Single Cell Analysis III

Spatial transcriptomics is a technique that assays gene expression at the RNA level while preserving the spatial context of the data.
It combines spatial information with transcriptomics, allowing us to understand gene expression in relation to its location within a tissue.

Location matters in understanding cell-to-cell communication and inferring function.
Spatial transcriptomics allows us to overlay gene expression over the entire tissue.
It provides supported validation for single-cell or single-nuclear RNA sequencing data.

No tissue dissociation is required, minimizing sample preparation.
Can use frozen samples to cut sections and maintain genetic material.
Can also use formalin-fixed paraffin-embedded sections for better morphology and histology.
Retains the organization of tissue and cells in a microenvironment.
Identifies cell types in a heterogeneous tissue and their gene expression.

Visium spatial slide has four squares, each called a fiducial frame.
Each fiducial frame contains spots or dots that are similar to the gems in single-cell RNA sequencing.
These spots contain spatial barcodes and unique molecular identifiers (UMIs).
Tissue sections are placed on top of these spots.

Capture spots contain primers with a barcode and a unique molecular identifier (UMI).
Tissue is placed on top of the dots, and genetic material interacts with the primers when permeabilized.
cDNA synthesis occurs, and then the tissue is washed off.
Dots are washed off and captured for library sequencing.

Outputs are similar to single-cell RNA sequencing but with spatial localization.
The image of the tissue section is maintained, and spots are superimposed on it.
Data is processed through a pipeline similar to single-cell RNA sequencing.

Read depth is not comparable to single-cell RNA sequencing because it involves a 10-micrometer section of cells.
Dot size and coverage can be an issue; the dot size may cover multiple cells, leading to genetic material from multiple cells in one go.

Deconvolution can be used if a single-cell RNA sequencing reference is available.
It is recommended to run single-cell or single-nuclear RNA sequencing with the same tissue to enhance spatial transcriptomics data.

Each dot can have 1 to 7 overlapping cells.
A reference single-cell or single-nuclear RNA sequencing dataset is used for deconvolution.
Deconvolution identifies cell types within each dot, providing an idea of cell localization and relevant gene expression.

Spatial transcriptomics is enhanced when combined with single-cell RNA sequencing data.
Single-cell RNA sequencing data provides a higher resolution reference.
Pipelines are used to compare single-cell data with spatial transcriptomics data using "anchors."
High scoring correspondence and low scoring correspondence are obtained to confidently identify cells of interest.
Unique gene signatures in the area can be identified due to the sequencing depth of the single-cell reference dataset.

Integration of single-cell data with spatial transcriptomics of the aorta.
Sections of healthy and hypertensive disease aortas were analyzed.
Fibro one population was identified in a diseased aorta but not in a healthy aorta.

Top 8 genes of the fibro one cell type were analyzed.
Healthy vessels showed minimal gene expression related to the top 10 genes.
Diseased vessels showed a dramatic increase in expression levels.
Localization was observed inside the vessel wall, but not predominantly on the outside as expected.

Unique cellular signatures can be identified.
Genes associated with the novel cell type (fibro one) were not expressed in healthy vessels but were expressed throughout the diseased vessel.
A fibrosis module (genes associated with fibrosis, including extracellular matrix genes) was created.
Healthy vessels showed low expression of the fibrosis module, while diseased vessels showed high expression.
Areas with high expression of the CTHRC1 module were identified.
The study suggests that these cells are associated with locations of fibrosis and may contribute to fibrotic disorders.

Strengths:
- Allows localization of genes on a single section with a certain limitation on resolution.
- Transcriptomic approach that covers reference genomes and single-cell gene analyses.
- Allows analysis of multiple genes in one section, unlike in situ hybridization.
- Combining spatial data with single-cell RNA sequencing enhances insights and provides spatial validation of single-cell hypotheses.
Limitations:
- Dot size limits resolution (1 to 7 cells per dot).
- Requires precision histology and careful sample preparation.
- Outputs are limited without a single-cell RNA reference.
- High cost.

10X Genomics has released Xenium or Visium HD, which allows much greater resolution (1 to 3 or 1 to 2 cells per dot).

The cost of the slide itself is significant.
Sequencing data from the slide must be run, and single-cell or single-nuclear sequencing is often run in parallel.
This more than doubles the transcriptomics price tag.

Spatial transcriptomics is a useful technique for sequencing at a tissue and cellular level.
Key steps include using the Visium slide and capture dots.
It is important to understand the strengths and limitations of spatial transcriptomics and the benefits of integrating a single-cell RNA sequencing reference database.