# A bioinformatics analysis revealing autophagy related DEGs in head and neck squamous cell carcinoma: expanding insights into lipid metabolism

## TL;DR

This study identifies a prognostic model for head and neck squamous cell carcinoma based on autophagy-related genes, showing its potential for predicting patient outcomes.

## Contribution

The novel contribution is the construction and validation of a prognostic model using autophagy-related differential genes in HNSC.

## Key findings

- A prognostic model based on 7 autophagy-related genes showed significant survival differences between high- and low-risk groups.
- Knockdown of GSK3B reduced cell migration in HNSC cell lines.
- GSK3B expression was significantly higher in laryngeal and pharyngeal cancers compared to benign vocal cord polyps.

## Abstract

•A successful establishment of a prognostic model for head and neck tumors.•Differential autophagy genes in tumor models.•A bioinformatics analysis about head and neck squamous cell carcinoma.

A successful establishment of a prognostic model for head and neck tumors.

Differential autophagy genes in tumor models.

A bioinformatics analysis about head and neck squamous cell carcinoma.

To identify value of prognostic model constructed based on tumor module-associated differential autophagy genes in Head and Neck Squamous Cell Carcinoma (HNSC), and verified through in vitro and tissue experiments.

After obtaining the relevant information of HNSC patients from the GEO and TCGA databases, we used Weighted Correlation Network Analysis (WGCNA), multivariate Cox regression analysis, and survival analysis to construct a prognostic prediction model, and validated the prognostic value of the model. We used RT-qPCR and cell scratch assays to validate in vitro for the biological results with TU686 and CAL-27 cell lines. We collected pathological sections from 30 patients with laryngeal cancer, hypopharyngeal cancer, and 10 patients with vocal cord polyps, and immunohistochemical DAB staining was performed to detect the expression of GSK3B in these tissues.

A total of 6163 differentially expressed genes were jointly identified between HNSC and control samples. WGCNA analysis revealed that the genes clustered in the brown module had the highest correlation with HNSC. Using the formula derived from the inclusion of 7 genes in the prognostic model, the risk scores were calculated for the training and validation sets of HNSC disease samples. Survival analysis results showed a significant difference in Overall Survival (OS) between the high and low-risk groups in both the training and validation sets (p < 0.05). The one-year, three-year, and five-year AUC values of the risk scores obtained from the prognostic model in the training and validation sets were all greater than 0.6. After knocking down GSK3B, the relative expression levels of si-GSK3B in the si-GSK3B1 and si-GSK3B2 groups were significantly lower than those in the NC group (both p < 0.05). We compared with the NC group, the migration ability of TU686 and CAL-27 cells was significantly decreased (both p < 0.05). We used immunohistochemical DAB method to detect and found that the content of GSK3B in laryngeal cancer was significantly higher than that in vocal cord polyps (p < 0.05), and the content of GSK3B also in pharyngeal cancer was significantly higher than that in vocal cord polyps (p < 0.05).

Prognostic model constructed based on tumor module-associated differential autophagy genes in HNSC holds prognostic value.

Level 3.