC-3' | |
| C101A | 5'-CTGCAGTGGGAAGGAGGTGACCGCTGAGATCTCCCGCTACTTTCTC-3' 5'-GAGAAAGATGCGGGAGATCTCAGCGGTCACCTCCTTCCCACTGCAG-3' | |
| C191A | 5'-GGTCCTCAGCCTCCTTGGACGCTGGCTTCTCCATGGCACCGG-3' 5'-CCGGTGCCATGGAGAAGCCAGCGTCCAAGGAGGCTGAGGACC-3' | IgV domain |
| C262A | 5'-CAGGACGAGGGGACCTACATTGCCCAGATCACCACCTCTCTGTAC-3' 5'-GTACAGAGAGGTGGTGATCTGGGCAATGTAGGTCCCCTCGTCCTG-3' | |
| C300A | 5'-GCTCTGCTGCCCACCCTCATCGCCGACATTGCTGGCTATTACC-3' 5'-GGTAATAGCCAGCAATGTCGGCGATGAGGGTGGGCAGCAGAGC-3' | IgC domain |
| C361A | 5'-CTGCAGGTGCAACTTACACCGCCCAGGTCACACACATCTCTC-3' 5'-GAGAGATGTGTGTGACCTGGGCGGTGTAAGTGGCACCTGCAG-3' | |
| C427A | 5'-GAACGCTGGGAGACCACTTCCGCTGCTGACACACAGAGCTCCC-3' 5'-GGGAGCTCTGTGTGTCAGCAGCGGAAGTGGTCTCCCAGCGTTC-3' | Cytoplasmic tail |
Additional files
-
Supplementary file 1
Primers used to generated TAPBPRUBS1- and TAPBPRUBS2-variant molecules
To create the UDP-glucose:glycoprotein glucosyltransferase 1 binding site mutants, site-directed mutagenesis was performed on untagged TAPBPR in pCR-Blunt II-TOPO using Quik-Change site-directed mutagenesis (Stratagene) together with the primers specified in this table. The resultant TAPBPRUBS1 (I83K and E87K) and TAPBPRUBS2 (E87K, L90K, H91S, and D93R) variants were subsequently cloned into pHRSIN-C56W-UbEM and transduced into HeLaM cells.
- https://doi.org/10.7554/eLife.23049.023
-
Source code 1
A MATLAB script was developed and applied in order to calculate the densities of the HLA-A68 and -B15 bands bound to tapasin in Pasta1 immunoprecipitation experiments.
The maximum peak corresponding to B15 was aligned relative to that of the wild-type (WT) track because the separation of B15 and A68 is most distinctive. The separation point between B15 and A68 was identified in the WT track as the minimum between the two peaks. This distance between the B15 peak and the B15/A68 separation point as found in WT was calculated and applied to the alignment positions on the knockout and C94A tracks in order to separate B15 and A68 in these two tracks. Areas under the respective curves generated the densities of the corresponding MHC class I molecules.
- https://doi.org/10.7554/eLife.23049.024
