An IgG‐based bispecific antibody for improved dual targeting in PSMA‐positive cancer

Abstract The prostate‐specific membrane antigen (PSMA) has been demonstrated in numerous studies to be expressed specifically on prostate carcinoma cells and on the neovasculature of several other cancer entities. However, the simultaneous expression of PSMA on both, tumor cells as well as tumor vessels remains unclear, even if such “dual” expression would constitute an important asset to facilitate sufficient influx of effector cells to a given tumor site. We report here on the generation of a PSMA antibody, termed 10B3, which exerts superior dual reactivity on sections of prostate carcinoma and squamous cell carcinoma of the lung. 10B3 was used for the construction of T‐cell recruiting bispecific PSMAxCD3 antibodies in Fab‐ and IgG‐based formats, designated Fabsc and IgGsc, respectively. In vitro, both molecules exhibited comparable activity. In contrast, only the larger IgGsc molecule induced complete and durable elimination of established tumors in humanized mice due to favorable pharmacokinetic properties. Upon treatment of three patients with metastasized prostate carcinoma with the IgGsc reagent, marked activation of T cells and rapid reduction of elevated PSA levels were observed.

12th Feb 2020 1st Editorial Decision 12th Feb 2020 Dear Gundram, Thank you for the submission of your manuscript to EMBO Molecular Medicine. Please accept my apologies for the delay in getting back to you, which is due to the fact that one referee did not return his/her report despite several reminders. In order to avoid delaying the process further, we prefer to make a decision now based on the two reports we received.
As you will see, while both referees mention the potential translational interest of the study, they also raise substantial concerns on your work, which should be convincingly addressed in a major revision of the present manuscript. In particular, both referees noted a lack of rigor in several places, such as in the characterization of the antibody, the choice of in vivo models, and the unclear correlation between preclinical and human studies.
Addressing the reviewers concerns in full (above points as well as other reviewers' comments) will be necessary for further considering the manuscript in our journal. Still, revising the manuscript according to the referees' recommendations appears to require a lot of additional work and experimentation, and 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 acceptance of the manuscript would entail a second round of review. EMBO Molecular Medicine encourages a single round of revision only and therefore, acceptance or rejection of the manuscript will depend on the completeness of your responses included in the next, final version of the manuscript. Should you find that the requested revisions are not feasible within the constraints outlined here and prefer, therefore, to submit your paper elsewhere, we would welcome a message to this effect.
When submitting your revised manuscript, please carefully review the instructions that follow below. Failure to include requested items will delay the evaluation of your revision: 1) A .docx formatted version of the manuscript text (including legends for main figures, EV figures and tables). Please make sure that the changes are highlighted to be clearly visible.
3) A .docx formatted letter INCLUDING the reviewers' reports and your detailed point-by-point responses to their comments. As part of the EMBO Press transparent editorial process, the pointby-point response is part of the Review Process File (RPF), which will be published alongside your paper.
4) A complete author checklist, which you can download from our author guidelines (https://www.embopress.org/page/journal/17574684/authorguide#submissionofrevisions). Please insert information in the checklist that is also reflected in the manuscript. The completed author checklist will also be part of the RPF. 5) Please note that all corresponding authors are required to supply an ORCID ID for their name upon submission of a revised manuscript. 6) Before submitting your revision, primary datasets produced in this study need to be deposited in an appropriate public database (see https://www.embopress.org/page/journal/17574684/authorguide#dataavailability). Please remember to provide a reviewer password if the datasets are not yet public. The accession numbers and database should be listed in a formal "Data Availability " section (placed after Materials & Method). Please note that the Data Availability Section is restricted to new primary data that are part of this study. *** Note -All links should resolve to a page where the data can be accessed. *** 7) We would also encourage you to include the source data for figure panels that show essential data. Numerical data should be provided as individual .xls or .csv files (including a tab describing the data). For blots or microscopy, uncropped images should be submitted (using a zip archive if multiple images need to be supplied for one panel). Additional information on source data and instruction on how to label the files are available at . 8) Our journal encourages inclusion of *data citations in the reference list* to directly cite datasets that were re-used and obtained from public databases. Data citations in the article text are distinct from normal bibliographical citations and should directly link to the database records from which the data can be accessed. In the main text, data citations are formatted as follows: "Data ref: Smith et al, 2001" or "Data ref: NCBI Sequence Read Archive PRJNA342805, 2017". In the Reference list, data citations must be labeled with "[DATASET]". A data reference must provide the database name, accession number/identifiers and a resolvable link to the landing page from which the data can be accessed at the end of the reference. Further instructions are available at . -For the figures that you do NOT wish to display as Expanded View figures, they should be bundled together with their legends in a single PDF file called *Appendix*, which should start with a short Table of Content. Appendix figures should be referred to in the main text as: "Appendix Figure  S1, Appendix Figure S2" etc.
-Additional Tables/Datasets should be labeled and referred to as Table EV1, Dataset EV1, etc. Legends have to be provided in a separate tab in case of .xls files. Alternatively, the legend can be supplied as a separate text file (README) and zipped together with the Table/Dataset file. See detailed instructions here: .
10) The paper explained: EMBO Molecular Medicine articles are accompanied by a summary of the articles to emphasize the major findings in the paper and their medical implications for the nonspecialist reader. Please provide a draft summary of your article highlighting -the medical issue you are addressing, -the results obtained and -their clinical impact. This may be edited to ensure that readers understand the significance and context of the research. Please refer to any of our published articles for an example. 11) For more information: There is space at the end of each article to list relevant web links for further consultation by our readers. Could you identify some relevant ones and provide such information as well? Some examples are patient associations, relevant databases, OMIM/proteins/genes links, author's websites, etc... 12) Every published paper now includes a 'Synopsis' to further enhance discoverability. Synopses are displayed on the journal webpage and are freely accessible to all readers. They include a short stand first (maximum of 300 characters, including space) as well as 2-5 one-sentences bullet points that summarizes the paper. Please write the bullet points to summarize the key NEW findings. They should be designed to be complementary to the abstract -i.e. not repeat the same text. We encourage inclusion of key acronyms and quantitative information (maximum of 30 words / bullet point). Please use the passive voice. Please attach these in a separate file or send them by email, we will incorporate them accordingly.
Please also suggest a striking image or visual abstract to illustrate your article. If you do please provide a jpeg file 550 px-wide x 400-px high. please see critique. figures lack rigor in presentation so it is unclear if results are internally consistent. the table demonstrating prevalence and intensity of staining is confusing. the murine model is not ideal.
Referee #1 (Remarks for Author): The authors generate a novel PSMA-specific antibody and describe its novelty in targeting both tumor and vascular components; they further construct a bsAb which they evaluate in a xenograft model as well as a pilot clinical trial for toxicity. There is certainly significant interest in developing optimal bsAbs that are both effective and safe, and the authors claim that their PSMA bsAb addresses both these challenges. However, there are several major flaws in their studies that undermine their conclusions and the manuscript, as presented, does not provide a coherent portrayal of the significance and broader applicability of their findings. Major criticisms 1. The authors claim that 10B3 exhibits dual targeting properties as it recognizes a conformational state of PSMA and is thus able to target tumor vascular endothelial as well as PSMA+ tumor cells (in contrast to the prototypic anti-PSMA antibody, J591 which only recognizes tumor cells). They provide an IHC of SSC of lung (fig 1 c) and demonstrate 10B3 staining of presumably vascular endothelial cells. In the same slide staining is shown for 10B3 staining of SSC lung tumor cells. a. No counter staining of Ve is performed confirming these are Ve cells (e,.g CD105 a marker of tumor vasculature, or any other Ve-specific marker) b. Absence of J591 staining in SSC is claimed to be the result of inferior recognition of PSMA by this antibody vs 10B3. The authors do not perform any other studies on the same slide demonstrating that staining of Tumor (and Ve) by 10B3 is the result of PSMA recognition. Fig EV4 is cited as evidence that 10B3 stains PSMA+ tumor cells by RNAscope, yet there is no indication that this is the same section as that presented in Fig 1c and no expression of PSMA in Ve is demonstrated here. That this is not presented in a clear and unequivocal fashion is concerning given that this is one of the major findings of the manuscript 2. The authors then proceed to use 10B3 as both Fabsc and IgGsc constructs. However, there is no demonstration that either form recognizes Ve cells. 3. A first-in-man clinical study is then performed. Understandably, the hurdles that must be overcome to move this into the clinic can be formidable and appear to have been addressed for the trial to be realized. However, adding this to this manuscript, due to lack of meaningful correlates, gives the impression that this was 'tacked on'. Expression of PSMA in the tumors and tumor vasculature of these patients is not presented. The PSA results presented do not give any idea of the velocity of PSA prior to treatment or extended results beyond the relatively short interval presented. It is unclear if PSA was already falling or the difference seen is due to fluctuations within a relatively narrow range of values. The toxicity mitigation with tocilizumab is not significant given the small number of patients nor surprising given its use in other CAR-T cell therapies. Minor criticisms 1. The wording in several key sentences is awkward; several run-on sentences make reading difficult (e.g. second para, page 7 "For comparative analysis..."). Redundant verb use is unnecessary (e.g. last sentence 3rd para, page 6). These are only some examples. A native English speaker to review the manuscript for readability and clarity is recommended. 2. Fig 5 c -appears to be mislableled -tumor eradication is seen at doses of 0.8, 1.0 and 2.0 but not at doses of 1.6, or 1.2 which is both internally inconsistent and contrary to the text in manuscript. 3. Table 1 is very difficult to read and understand. Perhaps an H-score to summarize both intensity and % cells + would have been more useful Referee #3 (Comments on Novelty/Model System for Author): Zekri et al claim to have made a novel antibody with improved dual-targeting capabilities, however it is unclear what the impact of this is. Their model system primarily looks at PSMA(+) tumor cells, and aside from IHC does not discuss the PSMA(+) vasculature enough. Although their antibody may indeed be a more clinically suitable therapeutic, the models used did not reflect this adequately.
Referee #3 (Remarks for Author): Zekri et al identified and engineered a bispecific antibody for the treatment of PSMA(+) cancer with T-cells. Patients with prostate cancer have limited treatment options an any novel approaches that limit toxicity while shrinking tumor are likely to have a major impact the field. In this manuscript, Zekri et al cover two important points: First they describe how this anti-PSMA clone differs from the conventionally used J591 clone, by showing how the epitope of 10B3 is more widely displayed in neovasculature of multiple cancer types. Second they engineer two T-cell bispecific antibodies using this sequence and evaluate their function in vivo, both pre-clinically and in three patients, providing first in human data for this T-cell bispecific antibody. Although two BsAb formats were compared, it is unclear why either was chosen among the big list of possible antibody platforms. The superiority of 10B3 over J591 was not adequately explained or evident in the data. Overall the manuscript provides interesting data regarding PSMA targeting and BsAb formats. With revisions it should be suitable for consideration.
The following changes are recommended to more clearly articulate the functional relevance of 10B3 compared to J591, in vitro: 1) Fc-attenuation is mentioned but not clearly cited or spelled out in the manuscript. Which mutations were used to ablate Fc binding? How did this attenuation affect antigen binding, pharmacokinetics, and T cell homing? 2) Reformat Table 1 to display the relative fraction of samples that stain at each intensity level. This can be both in a table, or even a graph, such that the differences between tumor types can be more clearly seen. In addition, statistical analysis between relevant groups should be presented to prove the novel reactivity with neovasculature. 3) Also reformat Table 1 such that the frequency of staining within each intensity level can more clearly compared between J591 and 10B3. For example, splitting each tumor type into a 2-cell wide by 4-cell long table for each tumor type and antigen type (Tumor, Vasculature) would allow one to compare J591 and 10B3 side by side for each intensity level. A similar change can be made for the aggregate staining scores to allow for statistical comparison. 4) it is not clear if these scores and the slides they were based on were evaluated by a trained pathologist. Although a pathologist is included in the author list. If he or she has evaluated these slides, it would be important to highlight this in the methods or even the text to strengthen the claims. 5) Figure EV3 shows very little difference between 10B3 and J591, relative to the slides shown in figure 1. Please comment on how representative these slides are of their respective staining intensity as displayed in table 1, and if necessary, provide additional examples of when these antibodies stain similarly or differently. 6) Figure EV4B does not clearly show the experimental intent of the authors. More robust quantification or single color controls or control tissues would helpful. 7) Figure 4B needs to have the legend updated to address more clearly what 1 and 2 are labeling.
The following changes are recommended more clearly explain the current in vitro and in vivo BsAb data, and more robustly evaluate its potential in other indications: 1) The figure legend of Figure EV5 C-E seems to be inconsistent with the text. In the figure, CD3 binding is clearly superior for the IgGsc, while PSMA binding is moderately improved with the Fabsc, in contrast to the text, which describes it in the opposite way. Please either correct or clarify. This is also seen on page EV7, which shows the same results as EV5.
2) Additionally, it would be useful to evaluate the CD3 binding to primary human PBMCs or T-cells, in addition to or instead of Jurkat cells.
3) The results of Figure EV5C-E are somewhat surprising either way. Could this be from the staining method used? The anti-Fab2 may bind differently to the Fabsc compared to the IgGsc. For example, could the IgG-sc molecule be bound by two different secondary molecules? Additional controls to evaluate this would be valuable. Additionally, while gel-filtration data looks clear, numerical summary of each peak for the SEC-HPLC would be helpful. If these values differ substantially it may be worth re-testing the staining using purer preparations, i.e. without aggregates. 4) Why does Fabsc show consistently better IFN and IL-2 release in vitro? This suggests that Fabsc may be aggregating in vitro, or may be contaminated in some way (endotoxin, etc). Please provide data supporting that Fabsc and IgG-sc have similar in vitro stabilities and endotoxin levels. Assuming they are the same, please provide a better explanation of why the Fab sc seems to perform so well. This would also impact why Fab-sc functions independent of tumor. 5) The binding data is referred to as affinity, which is slightly misleading. Given that PSMA binding kinetics were evaluated by SPR, it would be very helpful to both present this data earlier in the manuscript, (along with the other biochemical characterizations) and include the fitted affinity quantitation's. This should be done for both BsAb's and using CD3 as well. Doing so will allow for more appropriate conclusions to be made about binding affinities and binding kinetics (for comparison between different formats and antibody clones). 6) The LNCAP data appears compelling, but it would be helpful to compare this against a J591 BsAb as well. Although this is not entirely necessary in the case of LNCAP, it would be useful to show that 10B3 can indeed control or shrink non-prostate PSMA(+) tumors in vivo. It is understood that this level of biology may not be possible to model using a cell line or PDX xenograft system, but an explanation of such limits would be helpful too. In the absence of in vivo data, a clearer explanation of why the authors believed 10B3 to be superior to J591 would be helpful. For example, does the reduced specificity of 10B3 warrant concern that 10B3 would be more toxic to normal vasculature? Could it also reduce the tumor targeting of the antibody, and therefore reduce T-cell infiltration? 7) More details are needed in the method section regarding the outline of the animal models used. Routes of administration, timing and doses need to be more clearly spelled out. In addition, justification is needed for starting treatment so soon after tumor implantation (24hr, page 8). Please also update the figure axis labels to distinguish days since treatment or days since implantation. 8) On page 9, the following sentence is unclear: "In our hands, this antibody, in marked contrast to steroids, interferes with the anti-tumor activity of CC-1 neither in vitro nor in vivo". Please rephrase or clarify. 9) It was unclear how the clinical trial was designed. The exact regulatory authority who approved the study NCT 04104607) should be spelt out. For example, what was the justification for the dosing schedule and regimen. Additional, clearer quantitation of the CC-1 doses in Figure 6 would be helpful. In vitro cytokine release started at 0.01 nM (~2 ng/ml), peaking at 1 nM (~200 ng/ml) of BsAb. Serum concentrations of 200-500 ng/ml should hit the peak of cytokine storm -a dosing regimen seemingly incompatible with safe design in a phase I bispecific antibody trial. A serum level of 200-500 ng/ml for any bispecific antibody (whether BiTE or IgG formats) is very high and potentially lethal. It is possible that this construct was not very effective in activating T cells. The mention of steroids affecting anti-tumor activity should be referenced not as unpublished results but with citation from previous publications. 10) Regarding patient 2's response to tocilizumab, it is explained that they may have had a preexisting anti-human immune response. If so, shouldn't this have also impacted the level of CC-1 in the blood, not just Toci?

Referee#1 (Remarks for Author):
The authors generate a novel PSMA-specific antibody and describe its novelty in targeting both tumor and vascular components; they further construct a bsAb which they evaluate in a xenograft model as well as a pilot clinical trial for toxicity.
There is certainly significant interest in developing optimal bsAbs that are both effective and safe, and the authors claim that their PSMA bsAb addresses both these challenges.
However, there are several major flaws in their studies that undermine their conclusions and the manuscript, as presented, does not provide a coherent portrayal of the significance and broader applicability of their findings.    Figure   EV2D. The experiment showed that staining with 10B3 can be blocked by an excess of recombinant PSMA suggesting that it is PSMA specific. 3. A first-in-man clinical study is then performed. Understandably, the hurdles that must be overcome to move this into the clinic can be formidable and appear to have been addressed for the trial to be realized. However, adding this to this manuscript, due to lack of meaningful correlates, gives the impression that this was 'tacked on'.
Expression of PSMA in the tumors and tumor vasculature of these patients is not presented.
We thank the reviewer for raising this important issue and now provide additional information on the PSMA/PET scans of the three patients prior to CC-1 treatment. Respective images are included in the revised version, figure EV5A and clearly show for all patients, localization of a PSMA specific tracer to multiple tumor sites.
The PSA results presented do not give any idea of the velocity of PSA prior to treatment or extended results beyond the relatively short interval presented. It is unclear if PSA was already falling or the difference seen is due to fluctuations within a relatively narrow range of values.
We appreciate this valuable comment and have included additional information on the PSA values of the three patients before, during and after CC-1 treatment. This information is presented in Figure EV5B and confirms that, indeed, a rapid decline of PSA values was observed during CC-1 application, followed by a moderate increase after cessation of treatment.
The toxicity mitigation with tocilizumab is not significant given the small number of patients nor surprising given its use in other CAR-T cell therapies.
We certainly agree that data obtained from a limited number of patients are to be interpreted with due care. We also agree that the therapeutic activity of Tocilizumab for treatment of established CRS is well documented. In our paper, this antibody was used in a prophylactic setting. In a recently accepted manuscript from our group (Kauer et al.2020, in press), we have demonstrated that early Tocilizumab application prevents bsAb mediated CRS by maintaining a normal body temperature and relatively low CRP (C-reactive protein) levels at rather high serum concentrations of IL-6. If requested we could provide this manuscript.

Minor criticisms
1. The wording in several key sentences is awkward; several run-on sentences make reading difficult (e.g. second para, page 7 "For comparative analysis..."). Redundant verb use is unnecessary (e.g. last sentence 3rd para, page 6). These are only some examples. A native English speaker to review the manuscript for readability and clarity is recommended.
To address this issue of the reviewer, we have rephrased the paragraphs mentioned above in the revised manuscript as following: We sincerely apologize for mislabeling the figure 5c (old figure numbering) and have corrected this error (now presented as Fig 6C). Indeed, the tumor eradication was measured at doses of at 2μg, 1.6μg and 1.2μg (but not at a 1μg dose).

Referee #3 (Remarks for Author):
Zekri et al identified and engineered a bispecific antibody for the treatment of PSMA(+) cancer with T-cells. Patients with prostate cancer have limited treatment options an any novel approaches that limit toxicity while shrinking tumor are likely to have a major impact the field. In this manuscript, Zekri et al cover two important points: First they describe how this anti-PSMA clone differs from the conventionally used J591 clone, by showing how the epitope of 10B3 is more widely displayed in neovasculature of multiple cancer types. Second they engineer two T-cell bispecific antibodies using this sequence and evaluate their function in vivo, both pre-clinically and in three patients, providing first in human data for this T-cell bispecific antibody.
Although two BsAb formats were compared, it is unclear why either was chosen among the big list of possible antibody platforms.
We certainly agree with the reviewer that there are numerous different formats used for the construction of bispecific antibodies. It was not our intention to extensively compare a larger number of these. Rather, the "philosophy" of our group was and still is to construct reagents with a preserved affinity towards a target antigen and to avoid aggregation due to the use of single chains. With this guiding principle we have chosen for this paper to compare a representative "small" molecule with a rather short half-life and rapid elimination with a large molecule exhibiting relatively long serum half-life. With this we intended to contribute to an unresolved controversy as to the role of these parameters for in vivo antitumor efficiency.
While it has been argued that small molecules may be superior for tumor localization due to better penetration, our data clearly demonstrate superior tumor uptake of the large bispecific molecule.
The following changes are recommended to more clearly articulate the functional relevance of 10B3 compared to J591, in vitro: 1) Fc-attenuation is mentioned but not clearly cited or spelled out in the manuscript. Which mutations were used to ablate Fc binding? How did this attenuation affect antigen binding, pharmacokinetics, and T cell homing?
The mutations used to attenuate the Fc part are noted in the material and methods section.
That they result in a complete abrogation of Fc function is suggested by the data demonstrated in Fig. 4, as discussed in the paper. To additionally address the reviewers concerns concerning the effect of the mutations on pharmacokinetic, we compared the half-  2) Reformat Table 1 to display the relative fraction of samples that stain at each intensity level. This can be both in a table, or even a graph, such that the differences between tumor types can be more clearly seen. In addition, statistical analysis between relevant groups should be presented to prove the novel reactivity with neovasculature.
We thank the reviewer for this valuable comment, which was also raised by reviewer1.
Briefly, we replaced t table 1 by graph introducing an H-score and corresponding statistical analysis. We refer to our respective reply to comment "3 -minor criticisms" of reviewer 1 for details. Table 1 such that the frequency of staining within each intensity level can more clearly compared between J591 and 10B3. For example, splitting each tumor type into a 2-cell wide by 4-cell long table for each tumor type and antigen type (Tumor, Vasculature) would allow one to compare J591 and 10B3 side by side for each intensity level. A similar change can be made for the aggregate staining scores to allow for statistical comparison. figure 1D. We hope that this addresses the reviewer concerns.

As mentioned above we have now replaced table 1 by a graph included in the
4) It is not clear if these scores and the slides they were based on were evaluated by a trained pathologist. Although a pathologist is included in the author list. If he or she has evaluated these slides, it would be important to highlight this in the methods or even the text to strengthen the claims.
In fact, all slides were evaluated by the trained pathologist listed as an author (BS). The final graphs and figures of the revised manuscript were again controlled and approved by this author. A respective remark is now added in the material and methods section. Figure  iii-10B3 staining appears to be PSMA specific (EV2D) For more information and semi-quantitative analysis of immunohistological data, we refer to our reply to comment "1 and 2" of reviewer 1. Figure EV4B does not clearly show the experimental intent of the authors. More robust quantification or single color controls or control tissues would helpful.

6)
We thank the reviewer for this valuable remark.  Figure 4B needs to have the legend updated to address more clearly what 1 and 2 are labeling.
We thank the reviewer for pointing out this lack of clarity and have stated now in the figure legend that (1) represents the DU145 "PSMA negative" tumor and (2) the 22Rv1 "PSMA positive" tumor.
The following changes are recommended more clearly explain the current in vitro and in vivo BsAb data, and more robustly evaluate its potential in other indications: 1) The figure legend of Figure EV5 C-E seems to be inconsistent with the text. In the figure, CD3 binding is clearly superior for the IgGsc, while PSMA binding is moderately improved with the Fabsc, in contrast to the text, which describes it in the opposite way. Please either correct or clarify. This is also seen on page EV7, which shows the same results as EV5.
2) Additionally, it would be useful to evaluate the CD3 binding to primary human PBMCs or T-cells, in addition to or instead of Jurkat cells. 3) The results of Figure EV5C With the respect to SEC-HPLC data, we have included now the numerical summary of each peak presented in a small graph in figure 2C. Additionally, it should be noted that all the experiments performed in this manuscript were done with aggregate free proteins.

4) Why does Fabsc show consistently better IFN and IL-2 release in vitro? This suggests that
Fabsc may be aggregating in vitro, or may be contaminated in some way (endotoxin, etc).
Please provide data supporting that Fabsc and IgG-sc have similar in vitro stabilities and endotoxin levels. Assuming they are the same, please provide a better explanation of why the Fab sc seems to perform so well. This would also impact why Fab-sc functions independent of tumor.
We thank the referee for this comment. In accordance with the data presented we hypothesis that: i-The somewhat increased activity is due to increased CD3 affinity (as discussed above for the IgGsc-and Fabsc-molecules, respectively. 5) The binding data is referred to as affinity, which is slightly misleading. Given that PSMA binding kinetics were evaluated by SPR, it would be very helpful to both present this data earlier in the manuscript, (along with the other biochemical characterizations) and include the fitted affinity quantitation's. This should be done for both BsAb's and using CD3 as well.
Doing so will allow for more appropriate conclusions to be made about binding affinities and binding kinetics (for comparison between different formats and antibody clones).
We thank the reviewer for his comment and apologize for not including the affinity In the absence of in vivo data, a clearer explanation of why the authors believed 10B3 to be superior to J591 would be helpful. For example, does the reduced specificity of 10B3 warrant concern that 10B3 would be more toxic to normal vasculature? Could it also reduce the tumor targeting of the antibody, and therefore reduce T-cell infiltration?
We Thank you for submitting your revised manuscript to EMBO Molecular Medicine. Please accept my apologies for the unusual delay in getting back to you, which is due to the fact that I was expecting the report from referee #1, who was a critical referee during the first round of review. Despite several chasers and promises to provide a report, we still have not heard back from this referee, and thus prefer to make a decision now in order to avoid delaying the process further. Referee #3 kindly provided a report on your responses to both referee #1 and #3's comments that you will find attached below.
As you will see, referee #3 acknowledges your efforts to address the initial concerns, and recognizes that the manuscript has significantly improved. However, this referee also mentions issues that remain unanswered regarding both referees' initial reports, and additional experiments will be necessary to support the claims.
As EMBO Press encourages a single round of revisions only, we would normally reject the manuscript at this stage. However, as the reviewer recognizes (as we do) the potential clinical impact of the study and its interest for the community, we would like to exceptionally invite a second round of revisions. Please be aware that this will be the last chance for you to address the points raised by the referees. Particular attention should additionally be given to improve the flow and clarity of the manuscript. *** When submitting your revised manuscript, please carefully review the instructions that follow below. Failure to include requested items will delay the evaluation of your revision: 1) A .docx formatted version of the manuscript text (including legends for main figures, EV figures and tables). Please make sure that the changes are highlighted to be clearly visible.
3) A .docx formatted letter INCLUDING the reviewers' reports and your detailed point-by-point responses to their comments. As part of the EMBO Press transparent editorial process, the pointby-point response is part of the Review Process File (RPF), which will be published alongside your paper. 4) Please provide up to 5 keywords. 5) Please reformat the references so as to have 10 authors listed before et al. 6) Please remove "Data not shown". As per our guidelines on "Unpublished Data" the journal does not permit citation of "Data not shown". All data referred to in the paper should be displayed in the main or Expanded View figures (see also referee's comments). 7) Please note that references for Fig. 1B and for panels A and B of Fig. 4 are missing in the main manuscript text. 8) We would also encourage you to include the source data for figure panels that show essential data. Numerical data should be provided as individual .xls or .csv files (including a tab describing the data). For blots or microscopy, uncropped images should be submitted (using a zip archive if multiple images need to be supplied for one panel). Additional information on source data and instruction on how to label the files are available at . 9) Thank you for providing "The paper explained" section. Please insert it in the main manuscript file.
10) For more information: There is space at the end of each article to list relevant web links for further consultation by our readers. Could you identify some relevant ones and provide such information as well? Some examples are patient associations, relevant databases, OMIM/proteins/genes links, author's websites, etc... 11) Author contributions: please use initials instead of full names. Most of the contributing authors have not been entered in the submission system. This has to be done before the paper can be accepted.
12) Thank you for providing a synopsis image. We could unfortunately not open the .jpeg file, and the pdf was too small. Could you please double-check and provide a file (png or jpeg) 550 px-wide x 400-px high? 13) As part of the EMBO Publications transparent editorial process initiative (see our Editorial at http://embomolmed.embopress.org/content/2/9/329), EMBO Molecular Medicine will publish online a Review Process File (RPF) to accompany accepted manuscripts.
In the event of acceptance, this file will be published in conjunction with your paper and will include the anonymous referee reports, your point-by-point response and all pertinent correspondence relating to the manuscript. Let us know whether you agree with the publication of the RPF and as here, if you want to remove or not any figures from it prior to publication.
Please note that the Authors checklist will be published at the end of the RPF.
EMBO Molecular Medicine has a "scooping protection" policy, whereby similar findings that are published by others during review or revision are not a criterion for rejection. Should you decide to submit a revised version, I do ask that you get in touch after three months if you have not completed it, to update us on the status. The authors have made substantial effort to address the concerns of the reviewers. Critique: Referee #1: The IHC of blood vessel comparing anti-CD31, 10B3 and J591 was helpful, although the resolution was too low to delineate the vessels from the stroma. A much higher magnification or a spot view is necessary to make the point. To the untrained eye, the vascular staining between J591 and 10B3 was near identical (contrary to the statement that J591 was less). Based on these stainings, hard to understand figure 1C and figure 1D, where there was difference in staining of SSC, and evidence of staining of vessels in prostate CA. This raises the possibility that 10B3 was recognizing an epitope shared by proteins distinct from PSMA. The inhibition study with soluble PSMA should be done on tumor sections from the same SCC tumor. However, even if there was inhibition, the potential for a cross reactive epitope on proteins other than PSMA could not be ruled out. If 10B3 was specific for PSMA, the discordance between prostate (where J591=10B3), and lung cancer (where 10B3>J591) was hard to explain. This was especially troubling with 10B3>J591 for Ve staining in these same prostate tumors. Looking back at the BIACORE in figure EV1, 10B3 had substantially lower RU values (4-fold) than J591, inconsistent with the stronger staining of the SSC tumors if this binding was only to PSMA in figure EV2A or fig 1C, again raising the possibility that 10B3 was reacting with an epitope shared between PSMA and other proteins. The IHC in EV2C indicated staining, but again at such low magnification, its hard to make out details. The pattern of anti-CD31 also did not resemble those of 10B3 and should be stated clearly. The PSMA specific tracer should be disclosed. Was it PSMA-617? If it was, was it done with IRB approval or as part of standard clinical care using an approved agent? The authors provided the preceding PSA levels, 30 days, 120 days, and 250 days prior to their high level on the day of treat, with the implication that the patients were not on any therapy for metastatic prostate cancer during those periods. Otherwise as standard of care, they would have had PSA levels. The authors should state that clearly in the description. The rapid drop in PSA was followed by a rapid rise within 20-30 days. The authors should also make a note of it. In Fig1D, the use of unpaired T test requires the variance to be same between groups. Was that assumption valid? Referee #3 Regarding Fc attenuation using deletions E233P; L234V; L235A;ΔG236; D265G; A327Q; A330S (EU-index). The authors should explain why so many mutations for silencing Fc, when 3 of them should be more than enough. The number of mutations could potentially make the protein a lot more immunogenic, suggested by the authors in the treated patient. Fig 4 showed cytokine release in the absence of tumors: IgGsc had no cytokine release presumably because of lack of binding to PBMC. Surprising that the anti-CD3 scFv had none, not even a little bit of activation of T cells. A more definitive way to demonstrate Fc silencing is binding (ELISA or BIACORE) to FcR or C1q, or functional assays such as complement activation/cytotoxicity, or ADCC. The authors stated that FACS titration on CD4 and CD8 T cells in PBMC showed EC50 values similar to those obtained with Jurkat cells. But there was a clear difference in the curves between PBMC and Jurkat. The proposed explanation that "The difference in the plateau level can be explained by the different binding of the fluorescence labeled detection antibody to both formats,"does not make sense", since it was only found in Jurkat and not in PBMC. The statement that "The results demonstrate that the affinities of the chimeric 10B3 vs. chimeric J591 are comparable, in accordance with FACS titration data performed on PSMA positive cells ( Fig  EV1E)" was not accurate. The RU values were very different. Developing a non-prostate PSMA model to validate the claims for vascular reactivity is understandable. In the absence of data, the authors should not claim superiority of 10B3 versus J591, given the equivocal IHC data presented. The timing of treatment, 24 hours after tumor implantation was clearly before tumor establishment and should be stated clearly instead of using the mislabeling term "established" tumor systems. This has significant implications for the interpretation of the data. The term tumor regression was also misleadding sinnce there was no measurable tumor to regress from. A more appropriate description is "it suppressed tumor growth". The ability to achieve such a high serum level of BsAb is noteworthy since serum levels in clinical BsAb studies not easily measurable even at high doses. The absence of CRS was even more remarkable. They should make a note of it in the results and in the discussion.

Referee#3 (Remarks for Author):
The authors have made substantial effort to address the concerns of the reviewers.
We thank the reviewer for acknowledging our efforts and we think that the helpful comments have largely improved our manuscript. We hope that by conducting the additional amendments requested in the second round of revision, the issues of the reviewers are fully addressed.

The IHC of blood vessel comparing anti-CD31, 10B3 and J591 was helpful, although
the resolution was too low to delineate the vessels from the stroma. A much higher magnification or a spot view is necessary to make the point. To the untrained eye, the vascular staining between J591 and 10B3 was near identical (contrary to the statement that J591 was less). Based on these stainings, hard to understand figure 1C and figure   1D, where there was difference in staining of SSC, and evidence of staining of vessels in prostate CA. This raises the possibility that 10B3 was recognizing an epitope shared by proteins distinct from PSMA.
We apologize that the description of our results was obviously misleading in the previous version of the manuscript. We agree with the reviewer that in Fig EV2B, Fig 1D,  Again, we apologize that the initial description of the respective results was confusing. We are convinced that this circumstance has lead the reviewer to consider the possibility that showed a similar binding affinity to PSMA in vitro (Fig EV1E-F Fig   1B- D and EV2 revealed that (i) EV2D). While it cannot fully be excluded that 10B3 might recognize an epitope/protein distinct from PSMA, in our view the immunoprecipitation data shown in Fig EV1A- In addition, for information of the reviewer, we also uploaded high resolution images of IHC slides to further address this issue. Please see the attached file: SourceDataForEV2A-D.

The inhibition study with soluble PSMA should be done on tumor sections from the same SCC tumor.
We fully agree with the notion of the reviewer, and indeed we have used consecutive 3µm sections of the very same tumor in this experiment. We apologize that this was not described clearly in the previous version of the manuscript; it is now stated clearly in a respective paragraph in the methods section and the figure legend. Accordingly, the manuscript has been amended as follows: Method section, page 16, line 16: "…In the blocking experiment presented in Fig EV2D,  3. However, even if there was inhibition, the potential for a cross reactive epitope on proteins other than PSMA could not be ruled out. If 10B3 was specific for PSMA, the discordance between prostate (where J591=10B3), and lung cancer (where 10B3>J591) was hard to explain. This was especially troubling with 10B3>J591 for Ve staining in these same prostate tumors.
As stated above in our reply to comment 1, we apologize for not clearly describing the results in the previous version of the manuscript, which in our view has caused confusion. Please refer to comment 1 above for a detailed explanation how the manuscript was amended to clarify this issue.

Looking back at the BIACORE in figure EV1, 10B3 had substantially lower RU values (4-fold) than J591, inconsistent with the stronger staining of the SSC tumors if this
binding was only to PSMA in figure EV2A or fig 1C, Fig 1A, 10B3 binds to a conformational epitope of PSMA protein and shows a slow association kinetic (reflected by a low K on ) if compared to J591. This is compensated by the lower K off rate of 10B3 compared to that of J591, overall resulting in a comparable KD value.
To address this issue of the reviewer, the manuscript was amended as follows: Discussion section, page 12, line 1: "Analysis of binding by both, surface plasmon resonance and flow cytometry binding assays showed comparable affinities for J591 and 10B3 (Fig EV1E-F). This indicates that the differences observed by immunohistology might be attributable to a better accessibility of the 10B3-epitope once the cells are organized within a tissue." 5. The IHC in EV2C indicated staining, but again at such low magnification, it's hard to make out details. The pattern of anti-CD31 also did not resemble those of 10B3 and should be stated clearly.
We thank the reviewer for raising this issue and have, as already stated above, included spot views in Fig EV2C to better illustrate binding of the Fabsc and IgGsc constructs to tumor associated vasculature. We also agree thatin contrast to staining with monospecific 10B3 and J591 antibodies shown in Fig EV2B - Fig EV2B for  We apologize but we are not sure that we understand the issue of the reviewer correctly. The three patients were suffering from metastatic prostate carcinoma refractory to standard medical treatment regimes. This is reflected by the rise of PSA values prior to antibody therapy. While the patients had undergone treatment in the (depicted) time prior to CC-1 therapy, they had not received any disease specific treatment for at least 4 weeks. This information has now been included in the figure legend EV5.
To further address the rightful comment of the reviewer, we included a statement that PSA levels rose again within 20-30 days after cessation of CC-1 treatment.
Accordingly, the manuscript has been amended as follows: Results section, page 10, line 21: "In all patients, profound T-cell activation and a rapid and marked decline of PSA levels were observed, which rose again 20-30 days after cessation of CC-1 treatment (Fig 7, Fig EV5B)."

In Fig1D, the use of unpaired T test requires the variance to be same between groups. Was that assumption valid?
We thank the reviewer for raising this important issue. As outlined in our reply to comment 1 of reviewer 1 above, we apologize that we did not describe the results of IHC analysis clearly enough in the previous version of the manuscript.
We have now largely amended the description of the IHC data in the manuscript. We hope that the new description of these results and the amendments introduced in the manuscript are suited to sufficiently address this notion and rule out the doubts of the reviewer.
6. The timing of treatment, 24 hours after tumor implantation was clearly before tumor establishment and should be stated clearly instead of using the mislabeling term "established" tumor systems. This has significant implications for the interpretation of the data. The term tumor regression was also misleading since there was no measurable tumor to regress from. A more appropriate description is "it suppressed tumor growth".
We again apologize that obviously the description of our results was misleading and caused a misunderstanding. We have performed two different mouse models: (i) a metastasis mouse model ( Fig 6A) and (ii) an established tumor mouse model (Fig 6B-C Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed report from referee #3 who is supportive of publication pending minor revisions (see below). I am thus pleased to inform you that we will be able to accept your manuscript pending the following final minor amendments: 1) Referee's comments: We would like you to discuss the referee's points in writing. If you do have data at hand (clear example of prostate carcinoma vasculature and SPR data), we would be happy for you to include it, however we will not ask you to provide any additional experiments at this stage. Please provide a letter INCLUDING my comments and the reviewer's reports and your detailed responses to their comments (as Word file).
2) Main manuscript text: -Please answer/correct the changes suggested by our data editors in the main manuscript file (in track changes mode). This file will be sent to you in the next couple of days. Please use this file for any further modification.
-Please remove the yellow highlighted text. -Abstract: please remove "rather" from the first sentence. We would also encourage you to rephrase the second sentence to make it clearer. -Material and methods: o Cells: Please indicate the origin of cells (human vs. murine), and whether they were authenticated (if applicable) and tested for mycoplasma contamination. o Patients data: Please include a statement that informed consent was obtained from all subjects and the full statement that the experiments conformed to the principles set out in the WMA Declaration of Helsinki and the Department of Health and Human Services Belmont Report. (This also applies to patients' samples) -Please include a Data availability section: Primary datasets produced in this study need to be deposited in an appropriate public database (see https://www.embopress.org/page/journal/17574684/authorguide#dataavailability). If not applicable, the following sentence should be included: "This study includes no data deposited in external repositories".  EV2A 4) For more information: https://www.embopress.org/doi/full/10.15252/emmm.201910874 should be listed as a reference, not as a weblink in the FMI section. 5) Thank you for providing a synopsis. I slightly modified the text to fit our style and format, please let me know if you agree with the following: Insufficient penetration of immune cells and therapeutic antibodies into the tumor core is a major limitation in the immunotherapy field. This study reports the development of a novel bispecific antibody, named CC-1, for improved dual targeting of tumor-and vascular cells in PSMA positive tumors.
• A novel PSMA antibody (10B3) exhibiting enhanced reactivity with tumor-and vascular cells in samples from prostate carcinoma and squamous cell carcinoma of the lung was generated.
• Two different bispecific antibodies comprising 10B3 and anti-CD3 single chain in a Fabsc-and IgGsc-format were constructed and characterized.
• In vivo application of both bispecific antibodies revealed that only the IgGsc-molecule localized at a given tumor site, resulting in effective tumor cell destruction.
• A first-in-man application of the IgGsc-molecule, designated CC-1, in three patients with metastasized prostate carcinoma, demonstrated profound T cell activation and a rapid decline of elevated PSA levels.
• A first-in-man clinical study in patients with prostate carcinoma is currently ongoing (NCT04104607). 6) As part of the EMBO Publications transparent editorial process initiative (see our Editorial at http://embomolmed.embopress.org/content/2/9/329), EMBO Molecular Medicine will publish online a Review Process File (RPF) to accompany accepted manuscripts. In the event of acceptance, this file will be published in conjunction with your paper and will include the anonymous referee reports, your point-by-point response and all pertinent correspondence relating to the manuscript. Let us know whether you agree with the publication of the RPF and as here, IF YOU WANT TO REMOVE OR NOT any figures from it prior to publication. Please note that the Authors checklist will be published at the end of the RPF.
I look forward to receiving your revised manuscript.

Lise Roth
Lise Roth, PhD Editor EMBO Molecular Medicine To submit your manuscript , please follow this link:

Link Not Available
The system will prompt you to fill in your funding and payment information. This will allow Wiley to send you a quote for the article processing charge (APC) in case of acceptance. This quote takes into account any reduction or fee waivers that you may be eligible for. Authors do not need to pay any fees before their manuscript is accepted and transferred to our publisher. ***** Reviewer's comments ***** Referee #3 (Remarks for Author): Want to thank the authors for making substantial improvements in the manuscript. The following comments should be easily addressable. 1. It's still unclear which IHC images provide evidence that 10B3 stains prostate carcinoma vasculature better than J591. The text references Figure 1B and EV2 for this claim, but the image in 1B does not point out vasculature, and EV2 only shows slides from lung SSC. The authors should provide a clear example of the prostate carcinoma vasculature that contributed to the summarized results in 1D. 2. A few aspects of the SPR data in figure EV1F are worth addressing. For J591, the minor differences in RU between concentrations imply the chip is nearing saturation before switching to buffer. The SPR data for 10B3, however, shows a clear linear relationship between concentration and RU values, implying the chip is at an equilibrium instead of saturation prior to running buffer. Additionally, 10B3 seems to show less reduction in RU over time compared to J591. Considering the overall KD values are relatively similar, these observations imply different values in Kon and Koff between the two antibodies. It would be meaningful to show the Kon and Koff values of the two antibodies along with the KDs. Also, 10B3 showed such minimal loss in RU over 1000 seconds that the authors may not have adequately been able to determine the Koff. Performing SPR at 37*C or increasing the time for koff measurement should allow for a more reliable measurement of these kinetic properties.
Manuscript number: EMM-2019-11902 ***** Reviewer's comments ***** Referee #3 (Remarks for Author): 1. It's still unclear which IHC images provide evidence that 10B3 stains prostate carcinoma vasculature better than J591. The text references Figure 1B and EV2 for this claim, but the image in 1B does not point out vasculature, and EV2 only shows slides from lung SSC. The authors should provide a clear example of the prostate carcinoma vasculature that contributed to the summarized results in 1D.
To address this comment, we have now included an exemplary staining picture in the figure   EV2A reflecting the superiority of 10B3 staining in prostate carcinoma vessels over J591. Follow us on Twitter @EmboMolMed Sign up for eTOCs at embopress.org/alertsfeeds *** *** *** IMPORTANT INFORMATION *** *** *** SPEED OF PUBLICATION The journal aims for rapid publication of papers, using using the advance online publication "Early View" to expedite the process: A properly copy-edited and formatted version will be published as "Early View" after the proofs have been corrected. Please help the Editors and publisher avoid delays by providing e-mail address(es), telephone and fax numbers at which author(s) can be contacted.
Should you be planning a Press Release on your article, please get in contact with embomolmed@wiley.com as early as possible, in order to coordinate publication and release dates.

LICENSE AND PAYMENT:
All articles published in EMBO Molecular Medicine are fully open access: immediately and freely available to read, download and share.
EMBO Molecular Medicine charges an article processing charge (APC) to cover the publication costs. You, as the corresponding author for this manuscript, should have already received a quote with the article processing fee separately. Please let us know in case this quote has not been received.
Once your article is at Wiley for editorial production you will receive an email from Wiley's Author Services system, which will ask you to log in and will present you with the publication license form for completion. Within the same system the publication fee can be paid by credit card, an invoice, pro forma invoice or purchase order can be requested.
Payment of the publication charge and the signed Open Access Agreement form must be received before the article can be published online.

PROOFS
You will receive the proofs by e-mail approximately 2 weeks after all relevant files have been sent o our Production Office. Please return them within 48 hours and if there should be any problems, please contact the production office at embopressproduction@wiley.com.
Please inform us if there is likely to be any difficulty in reaching you at the above address at that time. Failure to meet our deadlines may result in a delay of publication.
All further communications concerning your paper proofs should quote reference number EMM-2019-11902-V4 and be directed to the production office at embopressproduction@wiley.com. Do the data meet the assumptions of the tests (e.g., normal distribution)? Describe any methods used to assess it.
Is there an estimate of variation within each group of data?
Is the variance similar between the groups that are being statistically compared?

Reporting Checklist For Life Sciences Articles (Rev. June 2017)
This checklist is used to ensure good reporting standards and to improve the reproducibility of published results. These guidelines are consistent with the Principles and Guidelines for Reporting Preclinical Research issued by the NIH in 2014. Please follow the journal's authorship guidelines in preparing your manuscript.

Journal Submitted to: EMBO Molecular Medecine
Corresponding Author Name: Gundram Jung

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).
Ex vivo experiments were usually performed with at least three independent experiments using three different PBMCs donor. Each condition was measured in triplicate within the same experiment. Sample sizes were selected on the basis of previous studies and to ensure robust statistical analysis. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

Captions
For in vivo experiments, a minimum number of animals (groups of 5 animals per condition) were used, allowing a reliable statistical analysis to be done. Patient data were presented on an individual basis. Respective statements are included.
Animals were excluded only if they became sick or loose lots of weight or did not develop a tumor (in the established mouse model).
The animal groups were assigned blindly.
Yes. The Data distribution was checked for normality by D'Agostino-Pearson normality test.
Yes. SD or SEM are also included.
Yes, homogeneity of variance was checked by Levene test.
Randomization was used in all animal experiments presented in the manuscript.
Ex vivo experiments were not blind. But occasionally the same experiment was conducted independently by two different scientists. For the In vivo experiments, animal group assignments and treatments were random. The in vivo experiments described in this paper were conducted in half-blinded way, since the person setting up the assay was not always the one analysis it.

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.