Mutant SF3B1 promotes malignancy in PDAC

Decision letter after peer review:

Thank you for submitting your article "Mutant SF3B1 promotes PDAC malignancy through TGF-β resistance" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Erica Golemis as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) Additional experimentation testing the validity of findings in vivo is required. All three reviewers requested that data beyond organoids and cell lines be added. For example, orthotopic implantation of the organoids or cells used to examine how the Sf3B1-K700E mutant or Map3k7 expression affects tumor growth and metastasis in vivo is highly recommended as suggested by the reviewers.

2) Increase the human relevance through use of human PDAC lines that have the K700E mutation.

3) Articulate more clearly how the work is novel and distinguishes itself from previous work highlighting Map3k7 as downstream of the Sf3b1-K700E mutation in other cancers.

Reviewer #1 (Recommendations for the authors):

Additional analysis of the invasion properties following both expression of Sf3B1-K700E and also following Map3k7 expression, enhanced analysis of the PanINs and PDACs formed in KC and KPC +/- the Sf3b1-K700E mutation, examining effects of Map3k7 on tumors in vivo expressing Sf3b1-K700E would all add to the work significantly.

Reviewer #2 (Recommendations for the authors):

Increasing the human relevance of the study would greatly broaden the impact. Some suggestions are listed above. Another possibility would be to seek out cell lines with the SF3BP1 mutation if they are available.

Comment on figures and writing:

A scheme depicting the genotype of the mouse models, specifically of the SF3B1K700E allele would help the reader quickly determine whether the model is a knock-in, a transgenic mouse and so on.

Reviewer #3 (Recommendations for the authors):

– The authors propose that SF3B1K700E functions much like SMAD4 mutations in blocking TGFb signaling. But the authors nevertheless propose that SF3B1K700E mutant tumors would nevertheless still benefit from further SMAD4 mutation, even though normally mutations that target the same pathway tend to be mutually exclusive, as SMAD4 may be a more potent block to TGFb signaling. From analysis of patient genomic data, can the authors determine if SF3B1K700E and SMAD4 mutations co-occur or if they are mutually exclusive? If they are mutually exclusive, would this not provide additional compelling patient in vivo evidence that the critical function of SF3B1K700E mutation is to tune TGFb signaling as SMAD4 mutations are no longer occurring in these samples?

– The authors show in figure 2F that KPC-SF3B1K700E cancer cells have a lower expression of EMT genes relative to KPC cell lines and further show in figure 3A that treatment with TGFβ is able to induce upregulation of some of these genes in KPC cells relative to control. As the main premise of the paper in the following figures becomes how the SF3B1K700E mutation can suppress and EMT lethal cell state promoted by TGFβ, it would be useful to see whether TGFβ treatment fails to induce upregulation of these genes in KPC-SF3B1K700E cancer cells relative to untreated control.

– While the authors show that lower expression of MAP3K7 can impair TGFB mediated cell death, a more nuanced discussion of the role MAP3K7 in programmed cell death is warranted, as there is ample literature on this subject, including how MAP3K7 inhibitor Takinib has been used in clinical trials against PDAC (PMID: 31695153, 25146924, 21743023).

– In figure 3B the authors show functional data that suggests SF3B1K700E pancreatic cancer cells have altered cell migration in a TGFβ-dependent context, likely related to the downregulation of EMT related genes observed before. If at all possible, whether these cells also exhibit an altered ability to metastasize in vivo (maybe by gross analysis of metastatic nodules or IHC of micro metastasis in other tissues) should also be commented on when describing the pathology of the mouse model.

– Figure 4 title should be modified to reflect the second main premise of the figure (4D-F), regarding regulation of MAP3K7.

Essential revisions:

1) Additional experimentation testing the validity of findings in vivo is required. All three reviewers requested that data beyond organoids and cell lines be added. For example, orthotopic implantation of the organoids or cells used to examine how the Sf3B1-K700E mutant or Map3k7 expression affects tumor growth and metastasis in vivo is highly recommended as suggested by the reviewers.

We thank the reviewers for raising this important point. To address their concerns, we established and analyzed three in vivo models:

• Orthotopic transplantation of tumor-derived KPC and KPC-Sf3b1^K700E cells (model of late-stage PDAC tumors)

• Orthotopic transplantation of KPC cells with shRNA-mediated knockdown of Map3k7

• Orthotopic transplantation of KPC-Sf3b1^K700E cells overexpressing Map3k7

To briefly summarize our results, tumors derived from orthotopically transplanted KPC and KPC-Sf3b1^K700E cells did not differ in size (see Figure 1—figure supplement 1M, N, and more detailed explanations in the response to Reviewer #1). Since orthotopic transplantation of KPC cells is a model of late-stage tumors (KPC cells are derived from fully developed PDAC tumors), these results are not surprising and support our initial hypothesis that Sf3b1^K700E is more relevant in early stage PDAC rather than late stage PDAC. As expected from these results, also modulating the levels of Map3k7 did not have an effect in tumor growth in orthotopically transplanted KPC or KPC-Sf3b1^K700E tumors (Author response image 1).

To further detangle the effect Sf3b1^K700E at different tumor stages, we re-analyzed our autochthonous model at an early and late stage during tumor progression. Histological examination at 5 weeks revealed a significant increase in ADM and PanIN incidence (shown by MUC5AC and CK19 IF staining) and a concomitant decrease of acinar cells (shown by b-amylase staining) in KPC-Sf3b1^K700E tumors (Figure 1G-J, see Figure 1—figure supplement 1J, K). Analyzing tumors at 9 weeks of age did not show differences in CK19 staining and fibrosis between KPC and KPC-Sf3b1^K700E tumors anymore (see Figure 1—figure supplement 1F-I). These findings are in line with our autochthonous KRAS-G12D pre-cancer mouse model, in which Sf3b1^K700E leads to a higher incidence of spontaneous progression of PanINs into PDAC (Figure 1L of the revised manuscript).

Taken together, orthotopic transplantation of KPC cells represents a late stage PDAC model, in which Sf3b1^K700E does not have an influence on tumor growth anymore. Of note, in a previous study we tried to transplant in vitro induced pancreatic KPC organoids, but failed to observe any tumor growth in vivo. Thus, while using KPC cells isolated from fully established PDAC seems to be a requirement for the orthotopic transplantation, these cells, unfortunately, cannot be used to test whether Sf3b1^K700E drives pancreatic tumor progression via MAP3K7. An autochtonous KPC model with downregulated Map3k7 or KPC-Sf3b1^K700E model with overexpression of Map3k7 would need to be generated to test this hypothesis in vivo. However, it would take approximately 2 years to generate and breed such models, which would be out of scope for a revision. Therefore, in the revised manuscript we discuss the limitation of our study. In short, we explain that we could only show in vitro in cell lines and organoids that Sf3b1^K700E leads to TGF-β resistance in pancreatic duct cells via missplicing of Map3k7, and that it would require further studies to proof that Map3k7 regulation is also relevant for PDAC progression in vivo (line 276-286).

2) Increase the human relevance through use of human PDAC lines that have the K700E mutation.

In the revised manuscript, we present multiple evidence showing the influence of SF3B1^K700E on human PDAC lines. To summarize, we found that SF3B1^K700E expression leads to Map3k7 missplicing in five human PDAC cell lines and one human pancreatic duct cell line (Figure 4F, see Figure 4—figure supplement 1D). Furthermore, human pancreatic duct cells that overexpress SF3B1^K700E demonstrated a higher susceptibility to TGF-β-induced growth restriction compared to cells overexpressing wildtype SF3B1 (Figure 3J). We also executed wound healing assays to empirically evaluate the effect of SF3B1^K700E on EMT in human PDAC cells. Consistent with our hypothesis, the results suggest that SF3B1^K700E negatively impacts the migratory abilities of PANC-1 cells when treated with TGF-β. We have included these findings in the manuscript (Figure 2I, J). Lastly, we demonstrate that preconditioning human PANC-1 PDAC cells with a Map3k7 inhibitor before exposure to TGF-β1 significantly reduces their migratory capabilities in wound healing assays (Figure 5L, M).

3) Articulate more clearly how the work is novel and distinguishes itself from previous work highlighting Map3k7 as downstream of the Sf3b1-K700E mutation in other cancers.

In the revised manuscript we more clearly state that we are the first to demonstrate that (a) SF3B1^K700E increases malignancy and reduces survival in a mouse model of PDAC (line 242-243) and (b) SF3B1^K700E decreases sensitivity of murine and human pancreatic duct cells to the tumor suppressive effects of TGF-β (line 246-248). Of note, previous work describing that SF3B1^K700E induces missplicing of Map3k7 did not provide a link to TGF-β-resistance and reduced survival. In addition, no other manuscript discussing SF3B1^K700E induced missplicing of Map3k7 described the impact of reduced Map3k7 on EMT and apoptosis.

Reviewer #1 (Recommendations for the authors):

Additional analysis of the invasion properties following both expression of Sf3B1-K700E and also following Map3k7 expression, enhanced analysis of the PanINs and PDACs formed in KC and KPC +/- the Sf3b1-K700E mutation, examining effects of Map3k7 on tumors in vivo expressing Sf3b1-K700E would all add to the work significantly.

For analyzing the invasion properties, we performed wound healing assays / scratch assays with PANC-1 cells with inducible SF3B1 WT / K700E overexpression. We observed a significant difference in migratory capacity between SF3B1 WT / K700E overexpressing PANC-1 cells when stimulated with TGF-β. We added this data to the revised manuscript (Figure 2I, J). In addition, chemical inhibition of MAP3K7 with Takinib led to a reduced migration capacity in wound healing assays when PANC-1 cells were stimulated with TGF-β (Figure 5L, M).

To enhance the analysis of PanINs and PDAC in early- and late stage KPC and KPC-Sf3b1^K700E tumors we assessed expression of CK19, MUC5A1 and b-Amylase by immunofluorescent histochemistry, as well as ADM structures. Significant differences were observed in early-stage tumors but not in late-stage tumors (Figure 1G-J, see Figure 1—figure supplement 1F-K).

To examine the effect of Map3k7 on tumors in vivo, we established orthotopic transplantation models with KPC and KPC-Sf3b1^K700E cells, with overexpression and knockdown of Map3k7 (see Author response image 1). However, in contrast to the autochthonous mouse model, orthotopically transplanted KPC and KPC-Sf3b1^K700E cells did not show differences in tumor size (see Figure 1—figure supplement 1I, J). These data are not surprising, as orthotopic KPC transplantation is a late-stage cancer model (KPC cells are isolated from fully establish PDAC tumors), and our data suggest that Sf3b1^K700E rather plays a role during early stages of PDAC development.

Importantly, our results that Sf3b1^K700E does not affect growth of orthotopically transplanted KPC cells means that manipulating Map3k7 levels in this experimental setup should also not affect PDAC growth. Indeed, downregulation of Map3k7 via shRNA in KPC cells yielded tumors of comparable in size to tumors derived from control KPC cells (Author response image 1A, B). Moreover, the EMT genes that were differentially expressed in our autochthonous (+/- K700E) mouse models were expressed at similar levels in shMap3k7 KPC tumors vs. control KPC tumors (Author response image 1C). Likewise, overexpressing Map3k7 (OE Map3k7) in transplanted KPC-Sf3b1^K700E cells led to the development of tumors of comparable size to control tumors without Map3k7 overexpression, again without differential expression of EMT genes between KPC-Sf3b1^K700E cells with- and without Map3k7 overexpression (Author response image 1D-F).

Author response image 1

Download asset Open asset

Reviewer #2 (Recommendations for the authors):

Increasing the human relevance of the study would greatly broaden the impact. Some suggestions are listed above. Another possibility would be to seek out cell lines with the SF3BP1 mutation if they are available.

In the revised manuscript we present data on missplicing of MAP3K7 in 5 human PDAC cell lines and one human pancreatic duct cell line stably overexpressing SF3B1-K700E (Figure 4F). We further demonstrate that the human pancreatic duct cell line overexpressing SF3B1-K700E is less sensitive to growth inhibition by TGF-β than SF3B1-WT overexpressing cells, while the PDAC cell lines are insensitive to TGF-β, regardless of their genotype (Figure 3J, see Figure 3—figure supplement I). These data support our hypothesis that SF3B1-K700E decreases sensitivity to TGF-β induced apoptosis/growth inhibition in pancreatic duct cells, but that in late stage PDAC misregulation of other TGF-β modulating factors plays a more important role.

Furthermore, performing wound healing assays / scratch assays with human PANC-1 cells with inducible SF3B1-WT/ K700E overexpression revealed a significant difference in migratory capacity when stimulated with TGF-β. We added this data to the revised manuscript (Figure 2I, J).

Comment on figures and writing:

A scheme depicting the genotype of the mouse models, specifically of the SF3B1K700E allele would help the reader quickly determine whether the model is a knock-in, a transgenic mouse and so on.

We added a schematic overview of the SF3B1^K700E knock-in to the revised manuscript (Figure 1A).

Reviewer #3 (Recommendations for the authors):

– The authors propose that SF3B1K700E functions much like SMAD4 mutations in blocking TGFb signaling. But the authors nevertheless propose that SF3B1K700E mutant tumors would nevertheless still benefit from further SMAD4 mutation, even though normally mutations that target the same pathway tend to be mutually exclusive, as SMAD4 may be a more potent block to TGFb signaling. From analysis of patient genomic data, can the authors determine if SF3B1K700E and SMAD4 mutations co-occur or if they are mutually exclusive? If they are mutually exclusive, would this not provide additional compelling patient in vivo evidence that the critical function of SF3B1K700E mutation is to tune TGFb signaling as SMAD4 mutations are no longer occurring in these samples?

We performed a mutual exclusivity analysis with SF3B1-K700E and components of the TGF-β pathway in published human PDAC databases (www.cbioportal.org), but did not observe mutual exclusivity between these genes. Of note, the SF3B1-K700E mutation is rare, and due to the limited sample size (only 7 samples contain the SF3B1-K700E mutation) the value of the analysis is limited.

– The authors show in figure 2F that KPC-SF3B1K700E cancer cells have a lower expression of EMT genes relative to KPC cell lines and further show in figure 3A that treatment with TGFβ is able to induce upregulation of some of these genes in KPC cells relative to control. As the main premise of the paper in the following figures becomes how the SF3B1K700E mutation can suppress and EMT lethal cell state promoted by TGFβ, it would be useful to see whether TGFβ treatment fails to induce upregulation of these genes in KPC-SF3B1K700E cancer cells relative to untreated control.

For the revised manuscript, we assessed the expression of the EMT-gene set of KPC and KPC-SF3B1^K700E cells stimulated with TGF-β. Indeed, 6 of 9 genes showed no or impaired upregulation in KPC-SF3B1^K700E cells in comparison to KPC-SF3B1 (see Figure 2—figure supplement 1G)

– While the authors show that lower expression of MAP3K7 can impair TGFB mediated cell death, a more nuanced discussion of the role MAP3K7 in programmed cell death is warranted, as there is ample literature on this subject, including how MAP3K7 inhibitor Takinib has been used in clinical trials against PDAC (PMID: 31695153, 25146924, 21743023).

In the revised manuscript, we discuss the multi-faceted role of MAP3K7 in apoptosis (line 269-275).

– In figure 3B the authors show functional data that suggests SF3B1K700E pancreatic cancer cells have altered cell migration in a TGFβ-dependent context, likely related to the downregulation of EMT related genes observed before. If at all possible, whether these cells also exhibit an altered ability to metastasize in vivo (maybe by gross analysis of metastatic nodules or IHC of micrometastasis in other tissues) should also be commented on when describing the pathology of the mouse model.

We analyzed macroscopic metastatic nodules in the liver of KPC and KPC-SF3B1^K700E mice. We found that 2 out of 32 KPC mice vs. 5 of 22 KPC-SF3B1^K700E mice developed macro-metastases in liver (of animals older than 56 days). However, this difference is statistically not significant (chi-square test, p=0.37), and the role of EMT in metastasis formation is complex and controversial. We therefore did not add this data to the manuscript.

– Figure 4 title should be modified to reflect the second main premise of the figure (4D-F), regarding regulation of MAP3K7.

We changed the title of Figure 4 accordingly (line 806).