Check for the number of cells from sample to sample. If cell number varies the staining conditions can vary a lot thus causing a different fluorescence signal. For example you can typically check it in the control: if you have not normalized the number of cells per sample late time-points can be shifted toward the zero (reduced fluorescence). The opposite is true also when you have less cells. It is a good practice before starting the processing the staining protocol to count and process the same number of cells for every time point in your kinetics. Take care that this reasoning is true not only for DNA dyes but also for antibodies: saturation must be always ensured.

Also check that the pressure on your flow is stable - if not it will cause sample to sample shift if you also see changes in forward/side scatter sample to sample then this may well be the issue

Thanks mario. I did think about doing that but decided not to, perhaps because of convenience :) I am performing screens and validations just now and thought I will do that once I have lesser number of samples. However, if the dye is in excess (if at all) then the cell number shouldn't matter fr optimum staining, right? I have been trying a few things and a few other problems that I was facing were resolved using more dye (resuspension of cells in a bigger volume). I guess I will dilute them further from now on and see if that works. But I do think that what you have suggested is the right thing to do and I intend to do it once I have short-listed my parameters and targets. Cheers.

I agree with Mario. You need to count the number of cells in each individual sample and then adjust the amount of PI accordingly. Since the PI you are adding is not excess as in case of antibody staining, it will intercalate into the DNA by the amount you are adding. Meaning if you have less cells and you nevertheless add the same amount of PI you will get the peak shift to the left and the other way round for more cells. I hope this helps! Good luck, Nicole

Agree with Mario and Beckie. Also set some minimal acceptance criteria to reduce inherent variation. say for example - if you are looking at lymphocytes - gate atleast 1000 cells etc.

Besides above suggestions I would suggest to consider below issues:

1. Do you use a negative and positive control?? This is necessary for objective evaluation of the shifts you are observing.

2. Also, check if your system is connected to a UPS line, +/- 15 A changes in electrical currency can cause this kind of changes.

Thank you all......I shall try out your suggestions.

Saurabh, first try to set your controls as suggested by mario, i.e. different number of stained cells of control and check. Later go for 1-2 samples with different samples. If you do not solve this problem first, what's the use of doing many samples?

Sometimes if you don't disconsider cell aggregates you can see some weird things during analysis. Beside all usefull info already posted, make sure these cell aggregates are not interfering in the end. You can do this simple by creating a dot plot FSC x FLW and gating your events in the FLW low population.

Shifting during the run could also be due to a problem with the detector itself. Had problems with PE stained samples (FL2) on a FACScalibur. Problem was fixed when the replaced the detector.

I am agree with Nicole and Mario. You should give appropriate incubation time also after adding the dye. and also run the QC Beads before running samples to check that your instrument is working well or not.

I agree with Govind Gupta, the first control is to add beads directly to your sample in order to check the detector and the stability of the laser beam

You don't mention details of your cell origin, staining protocol or whether you are interested in cell cycle arrest or apoptosis? The usual cause of this type of drift is variable cell numbers in samples and short PI incubation periods so that your DNA staining is variable as it has not reached saturation before starting acquisition. Aggregates blocking the SIP may also effect profiles but this can be prevented by filtering.

Have you considered including an internal standard control eg chicken erythrocyte nuclei ~2.5pg or trout erythrocyte nuclei ~5.2pg?

If you also have CEN or TEN stained separately you can run them when you see apparent instrument drift if they don't move on reacquisition it's a sample staining problem if they show drift during acquisition period it's the instrument.

With cell activation, there might be a shift in cell population in flow cytometry analysis with instument setting, controls etc. being the same (Chapoval SP et al. Clin Immunol. 2009 Sep;132(3):371-84).

1. I do have negative and positive controls. I also give ample time (overnight) for the dye to get incorporated. What is weird is that even when the cell numbers are similar, there is this shift. Now, most samples run fine, but in some cases, I see this shift along the axis. I get nice cell cycle profiles (eg G1, G2 peaks are fine), but the peaks 'move' along the axis in some cases, making it difficult to gate.

2.I am interested in apoptosis and it becomes difficult to maintain the gating for quantification.

3. The shift is not during the run per se, but between different samples. I have used this protocol in the past and it has been alright.

4.The cell line is murine mesothelioma.

Thank you all once again for your valuable suggestion.

Hi Saurabh

I've seen this phenomenon too back when I used to do CCA on Adeno infected cells, however then it certainly wasn't due to limiting PI nor variable cell numbers. I assumed it was some stability within the system but never got further than that. As it appears you may not get an immediate fix from this blog, I'd strongly suggest doing what Martin said and spike your samples with chick erythrocytes as an internal control (you can buy these from BD). Then you can simply 'align' each profile based on the internal control. This will have the advantage of working right now and will fit into what your're doing with no extra optimisation, and in the long run if you manage to sort out the drift issue you will be able to dispense with these. The only caveat I can think of is that if you're quantifying sub-G1 then the chick erythrocytes will contribute to this signal.

Good luck

Al

Hello all....solved the problem! Just thought of sharing my solution in case it may help someone else in the future. So it seems that the cell numbers, although seemingly similar, weren't so. I counted cells in every sample and processed the same number of cells for flow analysis and voila!! No shift in the peaks and cleaner profile. Thank you all once again for all your advice and suggestions; I did learn a lot :)

Hi All,

I am doing cell cycle analysis of yeast cells stained with sytox green. I see shift in the peaks along x-axis (sytox green Area) in wild type and mutant cells, with voltages, thresholds and other parameters constant. There is a similar shift in the peak in the FSC. The cell number is same for all the samples. Also, I analyze the frequency of G1 , S and G2 populations using flow jo software, do you let the software alone to fit in the data in different phases of cell cycle or should I do it manually. However, when I do it manually, it becomes difficult to decide at which position on the X-axis should I assign G1 and G2 peaks.

Thanks

Hello Mr. Saurabh Dayal . 

Parameters need to check - 1. Any technical error. 2.Pipetting error 3. PMT voltage 4.Cell number 5. Proper control. 

If still problem persist , please share in details after repeating the experiment with all parameter checked experiment.

Best regards, 

Babu