Dear all,
I am working on an antigen presentation assay now. The cells do proliferate well after 3 days of culture, esp. when we observed them under microscopy. We can see several cell clumps and the medium color also turned yellow after 3 days of culture. However, when I tried to quantify the dilutional effect by Celltrace violet or CFSE using flow cytometer, the peaks of different generations could not be distinguished well. I have tried different conditions, such as different CFSE or Celltrace violet labeling concentrations, but the peaks still lumped together (as is shown in the attached file). I am not sure if it's because of the parameters that I needed to set up during Flowjo analysis or it's because of the labeling is not good enough? The file shows the data from 0.5uM CellTrace violet labeled T cells. I was stuck by this CFSE/ CellTrace Violet dilutional analysis for more than 4 weeks. I will be super grateful if people can help on this issue.
Other specific questions related to the Flowjo analysis:
a) How many events will you usually collect for this type of FACS analysis? I usually collected around 8000-10000 T cell events in this assay. Should I collect less events so that will help to make the peaks distinguishable?
b) When using flowjo to analyze the data, how many peaks will you estimate for a 3-day culture? 4 or more or it does not matter?
Thanks for your help!!
Dear Ichun,
I don't know what your institutional access to journals is like, but this is a very nice thorough protocol for CFSE to monitor lymphocyte expansion in response to Ag presentation:
https://www.ncbi.nlm.nih.gov/pubmed/17853860
Unfortunately, the parameter I thought was most likely leading to your problem, changing the concentration of CFSE, you have already tried, and turns out is not the problem.
However, I can answer (a) and (b) for you. The number of expected peaks does not depend on the number of days, it depends on the average number of times your cells have expanded. So you need to know your approximate doubling time to predict peaks within a time duration.
As for (a), I would recommend collecting as much data as possible within reasonable duration of time. You can always analyze a smaller subset of data but you can never go back and collect more data on a sample you have discarded. Either way, fewer counts will not lead to better separation. What leads to good separation are the proper experimental parameters.
You mentioned your cells were clumped. How are you dissociating them before they go on the flow? Clumped cells need to be gated out, otherwise if two cells overlap, the fluorescence readout is ambiguous. Also, autofluorescence may be a problem from dead cells. Adding a viability dye will help you select only live cells.
Finally, it is not ideal that your media has turned yellow by the endpoint. perhaps you want to seed fewer cells at the start of your experiment or seed same number in a larger well/flask?
Good luck!
Hi Ichun
What is the CV of the original peak that you are sorting? There is a method called the sort gate reduction method which sets a channel width for sorting cells that do not vary more than two-fold in fluorescence intensity. So if your original peak has a large CV than it will be difficult to resolve progeny peaks post-culture. Also the more homogeneous your population the better your results will be. Keep us posted!
Analysis of growth kinetics by division tracking Nordon et al.
Firstly, when we treated purified primary T cells with CellTrace Violet (CTV) with 1 µM or less, we could not distinguish the division, separate peaks. But I increased the concentration of CTV to 5 µM we found 5/6 peaks as expected from the length of time incubation (5/6 days) in culture, and with no cell toxicity. Ideally, it is recommended to treat cells with CTV for 20 min in the dark at 37 ℃. After washing cells, distribute aliquots of stained cells into culture plates, personally I would prefer to use 96 well plate (1-2 x 105 cells/well) than 24 well plate (1-2 x 106 cells/well) because we observed the cells in 96-well plate have more proliferative capacity into divisions than they are in 24 or 12-well plates. For primary cells, you need to stimulate cells with a proper cytokines/growth factors, otherwise, cells may not undergo proliferation well. Another point, dead cells/ clumps hugely affect cell staining with CTV, and it is important to count cells and find out the percentage of live cells (recommended to have > 80%). Otherwise, you need to remove clumps/dead cells prior the staining or use Dead Cell Removal Kit. Moreover, from my experience, it is hard to see any cell generations/peaks in cell lines with this technique. This issue is not relevant to FlowJo, but rather with cell labelling with the stain. You may need to optimize the concentration of your stain and culture conditions. Good luck
What if the number of unstimulated T cells are less than 10000? Can I use the "relative to mode" feature to adjust the number of undivided population to the divided one?
I acquire say 1500 CD4 T cells from the unstimulated well vs 13000 from the stimulated one.