The human microbiome is a vast collection of microbial species that exist throughout the human body and regulate various bodily functions and phenomena. Of the microbial species that exist in the human microbiome, those within the archaea domain have not been characterized to the extent of those in more common domains, despite their potential for unique metabolic interaction with host cells. Research has correlated tumoral presence of bacterial microbial species to the development and progression of lung cancer; however, the impacts and influences of archaea in the microbiome remain heavily unexplored. Within the United States lung cancer remains highly fatal, responsible for over 100,000 deaths every year with a 5-year survival rate of roughly 22.9%. This project attempts to investigate specific archaeal species’ correlation to lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) incidence, patient staging, death rates across individuals of varying ages, races, genders, and smoking-statuses, and potential molecular targets associated with archaea microbiome. Archaeal species abundance was assessed across lung tissue samples of 527 LUAD patients, 479 LUSC patients, and 99 healthy individuals. Nine archaeal species were found to be of significantly altered abundance in cancerous samples as compared to normal counterparts, 6 of which are common to both LUAD and LUSC subgroups. Several of these species are of the taxonomic class Thermoprotei or the phylum Euryarchaeota, both known to contain metabolic processes distinct from most bacterial species. Host-microbe metabolic interactions may be responsible for the observed correlation of these species’ abundance with cancer incidence. Significant microbes were correlated to patient gene expression to reveal genes of altered abundance with respect to high and low archaeal presence. With these genes, cellular oncogenic signaling pathways were analyzed for enrichment across cancer and normal samples. In comparing gene expression between LUAD and adjacent normal samples, 2 gene sets were found to be significantly enriched in cancers. In LUSC comparison, 6 sets were significantly enriched in cancer, and 34 were enriched in normals. Microbial counts across healthy and cancerous patients were then used to develop a machine-learning based predictive algorithm, capable of distinguishing lung cancer patients from healthy normal with 99% accuracy.

Head and neck squamous cell carcinoma (HNSCC) tumor phenotypes and clinical outcomes are significantly influenced by etiological agents, such as HPV infection, smoking, and alcohol consumption. Accordingly, the intratumor microbiome has been increasingly implicated in cancer progression and metastasis. However, few studies characterize the intratumor microbial landscape of HNSCC with respect to these etiological agents. In this study, we aimed to investigate the bacterial and fungal landscape of HNSCC in association with HPV infection, smoking, and alcohol consumption. RNA-sequencing data were extracted from The Cancer Genome Atlas (TCGA) regarding 449 tissue samples and 44 normal samples. Pathoscope 2.0 was used to extract the microbial reads. Microbe abundance was compared to clinical variables, oncogenic signatures, and immune-associated pathways. Our results demonstrated that a similar number of dysregulated microbes was overabundant in smokers and nonsmokers, while heavy drinkers were characterized by an underabundance of dysregulated microbes. Conversely, the majority of dysregulated microbes were overabundant in HPV+ tumor samples when compared to HPV- tumor samples. Moreover, we observed that many dysregulated microbes were associated with oncogenic and metastatic pathways, suggesting their roles in influencing carcinogenesis. These microbes provide insights regarding potential mechanisms for tumor pathogenesis and progression with respect to the three etiological agents.

Human Papilloma Virus (HPV) is highly prevalent within the U.S., with studies estimating that over 80% of individuals will contract the virus in their lifetime. HPV is considered a primary risk factor for the development and progression of oropharyngeal cancers. The impact of the HPV virus’s E6 and E7 oncoproteins on cellular signaling pathways and genomic integration has been extensively characterized. Indirect genomic effects; however, remain relatively unidentified. In this study, we analyzed 83 HPV+ Head and Neck Squamous Cell Carcinoma (HNSCC) patients of varying HPV types. Expression counts of the HPV E6 and E7 oncogenes were estimated across samples and correlated with genomic mutational classes. High expression of E6 and E7 oncoproteins was associated with a greater number of total point mutations, especially on chromosomes 1, 11, and 17, which have been implicated in HPV-mediated cancers in previous studies. Samples with high E6 and E7 expression also exhibited more frequent non-clustered structural variation and a lack of clustered variation altogether. Copy number segments were present with fewer number of repeats in high E6 and E7 expression samples, which is known to correlate with decreased expression of affected genes. E6 and E7 expression was associated with increased activity of several cellular pathways associated in oncogenesis and telomere maintenance. In comprehensively characterizing the effects of the HPV oncoproteins on the human genome, potential mechanisms of HNSCC pathogenesis may be further elucidated.