Spatial Transcriptomics to define transcriptional patterns of zonation and structural components in the liver

Franziska Hildebrandt, Alma Andersson, Sami Saarenpää, Ludvig Larsson, Noémi Van Hul, Sachie Kanatani, J. Mašek, Ewa Ellis, Antonio Barragán, Annelie Mollbrink, Emma Andersson, Joakim Lundeberg, Johan AnkarklevbioRxiv 2021 · DOI

TL;DR

This paper uses spatial transcriptomics to map gene expression in the liver at a high resolution, revealing how transcriptional patterns define liver zonation and identify previously uncharacterized structures. This approach overcomes limitations of single-cell RNA sequencing by preserving spatial context, offering a powerful tool for understanding liver function and disease.

What Problem Does This Paper Address?

The liver is a complex organ with specialized functions distributed spatially within its lobules, a phenomenon known as zonation. While single-cell RNA sequencing (scRNA-seq) has provided insights into liver cell diversity, it often involves tissue dissociation, which can alter gene expression and lose crucial spatial information. This loss of spatial context makes it difficult to understand how transcriptional differences contribute to the overall heterogeneity of the liver and to map these differences to specific anatomical structures like lobules. The authors aim to address this by using spatial transcriptomics (ST) to map gene expression directly within intact liver tissue, allowing for a more accurate understanding of liver zonation and the identification of novel spatial structures.

Methods

The study employs Spatial Transcriptomics (ST), a technique that allows for gene expression profiling directly on tissue sections while preserving spatial information. Key methodological contributions include:

  • High-resolution ST: They generated ST data from 8 mouse liver sections, capturing expression from 19,017 genes across 4,863 individual capture locations (spots).
  • Unsupervised Clustering: They used a graph-based approach to cluster spots based on their gene expression profiles, identifying 6 distinct clusters.
  • Computational Annotation: They developed novel computational approaches to:
  • Spatial Annotation: Aligning gene expression clusters with histological images to annotate regions like portal and central veins.
  • Expression Gradients: Measuring transcriptional gradients along the lobular axis by analyzing gene expression as a function of distance to different vein types (portal vs. central).
  • Vein Type Prediction: Building a computational model to predict vein types based on the gene expression profiles of neighboring spots, complementing manual histological annotation.
  • Single-Cell Data Integration: They integrated their ST data with publicly available scRNA-seq data from the Mouse Cell Atlas (MCA) using the Stereoscope tool. Stereoscope models both single-cell and spatial data to deconvolve gene expression in each spot into cell-type proportions, allowing for validation and comparison.
  • Pathway Analysis: Gene Ontology (GO) enrichment analysis was performed on differentially expressed genes within clusters to identify enriched biological processes.

Datasets

  • Mouse Liver Tissue: 8 sections of wild-type adult mouse livers (caudate or right lobe) were used for spatial transcriptomics.
  • Mouse Cell Atlas (MCA): Publicly available scRNA-seq data from mouse liver cells was used for comparative analysis and integration with ST data.
  • Single-cell spatial reconstruction data: Another dataset was extracted for comparative analysis of zonation layers.

Key Insights and Ideas

  • Zonation is the Dominant Factor: The study strongly supports the concept that lobular zonation is the primary driver of spatial heterogeneity in the liver transcriptome.
  • Mapping Vein Structures: The ST data allowed for the precise mapping and computational annotation of portal and central veins, crucial for understanding zonation.
  • Novel Structures Identified: The analysis revealed transcriptionally distinct structures within the liver that were previously uncharacterized, suggesting a more complex spatial organization than previously understood.
  • Complementarity of ST and scRNA-seq: The integration of ST with scRNA-seq data highlights the power of combining spatial context with single-cell resolution, while also revealing the limitations of scRNA-seq alone for understanding tissue architecture.
  • Computational Tools for Annotation: The development of computational methods for vein annotation and gradient analysis offers a valuable resource for future studies.

Key Results

  • 6 Clusters Identified: Unsupervised clustering of the ST data revealed 6 distinct spatial expression clusters.
  • Cluster 1 & 2: These clusters were associated with periportal and pericentral regions, respectively, aligning with known liver zonation.
  • Vein Annotation: Computational annotation of veins based on gene expression showed high concordance with visual histological annotations.
  • Expression Gradients: The study successfully mapped expression gradients of zonation markers along the lobular axis, providing quantitative measures of spatial gene expression changes.
  • Cluster 5: This cluster showed enrichment for genes involved in "collagen fibril organization" and "response to cytokine," suggesting it represents a distinct structural component, possibly related to Glisson's capsule or mesenchyme.
  • Stereoscope Integration: Integration with MCA data confirmed that hepatocytes are the predominant cell type and that their spatial segregation aligns with zonation, but also highlighted that ST can reveal structures missed by scRNA-seq alone.

Limitations and Future Work

  • Resolution Limits: While ST provides high resolution, it still captures a mixture of cells within each spot. The authors note that rare cell types might be missed or their transcriptional profiles masked by more abundant cells, especially in scRNA-seq integration.
  • Tissue Dissociation Artifacts: The paper acknowledges that previous studies relying on tissue dissociation for sequencing may have introduced transcriptional artifacts. ST overcomes this by preserving spatial context.
  • Proposed Structures: The identification of cluster 5 as a potentially novel structure (e.g., Glisson's capsule) is based on initial indications and requires further functional validation.
  • Future Directions: The authors suggest that future advances in ST resolution will enable more detailed investigations of rare cell types. They also propose that ST is a valuable resource for studying liver development, immunity, and pathology, and that the developed computational tools can be applied to various annotation-based applications. The potential to study infection and inflammation effects on immune zonation is also highlighted.

Summary updated

Abstract

ABSTRACT Reconstruction of heterogeneity through single-cell transcriptional profiling has greatly advanced our understanding of the spatial liver transcriptome in recent years. However, global transcriptional differences across lobular units remain elusive in physical space. Here, we implement Spatial Transcriptomics to perform transcriptomic analysis across sectioned liver tissue. We confirm that the heterogeneity in this complex tissue is predominantly determined by lobular zonation. By introducing novel computational approaches, we enable transcriptional gradient measurements between tissue structures, including several lobules in a variety of orientations. Further, our data suggests the presence of previously transcriptionally uncharacterized structures within liver tissue, contributing to the overall spatial heterogeneity of the organ. This study demonstrates how comprehensive spatial transcriptomic technologies can be used to delineate extensive spatial gene expression patterns in the liver, indicating its future impact for studies of liver function, development and regeneration as well as its potential in pre-clinical and clinical pathology.
BibTeX
@misc{hildebrandt2021spatial,
  author = {Franziska Hildebrandt and Alma Andersson and Sami Saarenpää and Ludvig Larsson and Noémi Van Hul and Sachie Kanatani and J. Mašek and Ewa Ellis and Antonio Barragán and Annelie Mollbrink and Emma Andersson and Joakim Lundeberg and Johan Ankarklev},
  title = {{Spatial Transcriptomics to define transcriptional patterns of zonation and structural components in the liver}},
  year = {2021},
  doi = {10.1101/2021.01.11.426100},
  howpublished = {bioRxiv},
}
Figure