Blog
Oct 1, 2024 / Oncology / Neuroscience / Developmental Biology

High-plex spatial RNA: Adding an entire new dimension to your imaging studies

Josh Azevedo

mRNA profiling and high-content imaging were the two technologies that were most likely to be listed as game-changers by researchers in a 2020 survey. Now, massively multiplexed spatial RNA profiling—e.g., spatial transcriptomics—combines these two technologies and amplifies their strengths.

We wrote this blog not just to tell you how this technology will benefit your research, but to show you. Look at the image below:

Figure 1. FFPE section of whole mouse pup analyzed on Xenium with the Mouse Tissue Atlassing Panel.
Figure 1. FFPE section of whole mouse pup analyzed on Xenium with the Mouse Tissue Atlassing Panel.

That’s not immunofluorescence—that’s RNA analyzed with Xenium In Situ. Specifically, it’s 177 million transcripts from 1.4 million cells visualized across an entire single tissue section, combined to identify specific cell types (denoted by color) that then reveal specific tissues and structures. Now let’s take a closer look:

Figure 2. FFPE mouse colon analyzed using the Xenium Prime 5K assay with multimodal cell segmentation.
Figure 2. FFPE mouse colon analyzed using the Xenium Prime 5K assay with multimodal cell segmentation.

Each of those spots is an individual RNA transcript. Combine this with the ability to analyze hundreds to thousands of genes simultaneously with cell segmentation and subcellular resolution, and you have the ability to tease apart not just individual cells, but how they’re communicating with their neighbors, how specific cellular neighborhoods are reacting in health and disease, and so much more. 

Read on to see how Xenium provides deeper insights into each of your tissue sections with:

  • Greater characterization of cell types and states
  • Deeper understanding of how cells interact and communicate
  • More comprehensive data from your samples, without sacrificing your current analytes

High-plex spatial gene expression unlocks greater insights into cell types and states

High-plex RNA detection provides a clearer picture of cell types and states with hundreds to thousands of gene measurements per cell, helping you find the best markers for your cell types of interest to complement and augment your existing tools.

Cells are defined by the specific complement of genes they express. In the past, researchers typically relied on single (or low-plex) markers to characterize entire populations. However, this approach carries several risks, including the possibility of choosing suboptimal marker(s) for your cells of interest or combining heterogeneous cells together based on a single marker. 

For example, GFAP is a canonical marker of astrocytes, but astrocytes split into multiple populations in disease and inflammation (1). Single or low-plex protein markers may not resolve all these cell states, especially if you don’t know the markers for each cell (sub)type ahead of time. But high-content imaging with massively multiplexed RNA—which offers more simultaneous measurements of genes than proteomics approaches—can resolve those cell states (Figure 3).

Figure 3. In-depth characterization of cell types in a section of fresh frozen human lung cancer tissue using the Xenium Human Lung Panel. A. Cell types (denoted by individual colors) in their native spatial context plotted across a sample of lung cancer (lung adenocarcinoma in situ). B. Clustering and labeling of the cell types described in panel A. Image adapted from Figure 6I of Haga et al. 2023. (CC BY 4.0).
Figure 3. In-depth characterization of cell types in a section of fresh frozen human lung cancer tissue using the Xenium Human Lung Panel. A. Cell types (denoted by individual colors) in their native spatial context plotted across a sample of lung cancer (lung adenocarcinoma in situ). B. Clustering and labeling of the cell types described in panel A. Image adapted from Figure 6I of Haga et al. 2023. (CC BY 4.0).

What you see in Figure 3 is a lung cancer section from Haga et al. (2). The Xenium Human Lung Panel let researchers visualize dozens of transcripts in each cell, then use this multiplexed gene expression data to “bin” similar cells together. The researchers gave each cell type a unique color and plotted their localization across the tissue (Figure 3A). They were also able to see how different cell types clustered together (Figure 3B). This can be valuable in mapping cellular neighborhoods across tissue sections, but multiplexed RNA lets you go deeper (Figure 4).

Figure 4. High-resolution visualization of cell subtypes in lung cancer using Xenium. A. Subsection of the same lung cancer sample from Figure 2, showing spatial distribution and colocalization of various lung macrophage and immune markers. B. Subsection of tissue from Figure 2B, showing differential expression of various immune markers in lung macrophages (MARCO+ cells). Image adapted from Figure 6J of Haga et al. 2023. (CC BY 4.0).
Figure 4. High-resolution visualization of cell subtypes in lung cancer using Xenium. A. Subsection of the same lung cancer sample from Figure 2, showing spatial distribution and colocalization of various lung macrophage and immune markers. B. Subsection of tissue from Figure 2B, showing differential expression of various immune markers in lung macrophages (MARCO+ cells). Image adapted from Figure 6J of Haga et al. 2023. (CC BY 4.0).

Note: The cell segmentation used in Figure 4 represents an early version of segmentation on Xenium. The preferred cell segmentation approach now uses multiple morphological stains for the boundary and interior and leverages a purpose-built algorithm for accurate segmentation. More information on the current Xenium cell segmentation approach can be found here.

In the same tissue section, the researchers demonstrated how higher-plex RNA can better discriminate subpopulations of cells by using a marker of lung macrophages—MARCO (yellow)—as an example. Using their rich dataset, they were able to characterize, de novo, a variety of different MARCO+ lung macrophages in their tissue. Notably, two of these subtypes (FABP4+and SPP1+) had potential clinical significance, given they, “...affect fibroblasts in the vicinity and may cause alveolar collapse,” a more macro-level change linked to lung structure and function.

Expanding this approach to benefits for your own work: high-plex spatial RNA analyses can let you see more genes, giving you more points of data to identify cell types, states, and locations in your tissue. This can serve as an endpoint of your study by shedding light on the underlying biology of your tissue, or by helping you select better complements of markers for more focused proteomics studies.

High-plex RNA allows deeper characterization of how cells connect and communicate

RNA helps resolve complex biology with a richer characterization of what’s going on at the molecular, cellular, and tissue levels.

Cells that touch, talk. Understanding not just what’s going on in an individual cell, but in its neighbors—and how they all communicate as a unit—is critical to capturing the bigger picture of the underlying biology of your tissue.

While legacy methods permit assaying several markers to perform neighborhood and ligand–receptor analyses, high-plex single cell spatial RNA analyses help expand this by adding orders of magnitude more markers (Figure 5).

Figure 5. Xenium analysis of multiple sclerosis lesions in spinal cord. A. Spatial distribution of various healthy and disease-associated (DA) glia across the spinal cord. B. Cell–cell interaction networks for each cell in multiple sclerosis lesion “compartments” in the spinal cord. Image adapted from Figure 4A and Figure 5C of Kukanja et al. 2024. (CC BY 4.0).
Figure 5. Xenium analysis of multiple sclerosis lesions in spinal cord. A. Spatial distribution of various healthy and disease-associated (DA) glia across the spinal cord. B. Cell–cell interaction networks for each cell in multiple sclerosis lesion “compartments” in the spinal cord. Image adapted from Figure 4A and Figure 5C of Kukanja et al. 2024. (CC BY 4.0).

In this 2024 publication, researchers leveraged high-plex spatial RNA analysis to examine multiple sclerosis lesions in the spinal cord (1). First, they characterized cell states across the spinal cord (Figure 5A) and identified disease-associated (DA) glia. Next, they used both cell proximity and gene expression to create interaction networks for each cell in lesion “compartments” (Figure 5B). Ligand–receptor analyses in their findings gave rise to the hypothesis that DA microglia were regulated by DA oligodendrocytes and contributed to the dysregulation of these (and other) cell types in multiple sclerosis.

A second example came from Fynn Biotechnologies, where researchers compared immunotherapy responder and non-responder breast cancer patients using Xenium (3). With the several hundred markers afforded by the Xenium Human Breast panel, the researchers were able to characterize a multitude of cell types. Among these cell types were two populations, CTLA4+/CD8+ effector T cells and PD-L1+ macrophages, that only colocalized with breast cancer cells in responder patients (Figure 6).

Figure 6. Cellular colocalization reveals responder-specific immune cell interactions with cancer cells. Xenium analysis of various breast cancer tissues show that specific immune cell types (red, pink) only colocalize with tumor cells (blue) in patients that responded to immunotherapy treatment. Image adapted from Figure 5C of Wang et al. 2024. (CC BY 4.0).
Figure 6. Cellular colocalization reveals responder-specific immune cell interactions with cancer cells. Xenium analysis of various breast cancer tissues show that specific immune cell types (red, pink) only colocalize with tumor cells (blue) in patients that responded to immunotherapy treatment. Image adapted from Figure 5C of Wang et al. 2024. (CC BY 4.0).

Applying this to your own work: with clear segmentation of single cells, the ability to identify them as distinct cell types, and seeing how they colocalize with other cells, allows high-plex spatial RNA to give broader and deeper insights into cell–cell communication and a better understanding of the tissue microenvironment.

High-plex RNA offers more comprehensive datasets from your samples, without sacrificing your current analytes

Xenium is compatible with histology- and antibody-based detection methods—including H&E staining, immunofluorescence, and immunohistochemistry—on the tissue section for true multidimensional analysis.

High-plex RNA analyses are good. They’re even better when you can keep your current analysis methods and simply get more from your tissue. Let us show you how:

Imagine the above video is a Xenium analysis of your own tissue samples. Similar to your existing workflows, your tissue was previously stained with H&E and multiple morphological stains (nucleus, cell interior, cell membrane, etc.) and then imaged. Zoom in and you can switch H&E staining off or on to look at the overall morphology of your tissue, or toggle the different morphological stains.

Then, Xenium lets you build on your existing foundation with the ability to select regions of interest (up to and including the entire tissue section), individually toggle and visualize transcripts from up to 5,000 genes, view color-coded cell types and states of interest, and more. You can also combine this with the ability to get immunofluorescence* (IF) data (Figure 7), as well as  Visium HD sequencing-based spatial transcriptomics data, on a tissue section that’s already undergone both staining and the Xenium workflow. Uniting your typical workflows with RNA analyses can give you far richer datasets from your tissue samples without sacrificing your current analytes.

Figure 7. Brain tissue analyzed with both RNA and IF from a mouse model of Alzheimer’s disease. IF of amyloid plaques (white) surrounded by homeostatic and damage-associated astrocytes and microglia (shades of green and pink, respectively). Data taken from the Application Note, “Exploring Alzheimer’s-like pathology at subcellular resolution using Xenium In Situ.”
Figure 7. Brain tissue analyzed with both RNA and IF from a mouse model of Alzheimer’s disease. IF of amyloid plaques (white) surrounded by homeostatic and damage-associated astrocytes and microglia (shades of green and pink, respectively). Data taken from the Application Note, “Exploring Alzheimer’s-like pathology at subcellular resolution using Xenium In Situ.”

*Post-Xenium IF can be performed following a recent Technical Note, and up to 16-plex immunofluorescence staining was demonstrated on post-Xenium tissue in Watson et al. 2024 (4).

In summary, on a single Xenium-analyzed section, you can perform IF for protein detection, stain with H&E for morphological analysis, and then easily integrate these with assisted image registration to simultaneously visualize these outputs. This multiomics approach with RNA adds an entire new dimension to your imaging studies while letting you keep your traditional analytes, such as IF and H&E staining, intact.

Conclusions and capturing the bigger picture

The example experiments presented here showcase how incorporating high-plex spatial RNA profiling into imaging studies can offer transformative insights into tissue biology, giving researchers a richer understanding of cellular communication, cell types, and disease mechanisms without sacrificing current analytes and methods. We invite you to explore what else it can offer you with a sample interactive dataset, or you can try it for yourself with the Xenium Catalyst program.

Finally, while this piece has focused heavily on the benefits of imaging-based spatial transcriptomics methods (e.g., Xenium In Situ), for researchers looking to further enhance their tissue studies, we also recommend exploring Visium HD (Figure 8).This powerful NGS-based technology further complements imaging approaches, providing spatially resolved whole transcriptome data at single cell scale and offering even more comprehensive views of gene expression patterns across entire tissue sections. See for yourself with sample datasets, or get a broader overview of the technology.

Figure 8. Human FFPE colon cancer tissue analyzed with Visium HD and H&E staining. Visium HD spatial mapping of a human colorectal sample demonstrates fine resolution and highlights heterogeneous tissue structures throughout the tissue, with the ability to analyze the same section for gross morphology with H&E staining.
Figure 8. Human FFPE colon cancer tissue analyzed with Visium HD and H&E staining. Visium HD spatial mapping of a human colorectal sample demonstrates fine resolution and highlights heterogeneous tissue structures throughout the tissue, with the ability to analyze the same section for gross morphology with H&E staining.

References:

  1. Kukanja P, et al. Cellular architecture of evolving neuroinflammatory lesions and multiple sclerosis pathology. Cell (2024). doi: 10.1016/j.cell.2024.02.030 
  2. Haga Y, et al. Whole-genome sequencing reveals the molecular implications of the stepwise progression of lung adenocarcinoma. Nat Commun (2023). doi: 10.1038/s41467-023-43732-y
  3. Wang N, et al. Spatial single-cell transcriptomic analysis in breast cancer reveals potential biomarkers for PD1 blockade therapy. Research Square (preprint) (2024). doi: 10.21203/rs.3.rs-4376986/v2
  4. Watson S, et al. Fibrotic response to anti-CSF-1R therapy potentiates glioblastoma recurrence. Cancer Cell (2024). doi: 10.1016/j.ccell.2024.08.012

About the author:

Josh earned a PhD in Neuroscience from the University of Michigan. His thesis work focused on two separate projects: one on disruptions in the microRNA regulatory network in human mood and anxiety disorders, and the second on the cellular basis of transcriptional dysregulation in an animal model of depression. After a postdoc centered on microglial alterations in Alzheimer’s disease, he began his career as a technical writer in the biotech industry. A longtime advocate of better scientific communication, he now uses his skills to make single cell and spatial tools engaging and intuitive, with a particular emphasis on technological comparisons.
Josh earned a PhD in Neuroscience from the University of Michigan. His thesis work focused on two separate projects: one on disruptions in the microRNA regulatory network in human mood and anxiety disorders, and the second on the cellular basis of transcriptional dysregulation in an animal model of depression. After a postdoc centered on microglial alterations in Alzheimer’s disease, he began his career as a technical writer in the biotech industry. A longtime advocate of better scientific communication, he now uses his skills to make single cell and spatial tools engaging and intuitive, with a particular emphasis on technological comparisons.