Spatial Omics

The implementation of spatial technologies advances disease research, but challenges must be overcome before spatial multi-omics becomes broadly accessible

Christina Fan at Curio Bioscience (Takara Bio)

Prior to solidifying drug candidates, scientists perform extensive research to identify disease mechanisms and drug targets using model systems. Traditional model systems such as in vivo mouse models, ex vivo tissue slice cultures and organoids are cornerstones to these early stages of the drug discovery workflow. The recent emergence of single-cell sequencing, which involves breaking down tissues from these model systems into individual cells, helps researchers delve deeper into cellular functions in specific diseases. While single-cell analysis has allowed scientists to reveal the cell types in complex tissues and uncover previously unknown or rare cell populations, these analyses can only provide a one-dimensional viewpoint. A critical dimension is missing: the spatial orientation of cells and their interactions within a microenvironment.

Spatial multi-omics (omics) offers a new window into the second dimension and is being incorporated into the discovery phase, providing scientists with a new way to amplify the utility of traditional model systems. To get two-dimensional data, researchers take one-cell-thick tissue slices from organs, tumours or organisms and analyse the positions of their molecular components. These molecular components include DNA, RNA, protein and other molecules, which can be examined individually or in combination, to determine which cells are in the sample, how those cells are behaving and interacting, and where those cells are located in relation to each other (Figure 1). Spatial omics is transforming our understanding of the cellular niche and has far-reaching applications in disease research.

Figure 1: Single-cell and spatial transcriptomic analyses of a rodent brain, with colours indicating different cell types

Spatial omics is enhancing the discovery and function of biomarkers across many domains of biology, including neuroscience, oncology and immunology. For example, the inclusion of spatial context helped researchers better understand the progression of Alzheimer’s disease, which involves multiple cell types and is linked to microenvironmental changes in the brain.1 Just last year (2024), over 200 years after the initial discovery of glioma – a type of tumour that originates in brain or spinal cord cells – spatial transcriptomics and proteomics elucidated the organisation of cells in this disease, adding to our growing understanding of the glioma ecosystem.2,3 This year (2025), the spatial omics-assisted identification of cellular neighbourhoods associated with heart disease unveiled the roles of macrophages in heart injury and repair.4

These discoveries were all made possible by an ever-growing list of spatial omics techniques.

Dissecting spatial biology tools

A range of spatial omics methods exists to examine the interactions between cells and their natural environments. These methods incorporate microscopy, mass spectrometry and next-generation sequencing (NGS). In spatial transcriptomics analysis, researchers use a subset of these spatial techniques to glean information on where genes are expressed within a tissue. One class of spatial transcriptomics methods uses fluorescence microscopy and targeted probes to visualise and quantify a pre-fixed number of gene targets per individual cell. While microscopy-based methods have been foundational in mapping gene expression across tissue sections and can attain single-molecule, sub-cellular resolution, targeted probes only allow the detection of a limited set of transcripts, and custom panels need to be created to analyse tissues from different species.5

Another class of spatial transcriptomics methods uses an array of spatially barcoded features to capture and label RNA transcripts from a tissue section. Then, sequencing and computational tools are used to identify highly detailed gene expression patterns at specific spatial coordinates. Unlike microscopy-based approaches, this method allows for higher sample throughput and analysis of the entire transcriptome (the entire suite of RNAs in a cell at a given time), which provides the most informative, unbiased measurements of gene expression. However, despite advancements in improving resolution of the spatial feature to the size of a typical single mammalian cell, these methods do not provide true single-cell resolution because the captured gene expression profile per spatial feature cannot be easily attributed to an individual cell.

Navigating challenges to widespread adoption

While these classes of spatial transcriptomics methods have allowed significant progress, issues with accessibility, sample throughput, data complexity and achieving true single-cell resolution still exist. Since each spatial transcriptomics platform’s characteristics and data types are unique, researchers need to establish different bioinformatics workflows and determine how to compare data across platforms. Researchers may also need to optimise protocols for different tissue types on their specific equipment, adding extra time to results. Low sample throughput, which primarily plagues microscopy-based techniques, delays answers since preparing, imaging and analysing each sample can take days to weeks. Both data volume and complexity from combining location information with sequencing data make analysis extremely cumbersome, especially for people without substantial bioinformatics experience.

Approaches based on microscopy or array-based transcriptome capture rely on complicated computational methods, such as cell-type deconvolution and cell segmentation, to estimate the contribution of a single cell. These inferred measurements require single-cell RNA-sequencing reference data sets that may not be applicable to the sample at hand, leading to uncertainty in data interpretation. Direct measurements, instead of inferred measurements from complex computations, are the gold standard: they greatly simplify analysis; reduce computational time and needed expertise; and identify gene expression of a single cell at a specific location with higher accuracy.

Fortunately, spatial omics is an evolving field, and more technologies are being discovered.

Breaking down barriers to progress

In a third class of spatial omics methods, single-cell sequencing is coupled with spatial nuclei tagging to achieve a direct spatial measurement of single cells. Here, a thin tissue section is placed on top of a spatially indexed DNA-barcoded surface. Instead of capturing transcripts onto spatial features, spatial DNA barcodes are donated to nuclei, and tagged nuclei are dissociated and analysed. Beyond transcriptome analysis, this technique enables researchers to perform multiple spatial omic analyses. For example, single-nuclei DNA and mutation profiling paired with spatial nuclei positioning could be used to study the organisation of clonal expansion in tumours. Immune profiling could uncover patterns of interactions between tumour-infiltrating immune cell clones and tumour cells. Assays for chromatin accessibility, DNA methylation and DNA-protein interaction can also be coupled with spatial nuclei tagging to map these omic features in 2D.

This class of methods alleviates the issues of accessibility, throughput, data complexity and single-cell resolution. Because it does not require specialised spatial omics instrumentation, it significantly reduces barriers to entry. Researchers can prepare and analyse multiple tissue sections at once, and larger sections can be covered using bigger barcoded surfaces, increasing throughput as compared to microscopy-based techniques. Spatial tagging of nuclei directly establishes location information of each single nucleus, reducing data complexity and removing the need for data inference by complex algorithms. Data is more easily interpretable because each data point represents a single cell – the ultimate quantum unit of measurement. The data type and format are the same as existing single-cell techniques, enabling the use of well-established bioinformatics analysis workflows.

While the introduction of this technology allows for major advancement in the field, there are improvements on the horizon. First, better protocols will minimise nuclei loss during dissociation, increasing cell throughput. Additionally, the introduction of new single-cell assays for multi-omics analysis will further enhance the strength of this approach.

Looking into the future of spatial omics

As innovators continue to improve upon existing methods and launch new technologies, spatial solutions are bound to become more effective, efficient and accessible. As it stands, labs need deep resources and expertise to run complicated protocols on expensive spatial platforms and to analyse complex data.

Advancements in spatial approaches that incorporate simpler workflows and take advantage of existing single-cell profiling methods will help labs more readily gain access to spatial data. With improvements in cost and workflow, spatial will move beyond discovery and atlas building, making it possible to implement in studies such as clinical trials that require a larger number of samples, or even as a diagnostic tool. As these advancements unfold, the boundaries of what spatial technologies can achieve are expanding.

The window into the second dimension is quickly becoming a door into the third and fourth dimensions. Made possible by increases in sample throughput and affordability per sample, protocols have been developed to create 3D spatial transcriptomics models from serial tissue slices processed through 2D technologies.6 One group reported the resulting gene expression signatures, spatial clusters, morphological features and cell type composition changes at inflammatory sites of rheumatoid arthritis-affected synovial biopsies.7 It is even attainable to peer into the fourth dimension by incorporating a temporal component into the experimental design. In a 4D marmoset model of multiple sclerosis, researchers recapitulated the development and repair of white matter lesions and uncovered the roles of various cell types in lesion formation and healing.8 This could be especially useful in determining how an organ responds to a treatment over time.

The continuous improvement of single-cell spatial omics methods will undoubtedly bring tremendous insights into how diseases develop and lead to the discovery of new drug targets. This is an exciting and evolving field – stay tuned!

References:

1. Ma Y et al (2024), ‘Spatial Multi-Omics in Alzheimer’s Disease: A Multi-Dimensional Approach to Understanding Pathology and Progression’, Curr Issues Mol. Biol, 46(5), 4968-4990

2. Stoyanov and Dzhenkov (2018), ‘On the Concepts and History of Glioblastoma Multiforme - Morphology, Genetics and Epigenetics’, Folia Medica, 60(1),48-66

3. Greenwald A C et al (2024), ‘Integrative spatial analysis reveals a multi-layered organization of glioblastoma’, Cell 187(10), 2485-2501.E26

4. Yang S et al (2025), ‘Functional diversity of cardiac  macrophages in health and disease’, Nat Rev Cardiol

5. Covert I et al (2023), ‘Predictive and robust gene selection  for spatial transcriptomics’, Nat. Comm,14, 2091

6. Schott M et al (2024), ‘Protocol for high-resolution 3D spatial  transcriptomics using Open-ST’, STAR Protocols, 6, 103521

7. Vickovic S (2022), ‘Three-dimensional spatial transcriptomics  uncovers cell type localizations in the human rheumatoid arthritissynovium’,Commun Biol, 5,129

8. Lin J P (2023), ‘A 4D transcriptomic map for the evolution of multiple sclerosis-like lesions in the marmoset brain’, bio R xiv2023.09.25.559371

Christina Fan is co-founder and CTO of Curio Bioscience (Takara Bio), and a pioneer in spatial, single-cell and molecular-counting technologies. Previously at Cellular Research, she invented a high-throughput single-cell transcriptome analysis technique that led to the company’s acquisition by Becton Dickinson. Christina received her PhD from Stanford University, US, is a recipient of MIT’s top 35 inventors under 35 and is a two-time recipient of Forbes Magazine 30 under 30.