Unlocking the Potential of FFPE Samples in Spatial Transcriptomics

Spatial transcriptomics allows scientists to simultaneously profile hundreds to thousands of genes at subcellular resolution, unlocking new possibilities to explore the spatial architecture of cells and their interactions within tissue contexts. Sample quality is crucial in spatial transcriptomics to achieve high-quality data. This article explores novel solutions designed to enhance mRNA recovery from formalin-fixed, paraffin-embedded (FFPE) samples, ensuring the success of spatial transcriptomics experiments.

Challenges with FFPE samples

FFPE methods are typically preferred for tissue preservation in clinical biospecimens due to their cost-effectiveness, ease of use, and excellent preservation of morphological detail.¹ Biobanks worldwide house extensive collections of FFPE samples, which are invaluable for genomics research and large-scale longitudinal studies.

While FFPE preserves tissue morphology well, even after prolonged storage, it has significant drawbacks.¹ The formalin fixation process can severely damage tissues through methylene crosslinking, leading to compromised nucleic acid integrity and reduced gene information compared to fresh-frozen tissues. RNA molecules in FFPE samples are often fragmented, with degradation frequently affecting the polyA tails of the RNA.

Innovative approaches are essential to address these challenges and improve mRNA recovery from FFPE samples.

Direct probe-detection for fragmented and low-quality RNA

One of the novel solutions for spatial transcriptomics with FFPE samples is 10x Genomics FFPE assays designed to perform robustly for fragmented RNA and be resistant to some amounts of RNA degradation. “Our Visium and Xenium assays utilize direct probe-detection instead of reverse transcription, and therefore don’t require full-length RNA,” explains Julia Cowen, Associate Director, Product Management, Visium, at 10x Genomics. “We have also optimized our sample preparation protocols to be more gentle on nucleic acids to avoid RNA degradation and genomic DNA detection.”

“Our Visium v2 and Visium HD assays employ a whole transcriptome panel consisting of probe pairs, each with a left-hand side (LHS) and a right-hand side (RHS) probe, targeting a specific gene. On average, the panel contains 3 probe pairs per gene target,” elaborates Cowen. “Both the LHS and RHS probes must bind to the mRNA target site to form a ligation product, which is then captured by the Visium slide and spatially localized.” This sophisticated probe design, including 3x gene coverage and the double probe binding event, ensures high sensitivity and specificity to target RNA within FFPE tissues, allowing for the specific detection of degraded mRNA molecules.

The Xenium probes are another innovative solution, featuring two regions that independently hybridize to the target RNA while also containing a gene-specific barcode sequence. “The binding of the probe ends and their ligation to each other generate a circular DNA probe, which is then enzymatically amplified,” Cowen points out. “If one part of the probe experiences off-target binding, ligation won’t occur, thus suppressing off-target signals and ensuring high specificity.”

These Visium and Xenium assays have been meticulously engineered to work across a wide range of FFPE tissues without further optimization by the user. The probe panels were specifically designed to deliver robust performance even when dealing with samples with low RNA quality. Cowen recommends using tissue with DV200%≥30%, but lower DV200-scored blocks will still yield data though likely of lower sensitivity. Additionally, she suggests using standard histology (e.g., hematoxylin and eosin stain) to assess tissue block quality and ensure the integrity of tissue morphology.

Using the 10x Genomics’ Visium platform, Joakim Lundeberg’s group at KTH Royal Institute of Technology, Stockholm, Sweden, developed a new protocol for unbiased genome-wide spatial analysis of mRNA in various FFPE tissues, including mouse brain tissue, human lung and kidney organoids, and a few clinical samples.² The authors highlighted that their spatial transcriptomics data obtained with FFPE tissues strongly correlates with data from fresh frozen tissues and is complex enough to delineate anatomical features without bias.²

Combined workflow for RNA and protein co-detection

Standard BioTools™ and Advanced Cell Diagnostics have developed a novel workflow for the co-detection of RNA and protein markers in the same FFPE samples—without spectral overlap or background autofluorescence. Combining Imaging Mass Cytometry™ (IMC™) and RNAscope™ workflows, this integrated approach offers researchers deeper insights into complex cellular interactions within the tumor microenvironment. Notably, this workflow is optimized for use with FFPE samples, ensuring reliable high-quality results, and is compatible with spatial transcriptomics assays.

“By leveraging the subcellular resolution and multiplexing capabilities of IMC along with the high sensitivity and specificity of RNAscope, this approach enables the study of previously inaccessible targets and activation states of cells. This drives further biological insights within IMC experiments,” says James Pemberton, Ph.D., Application Scientist at Standard BioTools. “The workflow supports the analysis of over 40 markers, including 12 targeting RNA expression, facilitating the study of cell type-specific gene expression and identification of cellular sources of secreted factors.” Researchers can first use the RNAscope protocol to label mRNA, then apply antibodies on the same slide to simultaneously analyze mRNA and protein via metal tags with IMC.

Three-step protocol to detect RNA and protein on the same FFPE slides with IMC. Image provided by Standard BioTools™.

To achieve reliable results with this integrated workflow, it is essential to account for the low RNA quality that can be seen in FFPE samples. The robustness of the RNAscope assay enables the detection of even partially degraded or incompletely unmasked RNA targets. Further optimization can be facilitated by using serial sections. Pemberton recommends analyzing RNA quality on one serial section with positive control probes. Successful detection of the positive control probes provides confidence that adjacent sections have sufficiently good RNA quality for applying target probes for your experiment. If positive control probes are not detected, Pemberton advises using a new, freshly cut tissue section sample. Finally, using negative control probes can also help you verify the specificity of RNA staining.

Optimization tips for improving mRNA recovery from FFPE samples

“To improve RNA recovery from your FFPE samples, it is important to follow best practices for working with RNA, such as working with equipment decontaminated for RNase, using nuclease-free water, and ensuring all reagents are free from contamination,” notes Cowen.

When preparing tissue sections for Xenium and Visium spatial transcriptomics assays, Cowen additionally recommends:

• Cleaning sectioning equipment with RNase decontamination solution before sectioning
• Using a clean, new blade when sectioning tissue
• Using fresh reagents for tissue processing
• Storing FFPE blocks at 4°C and storing section slides as indicated by the relevant product user guide
• Using clean glassware: spray it with RNase decontamination solution and/or autoclave before use
• Using nuclease-free water and 10x-recommended third-party reagents for all workflow steps
• Avoid reusing aliquots of the same staining buffer for multiple experiments

Conclusion

The development of robust methods for sensitive spatial transcriptomics of FFPE samples has been challenging due to formalin-induced cross-linking and degradation of mRNA molecules. These novel solutions represent significant progress in overcoming the limitations of FFPE samples, making high-quality spatial transcriptomics data more accessible and reliable.

References

1. Mathieson, W., Thomas, G. Using FFPE Tissue in Genomic Analyses: Advantages, Disadvantages and the Role of Biospecimen Science. Curr Pathobiol Rep 7, 35–40 (2019). 

2. Gracia Villacampa E, Larsson L, Mirzazadeh R, et al. Genome-wide spatial expression profiling in formalin-fixed tissues. Cell Genom 8;1(3):100065 (2021).