First post of 2022 is a hot topic at the moment, given their scarcity during this current Omicron wave here in Australia.
If you’re lucky enough to have access to one of these tests, here’s a post to show you how it actually works (i.e. the theory behind it). If I feel up to it, I might even do a PCR one.
So, first thing’s first. The overview of a RAT.
RATs work on the principle that specific antibodies, bind very specific antigens.
What’s an antibody?
It’s a protein (usually depicted as a ‘Y’ shaped protein) that binds very specific target peptides (small proteins) called antigens.
So if there’s a free floating antigen, and it happens to get close enough to the binding regions of the ‘Y’ shaped antibody that it’s targeted by, then the antigen will get bound by the antibody. As below.
That’s the overall principle, anyway.
The strength of antibody-antigen interactions can vary depending on the antibody generated. Some are very specific, some are very tight binding (once it latches on it’s very difficult to detach). Some only loosely attach, while others bind lots of roughly similar antigens (but not identical ones – by the way this type of interaction is actually, legitimately described as interactions with ‘high promiscuity’). It just depends on the antibody and antigen combo.
But antibodies generated in kits have been designed and mass produced to be as specific to their target antigen as possible, and will bind relatively better (strongly and more specifically) than one that might have been made in the initial trials when they were in development.
Now, lets take a step back and have a think about what these antibodies bind on SARS-CoV-2.
Here’s a very simplified diagram of the SARS-CoV-2 viral particle (i.e. the individual virus, or, the virion).
We’ve all probably heard a lot about the spike protein (S protein), because that’s the most common portion of the virus that our antibodies will be made against. The simple reason for that is because it’s the outer most protein on the virus’ surface. It’s the first point of contact with the outside world, so naturally, our immune system will see this portion the most.
But if you’ve been listening/watching/reading about the spike protein, you’ll know that it’s also the area that’s most likely to mutate. If our antibodies can’t bind the spike protein anymore (i.e. it’s no longer being recognised/targeted), then the virus can escape detection and destruction by the immune system and go on to infect more efficiently (and for a longer time). The evolutionary pressure is for the virus to keep mutating areas that our immune system identifies the most, so the spike protein will continue to mutate so long as the virus keeps infecting us.
The nucleocapsid (N) protein packages up the single stranded RNA (ssRNA), or the genetic material of SARS-CoV-2. These types of ‘internal’ proteins, which aren’t exposed on the surface of the virus, don’t have the same sort of pressure to mutate. This is because our immune system doesn’t typically see these proteins- because they’re hidden inside the viral particle. If our immune system doesn’t target it during infection, then the virus doesn’t accumulate mutations that allow it to escape detection. So, again, internal proteins don’t mutate anywhere near at the same rate as proteins that are on the surface of the virus.
Now, with all of this in mind. If you were designing a test to detect viral antigens (i.e. proteins) from SARS-CoV-2, knowing that mutations will eventually arise and you don’t want to have to make new kits…
What antigen would you try to target with your antibodies?
Would you pick the spike protein?
Or maybe the nucleocapsid protein?
Have a think…
Hopefully you’ve figured out that you would generally try to target internal proteins like the nucleocapsid, because they’re less likely to mutate quickly. Especially compared to external proteins like the spike protein, which are constantly under pressure to mutate because they’re the components that our immune system detects during infection.
But how do you access these internal viral components? You have to somehow break the virus open to expose the nucleocapsid protein to the antibody.
They’re too small to mush up manually and break open the viral particle. We can’t put it in a blender and slice them open…
… But we can break them open using chemicals…
Detergents will break open a viral particle and break down any other debris in your swab sample, which then allows access for your antibodies to bind their target antigen.
Okay, still with me so far?
So that’s the overall theory of how the first part of the RAT works. The second part is how this theory now applies to your test device specifically. We might wait for this second part until I explain the test device in more detail below.
Let’s just jump back a bit to: What’s in a typical kit?
The basics should be the same across all kits, in that they should have the below items.
The tube usually has some sort of mechanism to allow you to add droplets of your virus and buffer mixture onto your test device. Whether that’s in the lid or the tube portion will depend entirely on your kit manufacturer.
The kit I have also has additional items such as a tube rack (to hold the tube in place while you deal with sample preparation), and a sealable bag to chuck all of your used kit components for easy (and relatively sanitary) disposal.
But so long as you have the items listed in the figure above, you should be right to go.
Let’s break each component down to explain what they do.
Firstly- the buffer isn’t just water. 😅
Hopefully that’s pretty obvious.
The basic components of a RAT buffer are as follows:
- Buffered water
- Whether it’s phosphate-buffered saline (PBS) or tris-buffered saline (TBS), it’s something saline that keeps the pH of the water ‘buffered’ from going up or down drastically
- Huge changes in pH can destroy proteins and make them warp or fall apart, which is a big issue if you’re relying on identifying set structures using your antibodies
- Might be marked as ‘Tween-20’, which is a common detergent when you’re working with protein-based assays/tests (which a RAT is)
- A positive control antigen (to bind to the Control (‘C’) portion of your test device)
- This is usually a completely unrelated antigen (antibody target) that has zero percent chance of being in a sample swab
- Companies never tell you what it is, specifically, but usually it’s ovalbumin, which is a protein found in egg whites
- Why is it egg protein? I dunno- but it’s used a lot because it’s mass produced in labs anyway
- Usually it’s 0.1% sodium azide, which is a common preservative when you’re working with proteins
- Definitely toxic if ingested, so please don’t drink the buffer
The dropper tube
Do make sure this (and the buffer) is all set up ready to go before you start swabbing.
This is the vessel that’s going to allow your sample and buffer to mix together (and work its magic).
Usually they come with a dropper portion so that you can carefully drop your buffer mixture onto the test device.
When you add your mixture, make sure the tube is as vertical as possible, so that the required size droplets come out. If you have the tube on a tilt, the droplet sizes can vary, which means that you might end up adding less (or more) sample than the test was designed for.
The swab (aka the brain tickler)
This is the part that collects the sample from the location identified in the kit. Whether it’s a nasal swab, nasopharyngeal swab, or throat swab, this allows you to collect potential viral particles (alongside gunk you’ve made yourself- ya grub).
It’s very important to make sure you’ve thoroughly cleaned your hands prior to swabbing (or glove up with a sterile glove). This is to make sure you don’t accidentally contaminate the swab with anything other than your actual, specified swabbing site- because then it’s not a ‘clean’ sample. If you’ve touched other surfaces in addition to, say, the insides of your nose, then you can compromise your test in a number of ways.
- Maybe your sample might wipe off before your do the test
- Especially if you didn’t have much virus to begin with, or if your swabbing was inadequate
- Maybe you might add debris from your hands or the bench/table
- This might clog up your buffer and make it more difficult to add droplets to your test device
- Maybe you might add water to your swab
- This will inadvertently dilute out your buffer by adding more liquid, which might make it more difficult for your test to detect virus (causing a false negative- but more on those terms later)
So you can see- you have to be careful where you put your swab!
As mentioned, make sure you’ve set everything else up in your test before you take the swab out of its sterile packet. Once it’s out, the only thing you should do with it is to stick it straight into the swabbing area, and then add it straight to the prepared buffer mix in your prepared tube. Give the tube a bit of a squeeze to try and wring out as much liquid from your swab as you can, before discarding it safely in a sealed bag- so that you don’t leave potentially infectious material out and about.
Most sticks have a thinned out section, which indicates the area of the swab you should snap after use. This makes it really obvious that the swab has been used, and that you definitely shouldn’t be touching it, let alone using it again.
Once the swab and buffer mixture meet, it’s time for the viral particles to break open and fall apart (exposing that juicy internal viral protein/nucleocapsid protein). The detergent will also break open our own skin cells in our nose/mouth (if they were on the swab) and break down any other debris that might have been attached to your swab. Breaking open our own cells might help expose more virus that might have been hiding inside the cell, too.
You might notice that your buffer gets goopier after adding sample because of this. Delicious.
Now we can drop our buffer/sample mixture onto the test device using the dropper on the tube.
The test device
Okay, this is where part two of the antibody-antigen interactions continue to be explained.
For a test device to work, we need the following things.
Obviously the presence of antigen will depend on whether you had virus in your sample to begin with, and whether you added your buffer correctly. But the antibodies themselves are already embedded in the test device in their respective sections, so even if you stuff up ‘sample loading’, the antibodies are still there.
This type of test can be called an ‘antigen capture assay’, because what you’re essentially doing is capturing an antigen using an antibody sandwich.
The antigen is the meat. Or vegan protein alternative.
The bread is the antibodies.
These test devices rely on this buffer/sample liquid mixture to travel across the test device towards the absorption pad in the opposite end of the device (‘lateral flow’ because it’s moving across the surface). As the liquid travels, it hits different section of the test device.
So you add the buffer/sample mixture with your dropper to the first section of the device, the sample pad (the small circular window). The buffer is breaking down the viral particle, any other gunk in there is being broken down, and your positive control (depicted as an egg 🥚 with a ‘+’ symbol) is also present.
The second section, which is hidden from view, is the ‘conjugation pad’, where the sample meets its first set of antibodies.
These are the antibodies that have the detector molecules (pink squares) stuck to them. The detector molecules have some sort of property that produces that characteristic pinkish/reddish/purpleish colour that you read on the test device.
There will be antibodies against the nucleocapsid protein (orange) and the positive control protein (green). If the antigen is present, then these antibodies will bind their targets and continue washing down the test device with them attached. Think of this as the first layer of bread with the meat sitting on top. We still need to put the top pieces of bread on for the sandwich to be complete.
The third section is the section containing the test and control lines.
These lines contain yet another set of antibodies against either the nucleocapsid protein (pale orange) or positive control (pale green), but this time, there’s no detector molecule, and they’re all stuck to the test device in a single line. Thousands upon thousands of them all glued to their respective areas.
These will also bind their respective antigens as they wash across the test device, but remember- these antigens already have the first set of antibodies from section two (conjugation pad) that have detector molecules on them. Once this second set of antibodies bind, then the sandwich is complete. The antigen (meat) is trapped between antibodies (the ones with the detector on them and the ones that are stuck on the test device). The detector molecule accumulates in one concentrated area, and lo and behold! You have a pink line (or two)!
All in all the whole process takes about 15 minutes (or as per your manufacturer’s instructions). You should aim to read the result exactly at the time listed on the instruction manual, and no more/less. Otherwise this might lead to false positives/negatives (more on those terms later).
This is, again, assuming you had virus to begin with.
If you had no virus, then the test will look something like this.
You can see that the unbound anti-nucleocapsid protein antibodies with the detectors (orange with pink rectangle) are just washing away, but the control antibodies are binding fine.
Because there’s no detector molecule accumulating along the test line, the test line won’t change colour.
And then if you’ve completely stuffed up the test, where even the positive control antigen isn’t present, then the test would look something like this.
If you’ve got no control band, then the test is invalid. You’ll have to repeat it again. Even if the test band comes up with a line, the result is invalid. If your controls aren’t working, then you can’t interpret the test results. That’s rule 101 of experiments.
Lastly, the final section of the test device (also hidden from view) is an absorption pad that absorbs all this excess liquid, unbound debris or antigens and antibodies. Essentially where all the rubbish goes.
So that’s the overall theory behind how a RAT works!
I hope it’s shed some light onto the cool science behind these tests.
Below I have some additional info for your interest.
False positives and negatives
I’ve mentioned these terms above, so I’ll explain what these are.
Let’s start with a false positive. This is when a test result shows you a positive result for something, but it’s not actually true (the true result is negative). In this context, a line might appear in your test line, but upon doing a PCR, your test result is negative. It was actually a negative result all along.
How does this happen? Well- in order for a line to appear, we know that the antibody with the detector molecule from the conjugation pad needs to make its way over to the test line. This can happen if you leave the test strip for long enough. If you read the result after the indicated time point, you can sometimes have enough detector molecule accumulation that you get a weak band. This is why it’s always best to follow the instruction manual to a tee.
With this in mind, a false negative is the opposite. It’s when the test result shows you a negative, but the real result was positive. In this context, you might not see a line appear in your test line, but upon doing a PCR, your test result is positive (and you were positive all along).
This is probably the worst scenario, because then you could be infectious and out and about (not isolating).
This can happen in a number of ways. Maybe you didn’t have much viral particles on your swab to begin with. Maybe your infection wasn’t severe enough to have lots of virus in your body.
Maybe you didn’t perform the test properly – but this can be easily solved by making sure the control works on your test, each and every time you do it. If the control works, then you can be more confident that your test has worked as best it could.
Both scenarios can happen, so while a RAT is useful, a PCR is by far the most reliable technique for virus detection.
How are all these antibodies made?
Artificially! You can have giant bioreactors that use bacteria (usually – you can use yeast as well) to artificially make tonnes of these antibodies- without the use of any humans or animals.
Once the bacteria make the proteins, you essentially kill them and harvest the antibodies using a process called protein purification. I’ve done it on a significantly smaller scale, but it’s one of my least favourite things to do (just because it’s finicky).
But if it’s at an industrial level, then the process has already been optimised, and usually it’s not that much of a hassle to do, because chances are you’ll have machines that do most of the hard work for you.
Why is the control (‘C’) line always furthest from the sample pad?
The test (‘T’) line is always closer to the sample drop area compared to the control, and there’s a very good reason why this is.
Have a think about what the issue might be if it were the other way around. If the control was closer and the test was further away. And then think about what the purpose of the control line is.
If the control line was closer, and colour appears along the line- hooray, your control worked!
But if the test line doesn’t have any colour appear, can you be confident in saying that your result was negative?
What if the buffer/sample mix never made it that far? What if it’s only negative because your buffer/sample mix dried up before it reached the test line, and not because there was no virus in your sample?
By making sure the control line is what’s furthest away, then you can be confident in knowing that if there’s colour along the control line, then it’s because the buffer/sample mix has reached that line.
And if that’s the case, then it’s also reached the test line. If there’s no colour along the test line, then it’s not because it wasn’t exposed to the buffer/sample mix.
You can use this theory to test for a myriad of different things.
A pregnancy test will test for hormones present in urine using the exact same theory (because hormones are also made of proteins).
I’ve done hundreds of RATs for influenza recently, and it’s the exact same theory as SARS-CoV-2- just the antigen is influenza nucleocapsid and the antibodies are against influenza nucleocapsid.
You can shuffle some stuff around and use this overall theory to check whether someone has antibodies against a pathogen.
You just need to swap out the antibodies along the test line to the antigen instead (usually spike protein this time because that’s what our body will recognise), and you add blood or serum to the test device.
The detector molecule will be attached to antibodies against human antibodies (yes, antibodies are made of proteins so they can be bound by other antibodies). So the sandwich now becomes meat, bread, and bread- essentially. SARS-CoV-2 spike antigen – patient sample – anti-human antibody with detector.
Basically, scientists can use this simple antibody-antigen binding phenomenon, which happens all the time in our body, to do all sorts of cool things.
(by the way this is the same underlying theory as western blotting, which I’ve written about before)
A Ph. D graduate in Microbiology, residing in Victoria, Australia. Currently working in multiple locations but still in the STEM field. 👀 🦠 🧫 🧬