I am analyzing and sorting human fibroblasts with GFP transfected into them. They are not fixed---just trypsinized, harvested, washed and suspended in PBS. Recently we have started to see an odd problem where as they are being run through the cytometer, the general cell population in the same sample shifts over time towards the right (y-axis APC-1 vs x-axis FITC ) making it seem like there are two (albeit not very distinct) populations. This makes it harder to gate the general population and quantitate the GFP positive population. We see that the second population is of smaller cell sizes that are at the bottom of the falcon tube which carries our sample. I don't know where this second population might be coming from. I am not a flow cytometry person at all---a cytometry core tech runs our samples and I don't quite understand why we are seeing this shift. Is this some sort of autofluorescence, something with the machine or my technique when harvesting the cells that damages or ruptures them? Any suggestions, pointers are welcome. My cells are not contaminated and another person in our lab is having a similar issue.
Just a proposition : try to find out which samples have been analysed before yours, some dyes pollute the pipes (bodipy for example) , and when you pass a new sample, cells are progressively labelled by these dyes...
Do you have EDTA in your buffer? could be cell clumping over time... Have your flow cytometry tech check for doublets (FSC-W vs FSC-H graph). also, autofluorescence is high in GFP channel so that if your cells grow or change shape over time, it is possible that they shift in the graph, going towards higher GFP values, while APC channel is not so much influenced by autofluorescence. Make sure your PBS is isoosmolar. Normally we use PBS/BSA 0.5% or FBS 2%/EDTA 2mM as a buffer for flow cytometry. Good luck!
Sounds like good advice (clumping). How long do your samples wait before being run - has this increased (giving more time for clumping)? Since FSC v SSc is nearly always collected for any sample, check for shifts in FSC (size). Also, are the samples mixed before running - maybe a vigorous vortexing might cure (if your cells are reasonably robust).
It could be clumping as mentioned (doublets). Control proper re-suspension or gate them out post FS/SS but pre-fluorescence. Additionally, there may be a reaction in your tube during collection. Quenching of GFP or APC by an ongoing/incomplete fixing process due to contamination of tubing with fixative or chelators. Fixation can modify GFP/EGFP signals. Finally, make sure you have not changed tube suppliers. We find cheap tubes leach fluorophores into samples leading to shifting populations during sample runs (1-5 min). Especially a problem using Pacific Blue collection channels.
Just a proposition : try to find out which samples have been analysed before yours, some dyes pollute the pipes (bodipy for example) , and when you pass a new sample, cells are progressively labelled by these dyes...
my suggestions would be:
1. Cell clumping, as it was noted
2. Unbound and unwashed fluorochrome in the sample, that goes always with 5-60 seconds lag depending on machine and aquisition params
3. Degradarion of any dyes. It can be degradatin of tandem dyes if present that happens very often under the light. It can be reddind of GFP in some photochemical reactions. In can be spectral shift due to pH change.
4. Autofluorescence change. For example, undound valencies of aldehide can bind some dust inside cytometer. Or fixative (such as glutaric aldehide) polymerization under light.
You can try analyze a different sample (eg fibroblasts stained with different Abs, or a different cell type). If problem persists, it should be the machine.
In any case I would suggest a thorough cleaning of the cytometer!
HI Aditi
If you could post a rough protocol and your gating strategy (and perhaps even a pic of this mysterious population) I think that would certainly help everyone here to rule out a few things (doublets etc).
Al
A guy in my lab suggested that the harvest protocol might be shearing the cells or rupturing them and those damaged cells/ debris can cause this 2nd population to appear towards the end of the run. Is this possible?
Al: Yes I will post a picture soon. My protocol is very simple. The cells are transfected with a GFP vector and 24 hrs later harvested with trypsin and media. Cells are pelleted for 5 mins at 700 rpm and media is aspirated. Cells are then resuspended in PBS buffer and taken over to flow core on ice. I finger flick the tube before loading sample into the cytometer for analysis and sorting. So there are not dyes or antibodies or staining process. It is endogenous GFP that has been transfected into the cells.
Thanks all for your responses. Will put up a picture soon.
Have you considered that the cells may be clumping due to secretion of DNA (which is like glue) by dead/dying cells. If you transfected by electroporation you may have considerable cell death still at 24h. Consider using some live/dead fluoresecent dye (maybe 7-AAD) and gate on your live cells only (although if cells are leaking DNA they will stick to live cells too)
A DNase I treatment prior to FACS analysis could solve a dead cell clumping issue.
If the samples are shifting intensity during running, it may be a problem with the detector. We had issues with our FACScalibur a few years ago where the PE population in a histogram would start shifting during a run. They had to replace the FL2 detector.
I suppose that if you sort them, you can then visualize the population under a fluorescence/confocal microscope to decipher whether a new population is, cell debris or degradation of dyes, etc…. as pointed previously. Furthermore, you can support the idea of cellular clumping under visualization of the different sorted gates in order to verify the previous issue. Good luck!
For better understanding of the problem pls attach few plots.
Just found the same problem with HEK293 cells sorted with 60 psi - cells are dying under pressure while not taken inside sorter’s tube. This caused intracellular pH shift and spectral drift of mTPF1. The problem was solved by decreasing pressure to 30 psi.
I have the same problem with hek 293 cell suspension using Novocyte instrument.
- Badges
- Science method