Beyond The Rainbow: From Reference Controls to Regulatory-Ready Unmixing
In the previous blog post, I spoke to how spectral flow cytometry uses a fluorophore's unique spectral fingerprint to deconvolute overlapping emissions, allowing researchers to design larger panels with more markers for deeper investigations into cellular populations. This deconvolution is achieved using an unmixing strategy, which can be thought of in the same way as compensation controls.
A set of reference controls consisting of single-stained beads and/or cells corresponding to each of the fluorophores in a panel are used to unmix a spectral flow experiment. Once the reference controls have been acquired, the researcher assigns each reference control to a fluorophore and denotes which marker it corresponds to on the instrument. These acquired reference controls are then used to deconvolute or unmix the raw data either as the samples are being run on the instrument (live unmixing) or once all the raw data has been acquired.
Reference controls can usually be stored in a library on the instrument software and can be applied to multiple different experiments, saving researchers valuable time and resources particularly when running the same panels on multiple different samples over multiple timepoints. That said, there are some regulatory compliance issues surrounding this which we will touch on later.
This post aims to provide some guidance around choosing the correct reference controls for your panel and samples, how to test and troubleshoot your unmixing strategy, and how to make your unmixing strategy regulatory ready.
Choosing the right reference controls
Much like compensation controls for conventional flow cytometry, reference controls can be made using either cells or beads. Compensation beads such as UltraComp eBeads are widely used for flow cytometry applications and give clean positive and negative peaks for most antibodies of multiple origins when the manufacturer's recommended staining protocol is followed.
It might be tempting to think that using beads alone would provide the cleanest spectra and simplest workflow. However, in spectral flow this is not always the case. In practice, a combination of beads and cells often provides the most robust unmixing strategy. This is because spectral flow cytometry is highly sensitive and spectra can be influenced by factors such as staining strategy and cellular autofluorescence.
When you compare your fluorophore reference control spectra to the manufacturer's spectra, you might notice subtle discrepancies. These differences can arise from using a different marker or sample type. For example, BV750 stained on CD80 may not look identical to BV750 stained on CD4 in PBMCs, as shown in the Cytek Guide. These small differences can become important when fine-tuning your unmixing strategy.
To address this, it is often useful to build multiple reference control libraries using different combinations of beads and cells. Running several unmixing strategies against these libraries gives you the best chance of identifying an approach that is both robust and reproducible.
Beyond the Basics: Specialised Beads and Autofluorescence
In addition to UltraComp eBeads, there are beads designed specifically for spectral flow cytometry. Slingshot beads are one such example, engineered to match the full spectral profile of a fluorophore across detectors rather than just providing a clean positive/negative peak. These are especially valuable when working with complex panels or problematic fluorophores.
Amine-reactive beads are another useful tool, particularly for live/dead staining reference controls. By removing the variability that can arise from using cells, they provide more consistent results and speed up the staining process.
Another factor to keep in mind is autofluorescence. In conventional flow, autofluorescence was often seen as an unwanted source of background noise. In spectral flow, however, autofluorescence can be treated almost like an additional "colour." With the right setup, it can be measured and computationally subtracted from your data. Cytek Biosciences provide excellent resources on how to implement autofluorescence extraction in your workflow. Whether you decide to remove it or not, remember that it will affect your unmixing matrix and ultimately the reliability of your results.
Putting It to the Test: Validating Your Unmixing Strategy
Once you have your candidate reference controls, the next step is to test which unmixing strategy works best. Start by creating single-stain controls for each fluorophore using both beads and cells. This allows you to compare the spectra directly and build different unmixing matrices based on the combinations that give the cleanest separation.
With these strategies in hand, apply them to your single-stain library. Compare the raw single-stain data with the unmixed data to see how well the strategy preserves positive and negative populations. The Relative Percentage Difference (RPD) is a useful metric here, and best practice is to accept only those strategies where the unmixed positives fall within 10% of the raw positives.
Once you've identified promising strategies, challenge them with a multicolour test article. This is where potential weaknesses often emerge. Even if a strategy passes all the RPD checks on single stains, it may distort population placement when multiple markers are combined, leading to smearing or shifts in the plots.
If this occurs, NxN analysis (comparing each marker against every other marker) can help pinpoint which colours are skewing the results. You may need to adjust your reference controls, alter your unmixing strategy, or in some cases revisit your antibody/fluorophore choices. Through this process, you will often land on a hybrid approach where both beads and cells form part of the reference control library.
From Bench to Regulation: Making your Unmixing Strategy GCP-ready
If your spectral flow assay is destined for clinical use, regulatory considerations must be built into the unmixing strategy. A key requirement is consistency. All antibodies used in both your test articles and your reference controls should come from the same lot number.
Once the unmixing strategy is finalised, intra- and inter-assay validations should be performed using lot-conserved antibodies and a bank of the same test article. Intra-assay validation involves staining replicates of the test article and calculating RPD between them. If all markers fall within the defined thresholds, the assay can be considered intra-assay validated. Inter-assay validation extends this process across different days, again in triplicate, to confirm reproducibility over time.
Lot bridging is another important consideration. If the original lot number is exhausted or a custom antibody becomes unavailable, new lots must be bridged to the validated assay to demonstrate equivalence. This is a detailed topic in itself and will be explored in a future post.
Designing a robust unmixing strategy in spectral flow cytometry requires careful thought at every step, from choosing reference controls to validating strategies and planning for regulatory compliance. While beads, cells, autofluorescence, and multicolour testing may each present their own challenges, a systematic approach allows you to develop an assay that is both scientifically rigorous and regulatory ready.