There is no way around cyclic imaging in spatial tissue analysis. Or is it?
von Christoph Enz
Beyond Cyclic Imaging limits
We´ve gotten so used to it. As soon as more than a handful of markers is desired in spatial biology tissue imaging, the current approach is to revert to cyclic workflows. Actually, sometimes even a handful of markers may require more than one cycle (see our blog on secondary antibody use).
And yes, 4x4 makes 16, and 3x5 is 15, as well as 5x3, there´s little doubt. But is the effort adequate? And, more importantly, is the quality of the results the same, beyond this basic math?
Figure 1: High-plex spatial imaging examples obtained with the RareCyte Orion platform
In any case, these cycling approaches come at a certain price. Whether you are going down the bleaching or the stripping path, the first thing you will encounter is the extra time it takes to analyze a given area of tissue sample for the full set of markers which you are after. A single sample´s data set can take many hours, if not days, to acquire. This might not be a big thing if samples are very few, but it may have a significant impact on the reproducibility of the results as well.
This leads us to results quality. The cyclic treatment of the sample, be it bleaching or stripping, is harsh and may compromise tissue morphology and epitopes. This degradation can:
- Reduce antigenicity
- Distort tissue structure
- Reduce data quality
Particularly after several cycles. Furthermore, antibody stripping is not always complete, which can result in residual signal, steric hindrance for subsequent rounds, and false positives/negatives in later cycles. Some methods also struggle to remove all dye reliably from previous rounds.
It is clear that these effects accumulate over the number of cycles needed to achieve the desired multiplex, making later cycles less reproducible and hence the results less dependable.
Now this is only the biochemical side of things. Because each cycle generates a separate image set, precise spatial alignment (image registration) is required to merge data. Misregistration leads to artifacts or inaccurate co-localization of markers. And, as mentioned previously, this challenge intensifies with multiple cycles. To make things worse, sequential use of fluorophores can exacerbate autofluorescence and spectral overlap challenges, which complicate signal discrimination as panel size increases.
While some of these effects can be dampened by elaborate (and costly!) tagging strategies, fundamental issues remain in different proportions, depending on the individual situation.
Figure 2: High-plex spatial imaging examples obtained with the RareCyte Orion platform
Now what is the way out? Is there any alternative?
As so often the answer is: it depends. It depends on your goals and experimental conditions. If you are after a whole lot of markers, there may be no way around some degree of cycling. However, the closer you get to truly translational work with higher sample numbers and a reduced set of truly relevant markers, the more modern "one shot" approaches can make a difference, considering time, effort, and reproducible quality of the results. Given this, the obvious first question is:
How many markers do you really need?
High-quality optical systems can detect many fluorophores at once via spectral unmixing in a single staining round. For lower two-digit panels, this avoids cyclic stripping entirely, minimizing tissue handling and registration issues. Here, Rarecyte´s Orion system could be the first choice.