bioinformatics.ca
Introduction to Spatial�Single Cell RNA-seq Technologies
Trevor Pugh, PhD, FACMG
Introductory Spatial 'Omics Analysis
Feb 20-21, 2025
bioinformatics.ca
Learning Objectives
We are all made of cells: tissues consist of immune, stromal & many other cell types that interact physically and functionally
Single cell analysis is not new…the revolution is in the scale, completeness, & quantitative nature of genomic technologies
One karyotype in one cell
In situ hybridization of one transcript
Quantification of all chromsomes in all cells
Visualization of 1,000s of genes expressed in all cells
Single-cell analysis reveals heterogeneity in molecular profiles�at resolution bulk analysis may not permit
e.g. Six cells with heterogeneous
expression of three genes
Bulk analysis detects uniform
expression of all three genes
Single-cell analysis directly
measures diversity of expression
“A wide variety of single-cell methods have now been developed to measure a broad range of cellular parameters”
Stuart and Satija. Nature Reviews Genetics. 20:257-272. May 2019
We will focus here
Lineage
State
Trajectory
“Exponential scaling of single-cell RNA-seq in the past decade”
Svensson, Vento-Tormo, and Teichmann. Nat Protoc. 2018 Apr;13(4):599-604.
New spatial technologies enable additional cellular metadata describing physical distances between cell types and cell states
https://www.10xgenomics.com/in-situ-technology
Stuart and Satija. Nature Reviews Genetics. 20:257-272. May 2019
10X Genomics
Xenium
Cell organization & cell state is now measurable at all biological scales:�cells, tissues, organisms, populations
Du, Yang et al. J Transl Med. 2023; 21: 330.
https://www.10xgenomics.com/products/xenium-panels
Mouse Tissue Atlassing (379 genes)
Several strategies to localize a transcript within a cell�a cell within a tissue; and a tissue within an organism
Rao, A., Barkley, D., França, G.S. et al. Exploring tissue architecture using spatial transcriptomics. Nature 596, 211–220 (2021).
10X Genomics as an example of commercialized spatial platforms that include droplets, sequencing, and probe panels
https://omicsomics.blogspot.com/2021/02/more-details-on-10xs-sample-profiling.html
Janesick et al. bioRxiv 2022.10.06.510405 doi: https://doi.org/10.1101/2022.10.06.510405
How do you choose a spatial transcriptomic method?
Gene throughput Unbiased sequence everything versus targeted panels? | Sequence information Do you need knowledge of mutations, fusions, splice variants? | Sensitivity Targeted approaches more sensitive than sequencing-based methods |
Resolution Limitations of spotted arrays versus optical diffraction limit (subcellular) | Area size Small 6-10mm2 areas versus full-slide imaging. Trade-off with time. | Feasibility Access to tissues, imaging, panels, bioinformatics and reference data |
Summarized from Rao, A., Barkley, D., França, G.S. et al. Nature 596, 211–220 (2021).
Cancer as an example
An example: glioblastomas contain self-renewing cancer stem cells that underlie tumour initiation & therapeutic resistance
Richards, Whitley et al. Nature Cancer. 2021 Feb. Gradient of developmental and injury-response transcriptional states defines functional vulnerabilities underpinning glioblastoma heterogeneity
Brain tumour stem cell cultures derived from primary GBMs
Single cell RNA-seq of
all cells within a tumour
Single cell RNA-seq of
only the brain tumour stem cells
scVelo. Bergen et al.
Nat Biotechnol 38, 1408–1414 (2020).
0 - |
2,000 - |
4,000 - |
6,000 - |
8,000 - |
10,000 - |
. 0 | . 2,000 | . 4,000 | . 6,000 | . 8,000 | . 10,000 |
2,000 um
Glioblastoma
Grade IV
Primary – Surgical specimen
Formalin-fixed paraffin embedded tissue
Applied custom 300 gene probe panel to profile transcriptional cancer stem cell gradients in GBM
(300 gene immune panel also available)
x position (um)
y position (um)
Shamini Ayyadhury,
Gary Bader, Peter Dirks
0 - |
2,000 - |
4,000 - |
6,000 - |
8,000 - |
10,000 - |
. 0 | . 2,000 | . 4,000 | . 6,000 | . 8,000 | . 10,000 |
2,000 um
209,944 cells detected
266 transcripts
per cell
Transcripts per bin
Size (um2)
Cell Size Distribution
0
500
1,000
4k
3k
2k
1k
0
Genes
per cell
0
100
200
12k
3k
2k
1k
0
Genes
95.9% transcripts within cells
0.308 cells per 100 um2
93 genes from
266 transcripts
per cell
x position (um)
. 0 | . 2,000 | . 4,000 | . 6,000 | . 8,000 | . 10,000 |
0 - |
2,000 - |
4,000 - |
6,000 - |
8,000 - |
10,000 - |
Transcriptional cluster
2,000 um
x position (um)
Likely blood vessels
(COL1A1, COL3A1, COL4A1, PDGFRB, ACTA2, TGBFI)
Likely immune cells
(CXCL8, HLAs, RGS2, ITGAX, TLR2)
Shamini Ayyadhury,
Gary Bader, Peter Dirks
Highly specific cell populations within all types of cancer characterized over the past 3 years during rapid adoption of spatial transcriptomics
Cilento, M.A., Sweeney, C.J. & Butler, L.M J Cancer Res Clin Oncol 150, 296 (2024)
Spatial transcriptomics in cancer research and potential clinical impact: a narrative review.
Single cell RNA-seq reference sets from all parts of the brain, healthy and diseased
Piwecka, M., Rajewsky, N. & Rybak-Wolf, A. Single-cell and spatial transcriptomics: deciphering brain complexity in health and disease. Nat Rev Neurol 19, 346–362 (2023).
The Allen Brain Map is a powerful source of reference data, biological validation resources, & computational tools for brain research
https://portal.brain-map.org
Revisiting the Learning Objectives