RNASeq as a tool to understand dysregulation of potential biomarkers in HNSC

Ezhuthachan Mithu, Madhvi Joshi, Ishan Raval, Chaitanya G. Joshi, Anirban Dasgupta, Sharmistha Majumdar, Siddharth ShahbioRxiv 2022 · DOI

TL;DR

This study leverages RNA sequencing to identify potential biomarkers for Head and Neck Squamous Cell Carcinoma (HNSC), a deadly cancer. By analyzing gene expression in tumor versus normal tissues and building protein-protein interaction networks, researchers pinpointed eight key genes that are significantly dysregulated and could serve as diagnostic or prognostic markers for improved patient outcomes.

What Problem Does This Paper Address?

Oral cancer, particularly HNSC, is a significant global health concern with high fatality rates. Current diagnostic methods, like biopsies, are invasive and often detect the cancer at later stages, leading to poor treatment outcomes and high costs. The paper addresses the critical need for less invasive, more effective methods for early detection, diagnosis, and staging of oral cancer. Specifically, it aims to identify novel biomarkers that can help in screening, diagnosing, and assessing the extent of the disease, particularly concerning margin clearance during surgical treatment.

Methods

The researchers employed a multi-pronged approach:

  • RNA Sequencing (RNASeq): This core technique was used to analyze the transcriptome of tumor tissues and adjacent normal tissues from four HNSC patients. This allowed for the identification of differentially expressed genes (DEGs) between the two conditions.
  • Gene Ontology (GO) Enrichment Analysis: This was performed to understand the biological functions and cellular locations of the DEGs.
  • Protein-Protein Interaction (PPI) Network Construction: Using the STRING database, the upregulated genes found in the plasma membrane were used to build a network.
  • K-Means Clustering and MCODE Analysis: These algorithms were applied to the PPI network to identify clusters of highly interconnected genes, from which "hub genes" (genes central to these clusters) were extracted.
  • TCGA Database Validation: The expression and survival data of the identified hub genes were validated using the publicly available The Cancer Genome Atlas (TCGA) database, specifically focusing on Head and Neck Squamous Cell Carcinoma (HNSC) data.

Datasets

  • Patient-derived tissue samples: Four pairs of Oral Squamous cell carcinoma tissues and adjacent normal tissues were collected from Zydus Cancer Centre, Ahmedabad, India.
  • The Cancer Genome Atlas (TCGA) database: This publicly available dataset was used for validating gene expression and survival analyses. The study specifically utilized HNSC data from TCGA, as it represents the vast majority of oral squamous cell carcinoma cases.

Key Insights and Ideas

  • Plasma Membrane as a Biomarker Hotspot: A significant finding was that a substantial number of upregulated genes (253) were localized to the plasma membrane. This suggests that cell surface proteins could be particularly valuable for developing non-invasive diagnostic tools.
  • Network-based Identification of Hub Genes: By constructing and analyzing PPI networks, the study moved beyond individual gene analysis to understand how genes interact within biological pathways. This network approach helped identify key "hub genes" that are central to disease processes.
  • Potential for Margin Clearance Biomarkers: The identified genes not only show dysregulation in HNSC but also have potential implications for assessing margin clearance, a critical factor in preventing cancer recurrence.

Key Results

  • Identification of Differentially Expressed Genes (DEGs): The RNASeq analysis identified 1518 upregulated and 2226 downregulated genes in tumor tissues compared to normal adjacent tissues.
  • Functional Enrichment: GO enrichment revealed that upregulated genes were significantly associated with processes like androgen catabolic process and interleukin-1 receptor binding, while downregulated genes were linked to supramolecular complex and cytoskeletal protein binding.
  • Identification of Eight Key Biomarker Candidates: Through network analysis and validation, eight genes were identified as significantly dysregulated in HNSC and showed potential as biomarkers: SLC2A1, ITGA6, LAMC2, COL1A2, COL1A1, TNC, THY1, and CD276.
  • Survival Analysis: Kaplan-Meier survival analysis indicated that the expression levels of some of these genes (e.g., CD276, LAMC2) were associated with patient survival outcomes in HNSC.

Limitations and Future Work

  • In Vitro Validation: The authors acknowledge that the study's findings require validation using robust quantitative assays in vitro. This would involve experimental validation of the identified genes' roles in HNSC progression and their potential as biomarkers.
  • Sample Size: The study utilized RNASeq data from only four patient samples. While TCGA data provided a larger validation cohort, a larger initial RNASeq dataset could have provided more robust initial findings.
  • Focus on Plasma Membrane: While the plasma membrane focus was insightful, other cellular compartments might also harbor important biomarkers.
  • Future Directions: The paper suggests that the identified genes can be further explored as potential biomarkers for margin clearance. Future research should focus on developing assays to detect these biomarkers in easily accessible samples (e.g., saliva, blood) and investigating their precise roles in HNSC pathogenesis and progression.

Summary updated

Abstract

Abstract With its rising fatality rates, oral cancer is one of the most concerning public health issues. To reduce disease-related mortality and morbidity, advancements in screening and detection are critical. Finding specific biomarkers is one of the most successful approaches for screening, diagnosing, and staging this dreadful disease. In this study differentially expressed genes associated with oral cancer were analyzed using RNASeq to find the potential biomarkers. Functional enrichment of upregulated genes found that 253 genes were present in the plasma membrane. Three clusters were formed using KMean Clustering from the PPI networks, and highly connected hub genes were identified from each cluster. Eventually, expression and survival analyses of hub genes were performed using The Cancer Genome Atlas (TCGA) database targeting Head and Neck Squamous Cell Carcinoma. Among those genes, expression levels of eight genes SLC2A1, ITGA6, LAMC2, COL1A2, COL1A1, TNC, THY1 , and CD276 have significantly changed in Head and Neck Squamous cell carcinoma. There are reports that suggest these genes were significantly dysregulated in Oral Squamous cell carcinoma and can be explored further as potential biomarkers for margin clearance.
BibTeX
@misc{mithu2022rnaseq,
  author = {Ezhuthachan Mithu and Madhvi Joshi and Ishan Raval and Chaitanya G. Joshi and Anirban Dasgupta and Sharmistha Majumdar and Siddharth Shah},
  title = {{RNASeq as a tool to understand dysregulation of potential biomarkers in HNSC}},
  year = {2022},
  doi = {10.1101/2022.09.20.508683},
  howpublished = {bioRxiv},
}
Figure