Figure 5. Mutation of the K22-D35 loop of TAPBPR changes the peptide repertoire presented on cells.

Peptides eluted from W6/32-reactive MHC I complex isolated from IFNγ treated HeLaM-TAPBPR^KO expressing either TAPBPR^WT, TAPBPR^Øloop or TAPBPR^ØG30L were analysed using LC-MS/MS. In dataset 1 (a–d), cells were frozen immediately post-trypsination while in dataset 2 (e–h) cells were allowed to recover in media for 30 min after trypsination, before freezing. The sequences of identified peptides are listed in Figure 5—source datas 2–7. The comparison of all five technical replicates for the two datasets is shown in Figure 5—figure supplement 1. (a,e) Venn diagrams compare all the identified peptides using a presence/absence approach. (b,f) Volcano plots graphically summarise label-free quantitation, displaying modulated peptides between two cells lines. Colour circles highlight the peptide which are differentially expressed between two cell lines after applying an adjusted p-value of <0.01. The list of these peptides is available in Figure 5—source datas 8 and 9. n = number of significantly modulated peptides, % demonstrates the fraction of significantly modulated peptides in a specific cell line compare to all peptides in the comparison. (c,d,g,h) Bar graphs summarise the MHC I molecules (HLA-A*68:02, -B*15:03 or –C*12:03) that the (c,g) identified peptides in a/e, and (d,h) the significantly modulated peptides identified in b/f were matched to using the NetMHCpan-4.0. In (c) and (g), peptides not successfully assigned are indicated in orange (rest). Analysis of the peptide repertoire from a further TAPBPR-loop mutant lacking L30 and from a third biological repeat can be found in Figure 5—figure supplements 2 and 3 respectively. Analysis of the predicted affinity of peptides differential modulated upon mutation of the loop (i.e those in b) and (f) can be found in Figure 5—figure supplement 4.

Figure 5—figure supplement 1. Technical reproducibility of LC-MS/MS measurement.

Peptide elution experiments were performed in five technical replicates for all cell lines. Column analysis reveals the percentage of peptides identified in 1–5 out of 5 technical replicates for (a) dataset 1 (cells frozen immediately post-trypsinisation) and (b) dataset 2 (cells allowed to recover for 30 min in media post-trypsinisation before freezing).

Figure 5—figure supplement 2. Mutation of residue A29-D35 in the loop of TAPBPR changes the peptide repertoire presented on cells.

In addition to the TAPBPR^WT, TAPBPR^Øloop or TAPBPR^ØG30L analysed in Figure 5e–h, we also performed peptidomic analysis for an additional loop mutation of TAPBPR in which residues A29-D35 where mutated from ALASSED to GGSGGAA (TAPBPR^M29). Thus, this mutant also lacks the leucine 30 residue. Peptides eluted from W6/32-reactive MHC I complex isolated from IFNγ-treated HeLaM-TAPBPR^KO expressing either TAPBPR^WT and TAPBPR^M29 were analysed using LC-MS/MS as in dataset two in Figure 5 (with recovery in media for 30 min after trypsination, before freezing). The sequences of identified peptides for TAPBPR^M29 are listed in Figure 5—source data 10. (a) Venn diagrams compare all the identified peptides using a presence/absence approach. (b) Volcano plots graphically summarise label-free quantitation, displaying modulated peptides between two cells lines. Colour circles highlight the peptide which are differentially expressed between two cell lines after applying an adjusted p-value of <0.01. n = number of significantly modulated peptides, % demonstrates the fraction of significantly modulated peptides in a specific cell line compare to all peptides in the comparison. (c) Bar graphs summarise the MHC I molecules (HLA-A*68:02, -B*15:03 or –C*12:03) that the (c) identified peptides in (a) and (d) the significantly modulated peptides identified in (b) and were matched to using the NetMHCpan-4.0. In c, peptides not successfully assigned are indicated in orange (rest).

Figure 5—figure supplement 3. 3^rd biological repeat.

Peptides were eluted from W6/32-reactive MHC I complexes isolated from a third set of IFNγ treated HeLaM-TAPBPR^KO expressing either TAPBPR^WT, TAPBPR^Øloop or TAPBPR^ØG30L and analysed using LC-MS/MS. The treatment of these cells is comparable to dataset two in the main Figure 5 (cells were allowed to recover in media for 30 min after trypsination, before freezing. The sequences of identified peptides are listed in Figure 5—source datas 11–13. The dataset is based on three technical repeats. (a) Venn diagrams compare all the identified peptides using a presence/absence approach. (b) Bar graphs summarise the MHC I molecules (HLA-A*68:02, -B*15:03 or –C*12:03) that the identified peptides in (a) were matched to using the NetMHCpan-4.0. In (b), peptides not successfully assigned are indicated in orange (rest). As shown in Figure 5e and g, this additional dataset suggests that TAPBPR molecules with a function loop strips peptides from HLA-A*68:02 molecules.

Figure 5—figure supplement 4. Peptide affinity predictions using NetMHC.

The affinity of peptides identified as being differentially modulated in the Volcano plots in

Figure 5b and f) were predicted using netMHCpan3.0. For each comparisons the peptides were sorted into the A-/B-/C-alleles using best_NetMHC_Allele. Peptides with an affinity above 500 nM were excluded. Box and whiskers plots show the peptide affinities for HLA-A*68:02, -B*15:03 and C*12:03 for peptides differentially modulated in (a) dataset one and (b) dataset 2. As the data-points were not normally distributed, the Mann-Whitney test was used to find significant differences between the comparisons. The number of peptides used in the analysis is show below each graph. We feel it is very difficult to draw any definitive conclusions regarding the effect of mutating the loop on the affinity of the peptides bound to MHC I from this particular analysis. Firstly, the peptide numbers corresponding to each HLA is low, especially for the HLA-B15 assigned peptides from cells expressing the TAPBPR loop mutant. This questions whether the significance observed is reliable, particularly for HLA-B15. Secondly, our experimental design was set up to explore whether there was a global change in peptide repertoire upon mutation of the TAPBPR loop. Thus, technical limitations regarding the need to assign peptides to particular MHC I (by applying thresholds or cut-offs to determine which peptide was bound to which MHC I) greatly reduce the ability to detect peptides exhibiting low affinity for a particular MHC I. Other studies (Garstka et al., 2015; Kanaseki et al., 2013; Nagarajan et al., 2016), exploring the affinity of peptides by immunopeptidomics use cells expressing only one MHC I. Therefore, the identified peptides do not need to be assigned to MHC I, thus improving the ability to see reliable changes (including subtle changes) in anchors and affinity. A third consideration is that trypsination of cells may be cause the loss of low-affinity peptide from MHC I molecules. One interesting observation from this analysis was the affinity of HLA-A*68:02 assigned peptides between dataset 1 and dataset two for cells expressing TAPBPR^WT appeared to increase. Thus hinting that surface expressed TAPBPR^WT may dissociate low-affinity peptide from HLA-A*68:02. However, to conclude, from these particular datasets, we are unable to address whether there is any change in affinity of MHC I bound peptides upon mutation of the TAPBPR loop. Improvement in experimental design will address this in the future.