Simple tips for reducing errors with panel prep

Running large panels consistently is all about error prevention. Today let's look at some tips for how to minimize mistakes and catch them before they become big problems. The first step in this is recognizing that you're going to make mistakes, accepting that, and trying to figure out what types of mistakes you're most likely to make. A lot of errors come from having to think about what you're doing when there are lots of details; I find that the more I can reduce the thinking required, the more reproducible the results become.

The most frequent errors with big panels typically come from diluting the antibodies for the master mix and from preparing the single-stained controls. Here are some tricks we use in the Liston-Dooley lab to help reduce and catch mistakes in these steps:

First, we put all the antibodies for a single panel in a single box:

Each antibody has a specific location in the box, which you can see from the lid. This system is meant to help people put the antibodies back in the same location after use, meaning the next user doesn't have to search for them.

Of course, that doesn't always work, as we can see from looking inside this particular box.

Additionally, each antibody in a given panel gets a colored sticker placed on the lid. This way the panels are color-coded, and you can easily determine whether all the antibodies you've got out belong to the same panel. You can get stickers like this at stationary stores or on Amazon. A diameter of 10mm is good.

We generally just put a number on the sticker as this is easier to read than the full antigen + fluorophore combination.

These numbers are then re-used in the dilution calculation worksheets. I've set ours up such that all you have to do is change the number of samples being run to get the amount of antibody required.

	Overnight mix				Samples
Vial	Channel	Marker	Dilution	ul to add	19
1	BUV395	CD103	500	4.56
2	BUV496	CD4	500	4.56
3	BUV563	NK1.1	500	4.56
4	BUV661	CD19	2000	1.14
5	BUV737	CD62L	1000	2.28
6	BUV805	CD8	500	4.56

Note: the staining buffer and other reagents for the master mix cocktail are not listed here. This is an incomplete example. I usually make up a 20% excess volume to allow for multi-channel pipetting without worrying about running out. So, if we go for a staining volume of 100ul, we'd have 19 * 120ul = 2280ul final volume (see how this fits with the 2.28 for a 1:1000 dilution?). That final volume needs to account for the Brilliant Stain buffer (if used), total antibody volume and any other reagents (blocking, viability). The rest is the staining buffer (FACS or perm/wash).

Something we've switched to recently is re-organizing the worksheets by dilution factor rather than fluorophore excitation/emission. This has the advantage of allowing users to dispense all antibodies that require the same pipetting volume in sequence, rather than jumping around. This means fewer changes to the pipette settings, so it's faster. It also requires less thought and double-checking, which is good for reducing errors.

Here's that same worksheet, re-ordered by dilution factor:

	Overnight mix				Samples
Vial	Channel	Marker	Dilution	ul to add	19
4	BUV661	CD19	2000	1.14
5	BUV737	CD62L	1000	2.28
1	BUV395	CD103	500	4.56
2	BUV496	CD4	500	4.56
3	BUV563	NK1.1	500	4.56
6	BUV805	CD8	500	4.56

Notice that we've got the antibodies with the lowest amounts first. If we start from least to most, then at any given point if we mess up irrevocably, the mistake is more or less the cheapest it could be at that point.

It's important to keep track of your progress with adding the antibodies to the master mix. It can take a while to add all the antibodies for a large panel, and if you get interrupted during this process, you want to be able to pick up where you left off with a minimum of thinking or second-guessing yourself. When adding the antibodies to the master mix, there are two main ways of keeping track of which ones you've done. The first is to have all the antibodies in one rack and move each one to a second rack after adding it. I like to use different colored racks.

The second option is to cross off the antibody from the paper printout of the worksheet after adding it to the mix. The nice thing about this method is that you can go back and check the printout the next day, even if the antibodies have been put away (I've done this many times when there's been an odd result, and, yes, usually I missed adding something). Either way, you should be able to visually confirm that everything has been added.

You can additionally get white sticker labels for the sides of the tubes for the master mixes. This can help with accurate labeling because it's easier to write (or print) on these stickers than curved plastic tubes. This is particularly important if you plan to store the master mix as you'll want more information than if you plan to use it immediately.

Preparing the single colour controls correctly is critical. We'll look later at an independent way of reducing mistakes for this, but let's focus here on error prevention during the actual preparation. First and foremost, plan to make mistakes. By this, I mean prepare more wells/tubes of unstained cells or compensation beads than you actually need. This way, if you mess up a control (or think you messed up), you can ditch that one and prepare a new one from one of the extras. This is particularly helpful if you're doing multi-step flow with fixation because otherwise you have to go back and start the whole protocol again for the control you messed up.

Next, while we arranged the antibody vials in order of dilution factor for preparing the master mixes, for the controls, if you're staining on a plate, you want to arrange them by excitation/emission (UV to red). This is the way you'll run them on spectral machines, so it will save you time, thinking and reduce confusion later if you set them up in the same order you're going to acquire them.

Channel	Marker
BUV395	CD103
BUV496	CD4
BUV563	NK1.1
BUV661	CD19
BUV737	CD62L
BUV805	CD8

Again, you'll want to use the same procedure of moving the antibodies from one rack to the other or crossing them off on your worksheet.

Finally, if you're working with plates for the controls, I recommend writing the fluorophore (or a short version of it, e.g., 395 for BUV395) on the lid of the well to be used. Hopefully your handwriting is better than mine.

When you go to add the antibody to its well, cover the part of the plate you aren't using with the lid, exposing the active row. This helps in two ways: 1) it provides a clear guide for which row you want and 2) it prevents stuff accidentally getting into the other wells. You can then scan along the row until you reach the right fluorophore well to deposit the antibody.

I like to put a checkmark on the lid on the well after adding the fluorophore. This isn't critical for simple controls, but it can really help if you're doing two-step controls such as biotin + streptavidin. In that situation, I draw two circles on the well, reminding myself that I need to add two things to that well. I can then verify that both have been added by ticking off each circle.

Some more tips:

Before you run your first sample on the Aurora (better yet, any cytometer), run water for a minute or two. If the previous user has cleaned the flow cell with bleach and water, the flow cell may still be full of liquids other than sheath fluid, and you’ll get turbulence and misalignment of the stream as it goes through initially (at least, I think this is what’s happening). The result is often unstable flow with problematic readouts, particularly for SSC, which drops. This can be fixed by just running some water for a couple minutes. It may also help to do a SIT flush or two.

You may want to split your single-color controls in two prior to putting them on the machine. If you separate out a third or a half of each sample into a second plate, you now have a back-up without much extra work. This way if something goes wrong on the machine (a clog, a mistake in settings), you can run using the back-up even if the entire sample you put on the machine has vanished.

Try to prepare as much as possible ahead of time. Limit the amount of stuff you have to do on the day when you're doing your big flow experiment. This is true for any complicated experiment. You'd be surprised how much it helps to label the tubes the day before. Lots of time savings.

Finally, do the complicated stuff at the time of day when you work best. This is one of the main reasons I started doing overnight staining--I just wasn't functioning well after long days of isolating cells from tissues and performing cytokine restimulation assays. I was making more mistakes in the evening than in the morning.

Hope that helps!

Suggested reading:

Liechti et al., A robust pipeline for high-content, high-throughput immunophenotyping reveals age- and genetics-dependent changes in blood leukocytes

As usual, the supplementary information is particularly useful.

Lilac-breasted Roller, South Africa