There are many technologies and tools available to analyse tissue sections for spatial omics. Core facilities often have several of these instruments to perform spatial transcriptomics, proteomics and multi-omics. Sequentially combining these platforms on single tissue sections can provide deep insights into biology not acquirable from one tool alone.
Jared K Burks, Professor, MD Anderson Cancer Center is co-director of the Flow Cytometry and Cell Imaging Core Facility at the MD Anderson Cancer Center and recently spoke on this topic at a Front Line Genomics webinar.
In this article, we will cover how one tissue section can be assayed with multiple spatial technologies and explore a specific example of this approach in ovarian cancer. We will provide images and quotes taken directly from the webinar. To hear Jared’s talk in full, as well as the other talks on Spatial Meets Cancer Biology: Uncovering Cell Types, Interactions and Drivers, please follow the webinar link.
TME as a spatial puzzle
For the tumour microenvironment (TME), cell locale matters, and spatial technologies have proved crucial to dissecting the TME. Cancer cells, in fact, build their own niche environment; they are active players in the construction of the TME. This environment is built to be growth promoting and when immune cells enter this foreign environment, they are now at a disadvantage. A big goal for the field is to stop this process of niche construction.
Hence, the aspiration for spatial biology in understanding TME construction is being able to identify who, what, when, where, why and how, in a way that was not possible with previous technologies. Simply, for the TME, we want to know:
- Who is present
- What they are doing
- When, Where and Why are they doing it
- How can we affect it and potentially stop them doing it
In the TME, to identify who is present, single-cell methodologies have been sufficient. However, to identify how the TME is built, we need spatial/imaging methodologies. And, to find out how cells are working to build the niche and how to stop them, we require the investigation of functional and cellular networks. This depth of investigation often requires more than one technology, making experiments expensive and requiring large numbers of samples.
Layer your methods
Furthermore, if you use only one technology, and hence only one way of looking at things, you risk biasing your conclusions. This can occur when we rely too heavily on specific ‘kits’.
Jared: “If you bring your bias, your favourite opinion, to your experiment, you will see what you want to see and not what’s actually present in the data. So, we need to try to compensate for our biases and at least be very cognizant.”
While ‘kit biology’ is incredibly user-friendly and accessible, a kit only allows you to do one specific thing. One solution is to produce more samples and use different kits. However, an alternative is to develop experimental workflows that layer technologies together on single samples. This produces a more comprehensive image than following kits and getting individual data points (Figure 1).
By layering technologies, we can also use the findings from one technology to guide the tool selection and marker choice in the next one, getting the most ‘bang for your buck’ from one section.
Figure 1. An array of feedback loops using different single-cell and spatial tools. Image taken directly from FLG webinar (September 26th, 2023). Full credit – Prof. Jared K Burks.
Jared: “If we were to use the serum from a PBMC isolation in a technology like the Isoplexus, we can identify putative cytokines that might be found in our tissue sections. This opens the door to targeting those cytokines … downstream. Once we find some of these details, we might go back to cell-sorting to isolate a specific subpopulation of cells that we would then use for other analyses, which we can use to understand how these cells are being influenced. We might also employ different spatial transcriptomics methods and layer them in, in a targeted manner.”
An example of the approach – Ovarian Cancer
We will now cover an example of this approach in ovarian cancer (Ferri-Borgogno et al., 2023).
A prior study had found that the stroma-tumour interface was enriched for the APOE-LRP5 cell-communication pathway in short-term survivor cancer patients, suggesting that this needed to be better understood. The hope is that disrupting the pathway could eventually turn short-term cancer survivors into long-term survivors.
The study began with ovarian cancer samples and the first tool in the pipeline was the Lunaphore COMET, which is a sequential immunofluorescent tool that allows a 40-plex protein panel. Using markers for CAF’s, T-cells and tumour cells alongside functional targets, the researchers were effectively able to identify the tumour-stromal interface and begin functional assessment. However, following the sequential treatment, the tissue was still in excellent condition. The natural question, can you do more with the same tissue?
Hence, the researchers stained the same slide with metal-based probes from the Standard Biotools Hyperion system. Time was taken first to review the results of the COMET panel to decide the exact markers to assess on the second panel (i.e., novel functional markers of interest). Additionally, confirmatory markers between panels ensured consistency between the technologies and added a level of reliability to the experiment.
Figure 2. COMET and HYPERION data from a single section. Image taken directly from FLG webinar (September 26th, 2023). Full credit – Prof. Jared K Burks.
For even more depth, in this example, RNA-scope was used on top of the same section to leverage a few more targets, specifically using RNA to assess levels of relevant proteins that may not be detectable with the previous two methods since they were secreted quickly.
To merge these images of the same section from different technologies, VIRTUALDOUBLESTAINING™ in Visiopharm was used. This allows serial sections to be combined and share annotations to view these markers together.
Ultimately, this approach allowed three different spatial platforms to be deployed on a single tissue section, allowing a layered deconvolution of functional markers in the stromal-tumour interface, helping to better understand this part of the TME for effective treatment.
Q&A Highlights (Summarized – not direct quotes)
Q: Could you just confirm to me what the thinking was in the 26-plex Hyperion panel, how many confirmatory markers did you have? What functional markers did you choose?
A: We used 8 markers as overlap between the COMET and Hyperion panels. This gave us confirmation of specific cell identities; it allowed us to go for specific functional markers to make sure that we were talking about the same cell in image A and image B. For functional markers, we went for every type of functional marker we could go after … anything related to T-cell, macrophage of CAF functionality were included.
Q: How confident are you that your technology using laser does not alter the (tertiary) protein structure and its interaction with the antibody epitope? Have you validated this?
A: The COMET uses an LED, not a laser source, so it’s fairly innocuous. By comparing markers between the technologies, one can confirm the laser source of the HYPERION was not interfering with the process.
To hear from the other speakers at this webinar and for the full Q&A please use this link.
References
Ferri-Borgogno, S., Zhu, Y., Sheng, J., Burks, J. K., Gomez, J. A., Wong, K. K., Mok, S. C. (2023). Spatial Transcriptomics Depict Ligand–Receptor Cross-talk Heterogeneity at the Tumor-Stroma Interface in Long-Term Ovarian Cancer Survivors. Cancer Research, 83(9), 1503-1516.
