cyTOF Multidimensional Analysis of Immune Cell Dysfunction and Obesity-Induced Tumor Microenvironment

Chun-Kai Chiu

Mentor: Dr. Miriam Jacobs, Georgetown University Medical Center.

Date/Time: August 22nd, 2024 at 4:40pm.

Abstract: Background: In recent years immunotherapy for solid tumors has moved into the front line setting with the large majority of approved drugs focusing on modulation of T cell response to achieve tumor control, specifically PD-1 blockade. Although significant improvements have been seen in overall response rate and survival in patients with metastatic solid tumors receiving anti-PD-1 therapy, PD-1 blockade in metastatic ER+ breast cancer has been widely ineffective despite infiltration of cytotoxic T cells and NK cells into the tumor. Strategies to amplify the immune response in metastatic ER+ breast cancer are needed to open up new treatment modalities and advance the field of immunotherapy in breast cancer.

Method: To gain a better understanding of immune cell dysfunction in ER+ breast cancer and obesity we conducted a series of in vivo mouse studies using the EO771 ER+ syngeneic breast cancer model. Mice received either estrone (E1), estradiol (E2) or placebo pellets to mimic postmenopausal (E1) or pre-menopausal (E2) hormonal states and were fed either a low-fat diet (LFD) or a high fat diet (HFD) to induce obesity. Based on these experiments we found that obese E1 mice had an increase in tumor growth. We collected single cell tumor suspension, blood and spleen from each mouse and performed antibody staining for cyTOF analysis generating a larger data set with over 300 samples. During this internship, I will focus on the tumor samples within the dataset (n=60).

To ensure data quality and assess batch effects, we employed various QC steps, including multi-sample comparison histograms, T-REX (Tracking Responders Expanding), and MEM (Marker Enrichment Modelling) analysis. For the identification of target cell types, we utilized multidimensional analysis methods such as UMAP, opt-SNE, and FlowSOM. Finally, we conducted statistical analyses using R, creating various plots (e.g., histograms, box plots), and calculated the significance of cell differences across different treatment groups.

Results: After QC, I identified primary immune cell populations, including macrophages, B cells, T cells, NK cells, and MDSCs, and further distinguished CD8+ cytotoxic T cells, CD4+ helper T cells, and T-reg cells.

Conclusion: Some trends were observed. For example, in NK cell proportions, the control groups showed the highest levels, followed by E1 and E2 in both HFD and LFD groups; HFD groups exhibited higher macrophage counts. However, the significant variability within groups led to large error bars, complicating significance calculations, which will require further investigation and adjustments.