# Identification and validation of integrated stress-response-related genes as biomarkers for age-related macular degeneration

## TL;DR

This study identifies SLFN11 and GRIN1 as potential biomarkers for age-related macular degeneration, using bioinformatics and validation techniques.

## Contribution

The study introduces a novel approach combining ISR-related genes and transcriptomic data to identify and validate AMD biomarkers.

## Key findings

- SLFN11 and GRIN1 were identified as common biomarkers for AMD with good diagnostic predictive abilities (AUC > 0.7).
- SLFN11 and GRIN1 were enriched in pathways like proteasome and lysosome, and showed higher expression in AMD patients.
- RT-qPCR validation confirmed elevated SLFN11 and GRIN1 expression in AMD patients compared to controls.

## Abstract

Age-related macular degeneration (AMD) is a prevalent ocular condition associated with aging, serving as a significant contributor to vision loss among middle-aged and older individuals. Studies have shown that AMD and integrated stress response (ISR) are associated with oxidative stress, but no specific molecular mechanisms have been identified. Therefore, this study aimed to identify potential biomarkers for AMD through bioinformatics analysis based on the transcriptome database and integrated stress response related genes (ISR-RGs).

Transcriptomic data GSE76237, GSE247168, and ISR-RGs were sourced from public databases and related literature. The biomarkers associated with AMD were identified by differentially expressed gene (DEG) analysis, intersection of common DEGs, and ISR-RGs machine algorithm. After that, nomograms, GSEA, and immune infiltration analysis were performed for the biomarkers. The effects of transcription factors (TFs) and miRNAs on biomarkers were then explored by constructing a TF-biomarker–miRNA regulatory network. In addition, potential effective drugs of the biomarkers were explored by constructing a biomarker–effective drug interaction network. Finally, we verified the gene expression of the biomarkers by RT-qPCR.

We obtained 2,567 and 1,454 DEGs in GSE76237 and GSE247168, respectively. The up- and downregulated genes shared in both datasets were intersected with ISR-RGs taken to obtain eight candidate genes. SLFN11 and GRIN1 were identified as common biomarkers for AMD. An analysis of the nomogram model of biomarkers revealed good diagnostic predictive abilities (AUC > 0.7). SLFN11 and GRIN1 were mainly enriched in pathways such as proteasome, lysosome, and neuroactive ligand receptor interaction. In addition, the disease group’s monocyte expression was significantly higher than that of the control group in GSE76237 (p < 0.01). We obtained thirteen relevant miRNAs and 27 TFs by prediction, with three shared TFs, and seventeen potentially effective drugs were predicted. RT-qPCR validation showed in AMD patients, and SLFN11 and GRIN1 expression was significantly higher than controls (p < 0.05). Only SLFN11 expression was consistent with the bioinformatics analysis.

SLFN11 and GRIN1 were identified as AMD biomarkers, exhibiting robust diagnostic performance and providing new insights into the condition.