Periodic Reporting for period 1 - BrcaSCope (Organ and mutation dependencies shaping the tumor microenvironment)
Reporting period: 2022-10-01 to 2025-03-31
Our overall objective is to define the factors that reshape normal tissues into pro-tumorigenic microenvironments. We identify cancer mutations and organ identity as main factors in this process and ask whether specific cancer-derived signals reshape different organs in similar manners.
To approach this, we chose to focus on major cancer mutations – in BRCA1/2 – and on the four organs where BRCA mutations drive cancer: breast, ovary, pancreas and prostate. We set to dissect microenvironments of BRCA-driven cancers in these organs using a wide arsenal of imaging, sequencing, and molecular tools. Our specific interest was to characterize cancer-associated fibroblasts (CAFs) in this context. CAFs are the most abundant cell type in the microenvironment of many carcinomas, their increased abundance correlates with worse prognosis and yet they are an understudied population and unexploited therapeutic target.
Building on our recent discovery of distinct cancer-associated fibroblast (CAF) compositions in BRCA-mutated breast cancer, we defined three aims:
Our first aim was to map the transcriptional and proteomic landscape of CAFs in BRCA-mutated and BRCA-WT tumors in the four different organs by bulk and scRNA-seq as well as multiplexed immunofluorescent imaging. Applying image and data analysis tools to integrate maps from different organs, this aim would generate a pan-cancer BRCA-associated map, which will dissect the effects of the mutation and the organ on the fibroblastic microenvironment composition.
The second aim was to model cell-cell interactions in the four organs using a reductionist ex-vivo coculture approach, functional assays, and mathematical modeling, to define interaction networks involving fibroblasts in physiological and pathological conditions.
The third aim was to integrate our CAF maps (from aim 1) with the tissue dynamics data (from aim 2) to define actionable nodes for intervention. We planned to design combinations of BRCA-targeted therapies with chemical or genetic inhibitors of specific CAF subsets and test their efficacy across BRCA-mutated cancers in cell coculture and mouse models, to exploit both mutation and organ-specific vulnerabilities.
Taken together, this plan is aimed at creating a dynamic map that will explain how the microenvironment is rewired, how variable this rewiring is across cancer mutations and organs, and which are the most vulnerable nodes in it.
To approach this, we chose to focus on major cancer mutations – in BRCA1/2 – and on the four organs where BRCA mutations drive cancer: breast, ovary, pancreas and prostate. We set to dissect microenvironments of BRCA-driven cancers in these organs using a wide arsenal of imaging, sequencing, and molecular tools. Our specific interest was to characterize cancer-associated fibroblasts (CAFs) in this context. CAFs are the most abundant cell type in the microenvironment of many carcinomas, their increased abundance correlates with worse prognosis and yet they are an understudied population and unexploited therapeutic target.
Building on our recent discovery of distinct cancer-associated fibroblast (CAF) compositions in BRCA-mutated breast cancer, we defined three aims:
Our first aim was to map the transcriptional and proteomic landscape of CAFs in BRCA-mutated and BRCA-WT tumors in the four different organs by bulk and scRNA-seq as well as multiplexed immunofluorescent imaging. Applying image and data analysis tools to integrate maps from different organs, this aim would generate a pan-cancer BRCA-associated map, which will dissect the effects of the mutation and the organ on the fibroblastic microenvironment composition.
The second aim was to model cell-cell interactions in the four organs using a reductionist ex-vivo coculture approach, functional assays, and mathematical modeling, to define interaction networks involving fibroblasts in physiological and pathological conditions.
The third aim was to integrate our CAF maps (from aim 1) with the tissue dynamics data (from aim 2) to define actionable nodes for intervention. We planned to design combinations of BRCA-targeted therapies with chemical or genetic inhibitors of specific CAF subsets and test their efficacy across BRCA-mutated cancers in cell coculture and mouse models, to exploit both mutation and organ-specific vulnerabilities.
Taken together, this plan is aimed at creating a dynamic map that will explain how the microenvironment is rewired, how variable this rewiring is across cancer mutations and organs, and which are the most vulnerable nodes in it.
To map the landscape of CAFs in BRCA-mutated and BRCA-WT tumors in the four different organs (Aim 1), we launched the assembly, together with our clinical collaborators at MSKCC (NY), of four cohorts of BRCA-mut and BRCA-WT patients. For each organ we aimed at identifying 50-80 cases for which resected tissue and patient data (clinical and demographic) exist, and samples could be used to prepare a tumor microarray (TMA). We have successfully assembled the PDAC TMA, we are at final stages of assembling the prostate TMA and are in the process of preparing the ovary and breast TMAs. We already analyzed an initial cohort of ovarian cancer samples (10 BRCA-mut and BRCA-WT patients), and have separately analyzed a cohort of 42 BRCA-mut and BRCA-WT PDAC patients. We have now assembled a MIBI panel imaging and analysis of 40 CAF, immune, and cancer markers. This panel will be used on all TMAs to define common and different markers across the different organs.
To study organ dependecies, we examined the stress response landscape in BRCA WT tumors in these four organs. Using a combination of experimental and computational methods, we found that stress responses vary within the TME and are especially active near cancer cells. Focusing on the non-immune stroma we found, across tumor types, that NRF2 and the oxidative stress response are distinctly activated in immune-regulatory CAFs and in a unique subset of cancer-associated pericytes (Lior et al, Cell Reports 2024).
We developed an experimental-mathematical approach to decompose the tumor microenvironment into circuits (Aim 2), and found a hierarchical network of interactions in breast cancer, with CAFs at the top secreting factors primarily to tumor-associated macrophages. We showed that this network is composed of repeating circuit motifs, studied the fibroblast-macrophage circuit in-vitro, and identifyed a lignad-receptor pair (RARRES2-CMKLR1) mediating this interaction. This study demonstrated that the complexity of the tumor microenvironment may be simplified by identifying small circuits, facilitating the development of strategies to modulate it (Mayer* and Milo* et al, Nature Communications 2023).
Based on our findings in PDAC we are combining small molecules targeting HSF1 (in CAFs) with PARP-inhibitors (targeting BRCA-mut cancer cells) in an injectable mouse model of PDAC to test the effect of this BRCA-specific composition targeting both the mutated cancer cells and the matching CAFs. We also engineered HSF1-Flox Col1A1 -CRE muce in which HSF1 is knocked-out in fibroblasts, to test the effect of HSF1 loss in the stroma in a complementary manner. This work is on going.
To study organ dependecies, we examined the stress response landscape in BRCA WT tumors in these four organs. Using a combination of experimental and computational methods, we found that stress responses vary within the TME and are especially active near cancer cells. Focusing on the non-immune stroma we found, across tumor types, that NRF2 and the oxidative stress response are distinctly activated in immune-regulatory CAFs and in a unique subset of cancer-associated pericytes (Lior et al, Cell Reports 2024).
We developed an experimental-mathematical approach to decompose the tumor microenvironment into circuits (Aim 2), and found a hierarchical network of interactions in breast cancer, with CAFs at the top secreting factors primarily to tumor-associated macrophages. We showed that this network is composed of repeating circuit motifs, studied the fibroblast-macrophage circuit in-vitro, and identifyed a lignad-receptor pair (RARRES2-CMKLR1) mediating this interaction. This study demonstrated that the complexity of the tumor microenvironment may be simplified by identifying small circuits, facilitating the development of strategies to modulate it (Mayer* and Milo* et al, Nature Communications 2023).
Based on our findings in PDAC we are combining small molecules targeting HSF1 (in CAFs) with PARP-inhibitors (targeting BRCA-mut cancer cells) in an injectable mouse model of PDAC to test the effect of this BRCA-specific composition targeting both the mutated cancer cells and the matching CAFs. We also engineered HSF1-Flox Col1A1 -CRE muce in which HSF1 is knocked-out in fibroblasts, to test the effect of HSF1 loss in the stroma in a complementary manner. This work is on going.
We comprehensively analyzed pancreatic and ovarian cancer samples from patients, revealing different CAF compositions in germline BRCA-mut vs. BRCA WT tumors. Using cancer organoids and mouse models we showed that these CAF subset transitions are actually driven by loss of functional BRCA in the cancer cells. Our findings unraveled, for the first time, a dimension of stromal heterogeneity influenced by germline mutations in cancer cells, with direct implications for clinical research, which we are currently pursuing in parallel in the four BRCA-associated organs.
To unravel organ dependencies, we examined the network of stress responses in the tumor and its microenvironment in the four organs using BRCA-WT tumors. We created an atlas of stress response activation patterns. Through our analysis, we discerned distinct subpopulations of fibroblasts and pericytes that exhibit a clear association with cellular stress, in particular oxidative stress. This comprehensive map and the molecular pathways discovered pave the way to elucidate the contribution of stress responses in the TME in a cell-specific manner. Moreover, it offers insights into how the stress response landscape might influence tumor progression and disease outcome.
Our mathematical modeling of cell-cell interactions in the TME took network-motif analysis to a new level, asking for the first time whether network-motif structure and circuit autonomy might occur between cells in the TME. We revealed a hierarchical network structure, with CAFs at the top of the hierarchy. We explored this network and identified a recurring two-cell circuit motif, the strongest instance of which was a CAF-TAM circuit. We studied the dynamics and functions of this circuit in-vitro. The interaction strengths and gene expression profiles of the in-vitro circuit recapitulated those of the in-vivo circuit, and enabled testing the effect of ligand-receptor interactions on cell dynamics and function, as we showed by identifying RARRES2, a potential mediator of CAF-TAM interactions, and its receptor CMKLR1. Thus, this work showed that the complexity of the TME may be amenable to reductionist analysis by identifying and isolating small cell circuits.
To unravel organ dependencies, we examined the network of stress responses in the tumor and its microenvironment in the four organs using BRCA-WT tumors. We created an atlas of stress response activation patterns. Through our analysis, we discerned distinct subpopulations of fibroblasts and pericytes that exhibit a clear association with cellular stress, in particular oxidative stress. This comprehensive map and the molecular pathways discovered pave the way to elucidate the contribution of stress responses in the TME in a cell-specific manner. Moreover, it offers insights into how the stress response landscape might influence tumor progression and disease outcome.
Our mathematical modeling of cell-cell interactions in the TME took network-motif analysis to a new level, asking for the first time whether network-motif structure and circuit autonomy might occur between cells in the TME. We revealed a hierarchical network structure, with CAFs at the top of the hierarchy. We explored this network and identified a recurring two-cell circuit motif, the strongest instance of which was a CAF-TAM circuit. We studied the dynamics and functions of this circuit in-vitro. The interaction strengths and gene expression profiles of the in-vitro circuit recapitulated those of the in-vivo circuit, and enabled testing the effect of ligand-receptor interactions on cell dynamics and function, as we showed by identifying RARRES2, a potential mediator of CAF-TAM interactions, and its receptor CMKLR1. Thus, this work showed that the complexity of the TME may be amenable to reductionist analysis by identifying and isolating small cell circuits.