Question: How are DV200% and Visium for FFPE assay performance correlated?
Answer: Formalin fixation followed by paraffin embedding (FFPE) is commonly used for long-term tissue storage as it preserves and stabilizes samples. But, due to fixation and RNA hydrolysis during storage, recovered RNA from FFPE blocks can be low yield and fragmented. Furthermore, the quality of RNA isolated from FFPE tissue can be influenced by such factors as cold ischemic time, tissue type, tissue age, and preservation methods. Our Visium Spatial Gene Expression for FFPE Solution uses a probe-based approach for RNA detection to overcome these obstacles. This method is poly(A) tail-independent and designed to detect all protein-coding RNAs in either human or mouse tissue.
This article provides an overview of Web Summary File metrics for Visium Spatial Gene Expression for FFPE libraries from different tissue blocks across various storage temperatures, ischemic times, and DV200's. Individual results may vary depending on the particular sample type and quality, but DV200 metrics and assay sensitivity generally appear to be correlated.
Accelerated aging experiment:
In this experiment, human ovarian cancer and mouse brain FFPE blocks were exposed to high temperatures to accelerate RNA degradation. At different time points, sections were taken from each block for RNA quality assessment and library preparation.
Graphs showing the correlation between sensitivity (Median UMI Counts per Spot) of Visium FFPE libraries and RNA quality (expressed as DV200) for a human ovarian cancer block (left) and a mouse brain block (right) showing that DV200 and assay sensitivity are positively correlated.
As shown above, a drop in DV200% was often accompanied by a drop in library sensitivity. However, a drop in the library sensitivity could not always be explained by changes in the RNA fragment distribution. Other heat-induced changes (such as RNA or tissue structure) could contribute to a lower amount of transcript captured.
Spatial clustering (top) and t-SNE projections (bottom) of mouse brain before the accelerated aging experiment (left; DV200 ~90%) and after (right; DV200 ~40%) showing that DV200 and assay performance are positively correlated.
As displayed above, a drop in DV200% was often accompanied by a drop in spatiality and library complexity. Similar results were observed for human ovarian cancer.
Delayed fixation experiment:
In this experiment, mouse kidney, spleen, and intestine were dissected and cut into pieces. One piece of each organ was immediately placed into formalin (time point 0 hr) while the remaining pieces were kept in PBS at 4°C or at room temperature (RT) for 24, 32, or 72 hr. When incubation was complete, tissues were fixed, processed, and embedded to generate FFPE blocks. Several sections were then taken from each block for RNA quality assessment and library preparation.
Table showing that delayed fixation impacts DV200% and assay performance. Delays in fixation negatively impacted data quality. Tissue stored at RT before fixation generated poorer data quality compared to 4°C storage, regardless of delayed fixation time.
Spatial clustering (top) and t-SNE projections (bottom) of mouse intestine before the accelerated aging experiment (left; DV200 61%), after 4°C storage for 32 hr (middle; DV200 66%), and after RT storage for 24 hr (right; DV200 23%) showing that DV200 and assay performance are positively correlated.
As displayed above, a delay in fixation was often accompanied by a drop in spatiality and library complexity. Similar results were observed for mouse kidney and spleen. Interestingly, in the above example, we see that delayed fixation did not always lead to changes in the RNA fragment distribution but consistently showed poorer spatiality and complexity than tissue immediately fixed.
Overall, we've found that suboptimal storage temperatures, long ischemic times, and lower DV200's all negatively impact assay sensitivity. In many cases, the assay performance can be predicted by assessing the DV200% value. In cases where FFPE blocks were heated for extended periods of time, the DV200% was a less accurate indicator of the expected assay performance. Taken together, DV200% values and assay sensitivity appear to be correlated and is a critical QC step for assessing data quality.
Products: Visium for FFPE