Tumor Cell Spatial Organization Directs EGFR/RAS/RAF Pathway Primary Therapy Resistance through YAP Signaling

Reviewer #1 (Public review):

Summary:

In this study, Nakagawa and colleagues report the observation that YAP is differentially localized, and thus differentially transcriptionally active, in spheroid cultures versus monolayer cultures. YAP is known to play a critical role in the survival of drug-tolerant cancer cells, and as such, the higher levels of basally activated YAP in monolayer cultures lead to higher fractions of surviving drug-tolerant cells relative to spheroid culture (or in vivo culture). The findings of this study, revealed through convincing experiments, are elegantly simple and straightforward, yet they add significantly to the literature in this field by revealing that monolayer cultures may actually be a preferential system for studying residual cell biology simply because the abundance of residual cells in this format is much greater than in spheroid or xenograft models. The potential linkage between matrix density and stiffness and YAP activation, while only speculated upon in this manuscript, is intriguing and a rich starting point for future studies.

Although this work, like any important study, inspires many interesting follow-on questions, I am limiting my questions to only a few minor ones, which may potentially be explored either in the context of the current study or in separate, follow-on studies.

Strengths:

The major strengths of the work are described above.

Weaknesses:

Rather than considering the following points as weaknesses, I instead prefer to think of them as areas for future study:

(1) Given the field's intense interest in the biology and therapeutic vulnerabilities of residual disease cells, I suspect that one major practical implication of this work could be that it inspires scientists interested in working in the residual disease space to model it in monolayer culture. However, this relies upon the assumption that drug-tolerant cells isolated in monolayer culture are at least reasonably similar in nature to drug-tolerant cells isolated from spheroid or xenograft systems. Is this true? An intriguing experiment that could help answer this question would be to perform gene expression profiling on a cell line model in the following conditions: monolayer growth, drug tolerant cells isolated from monolayer growth conditions, spheroid growth, drug tolerant cells isolated from spheroid growth conditions, xenograft tumors, and drug tolerant cells isolated from xenograft tumors. What are the genes and programs shared between drug-tolerant cells cultured in the three conditions above? Which genes and programs differ between these conditions? Data from this exercise could help provide additional, useful context with which to understand the benefits and pitfalls of modeling residual tumor cell growth in monolayer culture.

(2) In relation to the point above, there is an interesting and established connection between mesenchymal gene expression and YAP/TAZ signaling. For example, analyses of gene expression data from human tumors and cell lines demonstrate an extremely strong correlation between these two gene expression programs. Further, residual persister cancer cells have often been characterized as having undergone an EMT-like transition. From the analysis above, is there evidence that residual tumor cells with increased YAP signaling also exhibit increased mesenchymal gene expression?

Reviewer #2 (Public review):

The manuscript by Nakagawa R, et al describes a mechanism of how NSCLC cells become resistant to EGFR and KRAS G12C inhibition. Here, the authors focus on the initial cellular changes that occur to confer resistance and identify YAP activation as a non-genetic mechanism of acute resistance.

The authors performed an initial xenograft study to identify YAP nuclear localization as a potential mechanism of resistance to EGFRi. The increase in the stromal component of the tumors upon Afatinib treatment leads the authors to explore the response to these inhibitors in both 2D and 3D culture. The authors extend their findings to both KRAS G12C and BRAF inhibitors, suggesting that the mechanism of resistance may be shared along this pathway.

The paper would benefit from additional cell lines to determine the generalizability of the findings they presented. While the change in the localization of YAP upon Afatinib treatment was identified in a xenograft model, the authors do not return to animal models to test their potential mechanism, and the effects of the hyperactivated S127A YAP protein on Afatinib sensitivity in culture are modest. Also, combination studies of YAP inhibitors and EGFR/RAS/RAF inhibitors would have strengthened the studies.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

In this study, Nakagawa and colleagues report the observation that YAP is differentially localized, and thus differentially transcriptionally active, in spheroid cultures versus monolayer cultures. YAP is known to play a critical role in the survival of drug-tolerant cancer cells, and as such, the higher levels of basally activated YAP in monolayer cultures lead to higher fractions of surviving drug-tolerant cells relative to spheroid culture (or in vivo culture). The findings of this study, revealed through convincing experiments, are elegantly simple and straightforward, yet they add significantly to the literature in this field by revealing that monolayer cultures may actually be a preferential system for studying residual cell biology simply because the abundance of residual cells in this format is much greater than in spheroid or xenograft models. The potential linkage between matrix density and stiffness and YAP activation, while only speculated upon in this manuscript, is intriguing and a rich starting point for future studies.

Although this work, like any important study, inspires many interesting follow-on questions, I am limiting my questions to only a few minor ones, which may potentially be explored either in the context of the current study or in separate, follow-on studies.

We appreciate Reviewer #1's comments that our work is of importance to the field and particularly that it will "...add significantly to the literature in this field by revealing that monolayer cultures may actually be a preferential system for studying residual cell biology..." We have sought to highlight the importance of how our findings could be applied to study resistance mechanisms at various points in the manuscript.

Strengths:

The major strengths of the work are described above.

Weaknesses:

Rather than considering the following points as weaknesses, I instead prefer to think of them as areas for future study:

(1) Given the field's intense interest in the biology and therapeutic vulnerabilities of residual disease cells, I suspect that one major practical implication of this work could be that it inspires scientists interested in working in the residual disease space to model it in monolayer culture. However, this relies upon the assumption that drug-tolerant cells isolated in monolayer culture are at least reasonably similar in nature to drug-tolerant cells isolated from spheroid or xenograft systems. Is this true? An intriguing experiment that could help answer this question would be to perform gene expression profiling on a cell line model in the following conditions: monolayer growth, drug tolerant cells isolated from monolayer growth conditions, spheroid growth, drug tolerant cells isolated from spheroid growth conditions, xenograft tumors, and drug tolerant cells isolated from xenograft tumors. What are the genes and programs shared between drug-tolerant cells cultured in the three conditions above? Which genes and programs differ between these conditions? Data from this exercise could help provide additional, useful context with which to understand the benefits and pitfalls of modeling residual tumor cell growth in monolayer culture.

We thank the reviewer for suggesting valuable future studies. We agree that the proposed experiments represent important next steps in understanding the role of YAP and other pathways in primary resistance. We believe, however, these experiments are both beyond the scope of the current manuscript and beyond what can reasonably be addressed in a revision. The distinct challenges associated with comparing in vivo and in vitro conditions would require significant optimization of single-cell approaches, especially given the robust cell death driven by afatinib treatment in vivo. Given the complexity of in vivo experimentation, we are concerned that such studies may not guarantee biologically meaningful insights. Nonetheless, we agree that this is a compelling direction for future research. If common gene expression patterns could be identified despite these challenges, such studies could help validate monolayer culture as a relevant model for investigating residual disease.

(2) In relation to the point above, there is an interesting and established connection between mesenchymal gene expression and YAP/TAZ signaling. For example, analyses of gene expression data from human tumors and cell lines demonstrate an extremely strong correlation between these two gene expression programs. Further, residual persister cancer cells have often been characterized as having undergone an EMT-like transition. From the analysis above, is there evidence that residual tumor cells with increased YAP signaling also exhibit increased mesenchymal gene expression?

We agree with the reviewer that a connection between YAP/TAZ activity and EMT is likely, given prior studies exploring correlations between these two gene signatures. We believe, however, exploring EMT represents a distinct research direction from the primary focus of the current manuscript. We are concerned exploration of EMT, especially in the absence of corresponding preclinical models or mechanistic data directly linking EMT to therapy resistance in our models, could distract from the main conclusions of the manuscript. While we plan to stain for EMT-associated markers in the residual cancer tissue from the in vivo studies, it remains unclear whether such data would meaningfully contribute to the revised manuscript, regardless of the outcome.

Reviewer #2 (Public review):

The manuscript by Nakagawa R, et al describes a mechanism of how NSCLC cells become resistant to EGFR and KRAS G12C inhibition. Here, the authors focus on the initial cellular changes that occur to confer resistance and identify YAP activation as a non-genetic mechanism of acute resistance.

The authors performed an initial xenograft study to identify YAP nuclear localization as a potential mechanism of resistance to EGFRi. The increase in the stromal component of the tumors upon Afatinib treatment leads the authors to explore the response to these inhibitors in both 2D and 3D culture. The authors extend their findings to both KRAS G12C and BRAF inhibitors, suggesting that the mechanism of resistance may be shared along this pathway.

The paper would benefit from additional cell lines to determine the generalizability of the findings they presented. While the change in the localization of YAP upon Afatinib treatment was identified in a xenograft model, the authors do not return to animal models to test their potential mechanism, and the effects of the hyperactivated S127A YAP protein on Afatinib sensitivity in culture are modest. Also, combination studies of YAP inhibitors and EGFR/RAS/RAF inhibitors would have strengthened the studies.

We thank the reviewer for their insightful comments. In this manuscript, we present data from 5 cell lines representing the EGFR/BRAF/KRAS pathway, demonstrating the generalizability of YAP-driven decreased cancer cell sensitivity to targeted inhibitors when cultured in 2D compared to spheroid counterparts. While expanding this analysis to a larger panel of cell lines is beyond the scope of the current study, we believe our findings provide a strong rationale for future investigations, including high-throughput screens conducted by other research groups and pharmaceutical companies, to recognize the value in screening spheroid cell cultures. We hope this work helps shift the field of cancer therapeutics toward screening approaches that better reflect tumor biology into drug discovery pipelines and believe this could be one of the most impactful and enduring contributions of our study.

Reviewer #2 also mentions that "...combination studies of YAP inhibitors and EGFR/RAS/RAF inhibitors would have strengthened the studies..." The concept that YAP/TAZ inhibitors (i.e. TEAD inhibitors) could be additive or synergistic in 2D culture is one that is being actively tested across several groups and in pharma. Several recent examples include a publication by Hagenbeek, et al., Nat. Cancer, 2023 (PMID: 37277530) showing that a TEAD inhibitor overcomes KRASG12C inhibitor resistance. Additional, recent work by Pfeifer, et al., Comm. Biol., 2024 (PMID: 38658677) suggests a similar effect between EGFR inhibitors and a different TEAD inhibitor. While neither of these studies extensively probes cell death pathways in the way performed in our studies, they nevertheless provide strong evidence that indeed TEAD + targeted EGFR/RAF/RAS inhibition in 2D have additive, if not synergistic, effects. We feel that these recent published studies affirm our findings and repeating such experiments is unlikely to add much new information. We thus feel they are beyond the scope of our present studies.