Systems-level analyses and clinical validation highlight CD53 as a diagnostic and prognostic marker in lung
Jun Chen1, Weiming Tan2, Yichen Wei1
1Department of Pathology, The Second Affiliated Hospital of Guangxi Medical University, Nanning, China.
Background:
Lung adenocarcinoma (LUAD) is the most common subtype of lung cancer, often diagnosed at advanced stages with poor prognosis. CD53, a tetraspanin involved in immune regulation, has an unclear role in LUAD.
Methods:
Five GEO LUAD transcriptomic datasets were integrated and batch-corrected. Differential expression and weighted gene co-expression network analyses identified LUAD-associated genes. Machine learning (Elastic Net) selected core predictive genes. Immune infiltration analysis, single-cell RNA-seq analysis, regulatory and drug-gene network construction, summary-data-based Mendelian randomization (SMR), and colocalization analyses were performed. Public TCGA-LUAD data were further used to evaluate the association between CD53 expression and clinicopathological characteristics. CD53 expression was clinically validated by immunohistochemistry (IHC) and PCR in 30 paired LUAD and adjacent normal tissue samples.
Results:
Six genes (BTK, CD163, CD53, F13A1, HCLS1, MS4A6A) were consistently downregulated in LUAD, with CD53 as a central immune-related candidate. CD53 expression positively correlated with monocytes and neutrophils, and negatively with naïve B cells, T follicular helper cells, plasma cells, and M0 macrophages. Single-cell RNA-seq analysis showed that CD53 was mainly localized to immune cell populations, particularly T cells, macrophages, and monocytes. SMR prioritized CD53 as a genetically associated candidate gene, whereas colocalization analysis did not support a shared causal variant between CD53 and LUAD. IHC and PCR both confirmed significantly lower CD53 expression in tumor tissues versus adjacent normal tissues, with low expression associated with poor prognosis. Public TCGA-LUAD analysis further showed that the CD53-low group had a higher proportion of deceased patients, and deceased patients exhibited lower CD53 expression than alive patients. Drug-gene network analysis predicted several candidate compounds potentially associated with CD53 and other model genes, providing hypothesis-generating clues for future experimental studies.
Conclusion:
CD53 is downregulated in bulk LUAD tissues, correlates with immune cell infiltration, and may serve as a candidate immune-microenvironment-associated diagnostic and prognostic biomarker. Its direct tumor-suppressive function remains to be determined in future mechanistic studies. These findings provide a basis for further mechanistic studies and future validation of CD53-related diagnostic, prognostic, and therapeutic hypotheses in LUAD.
