Related Experiment Video
Updated: Aug 5, 2026
Mining Spatial Transcriptomics Datasets using DeepSpaceDB
Published on: September 5, 2025
Spatial transcriptomics supports a role for SOX4-driven signaling throughout the disease course of biliary atresia
Ioannis A Ziogas1, Katie R Conover2, Evgenia Dobrinskikh2
1Department of Surgery, University of Colorado School of Medicine, Aurora, Colorado, USA.
Insights
Biliary atresia (BA) is a serious infant liver disease. Children with worse BA outcomes show higher SOX4 gene expression, leading to liver cell changes and immune scarring, indicating a potential driver of disease progression.
Area of Science:
- Pediatric Gastroenterology and Hepatology
- Molecular Biology
- Translational Medicine
Background:
- Biliary atresia (BA) is a neonatal fibroinflammatory cholangiopathy and a leading cause of pediatric liver transplantation.
- Understanding molecular mechanisms driving BA progression is crucial for improving outcomes.
Purpose of the Study:
- To define molecular mechanisms underlying differential disease progression rates in pediatric biliary atresia.
- To investigate the role of SOX4 in hepatocyte-cholangiocyte reprogramming in BA.
Main Methods:
- Spatial transcriptomics (ST) analysis of liver tissue from BA patients with varying disease severity and controls.
- Comparison of transcriptional signatures by tissue region (scar, hepatocyte, cholangiocyte).
- Validation using immunohistochemistry and in situ mRNA hybridization in larger cohorts, including samples at diagnosis.
Main Results:
- Patients with the most aggressive BA phenotype (SNL <2 years) exhibited reduced hepatocyte zonation and increased scar heterogeneity.
- Enrichment of SOX4-associated genes involved in hepatocyte-to-cholangiocyte reprogramming was observed in aggressive BA.
- Increased SOX4 expression was detected in BA patients at diagnosis, particularly those with worse prognoses.
Conclusions:
- Worse outcomes in pediatric BA correlate with elevated SOX4 expression at diagnosis, loss of hepatocyte zonation, and immune-driven scar heterogeneity.
- SOX4-mediated biliary reprogramming may contribute to maladaptive repair processes in BA, warranting further investigation.
Background:
Biliary atresia (BA) is a neonatal fibroinflammatory cholangiopathy of infancy and the most common indication for pediatric liver transplantation. We aimed to define the molecular mechanisms responsible for differences in the rate of disease progression among children with BA.
Methods:
We performed spatial transcriptomics (ST) analysis on frozen liver tissue at transplant from 14 children: BA with survival with native liver (SNL) <2 years (BA1, n=3), BA with SNL >2 years (BA2, n=4), non-BA cholestasis (n=4), and non-diseased donors (n=3). Transcriptional signatures were compared between patient groups by tissue region (scar, hepatocyte, cholangiocyte). Findings were validated in larger patient cohorts that included BA samples at diagnosis.
Results:
ST analysis of patients with BA1 showed the most aggressive disease phenotype, characterized by reduced hepatocyte zonation, low expression of homeostatic metabolic signatures, and increased scar heterogeneity enriched for pathways including extracellular matrix remodeling, interferon response, and leukocyte activation. Notably, genes involved in SOX4 hepatocyte-to-cholangiocyte reprogramming were most enriched in patients with BA1. Liver immunohistochemistry with in situ mRNA hybridization showed that patients with BA at diagnosis had increased SOX4 quantification as compared with patients with BA at transplant. Lastly, previously published liver bulk RNA-sequencing data demonstrated higher SOX4 gene-set expression in patients with BA at diagnosis with SNL <2 years.
Conclusions:
Children with BA and worse outcomes exhibit increased SOX4 gene-set expression at diagnosis with greater loss of hepatocyte zonation and immune-driven scar heterogeneity at transplant. Further mechanistic studies are needed to determine whether SOX4-associated biliary reprogramming contributes to maladaptive reparative processes in BA.
