Friday 30 August 2024

Myth-busting DNA dyes and single cell sorts (UPDATED)

Sample preparation for single cell and isolated nuclei sorting: de-mystifying the use of DNA dyes for transcriptomics and genomics

Sample preparation is a critical step in the downstream molecular analysis of single cells and nuclei.  This is true for all biological analysis of course, so do the amount of sample preparation your end-point requires and no more!  RNA work especially means much to consider and the purpose of this article is to de-mystify the use of nucleic acid binding dyes in the sorting of target cells and nuclei, a widely used technique to deliver high quality, singlet events onwards for downstream analysis.

What's the problem..?

This article was stimulated by recent community discussion threads that pointed to technical support posts (then repeated and therefore somewhat carelessly validated elsewhere) on single cell sorts for molecular biology that solely recommend 7-AAD as a viability dye to exclude dead cells.  It would appear (strangely) that no others have been tested to support this assertion despite the wide use of other well-known and widely understood and trusted analogous modern reagents in countless peer-reviewed publications.  

Conveniently, this author has many years of experience during the early evolution of modern molecular biology in the 80s and 90s.  That led to a curiosity about the validity of these statements and what the published literature might have to offer in defence or otherwise!  Interestingly, it is, in fact, molecular biology research and some drug discovery that provides the background knowledge.

The take-home message (if you can’t wait) is that you, if you are a shared resource lab or a user, can freely choose from a wide range of DNA binding dyes, be they cell impermeant or permeant.  There are some caveats to this that are explored herein, so do please read on all the same!

What's the evidence..?

Molecular biology research basics of the last 20 years or more give us a very good steer on this topic.  Work from 2000 shows that the classical DNA dye ethidium bromide (Eth-Br) demonstrates interference of Taq polymerase (Taq pol) with an IC₅₀ of 2-200 µM, a level that is similar to other dsDNA mono-intercalators (1).

This concentration was found to be consistent with possible interference of Taq pol by humic acid.  Humic acid is a DNA intercalator, found in soil and consequently often present in scene-of-crime forensic samples (2). This was understandably a critical and obvious concern in the early days of forensic science, asking what in a sample (i.e. assay contaminants) might impact performance of amplification of DNA evidence.

DNA intercalators have a useful role in therapeutics, including cancer e.g. mitoxantrone.  In a search for intercalators as candidate druggable inhibitors of viral reverse transcriptases (RT) similar IC₅₀ were observed.  Meanwhile, in the same work, it was also found that dimers (i.e. so-called bis-intercalators such as ethidium homodimer-1/EtHD-1) potently inhibit RT (at the nanomolar level) especially with limited substrate (3) which one might observe in a poor quality sample.

Some years later the same authors re-confirmed their earlier micromolar IC₅₀ for monomers but were also able to achieve nanomolar RT inhibition with bis-intercalators now designed on very weak monomeric intercalators (4), giving further credence to the general concept of template stabilisation by dimers of intercalators.

In other more tangential work to potentially aid quantitative PCR (QPCR) for enumeration of viable bacteria ethidium & propidium (DNA mono-intercalator) monoazides were used as cell impermeant DNA crosslinkers.  These exhibited stabilising effect on the templates to inhibit PCR of the templates in dead (permeabilised) bacteria.  This was demonstrated at 20 µM, a covalent cross-linking of DNA to exclude those unwanted templates (5), consistent with the IC₅₀ for the early work on bis-intercalator stabilisation of templates.

Exploring further the greater potential impact of bis-intercalators, an Aarhus lab showed clearly that the homo-dimers (examples include TOTO-3, BOBO-3, POPO-3, EtHD-1) and interestingly also the mono-intercalators SyBr® Green and SYTOX Orange appear problematic for real-time PCR, likely non-covalently stabilising dsDNA templates.  In their work, the classical methodology of double-stranded template melting temperatures (Cᴛ) were used to measure the stabilising effect of intercalation, with increasing C at higher concentrations (ca. 2 µM). Critically, some were considered to be potentially directly toxic to Taq pol, where no harmless concentration could be determined (6).

These publications, therefore, set the background for the potential interference of DNA and RNA template amplification.

What's the reality in a cell or nuclei sort..?

In a suspension of cells or nuclei what is the likely concentration of a DNA intercalator delivered to a RT/PCR reaction in a single cell sort droplet?

From orthogonal methods, we have approximated dsDNA occupancy in isolated nuclei / intact cells for our own mono-intercalators – cell-impermeant viability dye DRAQ7™ and cell-permeant DRAQ5™ respectively.  This approximates to 6 x 10⁶ molecules per cell or nucleus “event” at saturating concentration - a maximal level that is needed for DNA content / cell cycle analysis - circa 4- to 10-fold above the typical concentration required in sorting. This accords to attomoles per nucleated or nucleus event, an insignificant amount in terms of likely contribution to any inhibitory concentration.  

Further, the typical sorting droplet volume (0.8 nl) (A) with dye originally present at 2-20 µM (B) in the sample stream is diluted 1 in 25 (C) by sheath fluid and then massively diluted into the RT reaction well volume of 8 µl (8000 nl) (D).

Thus, the final concentration of DNA intercalating dye:

= B x A/D x 1/C = 2-20µM x 0.8/8000 x 0.04 = 8-80 pM

Overall, here is a 250,000-fold dilution and a final concentration that is a factor of 10⁵-10⁶ below the “inhibitory range” described in the earlier research, and as illustrated in this representative IC₅₀ curve with the inhibitory region shaded in purple and the red arrow showing the relative position of the practical concentration of DNA intercalator in the typical experiment.  Even allowing for batch sorting of events it would need in the region of 10,000 events to begin to reach the inflection of the IC curve.

In our own experience as part of the flow cytometry community we know that this is not a new story for single cell analysis.  In 2004 cytometry pioneer Willem Corver and colleagues sorted tissue digest cells into PCR: DRAQ5 for cell cycle vs. a gene mutation surface marker (7); all done without mishap from a very complex tumour tissue digest milieu.

UPDATE (Aug. 2024): recent experiments at an expert SRL/training centre on the recently-released BD FACSDiscover S8 show that on this instrument the desired concentration of DRAQ5 is 0.5 - 1 µM, meaning a further significant dose reduction to 2-4 pM. See related article

Subsequently, leading labs in the field who routinely sort cells and nuclei in massive studies (e.g. EMBL, UCSD respectively) utilise a wide range of DNA intercalating dyes, including total/dead combinations such as Hoechst + DRAQ7 (8) or DRAQ5 + DAPI (9) for cell sorting, and DRAQ7 for isolated nuclei (10). The EMBL FACS core facility’s seminal paper (9) begs questions of what can happen to your cell sample – apoptosis, non-lethal stress, mitosis, and so on – that can dramatically impact the transcriptome.  UCSD’s numerous papers, with Sebastian Preissl as the nuclei sorting expert, have standardized robust practice utilizing DRAQ7 as the nuclear event trigger (10) and, prior, combined DRAQ7 with forward scatter to allow sorting of viable cardiac myocytes from a complex cell digest (11).

The take-home message..

Bis-intercalator DNA dyes and the mono-intercalators Sytox Orange and SyBr® Green should be avoided in single cell sorting workflows for genomics and transcriptomics based on earlier evidence that is orthogonal to the field of interest here.  

Similarly, one would avoid Live/Dead fixable dyes (unless required for fixed cell workflows) due to limited signal enhancement of positive events over negatives and the risk of possible internal cross-linking.  

The latter is likely implicated with Calcein AM which should be avoided as a positive marker of cell integrity as determined by the altered gene profiles of cells labelled with it (12).  Nonetheless, for off-line evaluation of cell health Calcein AM can be safely used - as combined with DRAQ7 on the BD Biosciences Rhapsody platform.

Otherwise, a user should, in principle, be able to use any other mono-intercalator DNA dye for dead cell exclusion, positive single cell event marking (and even cell cycle position sorting) and sorting of isolated nuclei.  Importantly, this allows the user the widest choice of reagents for best fit to the demands of i) sample ii) available platform(s) and iii) sorting panel design components.

Based on this literature review and an investigation of the dose of intercalator delivered to RT or Taq pol reactions in single cell sorting any claims about the requirement for the sole use of 7-AAD as a DNA-binding viability dye would appear to be unfounded.

A Little Reflection..

This blog was precipitated by the assertion of 7-AAD being the only recommended choice of DNA intercalating dye to exclude dead cells for single cell sorting. According to expert opinion, a European BMT reference lab, 7-AAD is now a poor choice for dead cell exclusion.  That lab compared far-red fluorescing viability dye DRAQ7 and 7-AAD in a critical assay - the ISHAGE protocol for CD34 stem cell enumeration - for their ability to clearly define the “snapshot” of three clusters: negatively stained intact cells, an expected intermediate population of momentarily/newly leaky cells and the bright fully-stained dead cells.  The ability of DRAQ7 and inability of 7-AAD in this respect was described in a poster presentation at EBMT in 2013 (13).

One suggestion for the use of 7-AAD might be the cost-saving in using a first- generation viability dye.  The reality however is that the marginal cost saving (likely to be less than one US Dollar) would be dwarfed by costs of any antibodies used and moreover the downstream sequencing procedures and subsequent data analysis.

Most of all, think carefully about the reagents in your workflow.  Could they impact on or interfere with your biology?  Are they compatible with instrumentation options available to you?  Ultimately, do you know what you’re using?  The downstream molecular analysis is very expensive, so avoid shortcuts!

Acknowledgements 

Specialist technical knowledge was kindly provided by Christopher Hall MSc, Babraham Institute and Paul J Smith Prof. Em., Cardiff University.

References

1. Nath, K., Sarosy, J. W., Hahn, J., & Di Como, C. J. (2000). Effects of ethidium bromide and SYBR® Green I on different polymerase chain reaction systems. Journal of biochemical and biophysical methods, 42(1-2), 15-29.

2. Thompson, R. E., Duncan, G., & McCord, B. R. (2014). An investigation of PCR inhibition using Plexor®‐Based quantitative PCR and short tandem repeat amplification. Journal of forensic sciences, 59(6), 1517-1529.

3. Jain, N., Francis, S., & Friedman, S. H. (2012). Inhibition of therapeutically important polymerases with high affinity bis-intercalators. Bioorganic & medicinal chemistry letters, 22(14), 4844-4848.

4. Jain, N., & Friedman, S. H. (2019). Multiple weak intercalation as a strategy for the inhibition of polymerases. Bioorganic & medicinal chemistry letters, 29(3), 424-429.

5. Krüger, N. J., Buhler, C., Iwobi, A. N., Huber, I., Ellerbroek, L., Appel, B., & Stingl, K. (2014). “Limits of control”–crucial parameters for a reliable quantification of viable campylobacter by real-time PCR. PloS one, 9(2), e88108.

6. Gudnason, H., Dufva, M., Bang, D. D., & Wolff, A. (2007). Comparison of multiple DNA dyes for real-time PCR: effects of dye concentration and sequence composition on DNA amplification and melting temperature. Nucleic Acids Research, 35(19), e127.

7. Douwes Dekker, P. B., Corver, W. E., Hogendoorn, P. C., van der Mey, A. G., & Cornelisse, C. J. (2004). Multiparameter DNA flow‐sorting demonstrates diploidy and SDHD wild‐type gene retention in the sustentacular cell compartment of head and neck paragangliomas: chief cells are the only neoplastic component. The Journal of Pathology: A Journal of the Pathological Society of Great Britain and Ireland, 202(4), 456-462.

8. Klingler, E., De la Rossa, A., Fièvre, S., Devaraju, K., Abe, P., & Jabaudon, D. (2019). A translaminar genetic logic for the circuit identity of intracortically projecting neurons. Current Biology29(2), 332-339.

9.     Ordoñez‐Rueda, D., Baying, B., Pavlinic, D., Alessandri, L., Yeboah, Y., Landry, J. J., ... & Paulsen, M. (2020). Apoptotic Cell Exclusion and Bias‐Free Single‐Cell Selection Are Important Quality Control Requirements for Successful Single‐Cell Sequencing Applications. Cytometry Part A, 97(2), 156-167.

10. Preissl, S., Schwaderer, M., Raulf, A., Hesse, M., Grüning, B. A., Köbele, C., ... & Gilsbach, R. (2015). Deciphering the epigenetic code of cardiac myocyte transcription. Circulation research, 117(5), 413-423. (See suppl. data).

11. Zhang, K., Hocker, J. D., Miller, M., Hou, X., Chiou, J., Poirion, O. B., ... & Ren, B. (2021). A single-cell atlas of chromatin accessibility in the human genome. Cell, 184(24), 5985-6001.

12. De Micheli, A. J., Laurilliard, E. J., Heinke, C. L., Ravichandran, H., Fraczek, P., Soueid-Baumgarten, S., ... & Cosgrove, B. D. (2020). Single-cell analysis of the muscle stem cell hierarchy identifies heterotypic communication signals involved in skeletal muscle regeneration. Cell reports, 30(10), 3583-3595.

13. Moshaver, B., Huys, E., Terwindt, E., Kramer, P. A., & Preijers, F. (2013, April). DRAQ7, a novel viability dye to determine the correct amount of existing dead and apoptotic cells in CD34+ stem cell enumeration. In Bone Marrow Transplantation (Vol. 48, pp. S182-S183). London, England: Nature Publishing Group.

Roy Edward, FRMS       

© Copyright BioStatus Limited 2024

SYTOX, TOTO, BOBO, POPO and SyBr are trademarks of Thermo Fisher Scientific, Inc. Rhapsody is a trademark of Becton, Dickinson and Company