Instrument malfunction?

[emallen]

Hello all, I'm having a doozy of a problem that I just can't solve. My CD45RA stain is tagged with Eu153 and this antibody has worked beautifully in the past. However in my two most recent experiments my positive pop has been way higher than usual and my negative population is also shifted to the right.

At first I thought this was an antibody issue, that maybe I had conjugated one of my other antibodies to Eu153 by mistake. However when I ran single-stained beads on the machine, they all checked out.

Next I noticed that my 'Bead neg' population (I use the Eu 151/153 Calibration beads) in Eu151 (a bead-dedicated channel) was higher than usual. So I thought maybe the metals were dissociating from the beads and adding background to Eu153. I tried running the same sample with and without beads, and got the same results each way.

However I did notice that even in the sample without beads that my 'Bead Neg' population was still shifted to the right in my bead-dedicated channel. This holds true for the surrounding channels as well. This indicates to me that the noise is probably being generated by the instrument somehow. I've also noticed that my 'Bead Pos' signal has been dropping steadily over these experiments (in the samples where beads were included). Does anyone have any ideas as to what these symptoms could indicate?

Elyse

[nicole]

Hi Elyse,

What you are seeing sounds similar to an issue we had with reagent contamination. In our samples, lot-specific PFA from EMS that was contaminated with samarium. This caused a very significant positive shift in the negative population in all of the samarium channels as well as some of the neodymium channels. We did not have antibody in Eu151 or Eu153 at the time so they may or may not experience the same issue. We did not see this in the overall background, just the cells. We ran unstained cells and still saw a shift in these channels that should have been negative. A few questions to clarify -

1. Did you run beads alone (at 1x)? If so are you still seeing the shift? If so, can you share your tuning and bead results?
2. Which channels specifically are you seeing the shift in? Is it running as background overall in these channels or is it specific to DNA+ cell events? Have you tried running an Ir191/193 only unstained control since seeing this problem?
3. Have you changed any of your staining buffers or reagents? Even if it is just a new lot, this can be an issue.
3. Have the files been normalized? If so are you seeing the shift in the raw file as well as the normalized file?

-Nicole

[mleipold]

Hi Elyse,

1. Do you have a CyTOFv1 or a CyTOFv2?

2. What do you have in Nd150 and Sm152?

3. Have you run any of the EQ (Ce-containing) beads during the time that this Eu decrease has been happening? If so, are you seeing a similar decrease in "M" and increase in M-1 or M+1 for the Eq bead channels?

4. When did you buy the Eu-beads you're talking about?


At any biologically relevant pH in aqueous solution, the polystyrene beads don't really "leak" metal ions (see some of the Abdelrahman papers for more info on their synthesis and stability). Therefore, it is highly unlikely that you're getting contamination from there.

The fact that your Beadneg signal is going up in a Bead-only channel makes me wonder if your TOF window setting is off. This is something that you look at in the TOF peak window: it's call "TOF graph" in a CyTOFv2, "TOF range per reading" in a CyTOFv1. In a CyTOFv1, there are instructions for checking this around pp24-26 of the manual on how to use the tuning solution's Cs133 and Ir193 to check this. I can't find instructions on how to check this for a CyTOFv2, so it might be something that's confirmed or set during daily tuning.

Basically, the CyTOF doesn't "know" what Eu151 "is". It knows that you have told it that "count all the ions arriving, starting at TOF=X and continuing for Y units, add them together, and call that signal Eu151". But if your ions start arriving a bit earlier (or usually, later), the machine doesn't know to adjust....it still just counts all the ions arriving between X and X+Y. Depending on how far this drifts, you can start having a significant decrease in M, and an increase in M-1 or M+1, due to the left or right leg of the ion peak creeping into the next TOF acquisition window.

It's something pretty easy and quick to check, at least on a CyTOFv1.

Mike

[emallen]

Hello all, thanks for the helpful responses! Here is a little more info:

1) I don't think this is a panel issue, since I've run this exact panel at least 10 times before with a control sample that always looks exactly the same. Until now.
2) It's unlikely to be a reagent contamination issue, since I didn't use any new reagents or buffers for this experiment (all of my reagents had been previously validated).
3) The files have not been normalized. I always do my QC pre-normalization.
4) I have a CyTOFv1
5) I haven't been running EQ beads in these experiments, only Eu151/153 beads (which were purchased from DVS in Feb)

If anyone wants to look at my fcs files, I can send them to you via another route. Any help you could give would be greatly appreciated!

Elyse

[mleipold]

Hi Elyse,

1. Take your bead data, and check 150/151/152/153/154 signals. Either from a bead-only sample, or after you gate Ir vs Eu and take only the Ir- Eu+ events. Checking all those channels should help you figure out any neighboring spillover.

2. Check the Cs133 and Ir193 values for TOF start as described in the manual. If necessary, adjust with new values in the Mass Calibration tab, hit save, and then rerun the Eu bead sample to see if that made a difference.


Mike