Overnight staining for scRNA-Seq?

Today I’m going to talk about an idea that arose during discussions at the Danish Flow Cytometry meeting. I don’t have the time or, really, the expertise to test it, but I think it might work really well.

The idea? Combine overnight staining with RNA sequencing.

BD has recently released this product called OMICS-Guard. It’s supposed to preserve cell surface markers and transcripts for up to 72hrs at 4C. I don’t know what it is, but it doesn’t seem to be a fixative. In theory, using this product (or other options detailed below), you might be able to stain your cells overnight either for cell sorting or for CITE-Seq without loss of your desired transcriptional profiles.

Could you stain overnight before encapsulating cells for single cell RNA-sequencing? I don’t know. In the Liston lab, we don’t run scRNA-Seq experiments every week—they’re more of a special occasion experiment for us. Normally, when we sort cells for RNA sequencing, we try to minimize the handling time of the cells to keep the transcriptional profile as close as possible to the in vivo “true” transcriptome. So, this would be a big change.

How might this work and what would be the challenges?

Well, you’d need to prevent activation of the cells. That means keeping them cold. That’s easy enough. Additionally, we can take cues from this paper and use anisomycin and triptolide to prevent activation (we now do this routinely for our sorts for any type of RNA sequencing).

Equally importantly, we need to prevent the transcriptional profiles from changing. The same paper also uses actinomycin D to do this. OMICS-Guard is meant to achieve the same effect (perhaps via the same mechanism?).

Finally, we’d like to preserve cell viability. This is going to be tricky because for a lot of people, the cells for scRNA-Seq come from difficult samples:  things like tumour isolates, aspirates, biopsies, tissues. Those all tend to have lower viability. As we saw in this post, cells that are suffering to start with tend to not do so well with prolonged incubations. You can read through the tips in the post, but it boils down to optimizing your isolation method and using DMEM with 2.5% FCS (optionally also 2mM EDTA) for the incubation. For samples that have lots of debris and dead cells, an apoptotic cell removal step might be beneficial, although this would require further optimization.

In short, if I were doing this, I’d start by testing these conditions:

1. Freshly isolated cells without overnight staining (control)

2. Cells stained in OMICS-Guard overnight

3. Cells stained in DMEM containing: 2.5% FCS, 2mM EDTA, triptolide, anisomycin and actinomycin D.

4. A combination of 2 & 3.

Alternatively, the cells could potentially be fixed with glyoxal, a chemical that doesn’t cause DNA crosslinking and permits transcriptional analysis on fixed and permeabilized cells. I’ve tried this for flow, and it works, but isn’t perfect: the scatter is greatly increased, autofluorescence increases and not all epitopes are preserved (Foxp3 works, though). Check out the paper from Jon Houseley's group.

Here's an example of what mouse splenocytes look like with overnight staining for Foxp3 and Helios on CD4+ viable T cells after fixation with either formalin or glyoxal. Glyoxal doesn't preserve scatter perfectly; at this concentration, the cells are a bit shrunken, and at higher concentrations, the scatter increases dramatically as seen with alcohol-based fixatives.

Mouse Treg staining with glyoxal

How do the cells tolerate the four conditions I listed above? Here are some mouse splenocytes stained for a simple panel under those conditions. I’ve stained the cells with a fixable viability stain before and after the overnight incubation so we can see how they fared. In the example below, the cells are gated on cells that were alive at the start of the staining period (negative for the fixable viability dye added prior to staining).

Mouse splenocyte viability in Omics-Guard

A bit better with DMEM added.

And here are some cryopreserved PBMCs that were thawed and stained for a simple panel either for 1hr at 4C in FACS buffer (PBS 2.5% FCS 2mM EDTA) or overnight in Omics-Guard.

Cryopreserved PBMC viability in Omics-Guard

Not too bad. I don't usually recommend doing overnight staining without fixation on cryopreserved PBMCs due to the fragility of the cells.

Why would you want to do this?

Well, there are a number of reasons. Perhaps most importantly, the workflow is much better. You can isolate your cells and then leave them staining overnight rather than having to continue on with the FACS sorting (hours) and scRNA-Seq encapsulation (hours, depending on the platform). These are expensive experiments, so doing them all in one very long day is not ideal as you’re more likely to make expensive mistakes.

Secondarily, you’d get the enhanced separation of your markers that comes from overnight staining. That’s helpful if you’re trying to sort cells with low expression markers (e.g., TIM-3+PD-1+ T cells).

Human CD4+CD3+ T cells stained for CCR7. CCR7 can be tricky without incubation at higher temperatures or overnight staining.

Mouse splenocyte staining works well overnight in Omics-Guard

Overnight staining should be exceptionally helpful for CITE-Seq resolution where we lack the advantage of bright fluorophores to assist us in detection. CITE-Seq (Total-Seq, AbSeq) antibodies are also really, really expensive, so using a bit less of each would mean you can instead profile a much larger set of markers in one go.

Anyway, I think somebody should try it.