Multimodal decoding of the human and mouse mesenchymal lineage in biliary fibrosis

Abstract

Fibrosis is known as the final common endpoint of chronic inflammation and failure of tissue repair. In the liver, this endpoint of fibrosis and scarring throughout the organ is unresolvable cirrhosis contributing to decompensation and the loss of organ function. In the human liver there are many aetiological drivers of injury and disease of which fibrosis is a consequence, Primary biliary cholangitis (PBC) is a chronic autoimmune liver disease characterised by progressive destruction of the small bile ducts within the liver. The estimated incidence of PBC is approximately 1 to 2 cases per 100,000 individuals per year. It primarily affects middle-aged women, with a female-to-male ratio of around 9:1. However, PBC can occur in individuals of any age, including children and older adults. The primary treatment is the use of ursodeoxycholic acid (UDCA), a medication that helps improve liver function and reduce the inflammation and damage to the bile ducts. However, the only solution to end-stage PBC is transplantation and suitable livers are in short supply. There is a significant clinical need to identify anti-fibrotic therapies to effectively treat disease.

Single-cell RNA sequencing (scRNA-seq) is a transcriptomic technique that enables profiling of the gene expression of individual cells. Previously, researchers could understand the changes in gene expression in tissue but not the cells contributing to the change. This advance in resolution allows for the identification of disease specific cell states and populations and can contribute to novel hypotheses as to the complex cellular networks and signalling pathways involved. Laterally, single nuclei RNA (snRNA-seq) sequencing has been developed allowing the use of frozen archival tissue and that of heavily fibrotic tissue to liberate all cell types.

I hypothesised that single-cell transcriptomics could be leveraged to identify disease-specific novel cell states and sub-populations driving fibrosis in PBC and that this would lead to new targets for potential antifibrotic interventions.

I generated a single nuclei RNA sequencing atlas of all the cells in the end-stage diseased PBC liver. Using this I could identify the mesenchymal lineage of cells. The mesenchymal lineage plays a significant role in tissue development, wound healing, and the pathogenesis of fibrosis. When compared to healthy liver tissue analyses identified the enrichment of a population of cells called Portal Fibroblasts (PF). PFs reside adjacent to the bile duct and are thought to respond to the initial challenge to the duct. As disease progresses to a chronic state it was believed that Hepatic Stellate cells (HSC) were recruited to the injury site and became the predominant contributor to fibrosis. As well as the enrichment of the PF population it was also clear that a subpopulation of PFs (marked by FAP, LTBP2, DCLK1, VCAN) were highly enriched for COL1A1 expression. COL1A1 is a major component of collagen, itself a fundamental part of fibrotic tissue. Furthermore, the transcriptomic data identified two further transcriptionally distinct subpopulations of PFs (marked by GFPT2 and SCN7A). These three populations of PF could be spatially validated and identified within the fibrotic tissue by the use of immunofluorescence and RNAscope (in situ hybridisation).

My second aim was to identify corollary populations of mesenchymal cells in a mouse model of biliary injury with a specific interest in the PFs identified in human disease. To do this I generated two datasets with which to interrogate the murine mesenchymal lineage. Firstly, a snRNA-seq dataset of four timepoints progressing through injury. From this I could, once again, subset the mesenchymal lineage for analysis. Secondly, a scRNA-seq data set of the mesenchymal lineage isolated from Pdgfrb-GFP mouse at day twenty one, peak induction, of a dietarily induced model of biliary fibrosis. These datasets showed that there were broadly transcriptionally similar cells present in the mouse model with correlation by gene signature and single markers. However, not all markers were expressed suggesting a degree of evolutionary divergence or ongoing transcriptional changes.

Finally, I attempted to lineage trace and knock-out several markers of interest. Lineage tracing is a technique used to track the progeny of specific cells or cell populations over time. I used the Pdgfra-eGFP, Lumᶜʳᵉ;Ai14 and Cthrc1ᶜʳᵉᴱᴿtdTomˡᵒˣᴾ transgenic mice subjected to a model of biliary injury as well as the Cthrc1ᶜʳᵉᴱᴿDTAˡᵒˣᴾ transgenic mouse to deplete the population. These models demonstrated no reporting or phenotype in the liver supporting an ongoing hypothesis that the PF subpopulation markers identified in the human end-stage PBC are not present in the mouse liver (LUM, CTHRC1). Together the findings and datas generated from this project continue to support the power of single-cell transcriptomics to identify previously unknown cell types and states in human and mouse models of liver disease. The conclusions achieved here open fresh discussion into the markers we use to identify mesenchymal cell in the liver and how we consider their behaviour during chronic disease.