Spatial Multiomics: The Need for Multidimensional Analysis with Single-Cell, Spatial Context

Researchers studying the dynamics of disease and development are increasingly relying on approaches that enable spatial analysis of RNA and protein markers within tissue. Tools like spatial proteomics and spatial transcriptomics are instrumental in answering complex biological questions.

However, biological processes are driven by changes that occur simultaneously across multiple modalities. Transcriptomics and proteomics provide unique information about the cell types, states, and functions present in a tissue sample. Merging these approaches provides a more nuanced view of tissue biology – uncovering insights that might have been missed with just a single source of data.

Spatial transcriptomics enables unbiased discovery

Studying the transcriptome provides a snapshot of the RNA molecules present in a cell, tissue, or organism at a given point in time. Gene expression patterns vary depending on cell type, and by capturing transcriptomic data, we can glean valuable information about which cell types are present in a tissue sample and their functional states.

Several methods exist which allow researchers to study the transcriptome at varying levels of analysis. Bulk RNA sequencing approaches capture gene expression at the population level, while single-cell RNA sequencing facilitates gene expression profiling in individual cells to capture the heterogeneity present within a sample.

Spatial transcriptomics is an emerging approach which provides gene expression analysis with spatial context. Mapping gene expression across tissue allows us to determine the spatial organization of cell populations. Current spatial transcriptomic approaches, however, lack true single-cell resolution, relying on region of interest (ROI) or spot-based capture methods which cannot accurately capture cellular diversity.

Nevertheless, transcriptomics is a powerful tool for discovery-driven research. Profiling hundreds of thousands of cells and thousands of genes in a single experiment enables unbiased interrogation of the tissue microenvironment and makes this approach highly complementary to imaging-based spatial proteomics

Spatial proteomics captures true cellular behavior

While there is a significant correlation between mRNA and protein, studies also report considerable variation in these correlations [1]. Transcriptomic analysis does not reflect protein degradation, protein-to-protein interactions, and post-translational modifications, all of which can instead be characterized by direct protein measurements [2].

Proteomics approaches measure the proteins present in a cell, tissue, or organism at a given point in time in specific environmental conditions. As the drivers of almost all biological processes, studying protein brings us closer to a true representation of cellular behavior. Cells respond to signals and changes in their surroundings, and those changes are reflected in the proteome, which varies depending on cell type, functional state, location, and interactions with other cells.

Just as with RNA, there are tools available to study the proteome at different levels. Mass spectrometry can identify and quantify the proteins expressed in a sample at the population level.  Single-cell approaches, like flow cytometry and mass cytometry, profile protein expression in individual cells.

Imaging-based spatial proteomics, enabled by multiplex immunofluorescence (mIF), captures the spatial distribution of proteins in whole tissue sections at single-cell resolution. Using spatial proteomics approaches, we can capture the variability in protein expression and localization resulting from environmental changes and drug treatment [3].

While spatial proteomic approaches may not yet be as high-dimensional as their transcriptomic counterparts, there are tools available which can profile dozens of markers to enable discovery-driven research. The CODEX® platform can capture over 40 protein biomarkers in a single tissue section via highly multiplexed imaging.

Proteomics also provides valuable information for translational and clinical research studies. Proteins are representative of the dynamic changes that occur in healthy and diseased tissue and tend to make up the majority of drug targets. Studying protein biomarkers can thus be particularly useful in developing therapies and diagnostic tools.

Spatial multiomics integrates and visualizes transcriptomic and proteomic data

With many patients not responding to immunotherapies, a one-size-fits-all approach to patient stratification is untenable. Individuals have unique genomic, transcriptomic, and proteomic profiles – all of which can play a role in disease progression and response to treatment.

Unsurprisingly, there is growing interest in more comprehensive investigation of biological pathways at multiple levels. Proteomic and transcriptomic data have complementary features, which give researchers the exciting ability to merge information about a cell’s proteome and transcriptome with single-cell, spatial context. Integrated spatial multiomic analysis has the potential to uncover novel associations, rare cell populations, and complex markers of disease, providing a systems-biology view of the tissue microenvironment. Several integrative analysis strategies have been developed to merge multiplex imaging-based spatial proteomic data from CODEX with single-cell RNA sequencing and CITE-seq data.

GLUER (inteGrative anaLysis of mUlti-omics at single-cEll Resolution) is one such tool which integrates single-cell multiomics data and imaging data. It was developed by Dr. Kai Tan and his team from the Children’s Hospital of Philadelphia and the University of Pennsylvania. In a recent preprint, the team described how they used GLUER to merge imaging-based spatial proteomic data from CODEX with single-cell RNA sequencing data to study the spatial distribution of transcript and protein expression in murine spleen tissue.

STvEA (Spatially-resolved Transcriptomics via Epitope Anchoring) was developed by Dr. Pablo Camara’s lab at the University of Pennsylvania to overcome the limitations of CITE-seq, which enables simultaneous analysis of RNA and protein expression, but lacks spatial information. In a paper published in Science, the researchers demonstrated how STvEA can map the CITE-seq transcriptome to spatially resolved CODEX imaging data and mass cytometry data.

Dr. Will Wang, a member of the Baxter Laboratory for Stem Cell Biology at Stanford University, has also been working on bioinformatics tools to combine proteomic and transcriptomic data with the goal of studying tissue regeneration and aging. Dr. Wang and his colleagues have developed a tool which aligns CODEX imaging data with CITE-seq data to project the spatial transcriptome, model the spatial accumulation of growth factors and ligands, and predict intercellular signaling.

Interested in learning more? Check out this on-demand webinar series where Kai Tan, Pablo Camara, and Will Wang review these three approaches to integrative spatial multiomics. Plus, download this white paper on analysis frameworks for spatial multiomics to learn how you can add spatial context to your sequencing work.

References:

1. Gry, M., Rimini, R., Strömberg, S., Asplund, A., Pontén, F., Uhlén, M., & Nilsson, P. (2009). Correlations between RNA and protein expression profiles in 23 human cell lines. BMC genomics, 10(1), 1-14.
2. Slavov, N. (2020). Unpicking the proteome in single cells. Science, 367(6477), 512-513.
3. Lundberg, E., & Borner, G. H. (2019). Spatial proteomics: a powerful discovery tool for cell biology. Nature Reviews Molecular Cell Biology, 20(5), 285-302.