Amniotic fluid cell-free RNA profiling reveals AGMAT as a biomarker for fetal renal abnormalities

ABSTRACT: Background Prenatal renal abnormalities comprise a heterogeneous spectrum of congenital kidney and urinary tract phenotypes detected on fetal ultrasonography and contribute substantially to long-term renal morbidity. Current prenatal assessment relies primarily on ultrasonography and genetic testing. However, approximately 60%-70% of fetuses with congenital anomalies remain without a definitive diagnosis. In fetuses with renal abnormalities, the diagnostic yield is even lower, with only about 13.1% receiving a confirmed molecular diagnosis. These limitations underscore the need for complementary molecular biomarkers to improve etiologic characterization in prenatal diagnosis. Amniotic fluid (AF) cell-free RNA (cfRNA) reflects contemporaneous, organ-specific fetal transcriptional activity and offers a unique opportunity for biomarker discovery in the prenatal evaluation of renal abnormalities. Methods AF cfRNA transcriptomes were profiled in fetuses with apparent ultrasound abnormalities and in propensity score-matched controls. Differential expression and functional enrichment analyses were performed, followed by weighted gene co-expression network analysis (WGCNA) to identify phenotype-associated modules. Candidate genes were prioritized by intersecting differentially expressed genes (DEGs) with the key module, and subsequently refined through feature selection using least absolute shrinkage and selection operator regression, random forest, and support vector machine recursive feature elimination. Functional validation was performed in a CRISPR/Cas9-generated agmat crispant zebrafish model. Results Analysis of 10 case-control pairs after 1:1 propensity score matching identified 546 DEGs, enriched in pathways related to renal epithelial differentiation and tubular function. Integration of WGCNA with differential expression analysis yielded 103 candidate genes, and convergent machine learning (ML) prioritized three core genes: AGMAT, GPX2, and TIMP1. Among these, AGMAT was selected for further investigation based on its kidney-enriched expression profile and limited prior characterization in renal development. In vivo, genetic perturbation of agmat in a zebrafish model recapitulated the RNA profiling results, leading to disrupted renal morphogenesis and significantly reduced early survival. Conclusions By integrating AF cfRNA transcriptomics with ML analyses, we identified AGMAT as a candidate diagnostic biomarker for prenatally detected renal abnormalities. These results position AF cfRNA as a real-time molecular complement to prenatal imaging and genetic testing and provide a framework for etiologic interpretation, risk stratification, and mechanistic interrogation of fetal renal abnormalities.

ORGANISM(S): Danio rerio Homo sapiens

PROVIDER: GSE325556 | GEO | 2026/03/25

REPOSITORIES: GEO