Cell Proliferation Measurement Using the Dye Dilution Method

Table of Contents


In my blog I’ll be discussing how you can monitor cell proliferation by Flow Cytometry.

How do we measure cell proliferation?

Cell proliferation can be measured in many ways. The most obvious one is to simply count the cells in your various test conditions and compare this to your control. Or, if you have too many conditions to count, you could add tritiated thymidine to your test conditions and then measure radioactivity incorporated into the dividing cells, when the free tritiated thymidine has been washed away.

These methods will tell you if your population has divided, and by how much. They measure the overall proliferation of the bulk population. However, you get no information on whether the entire population is dividing or that only a subset of cells is dividing.

Can we measure proliferation using flow cytometry?

Utilising the dye dilution method and flow cytometry you can detect individual cells and observe how many times that cell has divided. You can do this for thousands of cells per sample population. This is immensely powerful and informative. You can also label the cells for surface markers and functional proteins like cytokines, simultaneously. You can see which cell subsets in a population are dividing, how much, and what they’re making! It can be applied to cells in vivo by adoptively transferring cells to a new host. You can then follow their migration and proliferation in their new host. It can also be used in vitro under culture conditions as I will describe.

The ground-breaking publication by Lyons and Parish, in 1994, described the use of: 5(6)-Carboxyfluorescein diacetate succinimidyl ester (CFDA-SE) to label lymphocytes, which were tracked, and their subsequent proliferation was monitored. The acetate groups of CFDA-SE help the dye traverse the cell membrane. Once inside, esterases in the cytoplasm cleave the acetates to convert the non-fluorescent CFDA-SE to the fluorescent ester CFSE. The succinimidyl ester group covalently binds to amine groups on intracellular proteins and in so doing, fluorescently labels the entire cytoplasm of the cell. Once bound it is well retained and is not transferred to neighbouring cells. CFSE is excited by the 488nm laser and detected through a 520nm or similar optical filter (as normally used for FITC). A narrow range of fluorescence intensities is produced due to the similar size and morphology of lymphocytes. When the cell divides exactly in half, the dye contained in the daughter cell is half that of its parent. In each subsequent division the dye concentration is halved. Detection can be measured until the dye is diluted to the background fluorescence level of the cell. In my experience this is about 7 to 8 divisions. Although written for CFSE, many of the topics I’ve talked about here are pertinent when other CFSE alternatives are used. I’ve listed some of these at the end of this blog.

Here is a link to the Lyons & Parish publication:


Determination of lymphocyte division by flow cytometry. Lyons AB, Parish CR.

J Immunol Methods. 1994 May 2;171(1):131-7. Doi: 10.1016/0022-1759(94)90236-4.

There was an earlier publication by Parish (1990) which investigated several fluorescent dyes to track lymphocyte migration. The link to this article is:


What should flow cytometry plots look like with CFSE staining?

Figure 1

The flow cytometry histograms show representative proliferation data. The upper plot shows the initial fluorescence measured from a population of undivided lymphocytes labelled with CFSE and that of the same cell culture sampled 3 days later is shown on the lower plot. On day 3 you can see that some cells remain undivided, but others have divided once, twice, and even three times, diluting the fluorescence in each subsequent cell division.

My CFSE-labelling protocol:

I made up my own CFDA, SE stock solution. It is particularly important to calculate the correct concentration. I can’t stress this enough. CFSE is toxic to cells at high concentrations, so you need to get the stock concentration correct. If you get high toxicity after labelling, check the dye concentration you used in the labelling process. You may want to titrate the CFSE concentration to find the one that works best for your cells, in your particular assay. I used a loading concentration of 5µM in my assays.

The molecular weight is 557.47 g/mol. Vials are commonly sold in 25mg amounts. Dissolve 25mg in 8.96 ml DMSO to give a stock of 5mM.

 Dispense into aliquots of 20µl to 100µl and store in a -200C freezer. I stored mine for 2 years, and they can probably be stored for longer. Do not re-freeze small aliquots. Use what you need from an aliquot and then throw away the remainder.

Labelling your cell population:

In summary

Maria Protocol

I will now describe in detail the procedure I used to label human or mouse lymphocytes.

  • Chill at least 200ml of wash solution: PBS containing 2.5% FCS, at 40C in the fridge.
  • Warm tissue culture medium, containing 10%, to 370C.
  • Warm up to 20ml PBS (no FCS or protein) in a 370C water bath.
  • Thaw an aliquot of the 5mM CFSE stock solution and make a working solution by diluting 1:100 with prewarmed PBS. You now have a working solution of 50µM.
  • Count the cells to be stained. Centrifuge the required number of cells in a 15ml conical tube to pellet.
  • Carefully remove the supernatant. Add 900µl prewarmed PBS. Up to 1 x 107 cells can be labelled. Vortex to mix.
  • Add 100ul of the working CFSE stock to the 900ul  cell suspension to give a final concentration of 5µM CFSE. Vortex to mix and immediately transfer to a water bath.
  • Incubate at 370C in a water bath, lid closed to protect from light, for exactly 5 minutes. The CFSE is now loading into your cell suspension.
  • After exactly 5 minutes add 14ml of the COLD wash solution of PBS containing 2.5% FCS. Vortex to mix. Centrifuge to pellet the cells.
  • Carefully remove the supernatant. The cell pellet should appear a green colour (if you have enough cells present to see a visible cell pellet). Vortex to break up pellet.
  • Resuspend the cells in prewarmed tissue culture medium containing FCS.
  • Leave the cell suspension to one side, protected from the light until ready to set up in culture.
  • When ready to use, count the stained cells and dilute to the required cell number for your assay in complete tissue culture medium.

What controls do I need for measuring cell proliferation using CFSE?

Your analysis can only be done if you’ve got the correct controls for your experiment. Always include an unstimulated control population. This will give you the fluorescence of the undivided population. For a time-course, you will need the unstimulated control population for each day , due to small amount of dye leakage over time. Always include a positive control to prove that the cells can respond and can proliferate. Stimulate the population with an agent which will elicit a response. For example, PMA + Ionomycin for T lymphocytes. Initially, you will need to do a time course to define the optimal time(s) needed to measure and compare your proliferative response. If you leave cells in culture too long the dye may have undergone too many divisions and diluted to the level of background cellular autofluorescence. I recommend that you include a viability dye in your sample prior to analysis to exclude the dead cells present in your cultures. If not correctly excluded dead cells can cause the generational peaks to widen and become less distinct.

Cell cultures can be set up in 96-well plates, or in test tubes (loosen the lid so that the CO2 in the incubator can enter). I generally set my cultures up in 5ml sterile, round-bottomed test tubes and culture in the incubator at 370C. I could then remove a sample aliquot, at each time point, to further stain and analyse on the cytometer.

Do you want to monitor cell division and phenotype?

The power of the technique is that you can add other markers to identify subpopulations and observe their proliferation. In the diagram I show below, human lymphocytes labelled with CFSE can be counter-labelled with surface markers (CD4, CD8) prior to analysis.

Figure 2

This is what the flow cytometry plots looked like, when analysed, after 7 days of culture. The unstimulated population had been cultured in medium alone with 5% FCS. Staphylococcal Enterotoxin B (SEB) was used as a positive control culture of a replicate sample.

Figure 3

As you can see, there are a small percentage of cells responding in the unstimulated population. Therefore, this is not a true negative control. Something is causing the cells to respond by proliferating. This is observed in the CD4 subset, but not the CD8 subset. When I substituted Human AB serum for the FCS as a serum source, there was no proliferative response seen in either population. The CD4 cells were responding to the FCS (Foetal Calf Serum) as a foreign protein, but they do not respond to the “self” Human AB serum. The ability to gate on subpopulations and display the data in a two-dimensional format moves populations off the axes and allows easy identification of population clusters. This is so important in troubleshooting.

How can I monitor cell division, phenotype, and cytokine production simultaneously?

You can add more surface markers and intracellular markers. For example, intracellular cytokines will give information on the maturation and functional response of cells to antigens presented to them in culture.

Figure 4

The plots above show the analysis of CD4 lymphocytes, taken from a human blood sample, and cultured with/without PMA and Ionomycin for 7 days. CFSE labelling of lymphocytes was performed on Day 0. All plots are gated on the CD4 cells. Plot A1 shows the lack of proliferation in the unstimulated control sample on day 7 of culture. B1 shows the proliferative response of the stimulated sample on day 7. Plots B2 and B3 show the IL-2 and IFNγ production in CD4 cells respectively. B4 shows the IL-2 and IFNγ production of the CD4 population. Plots A2, A3, A4 are the corresponding day 7 unstimulated control populations. Note that the unstimulated population is not proliferating! Histogram/Quadrant statistics are displayed.

How can you analyse proliferation data?

Visually, you can see that a proportion of the population is responding, but you may want to quantify this response. There are several analysis packages which will do this for you. I have shown an example below. Note that the unstimulated population is not proliferating!  Precursor frequency is a useful way to compare proliferative responses. Proliferation Index can also be used. The Precursor Frequency is the proportion of the total cells calculated to have been present at the start of the experiment (derived by back-calculation according to the model) which have then gone on to “true proliferation” during cell culture.

The Proliferation Index is the sum of the cells in all generations divided by the computed number of original parent cells theoretically present at the start of the experiment. It is a measure of the increase in cell number in the culture over the course of the experiment.

Maria Graphics 3

Using Precursor Frequency, you can compare proliferative responses to a range of antigens or other chemical stimulants. You can quantify and rank responses. I have only shown the negative and positive control data, giving you the background and maximal responses.

What are the practical considerations and pitfalls in measuring proliferation using CFSE?

To achieve good flow cytometry data, you need to pay attention to your sample preparation and the way in which you set up your instrument. I always check my cell cultures, under the microscope, before I analyse them. By viewing the culture, you see how many cells you have, and how they look. Resting lymphocytes, freshly set up in culture, will appear small and round. Very uniform in shape. If unstimulated, they will stay like this all the way through the time course (7 days in my case). Stimulated “blasting” cells will appear larger, and irregular in shape. They will scatter considerably more light. When setting up your Forward Scatter (FS) vs. Side Scatter (SS) plot you will need to be aware of this. Using the instrument settings, set the “unstimulated” resting cells in the lower left-hand corner of the plot. Blasting cells will appear from around day 3 of culture onwards and will appear in the upper right side of the plot. I find it useful to use “Log SS” as this will allow blasting cells to appear as a cluster. Set a generous gate to encompass both the small and larger cells, excluding debris and clumps. Do not gate out your proliferating cells! By day 7 of culture the stimulated cells will be returning to the resting state and will scatter less light again. The microscope is essential as a reality check! Use it.

What proliferation dyes, other than CFSE, are available?

The examples I’ve shown are all using CFSE, but there are alternative dyes available on the market. These come in kit form and are easy to use. The Violet-excited forms of the dye generally show less “autofluorescence”.  This helps to distinguish the cells which have proliferated more (greater dye dilution) from the background fluorescence of cells. You need to try them out to see how they perform in combination with the other fluorescent tags you’re using in your assay.

I’ve listed a few suppliers of proliferation dyes:

ThermoFisher have a range of dyes, excited by different lasers.



BD offer also have some kits available.


CFSE                = Carboxyfluorescein succinimidyl ester

CFDA, SE          = 5(6)-Carboxyfluorescein diacetate succinimidyl ester

FITC                 = Fluorescein isothiocyanate

PBS                  = Phosphate Buffered Saline

DMSO              = Dimethyl sulphoxide

PMA                = Phorbol myristate acetate

IL-2                  = Interleukin- 2

FCS                  = Foetal Calf Serum

IFNγ                 = Interferon gamma

CD                   = Cluster of differentiation


Related Posts