Viability Dyes for Flow Cytometry: It’s Not Just a Matter of Life and Death

As Morrissey once sang “in the midst of life we are in death, etc.” A fact of life in the research Lab is that whenever we run an assay there will almost certainly be some cell death in our sample. This may be insignificant in some techniques but in others it can be problematical.

Why dead cells are a problem

Flow cytometry is one technique where dead cells can be very problematic; this is because:

  • they will take up the antibodies and probes that we use non-specifically
  • they may become more autofluorescent.

Both of these consequences will cause an increase in background fluorescence and therefore reduce the dynamic range and reduce our ability to see something that is weakly positive for a marker. If that wasn’t bad enough, dead cells may release DNA which is notoriously sticky and will lead to cell clumping (not useful in a single cell technique like flow cytometry), and in cell sorting you really don’t want to compromise your experiment by sorting non-viable cells.

Solving the dead cell issue

However fear not, we can target these dead cells and remove them, if not from the tube, at least from the analysis. Firstly we can make sure that the medium we keep the cells in contains a small amount of DNAse which will help keep cells apart. Secondly we can use a feature of dead cells that will allow us to specifically identify them. When cells die, they lose their membrane integrity and become leaky. This is what causes us problems with non-specific binding but this time we are going to use a specific staining to help us identify those dead cells. There are two ways to do this – by using dyes that bind to DNA or dyes that bind to protein.

DNA binding dyes

Dyes such as Propidium Iodide (PI), DAPI, DRAQ7, 7-AAD and TO-PRO-3 will only enter cells whose membranes have been compromised. Once they do, they will bind to DNA and fluoresce so that these cells can easily be detected and gated out of any analysis (Figure 1A). In most cases the protocols are very simple, just add an appropriate amount of DNA dye to the sample about 10 minutes before running them on the cytometer (be aware of leaving the dyes on too long before analysis because every cell that dies when the dye is present will be positive so if you wish to compare between samples, keep the staining time consistent!). There are a range of these dyes with different excitation and emission characteristics so it shouldn’t be a problem to find one that will fit with your experimental panel.


BitesizeViability (1)

Figure 1. Panel A shows the use of a DNA binding dye, propidium iodide. Dead cells are positive for the dye (upper region). Panel B shows a fixable amine-binding dye (eFluor780). Again dead cells are identified by positive fluorescence in the upper region.


You will notice though that here we have added the dyes to live cells. If we were to add say PI to a samples of fixed cells, it would stain all the cells as fixation will permeabilise the cells. This is obviously a drawback if cells have to be fixed for safety reasons or because you want to detect an intracellular target. It will come as no surprise to learn that we have thought of that and this is where the protein-binding dyes come in…

Protein-binding dyes

These are also sometimes known as amine-reactive dyes or live/dead fixable dyes. They are still based on the principle that dead cells have compromised membranes but work in a slightly different way. Instead of binding to DNA they bind proteins so if they are added to a sample before fixation, live cells will bind some dye – there will be proteins present on the surface of the cell, but dead cells will bind much more as the dye can enter the cell and have access to the abundance of intracellular proteins.

At this point cells maybe fixed using with formaldehyde or alcohol and the difference in staining level will be maintained (Figure 1B). Again, protocols are very simple, a 20 minute incubation with the dye and these dyes come in a variety of excitation and emission characteristics and again you would almost certainly find one compatible with your other fluorochromes. Examples of theses dyes are the live/dead fixable dyes, the eFluor fixable dyes, the Horizon dyes, the Biolegend Zombie dyes and the Ghost dyes. Although the dyes can be fixed, this is not mandatory so they can also be used to identify dead cells in sorting experiments.

In these days of multi-laser flow cytometers there should never be any excuse not to add a viability dye and now you know how simple it is, you can do it too!


  1. nidhi on January 26, 2017 at 2:32 am

    What channel do you use to detect PI by flow cytometry? I read that PI stain can be detected by many different channels. In that case, how do you control the spill over with other markers? I generally use compensation beads (species specific, Fc binding ones) to carry out compensation, so using PI with that wouldn’t make sense, right?

  2. Ashley on May 24, 2016 at 8:19 pm

    Would it be reasonable to assume that live/dead fixable dyes will also stain cells that are degranulating?

  3. George Elias on March 25, 2016 at 10:18 am

    Is it possible that using DNA binding dyes (PI or Sytox Blue) to exclude dead cells during sorting might interfere with downstream applications like RNAseq, RT-PCR and DNA sequencing?

    • Roy Edward on May 11, 2016 at 2:46 pm

      Hi George,
      Preissl et al. 2015 (DOI: 10.1161/CIRCRESAHA.115.306337) have demonstrated use of far-red DRAQ7 to exclude dead cells from a cardiomyocyte sort prior for RNA-seq.

      The related (parent) cell permeant DNA probe DRAQ5 has been used in live cell DNA content-driven sorts prior to PCR for mutation detection – Douwes Dekker (J Pathol 202.4 (2004): 456-462).

  4. Bitesize Bio on January 6, 2016 at 7:00 am

    […] player in signal transduction. It plays important roles in exocytosis, muscle cell contraction, and cell death cascades, among other things. It acts as an activator of numerous enzymes through calcium-binding proteins. […]

  5. Bitesize Bio on November 17, 2015 at 9:00 am

    […] the cell has affected the activation of the FP. By using a viability marker, such as DAPI, you can gate out the dead cells, and this will give you a much more accurate number of FP+ cells. If you are making a clonal cell […]

Leave a Comment