NK-cell receptor-ligand interactions in chemokine-dependant activity in TNBC and during viral infection

Abstract

Background: While the important role of natural killer (NK) cells in eliminating both cancerous and infected cells is well known, many of the specific receptor-ligand interactions that mediate homing, recognition, and killing of target cells via NK cells are still poorly understood. NK cells have been explored more recently and are showing promising success in adoptive cell transfer and immunomodulatory therapies for various cancers and viral infections. However, before being implemented, key aspects of NK-cell biology and challenges in the application of these therapies must be addressed. These include the poor homing of NK cells into the tumor microenvironment of solid tumors and immune evasive strategies that both cancer cells and virally infected cells employ to subvert NK-cell immunity. In this thesis, we explore chemokine and chemokine receptor expression in solid tumors (i.e., triple-negative breast cancer [TNBC]) and peripheral blood NK cells, as well as receptor-ligand interactions in the context of viral immune evasion (i.e., SARS-CoV-2 and mpox virus [MPXV]).

Methods: We analyzed scRNAseq data from 16 TNBC samples, 2 healthy breast tissue controls, and 2 peripheral blood NK (pbNK) cell samples to analyze chemokine and chemokine receptor expression. After identifying CXCL16 as a candidate chemokine expressed on tumor cells, we over-expressed and knocked out CXCL16 in MDA-MB-231 cells, a TNBC cell line, and performed flow cytometric phenotyping and in-vitro motility, proliferation, and NK-cell killing assays. Given that CXCL16 exists in membrane-bound and soluble (cleaved) forms, we also performed in silico analyses of its protein sequence to identify cleavage sites and create membrane-bound and soluble mutants for subsequent in-vitro testing. For virus-focused studies, we carried out bulk mRNAseq analyses in SARS-CoV-2- and MPXV-infected cell lines and scRNAseq analyses on primary respiratory epithelial cells from COVID-19 patients, with a particular focus on NK-cell ligand genes.

Results: Among all chemokines analyzed, we identified CXCL16 as the most abundant, differentially expressed chemokine in TNBC tissues. Correspondingly, there was an enrichment of CXCR6 (the chemokine receptor for CXCL16) among tumor-infiltrating T cells, but no tumor-infiltrating NK cells could be detected in TNBC tissues. Peripheral blood NK cells from healthy donors were analyzed and had undetectable CXCR6 transcript expression. Flow cytometric analyses of MDA-MB-231 cells showed no expression of membrane-bound CXCL16 at baseline but did show expression of CXCR6. Over-expression of CXCL16 in MDA-MB-231 cells via transduction resulted in enhanced motility and reduced proliferation, while no changes could be observed with CRISPR/Cas9 knock-out of CXCL16, consistent with lack of CXCL16 expression at baseline. Remarkably, co-culture of MDA-MB-231 cells with NK cells resulted in increased CXCL16 expression and mixing of NK cells with wild-type and CXCL16-overexpressing MDA-MB-231 cells showed preferential killing of CXCL16-overexpressing cells. CXCL16 sequence analysis predicted an ADAM10/17 cleavage site in the mucin-stalk region which may be responsible for tumor-cell regulation of CXCL16 surface expression and shedding. In our virus-focused studies, scRNAseq analysis of COVID-19 patients showed specific upregulation of NKG2D ligands and sheddases in ciliated respiratory epithelial cells, the target of SARS-CoV-2-infection, and bulk mRNAseq analysis of MPXV-infected A549 cells showed modulation of a variety of NK-cell ligands.

Conclusions: These results suggest a role for CXCL16 in tumorigenesis and susceptibility to NK-cell killing and may be a beneficial target for enhancing NK-cell homing and targeting of solid tumors. Furthermore, we uncovered NK-cell ligands that SARS-CoV-2 and MPXV modulate as a potential means of immune evasion.