Reanalysis of Immunopeptidomics Datasets Provides Mechanistic Insight into TAPBPR-Mediated Peptide Editing on HLA-A, -B and -C Molecules

Background Major histocompatibility class I (MHC-I, human leukocyte antigen [HLA]-I in humans) molecules present small fragments of the proteome on the cell surface for immunosurveillance, which is pivotal to control infected and malignant cells. Immunogenic peptides are generated and selected in the MHC-I antigen processing and presentation pathway. In this pathway, two homologous molecules, tapasin and TAPBPR, optimise the MHC-I peptide repertoire that is ultimately presented at the plasma membrane. Peptide exchange on HLA-I by human TAPBPR involves the flexible loop region K22-D35, with the leucine at position 30 (L30) involved in mediating peptide dissociation. However, our understanding of the exact molecular mechanisms governing TAPBPR-mediated peptide exchange on HLA-I allotypes remains incomplete. Methods Here, in-depth re-analyses of published immunopeptidomics datasets was used to further examine TAPBPR peptide editing activity and mechanism of action on HLA-I. The role of the TAPBPR editing loop in opening the HLA-I peptide binding groove was assessed using a molecular dynamics simulation. Results We show that TAPBPR shapes the peptide repertoire on HLA-A, -B and -C allotypes. The TAPBPR editing loop was not essential to allow HLA-I to adopt an open state. L30 in the TAPBPR editing loop was typically sufficient to mediate peptide repertoire restriction on the three HLA-I allotypes expressed by HeLa cells. TAPBPR was also able to load peptides onto HLA-I in a loop-dependent manner. Conclusions These results unify the previously hypothesised scoop loop and peptide trap mechanisms of TAPBPR-mediated peptide exchange, with the former involved in peptide filtering and the latter in peptide loading.

mechanism of action on HLA-I.The role of the TAPBPR editing loop in opening the HLA-I peptide binding groove was assessed using a molecular dynamics simulation.

Results
We show that TAPBPR shapes the peptide repertoire on HLA-A, -B and -C allotypes.The TAPBPR editing loop was not essential to allow HLA-I to adopt an open state.L30 in the TAPBPR editing loop was typically sufficient to mediate peptide repertoire restriction on the three HLA-I allotypes expressed by HeLa cells.TAPBPR was also able to load peptides onto HLA-I in a loop-dependent manner.

Conclusions
These results unify the previously hypothesised scoop loop and peptide trap mechanisms of TAPBPR-mediated peptide exchange, with the former involved in peptide filtering and the latter in peptide loading.

Introduction
Immunogenic peptide presentation on major histocompatibility class I (MHC-I) molecules (human leukocyte antigen [HLA] in humans) is essential for induction of cytotoxic CD8 + T cell responses against infected or malignant cells.Peptides are typically generated by proteasomal degradation and selected in the MHC-I antigen processing and presentation pathway.In this pathway, two homologous peptide editors, tapasin and TAPBPR, control and optimise the peptide repertoire that is ultimately presented at the cell surface.Tapasin functions as part of the multiprotein peptide loading complex (PLC) containing MHC-I, the peptide transporter associated with antigen processing (TAP), calreticulin, and ERp57 1,2 .Interactions within the PLC stabilises MHC-I in a peptidereceptive state and allows tapasin to load peptide on MHC-I [3][4][5][6] .More recently the tapasin-related molecule TAPBPR was discovered, which functions outside the PLC [7][8][9][10] .TAPBPR can directly catalyse peptide exchange on MHC-I and can recruit UDP-glucose:glycoprotein glucosyltransferase 1 (UGT-1), which subsequently reglucosylates peptide-receptive MHC-I to recycle it back to the PLC for tapasin-mediated peptide acquisition [11][12][13] .
HLA-I molecules are highly polymorphic, and it has previously been shown that some allotypes are more dependent on tapasin than others to acquire peptide 3,14,15 .Allotypes independent of tapasin generally present a broader repertoire 16 .While it has yet to be determined whether HLA-I molecules display any TAPBPR-dependency, TAPBPR binds more strongly to HLA-A compared to HLA-B and -C molecules 17 .Members of the HLA-A*02 and -A*24 superfamily are particularly susceptible to TAPBPR-mediated peptide editing in assays using recombinant TAPBPR 17 .Interestingly, allotypes that were identified as strong TAPBPR binders were previously identified as tapasin-independent 16,17 , suggesting there may be an inverse correlation between tapasin and TAPBPR-dependence in the generation of optimally loaded peptide-HLA-I complexes 18 .
Several structures of MHC-I in complex with peptide editors have provided insight into the mechanism of action (reviewed in 19,20).TAPBPR has been suggested to widen and stabilise the peptide binding groove to facilitate peptide exchange 21,22 .Molecular dynamics (MD) simulations indicated that TAPBPR is able to remodel the B pocket of empty MHC-I, causing the peptide binding cleft to enter a more peptide-receptive conformation and facilitating the binding of peptide 23 .A flexible loop region spanning residues K22-D35 has been identified in TAPBPR that resides at the interface of the MHC-I peptide binding groove 22 .It was proposed that this editing loop acts as a scoop loop and occupies the MHC-I F pocket in the peptide binding groove, where the C-terminus of a bound peptide would reside 22,24 .The leucine at position 30 (L30) in the TAPBPR editing loop was shown critical for peptide dissociation, potentially through interaction with the F pocket 25 .It was suggested that the TAPBPR scoop loop in the MHC-I peptide binding groove can only be outcompeted by a high affinity peptide, thereby filtering and optimising the peptide repertoire 22,24 .
Alternatively, it has been proposed that the TAPBPR editing loop does not enter the MHC-I peptide binding groove, but instead hovers above it to act as a peptide trap to promote peptide loading 26 .This was supported by MD simulations showing that the peptide C-terminus may be drawn into the F pocket by the arginine at position 27 (R27) in the TAPBPR K22-D35 loop interacting with the carboxyl group of a peptide 23 .These two seemingly contradicting hypotheses highlight that our understanding of the mechanisms of TAPBPR-mediated MHC-I peptide editing is incomplete.
Editing of the HLA-I peptide repertoire can experimentally be addressed using the mass spectrometry-based approach immunopeptidomics.This technique has been instrumental to enhance our understanding of TAPBPR as an HLA-I peptide editor, to explore the role of the TAPBPR-UGT1 interaction and to establish the importance of the editing loop 9,12,25 .For example, Hermann et al. used immunopeptidomics to show TAPBPR-mediated restriction of the peptide repertoire on HLA-A and -B molecules 9 .Furthermore, Ilca et al. produced immunopeptidomics datasets to explore the contribution L30 in the editing loop to TAPBPR-mediated peptide dissociation 25 .However, the datasets in these two publications also contained a wealth of additional information that remained unexplored.Here, we re-examine these published immunopeptidomics datasets to obtain a more comprehensive understanding of TAPBPR-mediated peptide selection on HLA-I.We show that TAPBPR shapes the peptide repertoire on HLA-A, -B and -C molecules, with the editing loop not only involved in filtering but also loading of peptides.While L30 was typically sufficient to mediate peptide filtering, we show additional or other residues in the TAPBPR editing loop likely contribute to peptide loading onto HLA-I.

HLA-I peptide analyses
HeLaM cells, either wild-type, depleted of TAPBPR or transduced to overexpress (OE) TAPBPR, were described previously 7,9 .HeLaM cells depleted of TAPBPR and transduced to express TAPBPR K22-D35 loop mutants were previously published 25 .Where indicated, cells were treated with 50 U/ml of IFN-γ (Peprotech, #300-02) at 37°C for 48-72h to induce endogenous TAPBPR expression.Cells were harvested (without or with 30min post-trypsin recovery in media at 37°C), followed by lysis and HLA-I-peptide complex affinity purification, ligandome detection by mass spectrometry, and initial dataset processing as previously published 9,25 .Peptide lists of predicted binders were further processed with original Python scripts using NumPy 27 , pandas 28 , matplotlib 29 , and seaborn 30 packages.

MD simulations
All-atom MD simulations were carried out with GROMACS v2021.5 31 .Simulations were carried out under the AMBER99SB-ILDN 32 protein forcefield and TIP3P 33 water model.SETTLE 34 and LINCS 35 constraint algorithms were applied to constrain water and protein molecules respectively.Systems were solvated to an overall neutral charge containing Na + and Cl -ions to yield a physiological concentration of 0.15M.Short range interactions were treated with a Verlet cut-off scheme, while van der Walls forces and long-range electrostatics were handled with the particle mesh Ewald method 36 with grid spacing of 1.2Å and cubic interpolation.A velocityrescaling thermostat 37 was used to maintain temperature constant at 300K with a coupling time constant of 0.1ps, using a thermodynamic ensemble of nPT.An isotropic Berensden barostat 38 was used to maintain a constant 1.0 bar pressure, with a coupling time constant of 0.5ps and 4.5×10 -5 bar -1 compressibility.The three-dimensional structure of the TAPBPR: MHC-I complex was obtained from PDB 5OPI 22 .PDB 3MRD 39 was aligned with 5OPI to generate the TAPBPR:HLA-A*02:01 complex.TAPBPR øloop was generated by the Rotamer tool of UCSF Chimera 40 , with each most probable rotamer (as determined by the Dunbrack library 41 ) implemented into the loop.The mutated residue and those within 5Å underwent 100 steps of energy minimization using an AMBER ff14SB forcefield 42 .

Datasets
The immunopeptidomics datasets have previously been described 9,25 and are available at the Dryad Digital Repository 43,44 .

Results
TAPBPR filters more peptides than it loads on HLA-A*68:02, -B*15:03 and -C*12:03 Although TAPBPR has been shown to preferentially bind HLA-A molecules, interactions with HLA-B and -C molecules have previously been observed 9,17 .To assess how TAPBPR impacts the peptide repertoire presented by each of the three allotypes, we re-examined a published immunopeptidomics dataset comparing wild-type (WT) HeLaM cells, which in the absence of IFN-γ treatment express negligible levels of TAPBPR, with cells overexpressing (OE) TAPBPR 9 .The total number of unique peptide sequences, i.e. the repertoire diversity, was decreased in TAPBPR OE cells compared to WT cells (Figure 1A).These findings were complemented by comparing the peptide repertoire eluted from WT HeLaM cells, treated with IFN-γ to induce endogenous TAPBPR expression, to IFN-γ-treated cells in which TAPBPR had been depleted (knock-out [KO] clone) 9 .TAPBPR-depleted cells showed an increased repertoire diversity compared to the WT cells, confirming that endogenous TAPBPR also restricts the HLA-I repertoire on HeLaM cells (Figure 1B).
We further assigned the peptides identified from these datasets into three different categories.Peptides detected in both WT and OE TAPBPR cells were considered TAPBPR-independent (Figure 1A).Peptides unique to untreated WT cells were presumed to be dissociated or prevented from loading (i.e.filtered) by TAPBPR as they were only detected in cells expressing negligible levels of TAPBPR (Figure 1A).Conversely, peptides unique to TAPBPR OE cells were assumed to require TAPBPR for loading onto HLA-I (Figure 1A).Similar categories were assigned in the comparison of IFN-γ-treated TAPBPR KO and WT cells (Figure 1B).Interestingly, in addition to peptide filtering, TAPBPR also appeared to assist in peptide loading on HLA-I.
The peptides dependent on TAPBPR for filtering or loading were subsequently assigned to HLA-A*68:02, -B*15:03 and -C*12:03.Both OE and endogenous TAPBPR filters more peptides than it loads, i.e. restricts the peptide repertoire, on all three allotypes expressed by HeLaM cells (Figure 1C-D).While TAPBPR-mediated restriction of the peptide repertoire on HLA-A*68:02 and -B*15:03 had previously been identified 9 , the role of TAPBPR in restricting the repertoire presented on HLA-C*12:03 was not originally characterised.

The TAPBPR editing loop contributes to peptide loading as well as peptide filtering
The role of the TAPBPR editing loop in peptide selection had previously been assessed by establishing the peptide repertoire from HeLaM TAPBPR KO cells transduced to express wild-type TAPBPR (TAPBPR WT ), TAPBPR with mutated residues of the loop domain (TAPBPR øloop ), or TAPB-PR øloop in which the leucine had been reconstituted at position 30 (TAPBPR øG30L ) (Table 1) 25 .Using two-way comparisons between the TAPBPR WT with either TAPBPR øloop or TAPBPR øG30L , a key role L30 in mediating peptide dissociation from HLA-I was revealed 11 .This was observed in two independent datasets which differed in the recovery period post-trypsin harvest (addressed in more detail below).As additional mechanisms for the TAPBPR editing loop in peptide editing were proposed more recently 26 , reanalysis of these datasets could shed further light into how the editing loop functions in the peptide selection.
In our reassessment, the HLA-I peptides shared between cells expressing TAPBPR WT or TAPBPR øloop were considered to represent peptides that were not dependent on TAPBPR and/or the editing loop (Figure 2A).Peptides unique to cells expressing TAPBPR WT were assumed to require the editing loop to be loaded on HLA-I since they were not eluted from cells expressing TAPBPR øloop .Conversely, peptides unique to cells expressing TAPBPR øloop were presumed to be filtered by the TAPBPR WT editing loop.While it was previously highlighted that the editing loop was involved in peptide filtering 24,25 , our reanalysis showed that it also contributes to peptide loading onto HLA-I molecules.Slightly more peptides were filtered than loaded by the editing loop (Figure 2A).Direct comparisons of the immunopeptidomes from cells expressing TAPBPR øloop or TAPBPR øG30L were not previously established.When we performed this analysis, we observed that the increase in repertoire size caused by mutating the residues in the editing loop (TAPBPR øloop ) was reverted by reconstitution of L30 (TAPBPR øG30L , Figure 2B).Thus, L30 appears to be sufficient to restrict the peptide repertoire diversity on the HLA-I molecules expressed by HeLa cells.
TAPBPR is still able to open the HLA-I binding cleft in the absence of the editing loop Peptide filtering by L30 likely requires binding of this residue in the HLA-I F pocket and is presumably dependent on opening of the peptide binding groove 22,24,25 .It was previously shown that TAPBPR causes HLA-I molecules to enter an open conformation and this conformational change was hypothesised to be mediated by the editing loop 21,22 .However, this hypothesis contrasts our observation that L30 can mediate repertoire restriction in absence of the other loop residues (Figure 2B).Therefore, an all-atom MD simulation was conducted to test whether the peptide binding groove was opened in absence of the wild-type editing loop residues.In line with our immunopeptidomics results, TAPBPR øloop retained its ability to open the peptide binding cleft of peptide-deficient HLA-A*02:01 (Figure 2C).Thus, regions of TAPBPR other than the editing loop likely contribute either independently or conjunctively to opening of the HLA-I peptide binding groove.

TAPBPR editing loop-dependent peptide filtering is predominantly mediated by L30
The previous publication only compared cells expressing TAPBPR WT to either TAPBPR øloop or TAPBPR øG30L -expressing cells 25 .To further dissect the role of the editing loop in shaping the diversity of the HLA-I peptide repertoire, we expanded this analysis by performing a three-way comparison of peptides eluted from cells expressing TAPBPR WT , TAPBPR øloop and TAPBPR øG30L .The peptides filtered by the editing loop (unique to TAPBPR øloop compared to TAPBPR WT ) where further subdivided by their presence or absence in the repertoire eluted from cells expressing TAPBPR øG30L (Figure 3A).Peptides unique to TAPBPR øloop (compared to TAPBPR WT ) and not identified in TAPBPR øG30L (n=295) were considered dependent on L30 for filtering (Figure 3A).The remaining 133 peptides in the unique to TAPBPR øloop subset, were assumed to be filtered in a TAPBPR editing loop-dependent but L30-independent manner (Figure 3A).The peptides in these subgroups were assigned to the HLA-I molecules expressed by HeLaM, which revealed that peptide filtering on all three HLA-I allotypes was mostly L30-dependent (Figure 3B).
L30 is typically not sufficient to mediate editing loopdependent peptide loading Next, the role of the editing loop in peptide loading was further assessed using the peptides unique to TAPBPR WT (compared to TAPBPR øloop ) (Figure 3A).Peptides unique to TAPBPR WT and not identified in TAPBPR øG30L (n=198) were considered peptides where L30 alone was not sufficient to load them onto HLA-I, i.e. additional or other residues in the editing loop contribute to peptide loading.For the remaining 147 peptides in the subset unique to TAPBPR WT , L30 was presumed sufficient for HLA-I loading (Figure 3A).Most peptides loaded onto HLA-A*68:02 and -B*15:03 required the TAPBPR WT editing loop, suggesting that L30 alone was not sufficient for loading (Figure 3C).This was not the case for peptides assigned to HLA-C*12:03, for which L30 seemed sufficient to load most peptides (Figure 3C).However, the difference and the peptide number are small.Taken together, in addition to filtering, the editing loop seems to contribute to peptide loading, particularly on HLA-A*68:02 and B*15:03, and typically required additional or other residues than L30.
The recovery period post-trypsinisation impacts the immunopeptidome It was previously reported that OE TAPBPR traffics to the cell surface where it can perform peptide exchange on surface expressed HLA-I 11 .Therefore, two datasets were originally established with the TAPBPR editing loop mutants in which cells were harvested immediately after trypsinisation (no recovery) or incubated at 37°C in media for 30min posttrypsinisation to permit the OE surface TAPBPR to perform peptide editing on surface HLA-I 25 .Curiously, when allowing the surface expressed TAPBPR to edit on HLA-I using a 30min recovery period after trypsinisation, the editing loop-dependent filtering seemed to be relatively less dependent on L30 (Figure 3D-E).Conversely, TAPBPR editing loop-dependent loading seems more dependent on the additional/other residues than L30 compared to directly after trypsinisation (Figure 3D/F).

Discussion
Here, in-depth immunopeptidomics and HLA-I structural analysis were performed to provide further insight into the mechanism of action of TAPBPR-mediated HLA-I peptide exchange.TAPBPR was previously shown to preferentially bind and edit the repertoire on HLA-A, although interactions with and editing on HLA-B molecules has also been described 9,17 .We showed that TAPBPR can shape the peptide repertoire on HLA-A, -B and interestingly also on HLA-C molecules.The previously observed preference of TAPBPR for HLA-A over HLA-B or -C 17 may reflect variations in the kinetics of TAPBPR binding.Since peptide binding to MHC-I drives TAPBPR dissociation by a negative allostery release cycle 45 , a stronger TAPBPR:HLA-A interaction may persist longer as it takes a high affinity peptide to outcompete TAPBPR.Alternatively, TAPBPR may remain bound until UGT1 has mediated recycling of the peptide-receptive HLA-A back to the PLC.This recycling seems to occur less frequently for HLA-B or -C molecules 12 .Furthermore, HLA-C is expressed at lower levels compared to HLA-A or -B 46 , further complicating detection of interactions.The impact of TAPBPR on the peptide repertoire eluted from HLA-C indicates that TAPBPR may edit on a wider variety of allotypes than initially thought.
The three HLA-I alleles expressed by HeLaM cells all accommodate hydrophobic amino acids in their F pocket.This was shown to be a key feature for TAPBPR editing loop, and specifically L30, -mediated peptide exchange 25 .There are currently two main hypotheses for the role that the TAPBPR editing loop plays in peptide exchange.First, the TAPBPR editing loop was suggested to act as a scoop loop and function as an internal peptide surrogate, with L30 as the critical residue (leucine lever), thereby filtering the peptide repertoire 22,24,25 .Secondly, the editing loop may hover above the peptide groove and act as a peptide trap 26 .Interestingly, immunopeptidomics using cells expressing the TAPBPR editing loop mutants suggests that the K22-D35 domain is involved both in peptide filtering and loading on all three HLA-I allotypes expressed by HeLa cells.
With both TAPBPR WT and TAPBPR øG30L mediating restriction of the peptide repertoire compared to TAPBPR øloop , we hypothesised that TAPBPR allows the HLA-I peptide binding groove to adopt an open state independently of the editing loop to facilitate peptide exchange.Alternatively, L30 may be involved in opening of the HLA-I peptide binding cleft followed by peptide dissociation.All-atom MD simulations showed that TAPBPR øloop retained its ability to open the α2-1 helix of peptide-deficient HLA-I, suggesting that none of the residues in the K22-D35 editing loop, including L30, were responsible for this outward shift.The opening of the HLA-I peptide binding groove may therefore primarily be mediated by other regions of TAPBPR such as the jack hairpin, which was previously be reported to interact with the α2-1 helix and cause a partial downward shift of the HLA-I peptide binding groove 20,22 .Of note, these results do not exclude a role for the TAPBPR editing loop in opening of the peptide binding cleft when a peptide is bound.
The reanalysis of immunopeptidomics datasets shows that the editing loop and specifically L30 are involved in TAPBPRmediated peptide filtering.Data from Ilca et al. suggests that L30 binds the F pocket of HLA-I molecules, outcompeting the bound peptide and causing its dissociation 25 .It appears that residues outside of the editing loop are involved in the opening of the peptide binding groove, however, this mechanism alone may not be sufficient to mediate peptide filtering.Entry of L30 into the F pocket permitted by opening of the peptide binding groove is likely required for the peptide filtering ability of TAPBPR.While the L30 residue was typically sufficient to mediate filtering, peptide loading typically required other residues in the TAPBPR editing loop.This complements published MD simulations showing that R27 in the TAPBPR editing loop may draw a peptide into the F pocket 23 .Taken together, our reanalysis suggests the TAPBPR editing loop is multifunctional.Thus, it could be involved in peptide selection by both the previously proposed scoop loop/ leucine lever and peptide trap models, with L30 as the critical residue for peptide filtering and additional or other residues in the editing loop assisting with peptide loading.
Immunopeptidomics of cells expressing TAPBPR editing loop variants was carried out in duplicate.Cells were lysed immediately after harvesting the cells (no recovery) or cells were incubated 30min at 37°C post-trypsinisation, allowing OE surface TAPBPR to perform HLA-I peptide editing.The overall conclusion of the effect of the TAPBPR editing loop, and specifically L30, remain the same regardless of the recovery period.However, by adding the recovery period, TAPBPR editing loop-mediated filtering seemed less dependent on L30 whereas loading seemed more dependent on the editing loop.The underlying mechanism for this phenomenon is not clear.Although it has previously been suggested that HLA-I surface levels are not altered by trypsin-treatment of the cells 47 , trypsin may disrupt the HLA-I immunopeptidome or the TAPBPR:HLA-I interaction at the plasma membrane.The relative proportion of TAPBPR-mediated peptide exchange occurring in different subcellular compartment post-trypsin recovery, i.e. at the cell surface versus intracellular, may have caused a shift in the peptidome.The use of suspension cells that do not require trypsinisation as part of the harvesting process may be useful in future immunopeptidomics experiments.
We have previously shown that soluble recombinant TAPBPR can be utilised to display immunogenic peptides on cell surface-expressed HLA-I which triggers T-cell-mediated killing of target cells 11,48 .This reveals the translational potential to use TAPBPR as an immunotherapeutic.Insight into TAPBPR-mediated peptide editing may aid rationale design and improve efficiency of TAPBPR-based therapeutic candidates on polymorphic HLA-I molecules.For example, insight into the mechanism of the TAPBPR editing loop may allow us to improve HLA-I peptide binding or alter TAPBPR activity to HLA-I molecules on which the editing loop does not work naturally.Increased fundamental understanding of the mechanisms governing TAPBPR-mediated peptide exchange could expand the range of HLA-I allotype and therefore potentially the range of patients that may be responsive to TAPBPR-based therapies.

Underlying data
The immunopeptidomics datasets used in this study are available from: Dryad: Data from: TAPBPR alters MHC class I peptide presentation by functioning as a peptide exchange catalyst.https://doi.org/10.5061/dryad.487j9 43.
Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

Rui Chen
National Research Council Canada, Ottawa, Ontario, Canada The manuscript submitted by Altenburg et al. described a novel approach to study the role of the TAPBPR editing loop in opening the HLA-I peptide binding groove by using molecular dynamics simulation with immunopeptidomics data from previous studies that generated at different conditions of Hela cells, i.e., wild-type, depleted TAPBPR and overexpressed of TAPBPR or TAPBPR K22-D35 mutant.Although this study delivered some insights into the relationship between TAPBPR editing loop and antigen presentation of different HLA alleles, there is serious concern about the immunopeptidomics dataset that the authors used, which need to be addressed.

Detailed comments
1 The two datasets used in this study were generated in 2013 and 2018, respectively, by a relatively older mass spectrometer compared to that has been used in recent publications, which leads to the lack of depth towards the coverage of whole immunopeptidome.There was no mention of technical or biological replicates from these two studies which compromised the statistical significance of the current study.Also, the immunopeptidomics data were not quantitative.It might be difficult to repeat the experiments, however, re-analysis of the mass spectrometric data with relative quantification (LFQ) is needed.
2 The analysis of the two datasets needs to be harmonized as different protein database search engines were used and different criteria of peptides identification filtering were used.After the generation of peptides identification, same tools need to be used to predict the binding affinity of identified peptides to HLA alleles.The use of re-analysis of immunopeptidomics datasets and molecular dynamics simulations is appropriate and robust for studying the mechanisms of TAPBPR.These methods are well-suited to uncovering detailed molecular interactions.The methods section could benefit from a more detailed explanation of the specific datasets and simulation parameters used.For instance, the criteria for dataset selection and the specifics of the molecular dynamics simulations (e.g., force fields, simulation times) should be clearly stated.
The discussion effectively integrates the results with existing knowledge, providing a cohesive understanding of TAPBPR's role in peptide editing.The unification of the scoop loop and peptide trap mechanisms is a particularly valuable contribution to the field.The discussion could be expanded to address potential limitations of the study.For example, the reliance on computational simulations and re-analysis of existing datasets might limit the generalizability of the findings.Discussing these limitations and suggesting future experiments to validate the results would enhance the discussion.
The conclusions are well-supported by the data and provide a clear summary of the key findings.The identification of L30 as a critical residue for peptide exchange is particularly noteworthy.Expand the conclusions to include a brief discussion on the potential therapeutic implications of manipulating TAPBPR-HLA-I interactions, which could be of interest for developing novel immune-based therapies.Reviewer Expertise: Immunogenetics, Hematopoietic Cell Transplantation, Alloreactivity, Immunopeptidomics I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Figure 1 .
Figure 1.TAPBPR restricts the peptide repertoire diversity on HLA-A*68:02, B*15:03 and C*12:03.Peptides were eluted from W6/32-reactive HLA complexes isolated from (A/C) wild-type (WT) HeLaM and cells transduced to overexpress (OE) TAPBPR, or (B/D) HeLaM TAPBPR knock-out (KO) cells and WT cells treated with IFN-γ to induce endogenous TAPBPR expression 9 .(A-B) Size and overlap of the eluted peptide repertoire.(C-D) NetMHC was used to assign the peptides to one of the three HLA-I allotypes expressed by HeLa cells 9 .The absolute peptide count per HLA-I allotype was shown.

Figure 2 .
Figure 2. L30 in the TAPBPR editing loop restricts the HLA-I peptide repertoire diversity on HeLa cells.(A-B) IFN-γ-treated HeLaMTAPBPR KO cells that were transduced to express TAPBPR WT , TAPBPR øloop (mutated residues of the editing loop domain), or TAPBPR øG30L (TAPBPR øloop in which the leucine had been reconstituted at position 30) were harvested immediately post-trypsinisation (no recovery) or allowed to recover for 30min at 37°C in warm media (30min post-trypsin)25 .Subsequently, peptides were eluted from W6/32-reactive HLA complexes.Size and overlap of the eluted peptide repertoire are shown.(C) Sort-time MD simulation of TAPBPR øloop in complex with peptidedeficient HLA-A*02:01 at 0ns, 25ns and 50ns.The distance between the HLA-I α1 and α2 helix was measured at two positions (indicated by the dotted lines) to show opening of the HLA-I peptide binding groove over time.

Figure 3 .
Figure 3. Different residues are involved in TAPBPR editing loop-mediated peptide filtering and loading.Cells were harvested immediately post-trypsinisation (no recovery, A-C) or allowed to recover for 30min at 37°C in warm media (30min post-trypsin, D-F).Peptides were eluted from W6/32-reactive HLA complexes isolated from IFN-γ-treated HeLaM TAPBPR KO cells that were transduced to express TAPBPR WT , TAPBPR øloop (mutated residues of the editing loop domain), or TAPBPR øG30L (TAPBPR øloop in which the leucine had been reconstituted at position 30) 25 .(A/D) Size and overlap of the peptide repertoire.(B-C/E-F) NetMHCpan-4.0 was used to assign peptides to one of the three HLA-I allotypes expressed by HeLa cells 25 .The absolute peptide count per HLA-I allotype from each of the subgroups indicated in A/D was shown.

©
2024 Rana P.This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Priyanka Rana Case Western Reserve University, Cleveland, Ohio, USA The article by Altenburg et al, investigates the molecular mechanisms by which the TAPBPR molecule influences peptide selection and exchange on major histocompatibility class I (MHC-I) molecules, specifically focusing on human leukocyte antigen (HLA)-I.MHC-I molecules are critical for immune surveillance, presenting peptides on the cell surface to allow the immune system to detect infected or malignant cells.TAPBPR, along with tapasin, is known to optimize the peptide repertoire presented by MHC-I, but the exact mechanisms have been unclear.The study utilizes in-depth re-analyses of immunopeptidomics datasets and molecular dynamics simulations to explore how TAPBPR edits the peptide repertoire on HLA-A, -B, and -C allotypes.The findings indicate that the editing loop of TAPBPR is not essential for opening the HLA-I peptide binding groove but plays a significant role in peptide repertoire restriction, particularly through the leucine at position 30 (L30).This residue is crucial for both peptide dissociation and loading, suggesting that TAPBPR can mediate peptide exchange via a combination of the scoop loop and peptide trap mechanisms.The introduction effectively sets the stage by explaining the role of MHC-I in immune surveillance and the importance of TAPBPR in peptide editing.The background information is well-articulated, providing a clear context for the study.
the work clearly and accurately presented and does it cite the current literature?YesIs the study design appropriate and is the work technically sound?YesAre sufficient details of methods and analysis provided to allow replication by others?YesIf applicable, is the statistical analysis and its interpretation appropriate?YesAre all the source data underlying the results available to ensure full reproducibility?YesAre the conclusions drawn adequately supported by the results?YesCompeting Interests: No competing interests were disclosed.Reviewer Expertise: Immuno-oncology,, Hematology, Oncology, TNBCIs the work clearly and accurately presented and does it cite the current literature?YesIs the study design appropriate and is the work technically sound?PartlyAre sufficient details of methods and analysis provided to allow replication by others?PartlyIf applicable, is the statistical analysis and its interpretation appropriate?Not applicable Are all the source data underlying the results available to ensure full reproducibility?YesAre the conclusions drawn adequately supported by the results?PartlyCompeting Interests: No competing interests were disclosed.

the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Yes Are sufficient details of methods and analysis provided to allow replication by others? No If applicable, is the statistical analysis and its interpretation appropriate? No Are all the source data underlying the results available to ensure full reproducibility? Partly Are the conclusions drawn adequately supported by the results? Partly
It is recommended that the author to use the lasted version of netPanMHC(https://services.healthtech.dtu.dk/services/NetMHCpan-4.1/).It is not clear what parameters were used and the authors need to add more details about the HLA-I peptides analysis.
Competing Interests: No competing interests were disclosed.Reviewer Expertise: proteomics, immunopeptidomics, systems biology, mass spectrometry I confirm that I

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Reviewer Report 22 May 2024 https://doi.org/10.21956/wellcomeopenres.22949.r82626