Our Research

Our lab investigates how peripheral nerve signaling remodels the tumor microenvironment to drive cancer progression and therapeutic resistance. We have identified the neuronal adhesion molecules NRXN1 and NLGN1 as critical mediators of nerve–cancer communication in prostate and pancreatic tumors, where their expression is enriched in high-grade and treatment-resistant tumors.

By integrating in vitro co-culture systems, ex vivo organ-on-a-chip models (developed in collaboration with Dr. Luiz Bertassoni and the Biofabrication Hub), and in vivo mouse models, we investigate how ligand-receptor interactions between nerves, immune cells, and cancer-associated fibroblasts shape the tumor ecosystem. These studies aim to define targetable mechanisms that disrupt neural control of tumor growth, restore immune surveillance, and ultimately improve patient response to therapy.

We develop experimental and computational methods to capture high-plex (40+ protein) spatial proteomic and morphological features of tumor innervation directly in patient tissues. By integrating multiplex imaging, single-cell analysis, and deep learning, we quantify how nerves, immune cells, and stromal populations are spatially organized within the tumor microenvironment.

A central tool developed in our lab is AxonFinder, a deep-learning framework for automated segmentation and profiling of tumor-innervating nerve fibers. Using this platform, we map nerve subtypes and nerve-associated microenvironments across disease states and treatment contexts.

Our goal is to identify spatial biomarkers and organizational principles that distinguish tumor subtypes, track disease progression, and inform strategies for improved patient stratification and therapy.

Together with our collaborators, we integrate spatial proteomics with clinical imaging modalities, including histopathology and multi-parametric MRI, to identify biomarkers that distinguish indolent from aggressive prostate cancers. Using longitudinal tissue samples from active surveillance cohorts, we are mapping molecular trajectories of disease progression and developing models that predict patient outcomes based on integrated spatial and molecular features.

By linking imaging, molecular, and histologic data, we aim to establish a new paradigm for early detection and risk stratification, enabling clinicians to identify high-risk patients earlier and guide timely, personalized treatment decisions.