Today I'm going to share my tips for successfully running lots of samples on the Aurora without having to babysit the machine.
This post deals with plates. In the Liston-Dooley lab, we use these V-bottom 96-well plates for staining and acquiring samples. They have the advantages of largish volume (300ul) and pelleting the cells well. They're a bit of a tight fit on the newer plate loader, so you may find a better option.
Plates are the recommended approach for blood samples. Running in a plate allows you to walk away and is approximately 1.5x faster than tubes. For tissue samples or other messy samples, the tube loader is a safer approach, if you have that. Similarly, if you are acquiring samples with about 10 million or more cells, you'll want to use tubes. For volumes above 200ul, the tube loader is strongly recommended to prevent cross contamination. If you have the older model plate loader with an immersion mixer, keep the volume to 150ul or less (see more on this below).
Note: if you use a plate for reference controls, all reference controls (unstained, AF, single-color) must be in a single plate. This will not be possible with a tube loader due to the number of controls exceeding the number of tube slots (40).
How to:
Video link here. For the older plate loader, see this video.
If it isn’t already on, turn on the plate loader using the switch in the back.
Remove any tubes from the SIT.
Move the sample loader arm into the forward position.
In SpectroFlo, select a sample in the plate that you want to run. The green arrow next to the sample should remain solid dark green after moving the cursor.
"Eject" in SpectroFlo to get the plate loader to the front.
For best results, resuspend the samples on the plate using a multichannel pipette to agitate them.
Resuspending samples for acquisition
Check for any clumps. Ideally, you always want to filter the samples prior to acquisition. Samples without visible clumps will generally run fine, although there may be minor reductions in data quality. If you do have visible clumps, you need to either filter the samples or fish out the clump(s) using a P20. This needs to be done after resuspending the sample to avoid disproportionate losses of cells.
Manually fishing out a clump. You're better off filtering, but this can work.
With the lid on, insert the back of the plate first, confirming it is well seated on both sides.
Push down gently but firmly to seat the front of the plate.
Check that all four corners are all the way down and that well A1 is in the front left position.
Remove the lid.
In SpectroFlo, load the plate. Eject and load again in order to avoid any homing errors caused by manipulation of the plate loader. This will prevent the SIT from bending or breaking.
Set the flow rate to high (usually okay for fully stained samples) and hit Record. For single color controls, acquire on low speed for best results (more on this later).
SpectroFlo will prompt you for SIT calibration on the plate. Select a well that has nothing in it, if possible. In rare cases, backpressure from failing filters may result in liquid falling onto the plate. If this occurs, seek assistance from core staff prior to running any more samples. As the cytometer performs the SIT calibration, verify that it is indeed using the well you’ve selected and that the SIT is going into the well and reaching the bottom.
Confirm that acquisition is proceeding smoothly for the first sample and that the cytometer is continuing to the second sample before leaving. Check that scatter profiles are consistent with the setup profile. A dummy (unstained or fully stained control) sample for the first well may prove helpful. Cover the sample intake port with the dark lids.
If you have both controls and samples on the same plate, you will first need to record the controls, unmix, then record the samples. You can acquire everything in Raw mode and unmix at the end, but it's faster to unmix on the fly rather than waiting for all the calculations at the end. Another option is unmix using library controls, then add the new reference control group to your experiment to run everything all at once as unmixed data.
Now, there are some important differences between the old immersion mixer (AMS) and new orbital shaker (ASL) loaders:
The old plate loader is a lot easier to get plates in and out of.
The old plate loader carries a much risk of cross-contamination between samples. There's a large part that gets put into every sample and although it flushes between samples, it doesn't seem to work that well. The mixing on this plate loader also has a disturbing tendency to spray sample into adjoining wells if the volume is above 150ul. Watch, you can see it happen.
The new loader can handle slightly larger volumes in the same wells. This is because the old loader again stuck something large in each well, displacing volume.
The new shaking loader is better at getting the cells into suspension if they've settled out.
For the orbital shaker plate loader (the new so-called ASL), here are my recommended settings:
For the older immersion mixing model (the AMS), you'll want to do a double flush to reduce cross-contamination, and you'll need more vigorous mixing. Alternatively, run fewer samples at a go and resuspend them manually with a multichannel pipette between batches. You may want to have empty wells around each sample (using only about a quarter of the plate), at least for the first few times you use it, to check that it isn't spraying your sample all over the place.
You may of course set a delay on the data recording for your fully stained samples to prevent acquisition of the first seconds, which are more prone to instability in the stream. I generally prefer to record everything and sort it out later rather than irrevocably lose the data. Depends on how many cells you got and how rare the ones are you need to analyze.
Finally, running compensation beads on either loader can be problematic. Beads tend to settle out of suspension faster than cells. If the beads are all at the bottom of the well when the sample starts getting drawn up, they will be lost in the dead volume (50ul). To avoid this, you want to resuspend with a pipette immediately prior to running the beads. It can help to run the beads in batches of 8-12 wells at a time, ejecting and manually resuspending in between batches. Shaking for longer or more frequently on the new loader, or mixing for longer on the old loader, may also help. It possible that running the beads in a U-bottom plate rather than a V-bottom would also prevent this issue, which I haven't tried yet.
Yellow-breasted Chat, Arizona, USA
Comments