Advantages of a Standardized Automated Cell Counting Method for Cell Therapy

A: We have seen it before, but usually it’s right after cell lysing you can still see free nuclei flowing around. But, if you’re doing a long culture, we have also seen that the AO signal often decreases. That signal becomes very dim, so there’s a difference between that signal and the actual cell. You can also gate those small objects out by size gating using the Cellometer, so you don’t pick those objects up when you’re trying to count the cells.

A: As there are many different parameters for doing qualification methods, it’s important that you sample properly. If you’re testing a regular cell line, make sure you have enough cells so you can make samples ranging from very high concentration to very low concentration, giving you a good dynamic range for your cell counting method.

Sample prep is very important. When you’re getting the sample from flasks or tubes, you have to make sure that every single step is consistent. If you pipette up and down 10 times, keep it the same. Make sure that your sample is mixed properly and vortex it so you’re always getting as uniform a sample as possible.

There are multiple publications that say pipetting is the number one source of error for this type of qualification. If your sample preparation is not done properly, then you’re going to have inconsistencies and variation that may disqualify your counting method. However, depending on how you do the testing, you have to look at all parameters and then look at what the best criteria are for your facility, for your CAR-T process.

A: Yes, we have also seen this. There are a couple of things that may be happening here. One is, after a long culture time, sometimes the pH may shift a little bit, affecting uptake of AO. We are working on another dye similar to AO that will take up much more readily than AO, stay more localized and be much brighter.

One other possibility that we have seen with long cultures is that there are objects that look like cells that have formed. We did AO/PI (acridine orange / propidium iodide) staining and Hoechst and PI staining. All these are nuclear dyes. We got similar cell counts with all of these nuclear dyes. We also did a manual counting of the bright field and trypan blue. For this method we got more cells, more objects counted. Those objects that are not taken up with the AO or the Hoechst dye, don’t seem to have nuclei. We did another stain called calcein AM to look at enzymatic staining. The objects without nuclei are enzymatically active so we were labeling those objects ‘ghost cells’, due to the apparent lack of nucleus.

A: We found that one problem was red blood cell contamination in the brightfield. Another is the morphology change and storage problem with trypan blue. If you leave it for a long time, it starts to crystallize in the solution. You then have to filter it, otherwise when you stain cells, you see crystals that may look like cells which you can mistakenly count.

Besides underestimating the dead cells, in a hemocytometer counting, we have also observed that if you have a lot of cells, then it takes a long time to count them. Over that period of time, the trypan blue will have a lot of effect. Maybe you can count them within five minutes, but even within that five minutes, it does affect the cells. And if you’re counting something very sensitive like hepatocytes, trypan blue can kill them within the first two minutes, causing cell viability to drop by 10% to 15%.

A: Of course, you can’t leave it with the cells for too long. However, we can stain cells with AO/PI for up to two hours. The viability throughout was about 80% to 85%. That’s within two hours. What we usually recommend if you have a lot of samples, let’s say if you need to count maybe 20 samples, is that you stain five samples first, put them into the counting chamber and count them, then do the next five and so on. Keep in mind there are differences between different cell types. If you’re doing primary cells, they may be a little bit more sensitive, but within two hours there’s not too much change in viability.