Fast results?! Not so fast… I’d like to know now how ID NOW Covid-19 tests work! Unfortunately the proprietary nature of Abbott’s “rapid” Covid-19 diagnostic tests makes knowing exactly how they work basically impossible for an outsider. Despite desperate digging, I couldn’t get near the details about the test, but – if they use the same technology as Abbott’s flu tests, they rely on something called NEAR which is the viral genome equivalent of cutting off a sea star’s arm and growing it back over and over again – and using the cut-off arms to form a product you can detect. Hopefully that’ll make sense soon… And speaking of soon, how do these tests give results so much sooner than conventional tests? And why are they not quite so game-changing as they sound at first? (hint: only 1 test at a time…)
Covid-19 is the disease caused by the “novel coronavirus” SARS-Cov-2, which is a single-stranded RNA virus (it holds the genetic blueprint for making more of itself (i.e. its genome) in a single strand of RNA instead of 2 strands of DNA like we have). Diagnostic tests look for regions of this blueprint that it has and we don’t (and neither do other viruses, bacteria, etc.). If doctors/scientists find this SARS-Cov-2 genetic information in a patient, they know that the patient has Covid-19. But the viral genome’s really small, especially compared to ours, and the tests are only looking for a small piece of it, so there’s not a lot of starting material to find. So, instead, tests rely on making lots and lots of copies of that specific viral RNA and then detecting the copies.
I don’t know if you’re Harry Potter fans, but it’s kinda like in the 7th book when they’re in Gringott’s Wizarding Bank and there’s that spell where if they touch something that something starts making copies of itself. In Harry’s case, that was an inconvenience because he wanted the original – but we don’t care about whether it’s the original or not. Because in order for there to be a copy there must have been an original there to copy – and that’s what we care about. So a copy tells us it’s there. But our instruments aren’t sensitive enough to detect just a single copy – that’d get lost in the background noise…. So instead we make lots and lots and lots of copies – and there are different ways to do this.
Thanks to complementary “base pairing” between the nucleotide “letters” of nucleic acids (DNA & RNA) (e.g. A to T (or U in RNA) and C to G), protein enzymes (reaction mediators/speed-uppers) called DNA Polymerases can use strands of DNA or RNA as “templates” to make copies of the complementary strand that can be used to make copies of the template strand which can be used to make copies of the complementary strand which can be used… you (hopefully) get the point.
Different enzymes are able to do different types of copying (e.g. DNA to DNA is done by a DNA-dependent DNA Polymerase whereas RNA to DNA is done by an RNA-dependent DNA Polymerase). The latter is aka a “Reverse Transcriptase” and, before doing any amplification, the tests start off by using one of these “RTases” to make a DNA version of the RNA. This DNA version is more stable and it also allows them to now use a DNA-dependent DNA polymerase to make more DNA copies from it.
These polymerases are frequently just called “DNA Pols” & they’re super useful, but they have a couple key limitations – they can’t start from scratch (they need a free 3’ -OH to build off of) and they can only copy in one direction (5’ to 3’). Those “ ‘ “ symbols are pronounced “prime” and they specify positions on the sugar base of the nucleotide letters (probably best explained through pics). By providing short pieces of DNA (oligonucleotides aka “oligos”) called primers that bind to specific complementary stretches of a template strand you can provide a “launch pad” for the DNA Pol to copy from.
And this is the basis behind PCR (Polymerase Chain Reaction), which is the basis behind traditional SARS-Cov-2 tests (not to mention the foundation of LOTS of molecular biology…). RT-PCR tests for SARS-Cov-2 use primers that are specific to the viral genome to make lots and lots of copies of a stretch of viral genetic information, and fluorescent probes that bind to the copied region allow us to see when copies are made in “Real Time.”
Problem is, once the copies are made, they stay stuck to the template they were made from, because that base-to-base-pair-ability that makes copying so “easy” comes from physical interactions between the bases of the 2 strands. These partial-charge-based attractions can be overcome, however, if you provide heat, giving the strands enough energy to wriggle free of one another (melt step). But then you have to cool things down so that new primers can bind (anneal step) and then change the temp again to one that DNA Pol likes best, so that it can do its copying (elongation step). But that was just one copy (per template). So you keep doing this over and over in a wild temperature ride to get an exponential increase (e.g. 1 -> 2 -> 4 -> 8…)
Not so for the ID NOW test. Abbott’s tight-lipped on the details, but one thing they do tell us about the test is that it uses “isothermal amplification” – so, when it comes to DNA copying, it “flattens the curve” temperature-wise. The whole thing is carried out at the same (iso-) temperature (-thermal) and it’s able to do this by relying on a different way of pushing the copied DNA off.
If similar to its “Alere influenza tests,” this “different way” is called NEAR, which stands for Nicking Endonuclease Amplification Reaction. It uses a pair of “DNA scissors” (an Endonuclease) which only cuts a single strand (hence “Nickase”), providing a free end for a DNA copier (DNA polymerase) to fill it back in, creating more and more copies of the cut off region in a Reaction that thus Amplifies it!
Instead of making “full copies” of a strand, as is done with traditional PCR, it makes “half copies” where, instead of the end of a primer, the free 3’ -OH comes from the cut spot (which is specified by a sequence the nickase likes). DNA Pol copies from that cut, filling the strand back in. And then the nickase cuts it again. And, since the cut piece is fairly short, it’ll fall off without you needing to raise the temps (especially if DNA Pol starts plowing into it). So you don’t need all that temperature cycling and Pol doesn’t have to “wait around” for the next cycle to get back to work – it can go straight back to copying once the cut piece comes off.
If you provide one such “starter piece” per strand (they call them templates which I found pretty confusing term-wise…) you can copy complementary pieces of each strand, which can then stick to “starter pieces” and get filled in to produce even more strands that can get cleaved to get more strands. Over and over until things run out.
Detection is done with “molecular beacons” – these are kinda “gagged fluorophores” – they’re stretches of DNA that have a fluorophore on one end that’s capable of letting off light if you shine a certain wavelength at it, and a quencher on the other end that’s able to prevent light from being given off (it “steals” the energy that would be released as light through something called FRET). When alone, the DNA folds up into a hairpin-like shape so that the fluorophore and quencher are close together. But when the DNA binds to a complimentary sequence, this pin gets pried apart, so that the quencher is no longer close enough to steal the fluorophore’s spotlight. So molecular beacons can be designed to only bind (and thus give off light) if they find one of the copied bits of viral genetic information.
Since you don’t have to go through all that cycling, the amplification process is faster. But the real gain speed-wise is in the prep step. For the traditional tests, you first have to isolate the RNA through a delicate “RNA extraction” procedure that has to be carried out in a lab. But with the ID NOW test, all the cell breaking open (lysis) and RNA isolating is done in the machine. So, instead of several hours start to finish (not to mention the time getting the sample to the lab, etc.) you’re looking at less than 15 minutes to get results (~5 for positive results, 13 for negative results). Sounds great, right? Well, to the individual patient it is great. But, on the bigger scale of things, that was only 1 test – 1 patient. And that’s all you can do at a time with a single machine.
With traditional tests on the other hand, you can run a lot at a time. They’re typically done in 96-well plates. Even if you do 2 reactions (targeting different viral genome regions) per patient (as is commonly done to ensure result reliability), and you include negative and positive controls to make sure the test works, you can still test a lot of patients at once. And the process can be automized with robots working nonstop, plate after plate.
So, when you go to one of those test centers, they’re probably gonna ship your sample off to one of those labs, leaving you to wait, which must really suck. And I sincerely feel for each and every person that’s in that situation. But one test at a time just can’t handle the demand for testing. So, while ID NOW is great for “Point Of Care” testing, where doctors can perform the tests right in their office or clinic, it’s just not scalable to the level of mass testing needed for the general public. But it can potentially be really helpful for testing frontline workers to make sure they’re safe to work.
Don’t confuse this Abbott “ID NOW” test with those high-throughput Abbott tests you might have heard of – those “m2000” machines are for running the “traditional” RT-PCR assays in a highly-automated fashion. According to Abbott, each machine can run 470 tests a day https://www.abbott.com/corpnewsroom/product-and-innovation/an-update-on-abbotts-work-on-COVID-19-testing.html
And don’t confuse the ID NOW-type rapid tests with the rapid serological tests. Serological tests look in the blood for little proteins called antibodies that an infected person’s body develops to specifically target the virus as part of the adaptive immune response. People produce these antibodies in the later stage of the infection, and some stick around to keep watch and (hopefully) provide immunity against re-infection. Serology tests look for these antibodies to see if people have had the infection in the past (or are in the later stages of the infection). These tests are fast too but, unlike the ID NOW tests, they DO NOT detect active infections until several days in. They do, however, have the advantage of detecting evidence of past infections, which ID NOW can NOT.
So, when it comes to the kind of results you might be able to get quickly:
- ID NOW “rapid tests” can detect ACTIVE (so potentially contagious) infection – even in people without symptoms
- serology tests can detect PAST and/or LATE-STAGE infection
Finally, other types of rapid tests being developed, so, well ID NOW might be a true game-changer, I’m hoping that something else will. For now, I hope you’re all staying as healthy and happy as possible.