Research Center

The CSBC Research Center for Cancer Systems Immunology at MSKCC


Center Title

The CSBC Research Center for Cancer Systems Immunology at MSKCC

Participating Institutions

Memorial Sloan-Kettering Cancer Center (MSKCC)
Dana Farber Cancer Center

Center Website

Center Summary

Exciting clinical breakthroughs with checkpoint blockade antibodies and adoptive T cell transfers demonstrate the power of harnessing the immune system to eliminate cancer. However, fundamental challenges remain. Only a subset of patients and cancer types—especially hematological malignancies and melanoma—show significant clinical responses. What properties of tumors determine clinical responses? How can we produce clinical responses in cancers currently unbeatable?

The CSBC Research Center for Cancer Systems Immunology at MSKCC addresses these challenges towards predictable and effective cancer immunotherapy. We assembled a multidisciplinary team of computational biologists, immunologists and cancer scientists; together we will deepen our fundamental understanding of cancer-immune system interactions at the molecular, cellular, and systems levels.
We investigate cancer-immune system interactions at three stages of disease progression: cancer initiation and early tumorigenesis (Project I); established and progressing tumors (Project II); and latent disease and metastasis (Project III). All projects will use cutting-edge single-cell droplet sequencing technologies and computational analyses (Shared Resource Core).


Principal Investigators

Dr. Christina Leslie, Ph.D.

Dr. Christina Leslie, Ph.D. is an Associate Member in the Computational Biology Program at MSKCC. She is an expert in using machine learning to study mechanisms of gene regulation and the epigenetics of cell fates in differentiation, and is widely known for introducing k-mer string kernels for support vector machines in diverse computational biology sequence classification problems. Dr. Leslie is the computational PI for this Research Center and will be the computational Co-Lead for Project I.

Dr. Alexander Rudensky, Ph.D.

Dr. Alexander Rudensky, Ph.D. is Chair of the Immunology Program and Director of the Ludwig Center for Cancer Immunotherapy at MSKCC and an HMMI Investigator. He studies the molecular mechanisms of function and differentiation of CD4 T cells and their role in immunity and tolerance, and he is one of the world’s foremost experts on regulatory T cells. Dr. Rudensky is a Member of the National Academy of Sciences, Member of the American Academy of Arts and Sciences, and Member of National Academy of Medicine, and a recipient of the Coley Award in Basic and Tumor Immunology and an American Association of Immunologists BD-Investigator Award among others. He is the experimental PI leading this Research Center as well as the experimental Lead on Project II.

Computational Investigators

Dr. Joao Xavier, Ph.D.,

Associate Member in the Computational Biology Program at MSKCC

Dr. Dana Pe’er, Ph.D.,

Program Chair and Member in the Computational Biology Program at MSKCC

Dr. Chris Sander, Ph.D.,

Director of the cBio Center at Dana Farber Cancer Institute

Experimental Investigators

Dr. Joan Massagué, Ph.D.,

Director of the Sloan Kettering Institute and Member of the Cancer Biology Genetics Program at MSKCC

Dr. Andrea Schietinger, Ph.D.,

Assistant Member in the Immunology Program at MSKCC.

Clinical collaborator

Dr. Jedd Wolchok, M.D., Ph.D.,

Chief of the Melanoma and Immunotherapeutics Service and Lloyd J. Old Chair for Clinical Investigation in the Department of Medicine at MSKCC


Project 1: Tumor-specific T Cell State Dynamics and Heterogeneity in Early Tumorigenesis

Project 1 aims to define the molecular and epigenetic characteristics and mechanisms that cause tumor-specific T cell to differente to a dysfunctional state during early tumorigenesis. In Aim 1 we will define the chromatin states and transcription factor networks that mediate the transition from plastic to fixed T cell dysfunction states. In Aim 2, we will elucidate the mutational landscape, stromal and immune population dynamics over the course of tumor development and assess how tumor-specific, dysfunctional T cell populations and cell states co-evolve with these changes; exploit single-cell technologies to dissect diversity of T cell states, and develop a novel single cell technology to capture TCR sequences together with gene expression profiles to connect tumor antigen specificity with cell states; ultimately, develop computational and mathematical models of T cell differentiation states to predict responsiveness to therapeutic interventions (e.g. checkpoint blockade). In Aim 3, we will define cell states and heterogeneity of tumor-specific T cells from human solid tumors, and predict and validate changes in their cell states in response to therapeutic interventions (e.g. checkpoint blockade) at a population and single-cell level. Identifying cell states and epigenetic programs of tumor-specific T cells that mediate plasticity or imprinting of cellular hyporesponsiveness will not only provide new insights into the genomic control circuitry of T cell differentiation and dysfunction but may point to novel strategies for cellular reprogramming of cancer-specific T cells for cancer therapy.

Project 2: The tumor ecosystem in cancer progression and immunotherapeutic response

The goal of this project is to identify the ecological interactions between cancer and immune cells that govern cancer dynamics and response to therapy. A tumor can be considered an ecosystem or organ, where multiple accessory cell types are interconnected and communicate with each other and with tumor cells, which serve as their clients. We seek to identify key cellular and molecular regulatory elements in the tumor microenvironment and potential means of their manipulation for therapeutic benefit through systems analysis and modeling of functional interactions in the tumor ecosystem on different scales including cellular, protein, metabolite, and gene expression dynamics at a population and single cell levels. In Aim 1, we will explore the multiple accessory cell types and their interactions with tumor cells using experimental models of skin and lung cancer in mice and in human cancer patients. The accessory cells include myeloid cells, dendritic cells, innate lymphoid cells (ILC), neutrophils, eosinophils, endothelial cells, fibroblasts, and regulatory T (Treg) cells. We will also investigate features of mediators of anti-tumor immunity including NK cells, CD4 and CD8 T cells and their specialized subsets. In Aim 2, we will use perturbation of the tumor ecology impacting its progression in mice and human patients by established and novel immunotherapeutic modalities including PD1 and CTLA4 blockade and Treg cell depletion. The impact of these perturbations will be assessed through comprehensive analysis of cellular dynamics and states in relation to biological and clinical outcomes to generate predictive models. In Aim 3, we will then validate key interaction components in the tumor ecosystem by modeling cell-cell interactions in vitro using tissue mimetic systems and in silico using agent-based models.

Project 3: Latent Metastasis: Immune Regulation of Disseminated Cancer Stem Cells

The goal of Project 3 is to discover mechanisms that critically regulate immune evasion by disseminated tumor cells (DTCs) and their evolution as latent metastatic entities. We will use an integrated approach that combines single-cell interrogation methods with unique biological models of latent metastasis from breast cancer and lung adenocarcinoma, and novel computational strategies. Distant metastasis underlies the overwhelming majority of cancer-related deaths and its inception is exceedingly variable. Residual DTCs may outgrow immediately or, more frequently, linger in a viable state of replicative quiescence or mass dormancy for months to years after infiltrating distant organs. This latency state of DTCs is accompanied with significant resistance to anti-neoplastic therapy, which typically targets actively dividing tumor cells. Moreover, latent DTCs somehow evade immune surveillance. The biology underlying these adaptive abilities remains poorly understood and factors governing DTC population dynamics, whether stochastic or deterministic, remain unknown. We aim to address this significant knowledge gap by combining massively parallel, single-cell RNA expression profiling using a bead-based molecular barcoding technology with unsupervised learning methods to identify stable/transitory cell states within latent, residual disease and their molecular control mechanisms. As a complementary approach, individual cell responses to molecular perturbations will be dynamically tracked by live cell imaging. In Aim 1 we propose to determine whether the latent state pre-exists in the primary tumor or is induced by the stress of immunosurveillance in a host tissue. In Aim 2 we will model the evolutionary dynamics of metastatic cells as they exit latency under growth permissive and immunoediting conditions. In Aim 3 we will identify key regulators of metastatic immune evasion by probing transcriptional heterogeneity in quiescent subpopulations differentially sensitive to NK-cell mediated elimination. The amalgamation of these approaches, combined with our deep understanding of the biology of cancer metastasis, will promote the discovery of therapeutic strategies to eradicate or control metastasis from its earliest stages of inception.


Shared Resource Core

Single-cell sequencing technology and computational analysis
The Research Center’s Shared Resource Core will provide a broadly applicable set of experimental technologies and computational tools for high-throughput single-cell RNA-seq (scRNA-seq) of heterogeneous populations. Single-cell resolution is crucial for elucidating the functional cell subpopulations in both the tumor and its immune microenvironment and can be harnessed to elucidate regulatory relationships within and between these cell subpopulations. To achieve our goals requires a technology that can collect many thousands of cells per tumor and computational methods to analyze and interpret the complex data collected. We will use an improved version of in-drop scRNA-seq, which provides an unprecedented increase in throughput for automated capture and library preparation for single cells in a cost effective manner. The technology will be used to interrogate cell populations for Projects I, II, and III. The Shared Resource will provide the computation, visualization and analysis methods needed to interpret the collected single cell data and integrate it with other data types.