Evidence for a Functional and Spatial Natural Killer Cell Dichotomy in Human Endometrial Cancer — ASN Events
  1. Schedule
  2. Lightning Session One
  3. Evidence for a Functional and Spatial Natural Killer Cell Dichotomy in Human Endometrial Cancer

Evidence for a Functional and Spatial Natural Killer Cell Dichotomy in Human Endometrial Cancer (#176)

Caprice D. Eisele 1 2 , Courtney J. Riedinger 3 , Paulina J. Haight 3 , Everardo Hegewisch-Solloa 4 , Wantong Li 5 , Jesse J. Kowalski 1 , Youssef Youssef 1 , Paige M. Porrett 6 , Emily M. Mace 4 , Bethany L. Mundy-Bosse 5 , Bradley W. Blaser 5 , Casey M. Cosgrove 3 , Aharon G. Freud 1
  1. Department of Pathology, The Ohio State University Comprehensive Cancer Center, The James Cancer Hospital and Solove Research Institute, Columbus, Ohio, USA
  2. Biomedical Sciences Graduate Program, The Ohio State University, Columbus, Ohio, USA
  3. Department of Gynecologic Oncology, The Ohio State University Comprehensive Cancer Center, The James Cancer Hospital and Solove Research Institute, Columbus, Ohio, USA
  4. Department of Pediatrics, Columbia University Irving Medical Center, Vagelos College of Physicians and Surgeons, New York, USA
  5. Division of Hematology, Department of Internal Medicine, The Ohio State University Comprehensive Cancer Center, The James Cancer Hospital and Solove Research Institute, Columbus, Ohio, USA
  6. Comprehensive Transplant Institute, University of Alabama at Birmingham, Birmingham, Alabama, USA

Endometrial cancer (EC) is the most common gynecological malignancy in the US and responds variably to immune checkpoint blockade therapy (ICBT), warranting further study of the tumor microenvironment. It is well established that during pregnancy, CD49a+CD103+/- tissue resident NK (trNK) cells, which are distinct from CD49a-CD103- conventional NK (cNK) cells, localize within the decidualized endometrium and perform unique regulatory functions required for successful pregnancy. However, the phenotypes and functions of NK cells in the malignant postmenopausal endometrium have been minimally explored. Using ex vivo flow cytometry and 5’ scRNAseq on primary EC samples obtained from pretreatment diagnostic hysterectomies, we identified both cNK and trNK subsets, with trNK cells representing the most abundant subset across all EC subtypes in comparison to both autologous blood and grossly uninvolved normal adjacent tissue (NAT). Additionally, the trNK cells in EC possessed features reminiscent of trNK cells in pregnancy, including minimal expression of granzyme B and perforin, strong expression of XCL1 and CCL5, and abundant regulatory transcriptional networks. Utilizing cyclic immunofluorescence on tissue sections from distinct EC subtypes, we observed that trNK cells were more frequently enriched within tumor glands while cNK cells were more localized to the stroma. Considering these findings, computational inference of cell-cell interactions in our sequencing data predicted abundant regulatory interactions driven by trNK cells, including GM-CSF signaling with macrophages and SEMA4A interactions with tumor cells. In contrast, the cNK cells had putative Fas/FasL interactions with CD8+ T cells, as well as L-selectin and DNAM1 interactions with tumor cells. These results suggest a dichotomy of NK cell phenotypes in EC, with the trNK cell phenotype potentially mimicking evolutionarily conserved pregnancy programs to promote tumor growth and survival. Thus, targeting trNK cell function may provide a novel approach to enhance the therapeutic response of EC to ICBT.