Page 1 of 1

Signal change with time

PostPosted: Fri Feb 12, 2021 4:35 pm
by TroyLim
(224.16 KiB) Downloaded 202 times
Hi everyone,

I am planning to use CyTOF to analyse primary Chronic lymphocytic leukaemia cells in the setting of a clinical trial. While optimising my workflow, I noticed a change in signal in my most recent experiment, particularly affecting 144Nd & 145Nd. Attached is 2 plots of the same cryopreserved CLL sample, analysed recently and several months ago.

In my more recent experiment, there appears to be several very bright CD81 and IgD events, more than the experiment I performed several months ago. I repeated the experiment again and fast spinned my antibody cocktail (I was initially worried that there could be Ab conjugates) but still have the same result with these highly intense events in CD81 and IgD. I should also mention that the antibodies and my workflow has been exactly the same in both experiments.

Am I being overly pedantic here, and this could just represent batch variation across runs, or could there be an issue with my antibodies (my antibodies are <1 year old)?

Thank you


Re: Signal change with time

PostPosted: Fri Feb 12, 2021 5:27 pm
by mleipold
HI Troy,

Basic question: what cells are we looking at here? Total Ungated? Or have you already gated down to Live Intact Singlets?

Also, I would probably recommend filtering your daily mixed antibody cocktail, rather than just centrifuging the stocks. 0.1um spin filters are often sufficient to remove any potential antibody aggregates.


Re: Signal change with time

PostPosted: Fri Feb 12, 2021 5:31 pm
by TroyLim
Hi Mike, these are primary CLL cells. I have gated on Gaussian parameters and live singlets for this.

Do you think these high intensy signal are antibody aggregates?


Re: Signal change with time

PostPosted: Fri Feb 12, 2021 5:39 pm
by mleipold
Hi Troy,

It's entirely possible that they are antibody aggregates.

One other thing you might examine: you have the CD19 on the y-axis, and then show CD81 and IgD on the x-axis. At least to my eyes,it looks like the CD81-bright and IgD-bright events might be the same cells, based on Frequency.

You might gate on those CD81-bright cells, and then see which other markers are positive on them. Is it only IgD? Or are there a lot of other markers that they're bright for, that might not make biological sense (eg, they might also be bright for CD56 or CD3)? If so, then that would strengthen your case that they're junk of some sort.

Finally: you say that your antibodies are the same between these 2 experiments. By that, do you mean that they are the same exact tube of antibody? And are these in-house conjugates, or are they from Fluidigm?


Re: Signal change with time

PostPosted: Fri Feb 12, 2021 5:59 pm
by GregBehbehani
Hi Troy,

My quick impression from your plots is that this is probably biologically real, but you're not really showing us enough to know. I would agree with Mike's comments as well. In addition I would say that if you do have aggregates, you should be able to see "free aggregates" (events that wouldn't get into your singlet gate due to low Ir or cell length), if you aren't seeing any "free aggregates," then I doubt that aggregates attached to your cells is the explanation.

Other things to watch for are dead cells and non-specific antibody staining. Dead cells can pick up non-specific antibody binding, and it can also rarely happen due to what is probably intracellular granules that interact with the labelling polymer; this can be blocked with heparin (Adeeb Rahman has a nice paper about this).

Just FYI: the diagonals at the high ends of both axies are due to spillover of 144 into 145 and 145 into 144, and are expected.

Good luck,


Re: Signal change with time

PostPosted: Fri Feb 12, 2021 9:31 pm
by TroyLim
Thanks Mike and Greg for your excellent advise and suggestions.

- when I plot Cd81 vs IgD, it looks to be 2 different cell populations with streaking at both ends - I believe this is some amt of spillover between both channels? (I believe this is what Greg is referring to but please correct me if I am wrong)
- Gating on the CD81hi population, they do seem biologically feasible, with markers expected of CLL cells (I have attached the bivariate plots if you would like to have a look).
- Yes, these are the exact same tube of antibodies, maybe 4 months older, they are bought pre-conjugated from Fluidigm
- Regarding your earlier advise on filtering my antibody mixes, refering to your 2018 paper in immunological methods, is it right that I should make up my antibody mix with my predetermined titrations + CSB to a total of 100ul, place on a spin filter for 10mins with RCF 14000 and top up back to 100ul with CSB if necessary before adding to my sample?

- I am not sure which events would you refer to as "free aggregates", I have attached my gaussian gating strategy for your reference. Do you think that there are these free aggregates here?
- Regarding your suggestion about heparin, as I am working with cryopreserved PBMCs I do not think that there will be many/any eosinophils to cause these background staining effects, happy to be corrected though

Thanks again for both of your time.