Decoding the role of H19 in cholestatic liver injury using snRNA-seq, spatial transcriptomics, and machine learning-based disease prediction.

backgroundDespite recent advances, Primary Sclerosing Cholangitis (PSC)-a chronic obstructive biliary disease-still lacks effective therapies to prevent disease progression or the need for liver transplantation. Moreover, up to 30% of transplant recipients experience recurrence. Long non-coding RNA H19 (H19) has been implicated in promoting PSC progression, yet the cellular and molecular mechanisms underlying its pathogenic role remain incompletely understood.

resultsLiver tissues from age- and sex-matched wild type (WT), H19 knockout (H19KO), Mdr2 knockout (Mdr2KO), and double-knockout (DKO; Mdr2KO/H19KO) mice were analyzed using single-nucleus RNA sequencing (snRNAseq) and GeoMx spatial transcriptomics to define the cell type and spatially specific effects of H19 deletion in cholestatic liver injury. Machine learning models were built to develop cell-type-specific gene prediction signatures and cross-validated using the human dataset GSE243981. A disease-associated cholangiocyte subcluster that increased in Mdr2KO but was markedly reduced in DKO mice was identified. SPP1 signaling was significantly dysregulated in cholestatic liver injury and mitigated following H19 deletion. Translationally conserved healthy (Clu, Spp1) and diseased (Csmd1, Slco3a1, Cftr) cholangiocyte markers were identified. When validated in a human patient dataset (GSE243981), our machine-learning prediction models achieved AUC values > 0.869. Finally, spatial analyses demonstrated that the mitigation of disease-associated gene expression following H19 deletion was specifically restricted to hepatocytes within the bile duct region.

conclusionsH19 deletion mitigates cholestatic injury by suppressing pathogenic cholangiocyte states, normalizing Spp1-mediated signaling, and shifting transcriptional programs specifically within the periductal niche. Furthermore, our machine-learning signatures demonstrate robust cross-species translation and may benefit future post-transplant analytics and disease recurrence predictions.