CosMx SMI is a spatial multiomics single-cell imaging solution. It enables rapid quantification and visualization of up to 1,000 RNA and 100 validated protein analytes at single-cell and subcellular resolution in intact Formalin-Fixed Paraffin-Embedded (FFPE) and fresh frozen tissue samples. With a tunable workflow from high-plex to high-throughput, CosMx enables not only discovery research by providing the highest plex and unbiased whole-slide analysis, but also translational research with low-plex, high-throughput biology-driven analysis.
The CosMx SMI platform utilizes easy-to-use sample preparation based on standard ISH/IHC protocols and offers both custom and pre-validated assay panels for reliable results. The combination of sensitive cyclic in situ hybridization chemistry, an ultra-high-resolution imaging readout instrument, and an interactive cloud-based data analysis and visualization software makes the data simple to collect and interpret.
NanoString Technologies recently generated an open-source dataset on non-small-cell lung cancer (NSCLC) tissues, which represents the largest single-cell and subcellular analysis on FFPE samples to date. This generated spatial cell atlas data, which utilized a 960-plex gene expression assay, identified 18 different cell types, including four sub-types of T-cells, and mapped cell types across eight FFPE tissue sections. (Figure 1)
Fig 1. Discover and map cell types. The map displays 135,707 cells across a 20mm2 NSCLC tissue section. A) UMAP Projection. B) Spatially-resolved cell-type map.
The spatial information resolved by CosMx SMI allowed the tissue to be partitioned into ten distinct neighborhood clusters, or “niches”, based on the relative locations of each cell’s 200 closest neighbors. (Figure 2)
Fig 2. Phenotype the tissue microenvironment. In this organizational map of NSCLC tissue, the color denotes which “niche” the cells belong to. A) UMAP Projection. B) Spatially-resolved neighborhood cluster map.
In addition to using gene expression, cell type, and spatial feature data, CosMx data also allows researchers to characterize tumors based on the identity of and degree to which immune cells have been able to invade into the tumor itself. (Figure 3)
Fig 3. Characterize the tumor microenvironment. A) Abundance of each cell type within each tissue. B) Abundance of each niche within each tissue.
CosMx data can also be used to analyze changes in gene expression patterns in different spatial locations and has been used to show how macrophages express more SPP1 in the tumor interior and tumor-stroma boundary than in more immune cell rich environments. (Figure 4)
Fig 4. Analyze changes in gene expression with spatial context. A) A heat map for the expression of all 960 genes across all niches. B) A spatial map showing the expression of SPP1.
Finally, CosMx enables researchers to analyze how the expression of up to 100 ligand-receptor partners may be enriched between different interacting cell types. In this study, 16 were found to be significantly enriched at the interface between tumor cells and T-cells. (Figure 5)
Fig 5. View ligand-receptor expression in different cell types. A) 100 unique ligand-receptor pairs were analyzed in this study. B) 16 ligand-receptor pairs exhibited spatial significance in lung cancer tumors.
Overall, CosMx enables researchers to define cell types, cell states, cell atlas and tissue microenvironment phenotypes. CosMx complements single-cell RNA-seq data and provides the location information for each cell. It helps researchers to better understand biological process controlled by ligand-receptor interactions, identify single-cell biomarkers, and characterize gene expression networks with spatial context.