FAQ  •  Register  •  Login

Question about barcoding "supervised" visualization

Forum rules
Please be polite and civil. We know that troubleshooting is vexing...
<<

sfauteux

Participant

Posts: 17

Joined: Tue Feb 23, 2016 2:18 pm

Post Tue May 10, 2016 11:56 am

Question about barcoding "supervised" visualization

Hi everyone,

We recently tried out Fluidigm's barcoding protocol which appears to work fine except for one thing. On one sample, antibody-negative cells do not seem barcoded... which seems odd.

Protocole:

Whole blood was ACK lyzed and 1.5*10^6 cells were barcoded and stained with either CD3 170Er/CD45 89Y/ CD45 141Pr. Cells negative for CD3 do not appear to be positive for Pd which makes me worying a bit. Anyone has experienced this?

Please find attached the Layout of this experiment. Used the #1-3 tags.

Best,

Sebastien
Attachments
Barcoding-Layout.pdf
Barcoding Layout_PDF
(371.68 KiB) Downloaded 329 times
<<

mleipold

Guru

Posts: 5796

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Tue May 10, 2016 3:05 pm

Re: Question about barcoding "supervised" visualization

Hi Sebastien,

I just want to confirm that I understand what your experiment was:

Whole blood, ACK lysed.
Sample 1: Barcode 1: 102/104/105
Sample 2: Barcode 2: 102/104/ 106
Sample 3: Barcode 3: 102/104/ 108

These three samples were then recombined into one sample, and stained with the following cocktail:
CD3-170Er
CD45-89Y
CD45-141Pr

In other words, two different CD45 antibodies and one CD3 antibody. Is this correct? Did you have the Ir (or Rh103) intercalator in your staining as well?


In the plots in your PDF, are you showing us Ungated cells (ie, still containing debris, doublets, etc), or are you showing de-barcoded data?


Mike
<<

sfauteux

Participant

Posts: 17

Joined: Tue Feb 23, 2016 2:18 pm

Post Wed May 11, 2016 8:40 am

Re: Question about barcoding "supervised" visualization

Hi Mike,

Thanks for taking the time to answer.

This is right for Whole blood ACK lysed and the barcodes.

However, we did not combined the samples at any moment (next step of validation). The markers were just to identify that cells Ab+ were barcoded. DNA was stained with Ir191/193 and gating was done only on DNA. The gating is pretty loose, but nothing extreme. I verified a thighter gating, this is not the issue here (File attached).

The dot plots are of course gated cells and the data was not debarcoded as samples were run separately.

Sebastien
Attachments
Barcoding-Layout-2.pdf
(61.15 KiB) Downloaded 325 times
<<

ChristophS

Contributor

Posts: 21

Joined: Thu May 07, 2015 2:11 pm

Location: Switzerland

Post Wed May 11, 2016 2:40 pm

Re: Question about barcoding "supervised" visualization

Hi Sebastien,

looking at your data I would think that your MCB has been working as it should (the intensities could be slightly higher possibly; what medians ar you getting for the nucleated cells in the Pd channels?). I took your image and added my interpretation. Probably you were focusing on the very bright small population that I suppose are dead cells (did you add CisPt for dead cell exclusion?). Hope this helps.

Christoph
Attachments
Barcoded PBMCs.jpg
Christoph Schwärzler
Director Cytometry
Flow Cytometry Facility (https://www.cytometry.uzh.ch/en.html)
University Zürich
<<

mleipold

Guru

Posts: 5796

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Wed May 11, 2016 3:11 pm

Re: Question about barcoding "supervised" visualization

Hi Sebastien,

Looking again at your initial flow plots slide, I want to confirm a couple more things.

1. You took whole blood, did ACK lysis.
2. Since you used the Fluidigm barcoding kit, the cells were then Fixed and Perm'd as on page 7 of the User Guide? http://www.dvssciences.com/products/pdf/UG_201060.pdf
3. You then did a single antibody stain separately on each sample. Your samples were:

Sample 5_P1.FCS (top of slide): Barcoded with 102/104/105. Antibody stained with *only* 89Y-CD45 (not CD3-170Er or CD45-141Pr). Stained with Ir. No live-dead (cisplatin, etc).
Sample 7_P3_01.fcs (middle of slide): Barcoded with 102/104/108. Antibody stained with *only* CD3-170Er (not 89Y-CD45 or CD45-141Pr). Stained with Ir. No live-dead (cisplatin, etc).
Sample 6_P2_01.fcs (bottom of slide): Barcoded with 102/104/106. Antibody stained with *only* CD45-141Pr (not 89Y-CD45 or CD3-170Er). Stained with Ir. No live-dead (cisplatin, etc).
-By the way: I assume it was CD45-Pr141 as you wrote in your original post. Your flow plots say CD196-141Pr

4. Then Ir staining, and running on the machine.

In other words, each samples was fixed, then perm'd, stained with a Barcode, stained with *one* antibody each (not a cocktail of 3 antibodies as I wrote in my initial reply yesterday), and then stained with Ir. Is this correct?


Mike
<<

sfauteux

Participant

Posts: 17

Joined: Tue Feb 23, 2016 2:18 pm

Post Wed May 11, 2016 4:00 pm

Re: Question about barcoding "supervised" visualization

Hi Mike,

Everything you say is correct. CD196_141Pr is from another panel, i just did not change it for CD45. My main concern is why are CD3- cells not barcoded?

Thanks again for taking the time.
<<

sfauteux

Participant

Posts: 17

Joined: Tue Feb 23, 2016 2:18 pm

Post Wed May 11, 2016 4:05 pm

Re: Question about barcoding "supervised" visualization

ChristophS wrote:Hi Sebastien,

looking at your data I would think that your MCB has been working as it should (the intensities could be slightly higher possibly; what medians ar you getting for the nucleated cells in the Pd channels?). I took your image and added my interpretation. Probably you were focusing on the very bright small population that I suppose are dead cells (did you add CisPt for dead cell exclusion?). Hope this helps.

Christoph


I did not add CisPt. They may be dead cells bur are also + for metals that are'nt supposed to be there, I think this is just a question of metal purity.

Seb
<<

ChristophS

Contributor

Posts: 21

Joined: Thu May 07, 2015 2:11 pm

Location: Switzerland

Post Wed May 11, 2016 5:08 pm

Re: Question about barcoding "supervised" visualization

Hi Sebastien,

the fact that this small population is positive in all kinds of things is more evidence for them being dead (more precisely: having been dead already prior to fixing). From the image you provided I can not follow your concern of CD3- cells not being Barcode positive. Am I missing something? Could you point out on the image what you refer to?

Best
Christoph
Christoph Schwärzler
Director Cytometry
Flow Cytometry Facility (https://www.cytometry.uzh.ch/en.html)
University Zürich
<<

mleipold

Guru

Posts: 5796

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Wed May 11, 2016 5:58 pm

Re: Question about barcoding "supervised" visualization

HI Sebastien,

I think I agree with Christoph.

Yes, there is an issue of purity with some of the Pd stocks: in particular, the 104/105/106 range tends to bleed into each other due to isotopic impurity.

If you look at your plots: Sample 7_P3_01.FCS is barcoded as 102/104/108. In those channels, the Pd signal on 90+% of your cells is *above* 10^1. In the other channels *not* part of your barcode (105/106/110), the Pd signal for 90+% of your cells is *below* 10^1. If you accept Christoph's reasonable idea that the Pd-bright population are dead cells, then your +/- cutoff for Pd signal would be at 10^1.

If that's the case, then both your CD3+ and CD3- would be barcoded at 10^1.

I think what's confusing you is that you don't have a *negative* in your samples. In this case, you said you added the barcoding agents, but did *not* mix them afterward.

I would repeat this experiment, but add a 4th sample where you mix your cells at the very end of your protocol (ie, do all the staining individually for your 3 samples, then take an aliquot each of your stained samples and mix them together to make a 4th, composite sample).

If you do that using the same Fluidigm Barcodes 1-3, then assuming a 1:1:1 ratio of Samples 1-3 in Sample 4, your Pd channels for Sample 4 should be:

102: 100% cells positive (~10^1 or above)
104: 100% cells positive
105: 67% cells negative, 33% cells positive (only in BC1, not BC2 or BC3)
106: 67% cells negative, 33% cells positive (only in BC2, not BC1 or BC3)
108: 67% cells negative, 33% cells positive (only in BC3, not BC1 or BC2)
110: 100% cells negative


Mike
<<

sfauteux

Participant

Posts: 17

Joined: Tue Feb 23, 2016 2:18 pm

Post Wed May 11, 2016 6:06 pm

Re: Question about barcoding "supervised" visualization

I was, indeed focusing on the dead cells.

Well you both cleared this up for me. I'm not used to look at barcoding plots, but the result is limpid now. Will do the mix experiment now.

I wish you a good day/night and thanks again.

Sebastien

Return to CyTOF troubleshooting

Who is online

Users browsing this forum: No registered users and 13 guests