One of the most widely used assays to determine apoptosis by flow cytometry is the estimation of fractional DNA content (aka sub-G1 assay). During apoptosis, genomic DNA is cleaved into smaller fragments, each approximately 180 bp (and multiples of it). This is a specific marker of apoptosis and therefore can be used to quantitate apoptosis using a number of assays. Using flow cytometry, propidium iodide stained cells will stain less intensely and show a peak below the G1 peak. It is the Sub-G1 peak.
The sub-G1 assay for measuring apoptosis is easy, rapid, reliable, reproducible, and cheap and is widely used. Sometimes we forget to think about what we are doing when using a seemingly simple assay. However, you need to understand the mechanics of the assay and the apoptotic process, otherwise you could over- or underestimate your apoptosis results!
But don’t freak out, now it is time to better understand your sub-G1 data.
Make sure they are dead cells: Fix the issue!
One of the most severe errors in sub-G1 assay is the misclassification of events with lower DNA stainability as single apoptotic cells. When performing the sub-G1 assay, you extract the fragmented, low molecular weight DNA from apoptotic cells, so they will be located below G1/G0 peak after propidium iodide staining. You may think that everything below the G1/G0 peak is an apoptotic cell…but this is WRONG!
When you use the detergent/hypotonic solution to extract DNA fragments, you lyse the cells, and you extract oligonucleosomal fragments (and multiples of it) from apoptotic cells. But that’s not all: you also extract other things from apoptotic (and non-apoptotic) cells. These “things” may be nuclear fragments, individual apoptotic bodies, chromosomes or chromosome aggregates from mitotic cells and can be mistaken for apoptotic cells.
To circumvent this problem, you could do a gentle permeabilization of the cells with detergent but in the presence of serum or serum albumin to protect cells from lysis. However, these cells are still extremely fragile and pipetting, vortexing, or even shaking the tube could cause lysis.
The solution? Fix the cells using 70% ethanol instead of detergent. Although this doesn’t extract the fragmented DNA, the degraded low molecular weight DNA will still be extracted during subsequent rinsing and staining steps.
Clean up your mess before acquisition
You can also improve your data analysis during acquisition. You can set gates to keep everything that is too small out of the analysis. You can also increase the threshold of DNA detection to prevent counting particles with a very low DNA content. Although both of these adjustments may lead to underestimating the percentage of apoptotic cells, this error would be constant and better than overestimating the percentage of apoptotic cells– counting non-apoptotic events as apoptotic cells.
The bias towards counting non-apoptotic events as apoptotic cells is particularly pronounced when a logarithmic scale is used to display DNA content on the histograms. This is due to the fact that the log scale allows you to detect events with minimal DNA content, let’s say 1 or even 0.1% of that of G1 non-apoptotic cells. To improve your results, use a LINEAR SCALE to exclude objects with minimal DNA content from the analysis.
Zombieland: is everything dead in sub-G1-land?
Other cells can appear to be apoptotic even when they are not. Cells with lower DNA content (hypodiploid cells) and cells with different chromatin structure (e.g., cells undergoing erythroid differentiation or even necrosis) will also stain less brightly. So, take that into account.
You can check if you have hypodiploid cells using a calibration standard (DNA check beads) or diploid cells as internal control. Morphological (microscopic observation of apoptotic bodies) or specific demonstration of DNA breaks (TUNEL assay) should be used to confirm apoptosis before quantitative analysis by flow cytometry.
Sometimes apoptotic cells are tricky
If apoptosis occurs during the G2/M or S phase the apoptotic cell could actually be positioned above the G1 peak after extraction of oligonucleosomal DNA fragments. This would make them indistinguishable from cells in phase G1 or S.
In this situation, a more extensive DNA extraction (using high molarity buffers) will improve the resolution of the sub-G1 peak. If that doesn’t work, it looks like it is time to move on and try a different apoptotic assay. For example, the TUNEL assay (based on in situ DNA break labeling) will tell you cell cycle specificity of apoptosis.
Duration can mess with your frequency. Know the limits and go beyond
Apoptosis is a kinetic event. You can not assume that the frequency of apoptotic cells in a cell population (by any given method!) is equal to the incidence of cells dying by apoptosis. For example, if you observe an increase in the percentage of apoptotic cells, it could be that the number of cells undergoing apoptosis was increased OR that the same number of cells underwent apoptosis, but the duration of apoptosis was longer. Or a combination of both. Alternatively, when the duration of apoptosis is shortened, the percentage of apoptotic cells will also be reduced even if the incidence of apoptosis remains the same.
If you really need to know the incidence of cell death, you should count the absolute number (not the percentage) of live cells in the culture and compare it with the absolute number of cells in the control. Remember to take into account the rate of cell proliferation and pray that the treatment that induces apoptosis does not affect the duration of any phase of the cell cycle.
It seems not so simple to solve the problem. Anyway, when interpreting and discussing your results, keep in mind this situation: the frequency of apoptotic cells is not always equal to the incidence of apoptosis.
Please, don’t lose your apoptotic cells!
Prolonged trypsinization, mechanical or enzymatic disaggregation from tissues, and extensive centrifugation steps can all lead to preferential loss of apoptotic cells during sample preparation. Don’t let this skew your results! If you have adherent cells, collect the cells left on the plate but don’t forget the ones that have been released into the media. Similarly, density gradient (e.g. using Ficoll-Hypaque solutions) separation of the cells may result in selective loss of apoptotic cells (live and dead cells have different densities). Make sure you collect both.
And what if you do not detect a sub-G1 peak?
“The lack of evidence of apoptosis, detected by a particular method is not evidence of the lack of apoptosis” (Zbigniew Darzynkiewicz).
The degree of low molecular weight DNA extraction varies markedly depending on the extent of DNA degradation, the number of cell washings, pH, and molarity of washing/staining buffers, type of cell, type of stimuli, and of course, duration of apoptosis.
In particular, timing is important. The time window that you use to analysis a hallmark of apoptosis is very important. Not all the hallmarks of apoptosis appear at the same time or stay there forever. If you don’t detect a sub-G1 peak, you may have collected your cells too early to detect the DNA fragmentation. Optimize the time collection using a time course. A positive control would be really useful for this.
In addition, not all apoptotic cells exhibit all typical features of apoptosis. And do not forget that there is not only one type of apoptosis, there are many – with different hallmarks! . In fact, a type of apoptosis has been described with DNA degradation that stops after generating 50-300 kb fragments meaning there is no internucleosomal fragmentation. In this case, the sub-G1 method is not the best assay to use. If you do not see a sub-G1 peak but believe apoptosis is occurring, analyze your cells with a different method that relies on other hallmarks of apoptosis.
As you see, there are many considerations to keep in mind when performing s sub-G1 assay. Although the sub-G1 assay is easy, rapid and cheap… use it with caution to make sure it is also reliable and reproducible.
- Riccardi C et al. (2006) Analysis of apoptosis by propidium iodide staining and flow cytometry. Nature Protocol. 1(3):1458-61.
- Wlodkowic et al. (2011) Apoptosis and Beyond: Cytometry in Studies of Programmed Cell Death. Methods Cell Biol. 103: 55–98.
- Darzynkiewicz et al. (1997) Cytometry in Cell Necrobiology: Analysis of Apoptosis and Accidental Cell Death (Necrosis). Cytometry. 27:1–20.
- Vermes et al. (2000) Flow cytometry of apoptotic cell death. J Immunol Methods. 243: 167–190
- Galluzzi et al. (2012) Molecular definitions of cell death subroutines: recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death and Differentiation. 19:107–120.
- Oberhammer et al. Apoptotic death in epithelial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation. EMBO J. 12:3679-3684