Blocking interaction between SHP2 and PD‐1 denotes a novel opportunity for developing PD‐1 inhibitors

Abstract Small molecular PD‐1 inhibitors are lacking in current immuno‐oncology clinic. PD‐1/PD‐L1 antibody inhibitors currently approved for clinical usage block interaction between PD‐L1 and PD‐1 to enhance cytotoxicity of CD8+ cytotoxic T lymphocyte (CTL). Whether other steps along the PD‐1 signaling pathway can be targeted remains to be determined. Here, we report that methylene blue (MB), an FDA‐approved chemical for treating methemoglobinemia, potently inhibits PD‐1 signaling. MB enhances the cytotoxicity, activation, cell proliferation, and cytokine‐secreting activity of CTL inhibited by PD‐1. Mechanistically, MB blocks interaction between Y248‐phosphorylated immunoreceptor tyrosine‐based switch motif (ITSM) of human PD‐1 and SHP2. MB enables activated CTL to shrink PD‐L1 expressing tumor allografts and autochthonous lung cancers in a transgenic mouse model. MB also effectively counteracts the PD‐1 signaling on human T cells isolated from peripheral blood of healthy donors. Thus, we identify an FDA‐approved chemical capable of potently inhibiting the function of PD‐1. Equally important, our work sheds light on a novel strategy to develop inhibitors targeting PD‐1 signaling axis.

Thank you for the submission of your manuscript to EMBO Molecular Medicine. As you will see from the reports below, the referees find the topic of your study of potential interest. However, they raise substantial concerns on your work, which should be convincingly addressed in a major revision of the present manuscript.
Importantly, additional experiments and controls are required to confirm that the observed antitumor effect of MB is indeed through blocking the interaction between PDL1 and Shp2. Further, referee #1 points out that some of the presented data are contradicting existing literature, which needs to be satisfactorily addressed.
Overall it is clear that publication of the manuscript cannot be considered at this stage. I also note that addressing the reviewers concerns in full will be necessary for further considering the manuscript in our journal and this appears to require a lot of additional work and experimentation. I am unsure whether you will be able or willing to address those and return a revised manuscript within the 3 months deadline. On the other hand, given the potential interest of the findings, I would be willing to consider a revised manuscript with the understanding that the referee concerns must be fully addressed and that acceptance of the manuscript would entail a second round of review. I should remind you that it is EMBO Molecular Medicine policy to allow a single round of revision only and that, therefore, acceptance or rejection of the manuscript will depend on the completeness of your responses included in the next, final version of the manuscript. For this reason, and to save you from any frustrations in the end I would strongly advise against returning an incomplete revision and would also understand your decision if you choose to rather seek rapid publication elsewhere at this stage.
antibody-mediated inhibition of PD-1. They elucidated the molecular mechanism, showing that PD-1/SHP2 protein/protein-interaction is inhibited, resulting in interruption of the efferent signaling cascade. Their experiments show effectivity of inhibition both in vitro and in vivo. The clinical relevance of these findings is obvious. I have hence very little criticism: Sometimes the authors use terms which are linguistically not correct, i.e. page 5: "...interactions (PPI) was previously ..." should be: "...interactions (PPI) were previously ...". or on the next page: "... but shaded light on ..." must read: "... but shed light on ..." On page 10 they use the word "trend". This makes the results weaker than they are, because trend is usually used when data do not reach statistical significance, although this is clearly the case here. I would suggest to use the word "pattern" instead. There is one issue I must unfortunately address here: In figure 5A, the 3rd and 4th panel in the top row (24h PD-L1/Pem and 24h PD-L1/Niv are identical i.e. copy paste. I assume that was a mistake when generating the figure. This must be corrected. Referee #3 (Comments on Novelty/Model System for Author): The overall argument and science in this article seems outstanding and worthy of publication. I don't have the technical skills to review the scientific methods and when I contacted my colleague to get her help, she had sent the manuscript to me because she felt that she didn't have the skill set to review this manuscript. I would say that if the experimental work can be appropriately reviewed and looks okay, then I would publish it on all other counts. The person I think would be able to review this manuscript is Anusha Kalbasi anushakalbasi@mednet.ucla.edu I don't think it is fair to the authors for the paper to not be accepted because of my inability to review it. But I will say that two of us have struggled with it and that may indicate that it is not written at the right level. Likely too much detail is provided on how the authors got to the SHP2 result and the paper needs to be written to be more readable and less of a detailed account of all the experiments that were done. I am not sure though. I truly apologize, especially for the delay.  We thank Reviewer for this great suggestion of using genetically engineered T cells to clearly show that MB enhances T cell activity by inhibiting PD-1 signaling. We therefore tested MB's impact on various genetically engineered T cells through the following experiments: Experiment#1: To clarify whether MB exerts its impact on T cells through inhibiting PD-1 function by blocking its recruitment of SHP2, we knockout or overexpressed PD-1 in Jurkat cells harboring NFAT-luciferase (designated J-luc) clones. Specifically, we knockout PD-1 through sgRNA/CAS9 system (Designated J-luc-sgPD-1), overexpressed PD-1 (designated JP-luc), or overexpressed PD-1 but knockout SHP2 (JP-luc-sgSHP2) (supporting Figure 1A-D). We then co-cultured these engineered Jurkat cells with Raji or Raji-PD-L1 and check the impact of MB on luciferase activity, with aPD1 as a positive control.
Our data showed that 1 μM of MB abolished PD-1's inhibitory effect on NFAT controlled luciferase activity in JP-luc cells. In J-luc-sgPD-1 cells, this effect is largely lost. Similarly, this effect is lost in JP-luc-sgSHP2 cell or JP-luc treated with 10 μM of SHP099, a small molecular SHP2-selective inhibitor (Fortanet et al.  Figure 1E).
We have updated these results in Figure 2F and Figure EV2E, EV3Q, EV3R & EV3S in our manuscript. We found that while aCD3/CD28 activated proliferation of WT CD8+ T cells, PD-L1 administration inhibited proliferation of these T cell. 1 μM of MB abolished PD-1's inhibitory effect of proliferation of WT CTLs. Most importantly, this effect was lost on PD-1KO CTLs. We also found that neither aPD-1 nor MB further significantly enhanced proliferation of PD-L1 treated CTLs in presence of SHP099 (supporting Figure 2E). We also found that the ability of aPD-1 or MB to enhance proliferation of PD-L1 treated CTL-sgSHP2 was significantly diminished in comparison to that of WT CTLs.

Experiment#3:
We tested MB's impact on IL-2 expression by CD8+ T cell of wild-type background (WT), PD-1 knockout (PD-1KO) (Nishimura, H. et.al. Int Immunol 10, 1563-1572 (1998)), SHP2 knockout T cell (sgSHP2), or T cells treated with SHP099 (supporting Figure 3). Our data showed that 1 μM of MB significantly ameliorated PD-1's inhibitory effect on IL-2 expression by WT CTL. However, this effect is lost in case of PD-1KO CTLs. We also found that the ability of MB to enhance IL-2 expression of CTL-sgSHP2 cells or by CTLs in presence of SHP099 was significantly inhibited in comparison to that of WT CTLs. We have updated our results in Figure EV3U in our manuscript.
We have updated our result in Figure 1E & 1F and Figure EV1O, EV1P & EV3V in our manuscript.
We have updated these results in Figure EV4I, EV4J & EV4K in our manuscript.
Supporting Figure 5. PD-1 is critical for mediating tumor-shrinking effect of MB in vivo A. Tumor growth curves of subcutaneous tumor allograft. EG7-PD-L1 cell (2 × 106 cells) were subcutaneously injected into the right flank of C57BL/6J mice (n = 4 per group). The mice were then injected with WT or PD-1KO CTLs (i.v.) on day 3 and 6 respectively and treated with vehicle, aPD1 (i.p.10mg/kg, every other day) or MB (i.g. 20 mg/ kg/ day). Tumor volume was shown as mean ± s.e.m. B. Image of the tumors. C. Bar graph of tumor weight shown in B. (n=4). Data are representative of three independent experiments. Statistics were analyzed by unpaired t-test. Error bars denote s.e.m. *P < 0.05; **P < 0.01; ***P < 0.001.
Taken together, these data strongly supported that MB enhanced T cell function and that this effect was mediated by PD-1. Fig. 2H, the authors showed that MB more effectively stimulates IL-2 production than anti-PD-1, but they also detected a similar effect in PD-L1 negative condition. In fact, there was little difference in IL-2 between PD-L1 negative and positive conditions, suggesting that MB effect is independent to PD-L1/PD-1 signaling.

2.
We apologized for our mistakes in preparing Figure 2H. Our experiment was conducted as following: 5×104 JP-luc were seeded into 96 well plate. These JP-luc cells were cultured by their own (grey group), co-cultured with parental 5×104 Raji cells (black group) or Raji-PD-L1 (red group). The cultured cells were treated with vehicle, 10 μg/ml of nivolumab PD-1 antibody, or 1μM MB respectively. IL-2 was quantified in the media to monitor T cell activation (Supporting Figure 6).
We apologized that we mis-labeled the data column when extracting data for making figure. The figure has now been corrected.
5×104 JP-luc were seeded into 96 well plate. These JP-luc cells were cultured by their own (grey group), cocultured with parental 5×104 Raji cells (black group) or Raji-PD-L1 (red group). The cultured cells were treated with vehicle, 10 μg/ml of nivolumab PD-1 antibody, or 1μM MB respectively. JP-luc: Jurkat cell harboring NFAT-luciferase transgene and overexpressing PD-1. Data are representative of three independent experiments. Statistics were analyzed by unpaired t-test. Error bars denote s.e.m. *P < 0.05; **P < 0.01; ***P < 0.001. Fig. 4, the authors showed MB can promote T cell mediated anti-tumor activity in a lung cancer model which does not respond to aPD-1. Wouldn't this data suggest that MB activates T cells at least partly independent of PD-1 pathway?

The authors suggested that MB blocks PD-1/Shp2 interaction. However, the roles of Shp2 in the PD-1 pathway has been questioned by Rota et al., which reported that Shp2 is dispensable for PD-1 function in vivo. This previous finding seems to contradict to the main claim of the manuscript that MB shrinks tumor by
blocking PD-1/Shp2 interaction. This concern needs to be addressed, perhaps using Shp2KO cells. If the authors' model is correct, MB should not have any effect in this background.
We are grateful for Reviewer for suggesting this great experiment. As shown in Supporting Figure 1-4, we did the experiment and found that the ability of MB to enhance luciferase activity of JP-luc-sgSHP2 or JP-luc treated by SHP099 was compromised in comparison to that of untreated WT JP-luc, to enhance proliferation and IL-2 expression of CTL-sgSHP2 or CTL treated with SHP099 compared to those of untreated WT CTLs, and to enhance the cytotoxicity of OT-1 CTL-sgSHP2 and OT-1 CTL treated with SHP099 compared to untreated WT OT-1. We extracted the related data of sgSHP2 or SHP099 treated cells and listed these data in Supporting Figure 7 below.
We think that the possible explanation lies in ways of PD-1's recruitment of downstream mediators. PD-1 predominantly recruits SHP-2. A very recent reports suggested that in the case of SHP2 knockout, PD-1 recruits SHP-1 and remains functional (Celis-Gutierrez J., Cell Rep. 2019 Jun 11;27(11):3315-3330.e7). In line with this rationale, SHP2 deficient T cells could behave similarly to WT T cells. SHP099 inhibited SHP2 protein prevent the compensatory recruitment of SHP1 by PD-1, thus TCR signaling is more robust.
We are so grateful to Reviewer for suggesting these experiments. We have embodied these findings and discussed the result in more detail in Discussion part (paragraph 3 in page 20).

impossble. However, to ensure it is not a fixation artifact, I'd like to see a timelapse movie of SHP2 recruitment upon the addition of PD-L1. formal quantitation with statistical test is required.
As suggested by Reviewer, we co-transfected Jurkat cell with constructs for expressing PD-1-EGFP and SHP2-mCherry. The engineered Jurkat cell were seeded onto poly-lysine coated cover slides. 10 μg/ml of PD-L1 was administered in media. Colocalization of EGFP and mCherry was recorded in a time-lapse recorder LSM880. We observed that EGFP and mCherry began to colocalize 3 minutes post PD-L1 administration and became obvious from minute 4 on (supporting figure 8). We have attached the video for Reviewer's reference.
Supporting Figure 8. PD-L1 induced colocalization of PD-1 and SHP2 in Jurkat cells. A. PD-1-EGFP and SHP2-mCherry cotransfected Jurkat cell were seeded onto poly-lysine coated cover slides. PD-L1 was administered in media. Confocal microscopy time-lapse imaging was conducted to record colocalization of PD-1 and SHP2. B, C. The percentage of cells positive of colocalization signal in the absence or presence of PD-L1.

No information on how MB inhibits PD-1/Shp2 interaction, but not EGFR/Shp2 interaction. Does MB binds to PD-1 or Shp2?
We have been pursuing the answer to this question ever since we identified MB as an inhibitor for PD-1/SHP2 interaction 3.5 years ago. Using BiacoreTM T200, we detected the binding of MB to SHP2 protein (Supporting Figure 9). In collaborating with Dr. Song Gao's lab in Sun Yat-sen University, a top lab for resolving crystal structure (Nature. 2017 Feb 16;542(7641):372-376, et.al.), we have been trying to resolve the co-crystal structure of MB-SHP2 complex. However, we have not been able to co-crystalize after trying many different methods. Without the structure information, we don't have binding details and are not able to validate them through point-mutating SHP2 gene.
We are still working on the co-crystal structure of MB-SHP2 complex. We hope that Reviewer could understand our situation.  Figure 9. MB directly binds SHP2 in vitro. SHP2-HIS protein was immobilized on a CM5 sensor chip by using standard amine-coupling at 25°C with running buffer HBS-P, a serial concentrations of compound MB was injected automatically. The binding signals were continuously recorded in response units (RU) and presented graphically as a function of time. The binding affinity of compound MB towards SHP2-HIS was assayed using the SPR-based Biacore T200 instrument. KD value was 56.45 μM for compound MB obtained by fitting the data sets to 1:1 Langmuir binding model using Biacore T200 Evaluation Software.
It could be that overexpressed PD-1 was eliciting baseline inhibitory signals in T cell. Indeed, we consistently found that PD-1 antibody, as well as MB, enhanced luciferase activity of just JP-luc seeded by its own in 96 well plate (Supporting Figure 10). This data suggested that overexpressed PD-1 underwent baseline activation, and that inhibition of this baseline inhibition leads to T cell activation. Along with this rationale, almost all of the PD-1 inhibitory activity in MB-treated T cells is cleared in comparison to baseline PD-1 activity in untreated T cells, which could explain stronger phos-CD28 signal in OT-1 CTL co-cultured with Raji-PD-L1 in the presence of MB than that in OT-1 CTL co-cultured with parental Raji cells.
Supporting Figure 10: MB and PD-1 antibody enhanced luciferase activity of JP-luc cells.

Corrected.
On page 10 they use the word "trend". This makes the results weaker than they are, because trend is usually used when data do not reach statistical significance, although this is clearly the case here. I would suggest to use the word "pattern" instead.
Thanks so much! We have adopted "pattern".
There is one issue I must unfortunately address here: In figure 5A, the 3rd and 4th panel in the top row (24h PD-L1/Pem and 24h PD-L1/Niv are identical i.e. copy paste. I assume that was a mistake when generating the figure. This must be corrected.
We apologize for our mistake. We have corrected it.

Referee #3 (Comments on Novelty/Model System for Author):
The overall argument and science in this article seems outstanding and worthy of publication. I don't have the technical skills to review the scientific methods and when I contacted my colleague to get her help, she had sent the manuscript to me because she felt that she didn't have the skill set to review this manuscript. I would say that if the experimental work can be appropriately reviewed and looks okay, then I would publish it on all other counts. The person I think would be able to review this manuscript is Anusha Kalbasi anushakalbasi@mednet.ucla.edu I don't think it is fair to the authors for the paper to not be accepted because of my inability to review it. But I will say that two of us have struggled with it and that may indicate that it is not written at the right level. Likely too much detail is provided on how the authors got to the SHP2 result and the paper needs to be written to be more readable and less of a detailed account of all the experiments that were done. I am not sure though. I truly apologize, especially for the delay.
We appreciate these highly positive feedbacks from Reviewer. Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed report from two referees who were asked to re-assess it. As you will see the referees are now supportive and I am pleased to inform you that we will be able to accept your manuscript pending the following amendments.
***** Reviewer's comments ***** Referee #1 (Remarks for Author): common tests, such as t-test (please specify whether paired vs. unpaired), simple χ2 tests, Wilcoxon and Mann-Whitney tests, can be unambiguously identified by name only, but more complex techniques should be described in the methods section; are tests one-sided or two-sided? are there adjustments for multiple comparisons? exact statistical test results, e.g., P values = x but not P values < x; definition of 'center values' as median or average; definition of error bars as s.d. or s.e.m.
1.a. How was the sample size chosen to ensure adequate power to detect a pre-specified effect size? 1.b. For animal studies, include a statement about sample size estimate even if no statistical methods were used.
2. Describe inclusion/exclusion criteria if samples or animals were excluded from the analysis. Were the criteria preestablished?
3. Were any steps taken to minimize the effects of subjective bias when allocating animals/samples to treatment (e.g. randomization procedure)? If yes, please describe.
For animal studies, include a statement about randomization even if no randomization was used.
4.a. Were any steps taken to minimize the effects of subjective bias during group allocation or/and when assessing results (e.g. blinding of the investigator)? If yes please describe.

B-Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner. a statement of how many times the experiment shown was independently replicated in the laboratory.
Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.

definitions of statistical methods and measures:
a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).

The data shown in figures should satisfy the following conditions:
Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name). samples size was set such that statistical significance can be reached. In experiment when we used methylene blue to treat transgenic mice, computed tomography was conducted to confirm lung cancer in mice. Mice bearing no noticable lung cancers before treatment were excluded from analysis. in our treatment studies, we chose a cohort of mice bearing similar burden of lung cancer. These mice were then randomized to different treatment.
not applicable no, there was no estimate of variation within each group of data. the mice were randomized for treatment. We have indicated this in main text.
ImageJ software was used to analyze the CT images for determining tumor burdens. statement was included in material and methods.

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way.