Developing inhibitor combinations through identification of drug tolerant cell populations

PROJECT SUMMARY/ABSTRACT Drug tolerant cells limit clinical success of many types of anticancer therapies. Drug tolerance has largely been associated with cell heterogeneity. Surprisingly, identifying successful drug combinations has rarely been done at the cell population level. We propose that the limited success of therapeutic strategies is due to the

gap in knowledge of the vulnerabilities of different cancer cell populations. We and others recently showed that during the initial response to cancer therapy, drug tolerant cells exhibit differences in gene expression across distinct cell subpopulations. The analysis of transcriptional phenotypes in lung cancer cell lines and xenografts

treated with a single agent has guided us to a successful targeting of the drug tolerant populations with the second agent. This suggested to us that describing different cell populations using single-cell RNA data may be used for identifying scorable drug response signatures. We put forward an innovative idea of assessing

druggable modulation in gene expression at the cell population level for identification of potential drug combinations. The goal is to find the drugs, each targeting distinct cell population/s tolerant to the first drug, to eliminate all cells with the prescribed combination treatment. This idea will be tested in three Specific Aims, by

drug scoring based on cell population signatures, drug validation in cell survival assays, and drug validation through the characterization of affected cell populations. The data from single-cell RNA sequencing will be used for in silico drug prescription, and the top drug combinations will be validated in vivo for inhibiting growth of

predicted cell subpopulations. To enhance translational value of this idea, we study the cell subpopulations that are tolerant to the treatment with standard-of-care cancer therapies; we use relevant cancer cell line models; and we determine applicability of our idea using large pan-cancer cohorts with available RNA-seq data. No drugs

that target cancer cell heterogeneity have yet reached the clinic. However, the inhibitors available in the databases that we use, with known safety profiles, have been approved or are in clinical trials for cancer treatment, and they could be candidates for halting drug tolerance. The proposed research helps find new

solutions and approaches for cancer treatment. It may enable oncologists to accurately tailor cancer care for individual patients with the drugs that target emerging tolerant cell populations. Patients stratified by markers predictive of drug response can enter clinical trials for combination therapy.