published_or_final_version / Anatomy / Doctoral / Doctor of Philosophy

Abstract not available

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide, with over 50,000 new cases in the United States annually. Major risk factors for HNSCC include chronic smoking and/or alcohol consumption, and more recently infection with human papillomaviruses (HPVs). HNSCC comprises a diverse collection of tumors, of which the oral cavity is the most affected site. Most HNSCC patients will die within the first 30 months of disease, an abysmal statistic largely reflecting a lack of effective treatment strategies. This challenge highlights the current unmet need to identify 1) distinct molecular subgroups of HNSCC and 2) prognostic biomarkers that can inform subgroup-specific treatment plans. The Lu Lab has identified a previously unappreciated HNSCC subgroup defined by alteration in NSD1, a histone methyltransferase enzyme mutated in up to 15% of HPV- HNSCC. NSD1 has specific di-methylase activity targeting histone H3 lysine 36 (H3K36). The formation of methylated H3K36 is associated with open chromatin and transcriptional activation. There are several distinct features of NSD1-mutant HPV- HNSCC tumors, including a) increased patient smoking history and mutational burden; b) a significantly better prognosis; and c) reduced immune cell infiltration in the tumor. These correlative findings may suggest a tumor-suppressive role of NSD1, yet the impact of NSD1-loss in HNSCC pathogenesis and the underlying mechanisms remain unclear. The scope of my thesis work aims to address this gap in knowledge through investigating the consequences of impaired NSD1 in normal epithelial tissue and in the setting of HNSCC tumors. In Chapter 2, we employed both a syngeneic tumor implantation model and a physiologic mouse model of HNSCC carcinogenesis to establish and characterize Nsd1-KO tumors. We found that mice harboring Nsd1-KO tumors are comparatively immune- ‘cold,’ a phenotype that persisted in human HNSCC patient samples. Our in vitro data suggests that depletion of NSD1 epigenetically silences the interferon response, an effect that was rescued through inhibition of epigenetic machinery that limits NSD1-deposited H3K36me2. These studies provide novel insight into the molecular underpinnings of observed immune-‘coldness’ in NSD1-mutated HNSCC. They also prompted us to probe the impact of NSD1-loss at early stages of tumor development. In Chapter 3, I describe a lingual-derived organoid system to assess unbiased transcriptional changes that occur upon Nsd1-loss during homeostasis, premalignancy, and overt tumor formation. Interestingly, we found that Nsd1-KO organoids harvested from stages preceding HNSCC featured downregulation in gene expression related to epithelial barrier formation and wound healing. Furthermore, organoids derived from Nsd1-KO tumors showed reduced expression of epithelial-to-mesenchymal transition (EMT) signature genes, potentially signifying a role of NSD1 in modulating cell adhesion and mobility. Several additional studies are needed to establish the functional basis of how NSD1 participates in these processes. Taken together, this body of work expands upon our current understanding of NSD1-loss in HNSCC development and presents a conceptual scaffold framing NSD1 as a multi-faceted player in the homeostatic maintenance and neoplastic events of the oral epithelium.

One of the most challenging questions in the study of cancer is the origin and the nature of the cells that initiate cancer. Accumulated studies have provided many molecular origins of cancers but we still do not know what kind of cells in the tissues transform to cancer cells. Therefore, identifying the cellular origin of these cells is critical for the development of better prognosis, diagnosis and treatment of cancer. A stem cell origin of cancer has been postulated over 150 years. Recent cancer stem cell studies have opened a new window on aspects of the cellular origin of cancer. In this communication, we will address two possible cellular origins of cancer in epithelial tumor development using mouse skin cancer model: tissue specific stem cells, and cells from other organs. To demonstrate contribution of the tissue specific stem cells in tumor development, we monitored the contribution of keratin-15 positive hair follicle bulge stem cells to skin tumor development in the multistage skin carcinogenesis model with Krtl- 15CrePRl;R26R transgenic mice. We found that labeled progeny of the keratin-15 positive bulge stem cells migrate into papillomas and these cells contribute to almost all papilloma samples by 20 weeks of promotion. Additionally, in contrast to the transient contribution of bulge-derived cells in skin wound healing, consistent percentage of the bulge-derived cells stay in the papillomas over 20 weeks. Furthermore, papillomas have heterogeneous expression of the codon 61 signature Ha-ras mutation, with approximately 30 percent of bulge-derived regions expressing the mutation. To determine the contribution of exogenous sources in skin tumor development, we examined bone marrow-derived cells (BMDCs) in the skin tumors from the allogeneic gender-mismatched bone marrow transplantation recipient mice after chemical skin carcinogenesis. We observed that genetically marked (EGFP) BMDCs were detected in the epithelial part of skin wounds and also skin tumors, and we found greater degree of BMDC contribution in chronic ulcer-related skin lesions. Lastly, an in-vitro assay demonstrated plasticity of BMDCs by inducing keratin-14 expressing cells from mesenchymal stem cells. These results demonstrated that hair follicle bulge stem cells and also BMDCs are able to contribute to skin tumor development.