Allosteric inhibition of the T cell receptor by a designed membrane ligand
Curation statements for this article:-
Curated by eLife
eLife assessment
The authors use a previously described technology of designing soluble transmembrane-targeting peptides, to interfere with the receptor function of the T cell receptor (TCR), which provides useful insights into the molecular mechanism of T cell activation. The designed PITCR peptide has functional effects, but the evidence for the proposed mechanism is still incomplete. With further data to support the conclusion, results from this study will be of interest to those studying the TCR as well as those seeking to use the TCR or its derivatives in synthetic biology studies and immunotherapy.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Structural Biology and Molecular Biophysics (eLife)
Abstract
The T cell receptor (TCR) is a complex molecular machine that directs the activation of T cells, allowing the immune system to fight pathogens and cancer cells. Despite decades of investigation, the molecular mechanism of TCR activation is still controversial. One of the leading activation hypotheses is the allosteric model. This model posits that binding of pMHC at the extracellular domain triggers a dynamic change in the transmembrane (TM) domain of the TCR subunits, which leads to signaling at the cytoplasmic side. We sought to test this hypothesis by creating a TM ligand for TCR. Previously we described a method to create a soluble peptide capable of inserting into membranes and binding to the TM domain of the receptor tyrosine kinase EphA2 (Alves et al., eLife, 2018). Here, we show that the approach is generalizable to complex membrane receptors, by designing a TM ligand for TCR. We observed that the designed peptide caused a reduction of Lck phosphorylation of TCR at the CD3ζ subunit in T cells. As a result, in the presence of this peptide inhibitor of TCR (PITCR), the proximal signaling cascade downstream of TCR activation was significantly dampened. Co-localization and co-immunoprecipitation in diisobutylene maleic acid (DIBMA) native nanodiscs confirmed that PITCR was able to bind to the TCR. AlphaFold-Multimer predicted that PITCR binds to the TM region of TCR, where it interacts with the two CD3ζ subunits. Our results additionally indicate that PITCR disrupts the allosteric changes in the compactness of the TM bundle that occur upon TCR activation, lending support to the allosteric TCR activation model. The TCR inhibition achieved by PITCR might be useful to treat inflammatory and autoimmune diseases and to prevent organ transplant rejection, as in these conditions aberrant activation of TCR contributes to disease.
Article activity feed
-
-
Author Response
Reviewer #2 (Public Review):
The authors present findings on a designed peptide, PITCR, and its role in inhibiting TCR activation through an extensive series of experiments. These include the measurement of phosphorylation in the TCR zeta chain and a number of associated signaling proteins such as Zap70, LAT, PLCg1, and SLP76. In addition, the authors measure the impact of PITCR on the TCR intracellular calcium response and examine the peptide-induced inhibition of TCR activation by antigen-presenting cells. They also present data indicating that the fluorescently labeled PITCR co-localizes with TCR in Jurkat cells and with ligand-bound TCR in primary murine cells. Overall the experiments provide useful insights into the mechanism of T cell activation and generally support an allosteric model of activation, while not …
Author Response
Reviewer #2 (Public Review):
The authors present findings on a designed peptide, PITCR, and its role in inhibiting TCR activation through an extensive series of experiments. These include the measurement of phosphorylation in the TCR zeta chain and a number of associated signaling proteins such as Zap70, LAT, PLCg1, and SLP76. In addition, the authors measure the impact of PITCR on the TCR intracellular calcium response and examine the peptide-induced inhibition of TCR activation by antigen-presenting cells. They also present data indicating that the fluorescently labeled PITCR co-localizes with TCR in Jurkat cells and with ligand-bound TCR in primary murine cells. Overall the experiments provide useful insights into the mechanism of T cell activation and generally support an allosteric model of activation, while not necessarily excluding alternative models.
However, some aspects of the study do need clarification.
- The authors do not provide a clear structural basis for their peptide design, which makes it difficult to understand the rationale for choosing this particular peptide. The use of a structural model based on the TCR zeta domain, for example, and how it becomes modified to generate PITCR would provide some clarity on what types of putative interactions are being engineered.
We thank the reviewer for giving us a chance to elaborate. We have expanded the results section to provide more information on the peptide design, where we now point out that the acidic residues in the TCR TM allow peptide design. We have also applied the artificial intelligence program AlphaFold-Multimer (AlFoM) to generate a structural model of the docking site of PITCR in the TCR (Figure 9), which informs on new mechanistic insights, as we describe in the updated results section and discuss below.
- The inhibitory effects of PITCR are not large. Measurement of dose dependence might improve confidence in the results.
As the reviewer points out, we have performed an extensive set of experiments to assess the inhibitory effect of PITCR. We have demonstrated that PITCR inhibits TCR phosphorylation. We have also tested all proximal signaling proteins: Zap70, LAT, SLP76, and PLC gamma. Critically, in all cases a statistically significant inhibition is observed. Furthermore, inhibition was additionally seen when TCR was activated by peptide presentation in antigen-presenting cells. Interaction between PITCR and the receptor is supported by co-localization, co-IP and the new AlphaFold-Multimer prediction. We are therefore confident in the results presented and that the inhibitory effect indeed exists. As we responded to reviewer 1 above, we discuss that inconsistent results were obtained with lower PITCR concentrations, suggesting that the use of a high peptide concentration is required for robust inhibition.
- Use of control peptides is not uniform. Control peptides similar to PITCR in Figure 1 and Figure 2 studies, for example, could strengthen the authors' arguments.
The original version of the manuscript contained two negative control peptides, the G41P mutant of PITCR, and pHLIP, another pH-responsive peptide which behaves as a conditional transmembrane peptide. However, for feasibility reasons we did not use all the negative controls in all different experiments, as we were satisfied when a negative control peptide acted as such in an experiment. However, because we agree that increased use of negative control peptides will strengthen the manuscript, we have expanded the use of negative control peptides. Specifically, the updated version of the manuscript contains a new section where AlFoM is used to predict the binding pose of PITCR and the structural consequences of interaction (see Figure 9 and the four new supplementary figures). AlFoM showed that PITCR binds with a large interaction interface, and peptide binding causes a large rearrangement of the two zeta chains in TCR. Importantly, neither of the two original negative control peptides (PITCRG41P or pHLIP) impacts the zeta chains. When we used a new negative control, the conditional transmembrane peptide TYPE7 developed by us, AlFoM did not predict it to bind to TCR, as expected, strengthening our argument.
Reviewer #3 (Public Review):
The use of pH-responsive TM-targeting peptides, which the authors previously developed, is a novel aspect of this study. Those peptides can be quite powerful for understanding molecular mechanisms of receptor signaling, such as the allosteric activation model as tested in this study. The manuscript contains several interesting approaches and observations, but there are concerns about the experimental design and interpretation of the results. More importantly, the authors' primary conclusion that the allosteric changes in the TM bundles determine TCR activation is not fully supported by the data presented. For example:
- The authors provided confocal fluorescence images showing the colocalization of fluorescently labeled peptides and TCR subunits. Based on the data, they concluded that "PITCR is able to bind to TCR". This is misleading, because given the spatial resolution of the imaging technique, "colocalization" does not indicate binding or interaction between molecules. Because the peptide binding to the TM region is the pillar of the primary finding of this study, direct evidence supporting the peptide-TM binding or interaction is essential.
We have to disagree that our statement is misleading: the section of the manuscript that the reviewer referred to, said “suggesting that PITCR is able to bind to TCR before it is activated by OKT3“. Therefore, we were not making a conclusion, just a mere suggestion, that we consider is justified, particularly as it is supported by data presented later. Nevertheless, we certainly agree with the reviewer that co-localization experiments fundamentally cannot indicate binding. We have modified the results (page 11) to follow the suggestion of the reviewer and indicate that co-localization data are not proof of interaction. In addition, we provide new AlphaFold multimer data, which supports that transmembrane binding indeed occurs.
- In calcium response experiments, the authors compared calcium influx (indicated by Indo-1 ratio) under different cell activation conditions (Figure 2). There are some concerns about how the authors interpreted the data: (1) The calcium plots from OKT3 activation in A-C panels are inconsistent. The plot in (A) showed a calcium peak after activation, which is not present in the plots shown in (B) and (C). There is no explanation or discussion on this inconsistency. (2) What is more concerning is that this prominent calcium peak in (A) was used to draw the conclusion that the designer peptide inhibitor effectively reduces calcium response. However, inconsistent with that conclusion, the calcium plots are indistinguishable for the three conditions: with PITCR (peptide inhibitor), with PITCRG41P (negative control that should not affect TCR activation), or no peptide. All three plots have similar magnetite and fluctuations. This does not support the authors' conclusion that the PITCR (peptide inhibitor) reduces calcium response in T cells.
We thank the reviewer for this comment. We have updated figure 3, which now contains a different replicate of the calcium assay, which we think it is more straightforward to analyze, and more clearly shows the calcium inhibition, as quantified in panel D of the figure.
- Different types of T cells were used for separate measurements: E6-1 Jurkat T cells were used for calcium influx experiments, J. OT.hCD8+ Jurkat cells were used for CD69 measurements, and primary murine CD4+ T cells were used for colocalization imaging experiments. Rationales for the choices of cells in different measurements are also unclear. This is different from the common practice where different cell types are used in repeated experiments to test the generality of a finding. Here, they were used for different experiments, and findings were lumped together as "T cells", without further evidence/discussion on how translatable the findings from different cell types are.
As the reviewer suggests, we have updated the manuscript to include discussion on the particularities of the use of the different T cells in pages 18 and 19. We envisioned this work as a proof of principle for the design of a peptide that can eventually be modified to be used for pre-clinical applications, and this paper is a first step. With this idea in mind, we wanted to test if this peptide can work in different types of TCR since: (1) TCR populations are diverse; and (2) our design is based on the transmembrane domain of CD3zeta chain, which is largely conserved among species. Using different types of T cells met this goal since they have different types of TCR, but the transmembrane domain of CD3zeta is conserved. In our paper, we used human Jurkat-TCR, OT1-TCR coupled with hCD8, and murine CD4-TCR. In addition, we not only used one activation marker to test the peptide’s inhibitory effect, we used three: phosphorylation, calcium influx, and CD69 activation. For the co-localization experiment, we not only use murine CD4 T cells, but we also tested it in Jurkat T cells with/without OKT3 stimulation as well.
We selected these T cells because they were particularly suited for the breath of different measurements that this manuscript contains, based on published reports. In our opinion this approach broadens the relevance of the work.
- The authors set out to test the model that TCR activation by pMHC occurs through allosteric changes in the TM region, but in most experiments, they activated Jurkat T cells by anti-CD3 antibody, not by antigen peptides. The anti-CD3 antibody activates TCR signaling through clustering. It is unclear whether TCR activation by anti-CD3 leads to the same allosteric changes in the TM region as activation by pMHC. As such, the main claim of the paper, namely that the designer peptide affects TCR signaling by disrupting the allosteric changes in the TM region, remains insufficiently supported by the data presented.
Figure 8 shows that the levels of co-IP in the presence of detergent are altered by OKT3 activation of TCR. It has recently been established (PMID: 34260912) that this assay allows the investigation of allosteric changes that contribute to activation of TCR. This evidence is supportive of allosterism in TCR activation. Additionally, the TCR proximal signaling is conserved between the Jurkat T cells activated by OKT3 and TCR activated by pMHC. We can reasonably argue that the peptide acts similarly in both conditions, since the peptide also exerts an inhibitory effect in T cells activated by antigen-presenting cells (Figure 4). The newly presented AlFoM model (Figure 9) predicts that PITCR binding displaces a zeta chain in TCR. This new result provides a plausible molecular rationale for the results in Figure 8, where we observe that PITCR changes transmembrane compactness, which has been linked to allosteric activation (Lanz et al., 2021; Prakaash et al., 2021).
-
eLife assessment
The authors use a previously described technology of designing soluble transmembrane-targeting peptides, to interfere with the receptor function of the T cell receptor (TCR), which provides useful insights into the molecular mechanism of T cell activation. The designed PITCR peptide has functional effects, but the evidence for the proposed mechanism is still incomplete. With further data to support the conclusion, results from this study will be of interest to those studying the TCR as well as those seeking to use the TCR or its derivatives in synthetic biology studies and immunotherapy.
-
Reviewer #1 (Public Review):
The authors design a peptide, PITCR, that is similar to the transmembrane domain of the TCR zeta, but is rendered soluble by adding an additionally charged residue to the TM domain and changing basic residues in the cytoplasmic juxtamembrane sequence to acidic residues. Some other bulky hydrophobic resides were made smaller. The strategy was based on earlier work with EphA2 sequences reported in elife in 2018. The TCRzeta conditional TM peptide was then tested for effects on T cell receptor signalling, co-localisation, and effects on TCR stability in biochemical assays. Significant effects were detected and these were eliminated by a strong helix-breaking mutation. There are currently some limitations with the interpretation of the signaling and co-localization studies. The results will be of interest to …
Reviewer #1 (Public Review):
The authors design a peptide, PITCR, that is similar to the transmembrane domain of the TCR zeta, but is rendered soluble by adding an additionally charged residue to the TM domain and changing basic residues in the cytoplasmic juxtamembrane sequence to acidic residues. Some other bulky hydrophobic resides were made smaller. The strategy was based on earlier work with EphA2 sequences reported in elife in 2018. The TCRzeta conditional TM peptide was then tested for effects on T cell receptor signalling, co-localisation, and effects on TCR stability in biochemical assays. Significant effects were detected and these were eliminated by a strong helix-breaking mutation. There are currently some limitations with the interpretation of the signaling and co-localization studies. The results will be of interest to those studying the TCR as well as those seeking to use the TCR or its derivatives in synthetic biology studies and immunotherapy.
-
Reviewer #2 (Public Review):
The authors present findings on a designed peptide, PITCR, and its role in inhibiting TCR activation through an extensive series of experiments. These include the measurement of phosphorylation in the TCR zeta chain and a number of associated signaling proteins such as Zap70, LAT, PLCg1, and SLP76. In addition, the authors measure the impact of PITCR on the TCR intracellular calcium response and examine the peptide-induced inhibition of TCR activation by antigen-presenting cells. They also present data indicating that the fluorescently labeled PITCR co-localizes with TCR in Jurkat cells and with ligand-bound TCR in primary murine cells.
Overall the experiments provide useful insights into the mechanism of T cell activation and generally support an allosteric model of activation, while not necessarily …
Reviewer #2 (Public Review):
The authors present findings on a designed peptide, PITCR, and its role in inhibiting TCR activation through an extensive series of experiments. These include the measurement of phosphorylation in the TCR zeta chain and a number of associated signaling proteins such as Zap70, LAT, PLCg1, and SLP76. In addition, the authors measure the impact of PITCR on the TCR intracellular calcium response and examine the peptide-induced inhibition of TCR activation by antigen-presenting cells. They also present data indicating that the fluorescently labeled PITCR co-localizes with TCR in Jurkat cells and with ligand-bound TCR in primary murine cells.
Overall the experiments provide useful insights into the mechanism of T cell activation and generally support an allosteric model of activation, while not necessarily excluding alternative models.
However, some aspects of the study do need clarification.
The authors do not provide a clear structural basis for their peptide design, which makes it difficult to understand the rationale for choosing this particular peptide. The use of a structural model based on the TCR zeta domain, for example, and how it becomes modified to generate PITCR would provide some clarity on what types of putative interactions are being engineered.
The inhibitory effects of PITCR are not large. Measurement of dose dependence might improve confidence in the results.
Use of control peptides is not uniform. Control peptides similar to PITCR in Figure 1 and Figure 2 studies, for example, could strengthen the authors' arguments.
-
Reviewer #3 (Public Review):
In this study, Ye et al investigated how a peptide that binds to the transmembrane (TM) domain of the T cell receptor (TCR) subunits affects TCR activation. The objective was to test the allosteric relaxation model of TCR activation. To this end, the authors leveraged their previously established strategy of designing TM-targeting peptides and studied how such peptide alters the TCR activation and downstream signaling cascades in Jurkat T cells. The authors found that the TM-targeting peptide inhibited phosphorylation of the TCR submits, phosphorylation of downstream signaling proteins such as ZAP70, and calcium influx in T cells. Using immunoprecipitation experiments, the authors proposed that the peptide binds into the membrane gap between CD3 and CD3 subunits in the TCR complex. The authors conclude …
Reviewer #3 (Public Review):
In this study, Ye et al investigated how a peptide that binds to the transmembrane (TM) domain of the T cell receptor (TCR) subunits affects TCR activation. The objective was to test the allosteric relaxation model of TCR activation. To this end, the authors leveraged their previously established strategy of designing TM-targeting peptides and studied how such peptide alters the TCR activation and downstream signaling cascades in Jurkat T cells. The authors found that the TM-targeting peptide inhibited phosphorylation of the TCR submits, phosphorylation of downstream signaling proteins such as ZAP70, and calcium influx in T cells. Using immunoprecipitation experiments, the authors proposed that the peptide binds into the membrane gap between CD3 and CD3 subunits in the TCR complex. The authors conclude that their data support the allosteric TCR activation model, in which allosteric changes in the TM bundle in the TCR complex determine the receptor signaling.
The use of pH-responsive TM-targeting peptides, which the authors previously developed, is a novel aspect of this study. Those peptides can be quite powerful for understanding molecular mechanisms of receptor signaling, such as the allosteric activation model as tested in this study. The manuscript contains several interesting approaches and observations, but there are concerns about the experimental design and interpretation of the results. More importantly, the authors' primary conclusion that the allosteric changes in the TM bundles determine TCR activation is not fully supported by the data presented. For example:
1. The authors provided confocal fluorescence images showing the colocalization of fluorescently labeled peptides and TCR subunits. Based on the data, they concluded that "PITCR is able to bind to TCR". This is misleading, because given the spatial resolution of the imaging technique, "colocalization" does not indicate binding or interaction between molecules. Because the peptide binding to the TM region is the pillar of the primary finding of this study, direct evidence supporting the peptide-TM binding or interaction is essential.
2. In calcium response experiments, the authors compared calcium influx (indicated by Indo-1 ratio) under different cell activation conditions (Figure 2). There are some concerns about how the authors interpreted the data: (1) The calcium plots from OKT3 activation in A-C panels are inconsistent. The plot in (A) showed a calcium peak after activation, which is not present in the plots shown in (B) and (C). There is no explanation or discussion on this inconsistency. (2) What is more concerning is that this prominent calcium peak in (A) was used to draw the conclusion that the designer peptide inhibitor effectively reduces calcium response. However, inconsistent with that conclusion, the calcium plots are indistinguishable for the three conditions: with PITCR (peptide inhibitor), with PITCRG41P (negative control that should not affect TCR activation), or no peptide. All three plots have similar magnetite and fluctuations. This does not support the authors' conclusion that the PITCR (peptide inhibitor) reduces calcium response in T cells.
3. Different types of T cells were used for separate measurements: E6-1 Jurkat T cells were used for calcium influx experiments, J. OT.hCD8+ Jurkat cells were used for CD69 measurements, and primary murine CD4+ T cells were used for colocalization imaging experiments. Rationales for the choices of cells in different measurements are also unclear. This is different from the common practice where different cell types are used in repeated experiments to test the generality of a finding. Here, they were used for different experiments, and findings were lumped together as "T cells", without further evidence/discussion on how translatable the findings from different cell types are.
4. The authors set out to test the model that TCR activation by pMHC occurs through allosteric changes in the TM region, but in most experiments, they activated Jurkat T cells by anti-CD3 antibody, not by antigen peptides. The anti-CD3 antibody activates TCR signaling through clustering. It is unclear whether TCR activation by anti-CD3 leads to the same allosteric changes in the TM region as activation by pMHC.As such, the main claim of the paper, namely that the designer peptide affects TCR signaling by disrupting the allosteric changes in the TM region, remains insufficiently supported by the data presented.
-
-