Lung cancer is the leading cause of cancer related deaths. In its most lethal form, small-cell lung cancer (SCLC), heterogeneity correlates with aggressiveness, however no driver mutations distinguishing SCLC subtypes have been identified. Another singularity of SCLC is that it responds well to initial treatment but quickly relapses into resistance, suggesting phenotypic plasticity. In this basic project, we will investigate the role of transcriptional and signaling mechanisms in promoting SCLC phenotypic heterogeneity and plastic state transitions, leading to aggressiveness and rapid relapse. Our preliminary results indicate that SCLC heterogeneity is more extensive than the canonical neuroendocrine (NE) and mesenchymal-like (ML) subtypes, and includes multiple hybrid states. Most significantly, we found that drug treatment results in phenotypic transitions toward the hybrid states, implicating them in resistance. Based on these data, our central hypothesis is that SCLC is a heterogeneous mix of NE, ML and hybrid phenotypic states and that, due to phenotypic plasticity, transitions between these states is a key mechanism of treatment evasion in SCLC. To test this hypothesis, we will combine computation and experiments to characterize the global landscape of phenotypes in SCLC, and define the impact of phenotypic transitions on resistance.
In Aim1, we will identify a regulatory transcription factor (TF) network that controls the differentiation of SCLC cells into NE, ML, and hybrid phenotypic states; validate model predicted phenotypes and quantify their drug sensitivity; and, define reprogramming pathways to drug- sensitive states. Our approach pipeline is comprised of phenotypic clustering and gene co-expression network analysis on SCLC tumor and cell line data, simulations of logic-based TF network models to prioritize TF targets for reprogramming, and experimental validation of model predictions in vitro and in vivo.
In Aim2, we will quantify phenotype sensitivity to chemotherapy and plasticity in response to signaling perturbations; identify perturbations that promote phenotype switching; and, test optimal drug/perturbagen combinations that maximize SCLC cell killing under treatment. Phenotypes and signaling pathways will be defined by flow and mass cytometry. SCLC clonal dynamics in response to perturbations will be quantified using a stochastic phenotype transition to prioritize drug/perturbagen combinations for experimental validation. Drug sensitivity and plasticity of SCLC phenotypes will be assessed with the drug-induced proliferation rate metric, which we recently described, and time series single-cell flow or mass cytometry. Success of this project will have translational impact by empowering searches for targeted therapies that reprogram drug-resistant cells toward drug-sensitive cells, which we anticipate will lead to significantly improved patient outcomes in SCLC. We further anticipate that this approach will be useful in other cancer types, opening the doors to a new paradigm of cancer treatment based on epigenetic tumor reprogramming.
Small-cell lung cancer (SCLC) is the most lethal form of lung cancer because of aggressiveness linked to heterogeneity and quick relapse after initial treatment response. We hypothesize that SCLC heterogeneous phenotypes are transcriptionally initiated, and that phenotypic plasticity contributes to relapse. We will model sources of heterogeneity and plasticity based on transcription factor driven gene regulatory networks, and on signaling pathways, to make and test predictions for actionable strategies to reduce SCLC aggressiveness and relapse.
| Jones, Zack W; Leander, Rachel; Quaranta, Vito et al. (2018) A drift-diffusion checkpoint model predicts a highly variable and growth-factor-sensitive portion of the cell cycle G1 phase. PLoS One 13:e0192087 |
| Paudel, B Bishal; Harris, Leonard A; Hardeman, Keisha N et al. (2018) A Nonquiescent ""Idling"" Population State in Drug-Treated, BRAF-Mutated Melanoma. Biophys J 114:1499-1511 |
| Harris, Leonard A; Nobile, Marco S; Pino, James C et al. (2017) GPU-powered model analysis with PySB/cupSODA. Bioinformatics 33:3492-3494 |
| Mitchener, Michelle M; Hermanson, Daniel J; Shockley, Erin M et al. (2015) Competition and allostery govern substrate selectivity of cyclooxygenase-2. Proc Natl Acad Sci U S A 112:12366-71 |
