IMR Press / FBL / Volume 29 / Issue 12 / DOI: 10.31083/j.fbl2912407
Open Access Original Research
Machine Learning Reveals Aneuploidy Characteristics in Cancers: The Impact of BEX4
Show Less
Affiliation
1 Department of General Surgery, The First Affiliated Hospital of Anhui Medical University, 230022 Hefei, Anhui, China
2 Department of General Surgery, Anqing Municipal Hospital, 246000 Anqing, Anhui, China
3 Clinical Pathology Center, The First Affiliated Hospital of Anhui Medical University, 230022 Hefei, Anhui, China
4 Clinical Pathology Center, Anhui Public Health Clinical Center, 230011 Hefei, Anhui, China
5 School of Basic Medical Sciences, Anhui Medical University, 230032 Hefei, Anhui, China
*Correspondence: 517563704@qq.com (Wenqi Yang); 13965053990@163.com (Chao Zhang); chongzhang93@126.com (Chong Zhang)
These authors contributed equally.
Front. Biosci. (Landmark Ed) 2024, 29(12), 407; https://doi.org/10.31083/j.fbl2912407
Submitted: 15 June 2024 | Revised: 23 August 2024 | Accepted: 30 August 2024 | Published: 29 November 2024
Copyright: © 2024 The Author(s). Published by IMR Press.
This is an open access article under the CC BY 4.0 license.
Abstract
Background:

Aneuploidy is crucial yet under-explored in cancer pathogenesis. Specifically, the involvement of brain expressed X-linked gene 4 (BEX4) in microtubule formation has been identified as a potential aneuploidy mechanism. Nevertheless, BEX4’s comprehensive impact on aneuploidy incidence across different cancer types remains unexplored.

Methods:

Patients from The Cancer Genome Atlas (TCGA) were stratified into high-score (training) and low-score (control) groups based on the aneuploidy score. Mfuzz expression pattern clustering and functional enrichment were applied to genes with BEX4 as the core to explore their regulatory mechanisms. Various machine learning techniques were employed to screen aneuploidy-associated genes, after which aneuploidy characteristic subtypes were established in cancers. Moreover, the aneuploidy characteristics across multiple cancer types were investigated by integrating the extent of tumor cell stemness acquisition and a series of immune traits. Immunohistochemistry and proliferation assay mainly verified the anti-tumor effect of different BEX4 level.

Results:

Functional clustering results showed that aneuploidy and stemness were significantly associated in kidney chromophobe (KICH) and thyroid carcinoma (THCA). And cell metabolism and cell cycle had key effects. Residual analysis indicates superior screening performance by random forest (RF). An aneuploid feature gene set with BEX4 as the core was screened to construct a Nomogram model. BEX4, calmodulin regulated spectrin associated protein 2 (CAMSAP2), and myristoylated alanine rich protein kinase C substrate (MARCKS) were identified as aneuploidy characteristic hub genes. Molecular subtypes in thymoma (THYM), thyroid carcinoma (THCA), and kidney chromophobe (KICH) showed significant differences in tumor cell stemness among different subtypes. The competitive endogenous RNA (ceRNA)-Genes network revealed that hub genes, co-regulated by hsa-miR-425-5p, hsa-miR-200c-3p, and others, regulate microtubules, centrosomes, and microtubule cytoskeleton. Furthermore, elevated BEX4 emerged as a significant protective factor in Pancreatic adenocarcinoma (PAAD), KICH, kidney renal papillary cell carcinoma (KIRP), and kidney renal clear cell carcinoma (KIRC).

Conclusions:

BEX4, CAMSAP2, and MARCKS specifically express in microtubules, centrioles, and cytoskeletons, influencing tumor chromosome division and inducing aneuploidy. Additionally, the relationship between the acquisition of tumor cell stemness and the severity of aneuploidy varies significantly across tumor types, displaying positive and negative correlations.

Keywords
aneuploidy
BEX4
machine learning
pan-cancer
biomarkers
Funding
82303475/ National Natural Science Foundation of China
2022AH040161/ University Research Project of Anhui Province
Figures
Fig. 1.
Share
Back to top