Hi Brenda,

The gates are set only on light scatter properties. In the DLD1 line, the gate is on the population with higher Front Scatter. In the HT29, on the lower Front and Side Scatter. I am not able to help in any sort of explanation as to why that gating strategy would be used.

Not sure if that's any help at all.

Hi Brenda,

I have not handled these mentioned celllines. But have encountered similar issues with cellline that differentiates. 

You see to standardize 2 important parameter.

1. plan your experiments with either differentiated or non-differentated cells (choose according to your need).

2. adjust your voltage settings such that it fits into your dot plot. Ideally you can plot Contour plot which would enable you to select the dense region in your plot.

Since you have already acquired the cells.

you can change your dot plot to contor plot and then analyse for your desired markers.

Basically your get 2 populations, but your desired marker would be present in only one population.

Note: Such heterogeneous populations may also arise if your trypsinization is not good.

Do you have another picture, maybe with different colors (i.e. a pseudocolor plot)? That would help you comment on how you chose the cells of interest based on the population distributions.

Otherwise, with what's above, I agree with Elizabeth, you can really only describe it by saying that the cells were gated based on their light scatter properties. Intact, DLD1s were defined as FSC hi cells and HT29s were defined as FSC lo cells.

You could also explain that the samples were from cell culture and you (presumably) analyzed the cells toward the end of the growth phase. You were most likely just looking for the cell population and you wanted to gate out the debris and clearly dividing cells using light scatter properties so the later antibody gates wouldn't include the debris and doublets.  This is just speculation as to your past intentions, though...

Hi Brenda,

I am agree with the  Mr. Dias suggestions.

This might be a problem of differentiated and non-differentiated population which is due to non-synchronized culture because in the given figure the FSC value (which tells about size of the cell) is just double in gated population of plot one than two that may be due to the cells are in differentiation stage.

if these plots are control and treated separately then may be your compound which you are using for treatment is effecting the generation time of your cells.

you can convert SSC and FSC parameter in the log scale when you set the population. it will constrict your population in one place.

You can use any cell specific fluorescent markers to find out your population of interest.

May these suggestions will help you.

All the Best

Regards,

Govind

Thank you all for your answers.  

The experiment was based on showing the purity of CD133 positive cancer stem cells obtained after magnetic separation.  Ideally I should have cancer stem cell markers to help me decide where to gate in the cell lines.  Nonetheless, the data is what it is.  I do have a question though, since my study is based on cancer stem cells (which are undifferentiated cells) I would expect them to be smaller in size than their differentiated cells, am I correct? and if so, it would make sense to gate on the population with a smaller forward scatter as it is more likely that the cancer stem cells reside in this population, as opposed the the population that is currently gated with a high forward scatter?

Please let me know what your thoughts are.

Thanks again

CD133 is a marker that is expressed in stem cells, endothelial cells and epithelial cells.

you can select other cellline (positive control) which do not differentiate as your positive control. 

Mahtab, unfortunately not :( This work was done five years ago, if I had known then what I know now I definitely would have included live/dead cell stains and used more markers to help me decide where to gate. However, this is what I have and somehow I have to try and justify what I have :(

Hi Brenda,

Without markers, it's pretty hard to say what cells are what in the scattergrams.  I assume from the light scatter that these cells are not perfectly spherical in solution? That would add to the disperse scatter properties.

Were your magnetic beads nano- or micro-?

If microbeads, then the scatter could also be influenced by beads which were not removed. I've had even nice, proper lymphocytes show up with extremely high side scatter after use of micro beads. No bead removal process is 100% (chemical equilibrium!).

Nanobeads should not affect scatter much at all. I've seen them show differences in receptor density on cells when examining the target receptor in a positive selection.

I am not convinced that your assumptions about size and non-differentiation are correct in cultured cell lines, but I am not an expert in cancer stem cells. Were these scattegrams from pure cell line immunomagnetic selections, or spikes in a different cell line?

Soap box lecture: My first research technician position out of college was with a Professor of Biochemistry who insisted that I keep my laboratory notebook "like chemistry 101, with a title, objective, materials and methods, data, results, conclusions". I have never regretted adhering to his words in the 27 years since his instructions. Even if at the end of the day my conclusion is "this was a failed experiment, look at (blah, blah)", it helps years later when I look up something in my lab notebook. I see now that I have not been writing enough about my gating strategy, either. ;)  End of soap box.

Hi Elizabeth, 

I used microbeads for my experiment.  The cells I have shown do not have beads attached though. The cell line on the left represents a Dukes' stage C tumour and is highly differentiated that's why I thought that perhaps the one population with a small forward scatter might represent lesser differentiated cells than those with a high forward scatter. It's difficult to say without differentiation markers. Do you think I should gate both the populations for DLD1? Perhaps it would be better as it can't be said what cells are what.

Thanks 

B

Sorry, I really have no idea, Brenda. I have no experience with non-hematopoietic tumours/cell lines.

best,

Elizabeth

Hi Brenda, i think this type of "double" population plot is quite common for cell lines. That said, assuming these are homogenous cultures, the leftish population can be a group of apoptotic cells. The apoptotic phenotype is causing the side scatter to increase and forward scatter to decrease. Typically seen in compromised membranes and nuclear fragmentation. It is therefore difficult to rule out without any live dead stain.  Did you do any markers for these runs? If you can back gate, meaning look at marker first then size, it might be easier to justify.

Hi Brenda, Looks like that there are two cell populations in the first dot plot. A density plot would have given a better discrimination of the two populations. Also the cell concentration is quite high and a diluted specimen may be of use.  There seems to be a larger and a smaller cluster of cells (first dot plot) If you have the original files try creating daughter plots from the gated populations and see what kind of histograms and dotplots they derive. Are there any fluorescent dyes used in the experiment. If so have a look at the data in those backgated from the original (shown above) dot plots

Hi Brenda, Do you happen to have data set for cells analysed before CD133 MACS and after CD133 MACS, if you can see a difference in the dotplot for FSC vs SSC then you can say that these cells have a different size/complexity. MACS sorting (so far in my experience) has never been 100%, so there will be residual cells in every sort.

Hi Michelle

Yes I do have these data (: the DLD1 forward versus side scatter is still quite heterogeneous after separation however, the population on the left becomes promiment (I have attached an  image). Could I then justifying backgating of the left population as my population of interest? 

Thanks

B

Hi Brenda, yes but you have to show clearly that it is an enrichment by a numbered percentage. Ideally, you will need to apply the same gating template on both data set (pre and post MACS), and the same total number of events were recorded. Hope it helps.

Thank you Michelle, your suggestions have helped a lot.  Thank goodness I was able to access the FCS files so that I could regate everything. 

To everyone who has provided answers and suggestions, a huge thank you to you all. It has been an extremely stressful week but you have all helped me sort out this problem. Thanks again (: 

Hi, an interesting topic, I think. Though you got a lot of answers and hints I still would like to ask you, did you do a doublet-exclusion?! Because after sorting just the left population mostly appear, the right population could be clumps of cells and therefore doublets?! FSC-A vs. FSC-H or if you used a color you can gate that one against SSC-W and clearly can see doublets.

Good luck with analysing.

Hi Jochen,

Unfortunately the parameters were not set at the time the data was acquired to allow me to analyse the singlet population afterwards. I acquired the data on the BD Calibur. However,  for my other experiments I acquired data on the Fortessa and gated on just the singlet cells. 

B

Hi Brenda,

that´s good, making it later easier to analyze :-)

cheers