Studying how cells interact within their native tissue environment is fundamental to understanding both healthy physiology and disease processes. However, until recently, it was challenging to profile all the genes a cell expresses while also preserving the precise location of those transcripts in tissues. Traditional single-cell RNA sequencing (scRNA-seq) methods lose spatial information, and prior spatial transcriptomics techniques have been limited by low plexity and/or regional resolution. Now, we have developed a new solution that brings together subcellular resolution, comprehensive coverage of the protein-coding human genome, and a straightforward workflow compatible with standard archival samples.

Our recent pre-print in bioRxiv describes the CosMx Whole Transcriptome (WTX) assay, demonstrating robust subcellular imaging of essentially the entire human protein-coding transcriptome (>18,000 targets) in formalin-fixed, paraffin-embedded (FFPE) tissues. By combining ultra-high-plex barcoded in situ hybridization (ISH) probes with advanced imaging and computational decoding, our new approach can identify and count transcripts from thousands to millions of cells on a single slide while retaining each transcript's exact x- and y-coordinates.

Going Beyond the Limits of Single-Cell RNA Sequencing

Single-cell RNA sequencing has revolutionized biomedical research by allowing us to explore and dissect cellular heterogeneity. Unfortunately, scRNA-seq requires enzymatically dissociating a tissue into individual cells, which can inadvertently lose fragile or tightly packed cells and always sacrifices the tissue's spatial organization. By contrast, the CosMx Spatial Molecular Imager (SMI) workflow keeps tissue architecture intact, analyzing every cell in its native place, enabling researchers to explore cell-to-cell interactions, microenvironmental gradients, and tissue anatomy without losing transcriptomic depth.

In our side-by-side comparison of matched FFPE colorectal carcinoma samples, we found that scRNA-seq underrepresented certain cell types, particularly those that were large or more difficult to dissociate. In contrast, the CosMx WTX panel revealed a richer cellular composition, capturing epithelial, stromal, immune, and tumor cells in their native positions. This underscores one of the most impactful benefits of spatial methods: every cell stays put and is counted.

Comparison of identified cell composition from scRNA-seq vs CosMx WTX.

A Broad Toolbox for Complex Biology

The new preprint demonstrates the CosMx WTX assay's ability to detect thousands of transcripts per cell across multiple FFPE tissue types, including colon, pancreas, breast, brain, kidney, and skin. Uniform and reproducible data quality was observed, with four logs of dynamic range and a high correlation to bulk RNA-seq for genes above a practical detection threshold.

Whole transcriptome spatial imaging from six different FFPE tissues.

Powering Whole Transcriptome Coverage with Subcellular Resolution

The WTX workflow uses barcoded probes and a robust image-based readout to map each gene's expression down to subcellular coordinates. The AtoMx® Spatial Informatics Platform (SIP) was then used for streamlined data analysis, processing billions of transcripts across millions of cells. By harnessing the scalability and power of cloud computing and machine-learning cell segmentation algorithms, AtoMx SIP supported the rapid analysis of multi-sample experiments and enabled us to examine subcellular spatial patterns without the need for extensive coding. This informatics approach helped produce the rich, whole-transcriptome maps that can enable advanced analyses including pinpointing cell type-specific pathways, finding rare cell populations, and visualizing ligand-receptor interactions within distinct tissue niches.

High assay reproducibility

Strong assay reproducibility was demonstrated by running three independent CosMx WTX experiments on a 36-cell-line FFPE array and comparing the results across different imaging scales. Regardless of number of fields of view, all three runs maintained high consistency, with correlation coefficients around 0.97, indicating that scaling up the imaging area does not compromise the reproducibility of WTX. The study also compared pseudo-bulk WTX profiles to Cancer Cell Line Encyclopedia (CCLE) bulk RNA-seq data and found strong concordance above the detection threshold, indicating that the WTX assay can reliably quantify gene expression over four orders of magnitude.

Three replicate CosMx runs on different imaging scales demonstrating consistent transcript totals.

Illuminating Tumor Complexity

One highlight in the paper focuses on colon adenocarcinoma, where we combined CosMx WTX data with pathway analyses. We discovered increased activity of pathways such as EGFR, TGF-beta, and VEGF within the tumor nest, coupled with decreased expression of certain apoptotic signals (e.g., TRAIL). This reflects an aggressive tumor phenotype, consistent with pro-survival and pro-angiogenesis signals. By localizing these expression patterns at a single-cell scale, the study showed how spatial context reveals distinct "pockets" of tumor microenvironment heterogeneity, which is key information that is impossible to detect by methods requiring cell dissociation. Additionally, monitoring all biochemical pathways at once with the WTX panel enables a completely novel "pathways-first" approach to spatial biology research.

Pathway and ligand-receptor analyses revealing intratumor heterogeneity in colon cancer.

Conclusion

The findings in this study point toward a future in which whole-transcriptome spatial profiling can become the go-to approach for discovery research, bridging the previous gap between comprehensive gene coverage and intact tissue context. As demonstrated in our pre-print, CosMx WTX eliminates the need to choose between high-plex coverage and subcellular resolution.

Further Reading

• Sub-cellular Imaging of the Entire Protein-Coding Human Transcriptome (18933-plex) on FFPE Tissue Using Spatial Molecular Imaging
• Spatial biology: What it is, why it matters, and how to access it
• Tips for successful CosMx SMI single-cell spatial runs at 1000-plex