Explore Spatial Biology at Ultra-High Resolution Using Stereo-seq

It has long been recognized that the sub-cellular location of genes impacts the regulation and cellular function of biological systems. Since its inception in 2009, single-cell RNA sequencing (scRNA-seq, Tang F 2009), has improved the understanding of cellular diversity, from detecting novel cell types (Fu Y 2020), to uncovering complex regulatory networks (Shalek AK 2013), to identifying key processes involved with cell development and differentiation (See, P 2017). Still, these approaches can introduce cellular stress and cell death during processing, prevent analysis of some cell types, and destroy the spatial context of the original tissue which can provide information on cellular identity and function (Williams CG 2022). Thus, the ability to preserve the spatial context of biological systems will provide a deeper understanding of cellular function, as well as interactions with other cell types that coalesce to regulate tissue function (Asp M 2019), which could lead to new strategies for the prevention and treatment of several diseases, including cancer, neurological, and metabolic disorders (Williams CG 2022). Several commercially available platforms currently exist for spatially resolved transcriptomics or spatial transcriptomics (ST), which combines the positional context of cells with transcriptome data to provide an unprecedented view of spatial biology (Stark R 2019). In this post, we discuss some key areas where ST has been used and evaluate some technical considerations when implementing ST in your research. We also introduce readers to Stereo-seq, an ultra-high-resolution ST approach enabling “tissue-to-data” solutions at nanometer resolution (Chen A 2022).  

Keywords: spatial transcriptomics, gene expression, untargeted/unbiased genome-wide 

Why Study Spatial Transcriptomics? 

You may be asking, “Can I use ST in my research?” Generally, any intact tissue with viable mRNA can be used for ST, having extensively been applied to tissues originating from humans and mice (Moses L & Pachter L 2022) to deepen the understanding of many fields including developmental biology, neuroscience, immunology, and cancer (Williams CG 2022).  

• Developmental Biology: The use of human tissue is imperative in studies of developmental biology, as model systems often fail to capture the minute variations that exist between species (Asp M 2019). For example, Asp and colleagues used ST (MERFISH, definition below) and scRNA-seq to evaluate gene expression profiles genome-wide, in embryonic human hearts at three developmental stages to show global spatial gene expression patterns were established early, but also identified unique transcriptional profiles that corresponded to specific anatomical locations at each developmental stage (Asp M 2019). The data was also used to develop a publicly available web resource for future investigations in human cardiogenesis.  

• Neuroscience: ST has improved investigations in neuroscience, as some cell types are less amenable to scRNA-seq analysis (i.e. neurons in the brain) and require specialized tissue dissociation protocols during tissue harvest (Jung N & Kim TK 2023). Chen and colleagues used two ST approaches, Visium & ISS (definition below), to examine hundreds of tissue domains in situ in proximity to amyloid plaques seen in Alzheimer’s disease (AD) using an AD mouse model (Chen WT 2020). The team demonstrated elevated transcriptional profiles in 57 plaque-induced genes (PIGs) that gradually increased co-expression as the plaque load increased, which was seen across several cell types and primarily associated with genes involving inflammation, lysosomal degradation, & oxidation-reduction.   

• Immunology: The impact of human inflammatory skin diseases on systemic health remains poorly understood, including human psoriasis (PsO) which is associated with cardiovascular disease, depression, metabolic syndrome, but also psoriatic arthritis (PsA, Castillo RL 2023). Combining scRNA-seq and ST (Visium), Castillo and colleagues profiled 25 subjects that were either healthy, displayed active lesions, or had clinically uninvolved skin biopsies to show stark differences in immune microniches between the healthy & inflamed tissue, which also facilitated stratification of PsO samples vs. PsA samples by disease severity (Castillo RL 2023). The data suggests distinct molecular features between mild and severe forms of PsO that may alter the metabolic and cellular composition of unaffected skin. 

• Cancer: Understanding tumor heterogeneity can improve the diagnosis and treatment response of various cancers (Lewis SM 2021). For example, Arora R and colleagues used scRNA-seq and ST (Visium) to evaluate HPV-negative oral squamous cell carcinoma (OSCC) and revealed unique transcriptional profiles, ligand-receptor (L—R) interactions, and neighboring cellular composition between malignant cells located in the tumor core (TC) and those in the leading edge (LE) cells, where the gene signatures in LE cells was also associated with poor clinical outcomes compared to TC gene signatures (Arora R 2023).  

ST has numerous other applications for downstream analysis such as identifying cell-to-cell communication, inferring gene-gene interactions, and localizing subcellular transcripts (Moses L & Pachter L 2022). Thus, ST can improve investigations of both normal and diseased tissue that may lead to new and improved strategies in disease prevention or treatment across many health disorders (Williams CG 2022). 

ST in the Past 

The rise in ST stems from a convergence of several factors, including the decreased cost of sequencing, improvements in computing infrastructure, and the creation of more quantitative genomic datasets (Moses L & Pachter L 2022). Broadly, ST technologies can be divided into five categories: region of interest (ROI) selection, single-molecule fluorescence in situ hybridization (smFISH) techniques, in situ sequencing (ISS) approaches, next-generation sequencing (NGS) methods, and techniques that lack a priori determination of spatial location (Moses L & Pachter L 2022). Due to space constraints, methods that permit de novo reconstruction of spatial information will not be discussed. Early spatial encoding approaches utilized laser-capture microdissection (LCM) for isolation of specific ROIs within tissues, which can later be analyzed by bulk RNA-seq, dissociated into single tubes for scRNA-seq or analyzed in 3D via geographical position sequencing (Geo-seq, Peng G 2020). Microdissection-based techniques can be limited by potential mRNA degradation during harvest and achieve comparably lower spatial resolution compared to more current approaches, which have limited more widespread usage. Still, these approaches do provide a practical first step for many labs and enable unbiased ST (Moses L & Pachter L 2022). Chronologically, the next broad ST approach stems from in situ hybridization (ISH) techniques that utilize fluorescently tagged, gene-specific probes, to color transcripts with different colors that are later imaged using microscopy (Williams CG 2022), like smFISH. Later ISH techniques (e.g. seqFISH, MERFISH [multiplexed error-robust FISH]) use more sophisticated barcoding strategies over multiple rounds of hybridization to adequately encode ≥10,000 genes, sufficient to encode all genes in the human genome (Moses L & Pachter L 2022). Another image-based approach, ISS (e.g. combinatorial probe anchor ligation [cPAL]) uses either gene barcodes (for targeted) or short cDNA fragments (for untargeted) for in situ capture of mRNAs, which are typically amplified using rolling circle amplification (RCA) and directly sequenced (typically using sequencing by ligation [SBL]) inside a tissue block or section (Williams CG 2022). Both ISH & ISS based ST approaches provide high-resolution—even subcellular—ST analysis, genome-wide; however, ISH & ISS based techniques can be limited in throughput, require long imaging times, and suffer from poor signal-to-noise ratios as more genes are profiled (Williams CG 2022).  

ST Now 

ST has dramatically expanded over the past decade with numerous commercial technologies driving its growth (Moses L & Pachter L 2022). Current NGS techniques capture mRNAs in situ using poly-T oligonucleotides and encode spatial information using in situ arrays (Visium), Drop-seq-like beads (Slide-seq), or microfluidic channels (deterministic barcoding in tissue for spatial omics sequencing [DBiT-seq], Moses L & Pachter L 2022). In this way, NGS-based techniques provide untargeted, genome-wide coverage (Williams CG 2022). NGS approaches also typically use unique molecular identifiers (UMIs) which are critical in RNA-seq and scRNA-seq studies for reducing sample variance and false-discovery rates, common to these datasets (Stark R 2019, Williams CG 2022). NGS-based approaches also tend to assay larger tissue areas or can be used to process multiple sections (samples, ROIs, etc.) per slide (Williams CG 2022). A popular current platform is Visium (see examples above) which arranges spatial barcode spots in a hexagonal grid for ST at micrometer scale (Moses L & Pachter L 2022). 10x Genomics has improved the spatial resolution with Visium HD to enable true, single-cell scale resolution, using a continuous grid of 2 µm barcoded squares for whole transcriptome analysis from most tissues, including from FFPE (formalin-fixed paraffin-embedded) samples (genengnews.com). In 2022, Stereo-seq reduced the spatial resolution to nanometer scale using DNA nanoballs (DNB) deposited into wells 500 or 715 nanometers apart and preserved cellular location with spatial coordinate identity (CID) barcoding (Chen A 2022). However, NGS-based approaches can be limited as array-based methods can have low capture efficiency vs. ISH & ISS methods and tend to profile at higher resolution, where currently only Stereo-seq has shown comparable resolution (e.g. <1 µm) (Williams CG 2022). More broadly, ST can be limited by the high cost of the specialized equipment and training required to perform these experiments and is typically limited to only fresh frozen tissue, with some exceptions (e.g. Visium HD) (Moses L & Pachter L 2022). Ultimately, there is no ‘best’ ST option in practice and a trade-off exists between the quantity of genes recovered vs. the quality of analysis (Moses L & Pachter L 2022). In general ISH and ISS-based approaches are better suited for targeted hypothesis testing experiments whereas NGS-based methods (Stereo-seq, Visium) are better suited for unbiased hypothesis generation experiments (Williams CG 2022).  

Superior, end-to-end Spatial Transcriptomics with Stereo-seq 

Admera recently partnered with STOmics, an advanced multi-omics biotechnology company and developer of Stereo-seq, to become the first lab in North America to participate in STOmics’ Certified Service Provider (CSP) program. Stereo-seq (SpaTial Enhanced REsolution Omics-sequencing) combines DNB-patterned arrays and ISS to explore spatial biology of fresh frozen tissue at nanometer resolution in a centimeter-sized field of view, enabling complete STs at the tissue-, cellular-, subcellular- and even molecular-level (Chen A 2022). Sample preparation begins with tissue embedding which is later cryosectioned and mounted onto Stereo-seq chips that utilize spatially-barcoded probes to capture polyadenylated mRNAs from in situ tissue sections, which facilitates subsequent spatial gene expression mapping following sequencing. The Stereo-seq workflow can also be combined with CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing, Stoeckius M 2017) for multiomic analysis. For data processing, Admera’s bioinformatic team utilizes the Stereo-seq Analysis Workflow (SAW, Gong C 2024), a suite of computational tools to map sequence reads to their spatial location (CID), quantify the corresponding gene expression, and visualize the spatial gene expression distribution tissue-wide using the StereoMap. In partnership with STOmics, Admera provides clients with a quality ST workflow using Stereo-seq for in-depth investigations of gene expression, cell morphology, and cellular microenvironments. Check out Admera’s demo report of Stereo-seq and related scRNA-seq blog post for more information. ST has already contributed to knowledge across many fields and may lead to new strategies for the prevention and treatment of many human diseases including cancers, metabolic, neurological, infections and, other disorders (Williams CG 2022). 

References 

1. Arora, R. et al. Spatial transcriptomics reveals distinct and conserved tumor core and edge architectures that predict survival and targeted therapy response. Nature Communications (2023). https://doi.org/10.1038/s41467-023-40271-4 

2. Asp M. et al. A spatiotemporal organ-wide gene expression and cell atlas of the developing human heart. Cell (2019). https://doi.org/10.1016/j.cell.2019.11.025 

3. Castillo, R.L. et al. Spatial transcriptomics stratifies psoriatic disease severity by emergent cellular ecosystems. Sci Immunol (2023). https://doi.org/10.1126/sciimmunol.abq7991 

4. Chen, A. et al. Spatiotemporal transcriptomic atlas of mouse organogenesis using DNA nanoball-patterned arrays Cell (2022). https://doi.org/10.10.16/j.cell.2022.04.003 

5. Chen, W.T. et al. Spatial transcriptomics and in situ sequencing to study Alzheimer's Disease. Cell (2020) 

6. Fu Y. et al. Single-cell RNA sequencing identifies novel cell types in Drosophila blood. J Genet Genomics (2020). https://doi.org/10.1016/j.jgg.2020.02.004 

7. https://www.genengnews.com/topics/omics/as-10x-launches-visium-hd-will-single-cell-resolution-draw-new-customers/ 

8. Gong, G. et al. SAW: an efficient and accurate data analysis workflow for Stereo-seq spatial transcriptomics. GigaByte (2024). https://doi.org/10.46471/gigabyte.111 

9. Jung, N. & Kim, T.K. Spatial transcriptomics in neuroscience. Exp. Mol Med (2023). https://doi.org/10.1038/s12276-023-01093-y 

10. Lewis, S.M. et al. Spatial omics and multiplexed imaging to explore cancer biology. Nature Methods (2021). https://doi.org/10.1038/s41592-021-01203-6 

11. Moses, L. & Pachter, L. Museum of spatial transcriptomics. Nature Methods (2022). https://doi.org/10.1038/s41592-022-01409-2 

12. Peng, G. et al. Spatial transcriptome for the molecular annotation of lineage fates and cell identity in mid-gastrula mouse embryo. Developmental Cell (2020). https://doi.org/10.1016/j.devcel.2020.11.018 

13. See, P. et al. Mapping the human DC lineage through the integration of high-dimensional techniques. Science (2017). https://doi.org/10.1126/science.aag3009 

14. Shalek, A.K. et al. Single-cell transcriptomics revelas bimodality in expression and splcing in immune cells. Nature (2013). https://doi.org/10.1038/nature12172 

15. Stark, R et al. RNA sequencing: the teenage years. Nature Review (2019). https://doi.org/10.1038/s41576-019-0150-2 

16. Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in single cells. Nature Methods (2017). https://doi.org/10.1038/nmeth.4380 

17. Williams, C.G. et al. An introduction to spatial transcriptomics for biomedical research. Genome Medicine (2022).
    https://doi.org/11.86/s13073-022-01075-1