Comprehensive analysis of brain tissue transcriptome and serum miRNA reveals molecular signatures of cognitive impairment after traumatic brain injury.

BACKGROUND

Cognitive impairment after traumatic brain injury (TBI) lacks early biomarkers. Integrating brain transcriptome and serum miRNA data may reveal molecular signatures for early prediction.

METHODS

We analyzed GEO datasets GSE104687 (brain transcriptome, n = 376) and GSE123336 (serum miRNA, n = 218). Brain samples were grouped into TBI with cognitive impairment (TBI-CI, n = 101), TBI without cognitive impairment (TBI-nonCI, n = 93), and healthy controls (n = 103). Differential expression and GSEA were performed. miRNA-target gene networks were constructed using TargetScan and miRTarBase. An independent cohort (n = 64) validated key miRNAs.

RESULTS

Ten core genes differentiated TBI-CI from TBI-nonCI, enriched in NF-κB/TNF immune pathways and ribosomal inhibition. Serum miRNA analysis identified 35 differentially expressed miRNAs, with miR-146a-3p and miR-203a-3p significantly downregulated. Integrated network analysis highlighted miR-146a-3p targeting PCK1, RPS28P7, and ANKRD20A4. In the validation cohort, both miRNAs correlated positively with MoCA scores (rho = 0.748 and 0.770, P < 0.001) and distinguished TBI-CI from TBI-nonCI with AUCs of 0.82 and 0.79 (combined AUC = 0.88).

CONCLUSION

Downregulation of miR-146a-3p and miR-203a-3p and upregulation of their targets characterize post-TBI cognitive impairment, involving immune-inflammatory activation and ribosomal dysfunction, offering potential diagnostic biomarkers.