Periodic Reporting for period 1 - Tx-phylogeography (High throughput phylogeography of tumors: how the tissue environment influences cancer evolution?)
Période du rapport: 2023-04-01 au 2025-09-30
The tumor microenvironment (TME) is a dynamic ecosystem wherein tumor cells, immune cells and stromal cells interact and evolve. One of the primary characteristics of cancer progression is the diversification, selection and expansion of tumor sub-clones in a manner that largely depends on the TME, e.g. through immune responses or metabolic crosstalk. Advances in single-cell and spatial genomics provide exciting opportunities for studying how tumors evolve in their tissue context, by lending an information-rich perspective on which cell subsets are present and how they might interact. However, fundamental questions, which require the knowledge of how single tumor cells are related sub-clonally, remain challenging to study through that perspective. Fundamentally - what characterizes the cellular environment of successful subclones?
The goal of this ERC-funded project is to develop transcriptional phylogeography - a paradigm for studying tumor development in-situ, on a transcriptome-wide scale, and at single cell resolution. Our work builds on previous studies by us and others that utilized CRISPR/Cas9 mutations coupled with single cell RNA sequencing to infer the lineage structure of thousands of single cells at a time. Adapting this system for in-situ profiling, we are generating a first of its kind resource of phylogeography of tumor models. This data allows us to identify tissue locations that harbor critical sub-clonal properties such as unrestrained growth, dedifferentiation, and metastatic seeding, and thereby investigate - through the lens of high throughput genomics- how the TME is associated with these properties. We also identify and offer solutions for outstanding analytical questions in this nascent area, from lineage inference given sparse data to characterizing the metabolic aspects of tumor-TME cross talk. Together, these studies will help establish new causal links between the TME and tumor evolution and lay the foundation for transcriptional phylogeography analysis.
The goal of this ERC-funded project is to develop transcriptional phylogeography - a paradigm for studying tumor development in-situ, on a transcriptome-wide scale, and at single cell resolution. Our work builds on previous studies by us and others that utilized CRISPR/Cas9 mutations coupled with single cell RNA sequencing to infer the lineage structure of thousands of single cells at a time. Adapting this system for in-situ profiling, we are generating a first of its kind resource of phylogeography of tumor models. This data allows us to identify tissue locations that harbor critical sub-clonal properties such as unrestrained growth, dedifferentiation, and metastatic seeding, and thereby investigate - through the lens of high throughput genomics- how the TME is associated with these properties. We also identify and offer solutions for outstanding analytical questions in this nascent area, from lineage inference given sparse data to characterizing the metabolic aspects of tumor-TME cross talk. Together, these studies will help establish new causal links between the TME and tumor evolution and lay the foundation for transcriptional phylogeography analysis.
We have made signifianct progress in both major areas of the our proposed program.
In one set of projects, we are advancing the set of analytical tools that are required for processing and making sense of our data, namely: phylogenies (family trees) of tumor cells, in which each “leaf” (cell) comes with its gene expression profile (transcriptome) and possibly also its location in the tissue (conduting our measurments of gene expression in an intact tissue and at a high spatial resolution). The work (three papers thus far; two published and one under review) includes new algorithms for (i) inferring lineage trees [Prillo et al REOMB 2025]; (ii) deciding whether an inferred lineage tree is accurate enough (without knowing the “correct” tree) [Zilber et al., bioRxiv], and (iii) estimating the divergence times between cells in a lineage tree [Prillo et al Systematic Biology 2025]. This work relies on the vast body of work in the field of phylogenetics (e.g. for studying divergence between species). However, it is geared toward the special case of cell lineages trees and accounts for the nuances of the respective experimental protocols. As such, the work in the three papers relies on solid mathematical foundation and its evaluation includes both theoretical and empirical results. In addition, we developed an algorithm that leverages lineage trees to approximate the transcriptional changes that the observed tumor cells will go through after a set time interval. This procedure helps shed new light on the ways by which tumors develop over time and enables predictions about their future state [this project was the subject of an MS thesis, and the manuscript is in preparation; early version of this idea appears in our work at Yang et al., Cell 2022].
In the second set of projects, we focus on empirical studies of tumor evolution, using mouse models and cell lines that are endowed with the lineage tracing system. Our first product, in collaboration with Jonatan Weissman (MIT) and Dian Yang (Columbia), is the first proof of concept study of high-resolution transcriptional phylogeography in tumors [Jones et al., bioRxiv]. By coupling spatial transcriptomics with Cas9-based lineage tracing in a mouse model of lung adenocarcinoma, we were able to study how the tissue environment of tumor cells is associated with critical sub-clonal properties, namely: expansion, plasticity, and metastases seeding. We found that rapid tumor expansion contributes to a hypoxic, immunosuppressive, and fibrotic microenvironment that is associated with certain types of macrophasges and fibroblasts. We also found that metastases arise from spatially-confined subclones of primary tumors and remodel the distant metastatic niche into a fibrotic, collagen-rich microenvironment. Together, this first study presented a comprehensive dataset integrating spatial assays and lineage tracing to elucidate how sequential changes in cancer cell state and microenvironmental structures cooperate to promote tumor progression.
In one set of projects, we are advancing the set of analytical tools that are required for processing and making sense of our data, namely: phylogenies (family trees) of tumor cells, in which each “leaf” (cell) comes with its gene expression profile (transcriptome) and possibly also its location in the tissue (conduting our measurments of gene expression in an intact tissue and at a high spatial resolution). The work (three papers thus far; two published and one under review) includes new algorithms for (i) inferring lineage trees [Prillo et al REOMB 2025]; (ii) deciding whether an inferred lineage tree is accurate enough (without knowing the “correct” tree) [Zilber et al., bioRxiv], and (iii) estimating the divergence times between cells in a lineage tree [Prillo et al Systematic Biology 2025]. This work relies on the vast body of work in the field of phylogenetics (e.g. for studying divergence between species). However, it is geared toward the special case of cell lineages trees and accounts for the nuances of the respective experimental protocols. As such, the work in the three papers relies on solid mathematical foundation and its evaluation includes both theoretical and empirical results. In addition, we developed an algorithm that leverages lineage trees to approximate the transcriptional changes that the observed tumor cells will go through after a set time interval. This procedure helps shed new light on the ways by which tumors develop over time and enables predictions about their future state [this project was the subject of an MS thesis, and the manuscript is in preparation; early version of this idea appears in our work at Yang et al., Cell 2022].
In the second set of projects, we focus on empirical studies of tumor evolution, using mouse models and cell lines that are endowed with the lineage tracing system. Our first product, in collaboration with Jonatan Weissman (MIT) and Dian Yang (Columbia), is the first proof of concept study of high-resolution transcriptional phylogeography in tumors [Jones et al., bioRxiv]. By coupling spatial transcriptomics with Cas9-based lineage tracing in a mouse model of lung adenocarcinoma, we were able to study how the tissue environment of tumor cells is associated with critical sub-clonal properties, namely: expansion, plasticity, and metastases seeding. We found that rapid tumor expansion contributes to a hypoxic, immunosuppressive, and fibrotic microenvironment that is associated with certain types of macrophasges and fibroblasts. We also found that metastases arise from spatially-confined subclones of primary tumors and remodel the distant metastatic niche into a fibrotic, collagen-rich microenvironment. Together, this first study presented a comprehensive dataset integrating spatial assays and lineage tracing to elucidate how sequential changes in cancer cell state and microenvironmental structures cooperate to promote tumor progression.
Our computational work tackles outstanding questions that are key for studying tumor evolution through the lens of single cell genomics and lineage tracing. In all our contributions, we used simulations and real data to compare to the current approaches and demonstrated state of the art performance. For instance, for the proboem of distinguishing accurate tree- reconstructions (given lineage traces) without knowing the ground truth tree, we compare our new algorithm to the prevalent approaches (based on parsimpny or likelihood) and show that we can achieve a much higher level of accuracy (Zilber et al., bioRxiv).
Our collaborative work on lineage tracing in a mouse model of lung cancer provides the first proof of concept for large-scale transcriptional phylogeography of tumors. Namely, it is the first demonstration of tumor-wide analysis of lineage trees in an intact cancer tissue at high spatial resolution that is coupled with transcriptome-wide quantification of gene expression. This study opens the way for in-depth interrogation of how interactions within the TME inluence the evolution and growth of tumors, and in turn, how the evolution of tumor sub-clones helps shape their tissue environment.
Our collaborative work on lineage tracing in a mouse model of lung cancer provides the first proof of concept for large-scale transcriptional phylogeography of tumors. Namely, it is the first demonstration of tumor-wide analysis of lineage trees in an intact cancer tissue at high spatial resolution that is coupled with transcriptome-wide quantification of gene expression. This study opens the way for in-depth interrogation of how interactions within the TME inluence the evolution and growth of tumors, and in turn, how the evolution of tumor sub-clones helps shape their tissue environment.