Spatially coordinated RTK-ERK signaling dynamics shape osteosarcoma single-cell drug response in the lung microenvironment↗

Rawan Makkawi , Carol Halsey , Vaibhav Murthy , Elise C. Manalo, Hugo Cros , Xin Yi, Jessie May Cartier, Mathew Chang, Mark Flory, Kelly Guptell, Amy Gross, Ryan D. Roberts, Jeremy Copperman , Alexander E. Davies

Lung metastases are the leading cause of death in osteosarcoma, where treatment response reflects interactions between genetic, phenotypic, and tumor-host environmental influences. This study investigated how dynamic receptor tyrosine kinase (RTK)-ERK signaling contributes to phenotypic drug response heterogeneity. Using live-cell time-lapse microscopy and data-driven dynamical modeling, we resolved ERK activity, FOSL1/Fra-1 reporter expression, and cell fate in biosensor-expressing osteosarcoma lung metastases ex vivo following MCL1 inhibitor exposure. RTK inhibitors were subsequently evaluated for their ability to suppress adaptive signaling and enhance therapeutic response ex vivo and in vivo. MCL1 inhibition induced extensive tumor-cell death but also enriched for persistent ERK-high/Fra-1 high reporter cells enriched at the tumor-lung interface. Experimental and computational analyses revealed that tumor-cell lysis amplified fibroblast growth factor receptor (FGFR)-mediated ERK signaling in neighboring cells, promoting an adaptive drug-resistant state associated with FOSL1/Fra-1 expression and upregulation of the drug-resistance pathways Nrf2 and HMOX1. Combining MCL1 inhibition with the FGFR inhibitor fexagratinib or broad-spectrum RTK inhibitor pazopanib suppressed adaptive ERK activation and enhanced treatment efficacy ex vivo and in vivo. Together, these findings identify tumor-host interactions and lysis-induced FGFR-ERK signaling as drivers of spatiotemporally heterogeneous adaptive resistance in metastatic osteosarcoma, supporting combinatorial RTK targeting to improve therapeutic response.

Single-cell morphodynamical trajectories enable prediction of gene expression accompanying cell state change↗

Jeremy Copperman , Ian McLean, Sean Gross, Jalim Singh, Vaibhav Murthy , Young Hwan Chang, Alexander Davies , Daniel Zuckerman, Laura Heiser

Cell Systems

Here, we develop MMIST, molecular and morphodynamics-integrated single-cell trajectories. This analysis predicts expression of thousands of RNA transcripts through extracellular signal-induced epithelial-mesenchymal transition (EMT) and mesenchymal-epithelial transition (MET) with near-continuous time resolution. The MMIST framework leverages true single-cell dynamical behavior to generate molecular-level omic inferences and is broadly applicable across biological domains, imaging approaches, and molecular snapshot data.

Serial Imaging of Tumor and microEnvironment (SITE) platform for live-cell ex vivo modeling of primary and metastatic cancer dynamics↗

Vaibhav Murthy , Rawan Makkawi , Francis Anderson , Carol Halsey , Elise Manalo-Hall, Sanjay Srikanth, Gabriella Tangkilisan, Ting Zheng, James McGann, Mark R. Birtwistle, Jeremy Copperman , Alexander E. Davies

Understanding cancer onset and progression is extremely challenging, in part due to experimental limitations in measuring and interpreting key signaling and tumor-host interactions that determine cancer behavior over time, across cell and tissue scales. Here we developed SITE (Serial Imaging of Tumor and microEnvironment), a spatially and temporally integrated platform combining ex vivo culture, biosensors, live imaging, and computation. Applied to modeling primary and lung metastatic breast cancer, SITE revealed tissue-specific tumor–host interactions and ERK signaling patterns linked to distinct single-cell behaviors. We found that multicellular niche formation involved active protrusion and cell–cell contact driven by both cancer and host cells. Mathematical modeling showed ERK signaling was co-influenced by neighboring cancer and host cells. Paracrine signaling among cancer cells increased signaling in a cluster-size–dependent manner, while disruption of cancer–cancer signaling loops amplified tissue-specific differences in tumor architecture. Applied specifically to breast cancer, we demonstrated the utility of the SITE platform, enabling quantitative exploration of ex vivo signaling and tumor-host interaction dynamics.

Featurizing Koopman mode decomposition for robust forecasting↗

David Aristoff, Jeremy Copperman , Nathan Mankovich, Alexander Davies

The Journal of Chemical Physics

This article introduces an advanced Koopman mode decomposition (KMD) technique—coined Featurized Koopman Mode Decomposition (FKMD)—that uses delay embedding and a learned Mahalanobis distance to enhance analysis and prediction of high-dimensional dynamical systems. We show that FKMD improves predictions for a high-dimensional linear oscillator, a high-dimensional Lorenz attractor that is partially observed, and a cell signaling problem from cancer research.

Host-derived growth factors drive ERK phosphorylation and MCL1 expression to promote osteosarcoma cell survival during metastatic lung colonization↗

Camille A. McAloney, Rawan Makkawi , Yogesh Budhathoki, Matthew V. Cannon, Amy C. Gross, Emily M. Franz, Maren Cam, Tatyana A. Vetter, Rebekka Duhen, Alexander E. Davies , Ryan D. Roberts

Cellular Oncology

For patients with osteosarcoma, disease-related mortality most often results from lung metastasis—a phenomenon shared with many solid tumors. While established metastatic lesions behave aggressively, very few of the tumor cells that reach the lung will survive. By identifying mechanisms that facilitate survival of disseminated tumor cells, we can develop therapeutic strategies that prevent and treat metastasis. Here, we analyzed single cell RNA-sequencing data from murine metastasis-bearing lungs to interrogate changes in both host and tumor cells during colonization. We used these data to elucidate pathways that become activated in cells that survive dissemination and identify candidate host-derived signals that drive activation. We validated these findings through live cell reporter systems, immunocytochemistry, and fluorescent immunohistochemistry. We then validated the functional relevance of key candidates using pharmacologic inhibition in models of metastatic osteosarcoma. We show that niche-derived growth factors drive MAPK activity and MCL1 expression in osteosarcoma, promoting metastatic colonization. Although later metastases produce less MCL1, they remain dependent on it. Furthermore, combining MCL1 inhibition with chemotherapy both prevented colonization and eliminated established metastases in murine models of osteosarcoma. Taken together, these data show that MCL1 is a promising target for clinical trials in both human and canine patients.