76: Single nuclei RNAseq Reveals Dedifferentiation States from Human ICCs to GIST

Thursday, March 23, 2023

11:16 AM – 11:26 AM

Location: Ballroom A

Abstract Presenter(s)

Ashwyn K. Sharma, MD (he/him/his)

Resident Physician
UC San Diego, United States

Disclosure(s): No financial relationships to disclose

First Author(s)

Ashwyn K. Sharma, MD (he/him/his)

Resident Physician
UC San Diego, United States

Disclosure(s): No financial relationships to disclose

Author(s)

JJ

John Jun, B.S.

Bioinformatics Scientist
UC San Diego, United States

Disclosure information not submitted.
AW

Alexander T. Wenzel, B.S.

PhD Candidate
UC San Diego, United States

Disclosure information not submitted.
CT

Chih-Min Tang, PhD

Staff Research Associate
UC San Diego, United States

Disclosure information not submitted.
DG

Diwakar Guragain, PhD

Postdoctoral Employee
UC San Diego, United States

Disclosure information not submitted.
JM

Jill P. Mesirov, PhD

Professor of Medicine
UC San Diego, United States

Disclosure information not submitted.
Jason Sicklick, MD (he/him/his)

Professor of Surgery and Pharmacology
Moores Cancer Center | UC San Diego Health , California, United States

Disclosure information not submitted.

Disclosure(s):

Ashwyn K. Sharma, MD: No financial relationships to disclose

Introduction: Gastrointestinal Stromal Tumor (GIST), the most common sarcoma of the GI tract, is most often driven by gene mutations in KIT, PDGFRA, SDH, and NF1. GIST is hypothesized to originate from the interstitial cells of Cajal (ICCs), i.e., the pacemaker cells of the GI tract. We sought to characterize the heterogeneity of GIST and to investigate their evolution from ICCs.

Methods: We performed single nuclei RNA sequencing (snRNAseq) on 16 resected human GISTs. Unsupervised clustering was performed to identify major cell types. GIST cells were identified using expression of KIT, ANO1, and ETV1. ICCs were isolated using published single cell RNAseq datasets of the normal GI tract. For each ICC and GIST tumor cell, CytoTRACE (CT) scores were computed to measure the cell’s degree of differentiation from well differentiated to dedifferentiated (0£CT£1). GIST cells were binned into CT score quartiles, and differential gene expression (DGE) analysis was performed. Finally, single sample Gene Set Enrichment Analysis (ssGSEA) was performed to evaluate for enrichment in "cancer stem cell (CSC)" gene sets.

Results: Unsupervised clustering of n=16 human GISTs (KIT: n=10; PDGFRA: n=3; SDHB: n=2; NF1: n=1) revealed a complex tumor microenvironment, including tumor, stromal (e.g., cancer associated fibroblasts), and immune cells (Panel A). Integrating GIST cells with ICCs demonstrated that some, but not all, ICC subpopulations co-cluster with tumor cells (Panel B). To assess these differences, CT scores were computed. GIST had higher mean scores compared to ICCs (0.53 vs 0.16, p< 0.001), indicating that tumor cells are in a less differentiated state than ICCs (Panel C). DGE analysis showed ICCs had minimal expression of CD34 (GIST & CSC marker), PROM1 (CD133, CSC marker) and MGAT5 (glioma & liver CSC regulator of growth & metastasis), whereas GIST cell expression of these genes increased in each CT quartile (Panel D). Consistent with this, ssGSEA analysis revealed that dedifferentiated GIST cells had higher enrichment in CSC gene sets.

Conclusions:

We report, to our knowledge, the largest single cell transcriptomic analysis of GIST and demonstrate that individual GIST cells have a wide range of differentiation states. As tumor cells become dedifferentiated, there is concomitant enrichment in genes associated with cancer stemness, tumor growth and metastasis. Taken together, we identify significant GIST tumor cell heterogeneity, as well as define a lineage of dedifferentiation that may serve to identify novel predictive biomarkers of aggressive tumor biology, as well as new therapeutic targets.

Learning Objectives:

Define the tumor microenvironment of GIST, and list other subpopulations within these tumors.
Describe and interpret methods of lineage tracing for GIST.
Identify possible biomarkers and novel therapeutic targets of GIST.