# Construction of a Mitochondria‐Related Gene Diagnostic Model Based on Integrated Multiomics Data and Functional Validation of ANK2 as a Key Regulator in Colorectal Cancer

## TL;DR

This study develops a diagnostic model for colorectal cancer using mitochondrial genes and validates ANK2 as a key tumor suppressor.

## Contribution

A novel CRC diagnostic model using mitochondria-related genes and functional validation of ANK2 as a tumor suppressor.

## Key findings

- Eight key mitochondrial-related genes were identified and used to construct a CRC diagnostic model with high accuracy.
- ANK2 was confirmed to promote apoptosis and inhibit angiogenesis in CRC, indicating its role as a tumor suppressor.
- The model's genes showed correlations with immune cell infiltration and specific expression patterns in single-cell data.

## Abstract

Colorectal cancer (CRC) is one of the most common malignancies of the digestive tract globally, characterized by high incidence, difficulty in early diagnosis, and poor prognosis. Traditional screening methods have limitations in sensitivity and specificity, thus necessitating the development of novel, efficient molecular diagnostic approaches. Recent studies have highlighted the crucial role of mitochondrial dysfunction in the initiation and progression of various cancers, suggesting that mitochondria‐related genes (MRGs) could serve as promising diagnostic targets for CRC. In this study, we integrated transcriptomic data from 1174 samples across The Cancer Genome Atlas (TCGA) and multiple Gene Expression Omnibus (GEO) public datasets (GSE21510, GSE44076, and GSE9348) and combined it with MRG data from the MitoCarta3.0 database for a systematic analysis of differentially expressed genes (DEGs). Using LASSO regression and SVM‐RFE, two machine learning algorithms, we identified eight key MRGs (ABCG2, ANK2, MACC1, PMAIP1, SLC22A5, SLC25A34, ACAT1, and PDK4) and constructed an early diagnostic model for CRC. Receiver operating characteristic (ROC) curve analysis confirmed the diagnostic efficacy of the model. Gene interaction networks were constructed using GeneMANIA, demonstrating the potential synergistic roles of these genes in regulating cellular metabolism, drug efflux, and immune modulation. CIBERSORT immune cell infiltration analysis revealed significant correlations between these genes and various immune cell subtypes, including T cells, macrophages, and dendritic cells. Further integration of single‐cell RNA sequencing data (GSE245552) identified the specific expression patterns of the diagnostic model genes across different cell types. Additionally, we conducted an in‐depth investigation of the ANK2 gene. Immunohistochemistry (HPA database), qRT‐PCR, and western blotting confirmed the significantly low expression of ANK2 in CRC tissues and cell lines. Moreover, TUNEL and angiogenesis assays showed that overexpression of ANK2 significantly promoted cell apoptosis and inhibited angiogenesis, suggesting that ANK2 may function as a key tumor suppressor in CRC. In conclusion, this study proposes and validates a CRC diagnostic model based on differentially expressed mitochondrial genes. We systematically explored the molecular mechanisms and immune microenvironment correlations of the model and confirmed the biological effects through single‐cell and molecular biology experiments. Notably, we highlight the potential regulatory role of ANK2 in the progression of CRC. This research provides theoretical support and new directions for early screening, diagnostic biomarker identification, and targeted therapy strategies for CRC.