Project 2: Human tumor analysis of lymph node and distant metastasis

Sylvia Plevritis, PhD, Lead

Using human tissues, we can now quantitatively compare the spatial biology of uninvolved LNs and their concurrent primary tumors by applying recent advances in multiplexed in situ imaging with single-cell sequencing technologies. 

We hypothesize that spatially resolved stromal-immune interactions in LNs, together with stromal-malignant properties in a primary tumor, set the stage for metastasis. We will test this hypothesis by spatially phenotyping human tumor samples across a spectrum of clinical stages, including early stages. 

1. We will reconstruct and compare spatially resolved tumor-stroma-immune colocalization patterns in samples from HNSCC and LUAD cancer patients of uninvolved LNs, involved LNs, and primary tumors. We will also probe cell composition and colocalization patterns within the extracellular matrix (ECM) to ascertain its role in establishing and maintaining a pro-metastatic microenvironment in uninvolved LNs. 
2. We will discover cell-cell interactions in uninvolved LNs using novel biocomputational approaches to integrate spatial features from CODEX with single-cell RNA sequencing data, toward identifying proximal cell-cell interactions among tumor-stromal and stromal-immune cell types associated with LN metastases. We will then evaluate selected cell-cell interactions in organoid models of human-derived cells, including perturbation with CRISPR-facilitated gene editing to reveal mechanistic insights. 
3. We will use spatially-aware Markov modeling to predict tumor-stromal and stromal-immune colocalization patterns in human-derived LNs and primary tumors associated with metastatic progression.