Tomorrow (2020 05 14) I am scheduled to be tested for the SARS-CoV-2 virus as I am among the first wave of employees to be allowed back to the research facility where I work. If I am healthy.
I’ve watched a number of these tests from afar as I parked my car in the garage so I know exactly how it goes. It's like this...
I am to drive up to the test station – really just a couple of hastily erected tents and tables just inside the parking garage – I’ll wait behind dozens of other cars and then pull up a little more and get my appointment confirmed and wait some more – and eventually one of the gowned, shielded, and masked nurses will approach me and ask some questions and fill out a form and I’ll wait some more and then just when I am about to drive away in frustration the nurse will commandingly yell at me to stop! and then draw a 10 foot long cotton swab from a shoulder holster like Angelina Jolie in a kickass movie and twirl it martially and start running towards me, then plant the swab in my nose as she pole vaults over my car.
Really.
I am terrified of that swab. It is as long as my head is wide. They could swab both ears at the same time.
But once they finish that swabbing, what happens next in the testing after I drive away with my head mummified in gauze to stanch the mortal swab wound?
The very first thing is that the swab immediately goes into a plastic tube and a screw cap tightened in place to keep out contaminants. The tube may contain saline or it may contain the reagents to extract RNA from all the other gook that came out of my nose. I don't really know but I am voting for the RNA-extraction reagents in the tube and I'll explain why shortly. Feel free to correct me if you know otherwise.
Before we go any further, let’s take a quick look at the coronavirus afflicting us right now. Here’s a cartoon image of a coronavirus and I want to focus on the cross-section image on the right.
That figure is essentially a parts list for a virus. To test for any pathogen we would ideally like to detect whatever makes it unique and is easy to find. One possibility is that we can test for the unique or distinguishing proteins such as the large spike proteins protruding from the virus forming the pink corona in the cartoon. Speaking of pink coronas, my kid asked if viruses have color - a great question. But unfortunately for those with a great fashion sense, viruses are smaller than the wavelength of visible light, so they do not exhibit color in any way we can perceive. Back to detecting proteins uniquely... a great way to do that is to use effective biological detectors from our immune system – antibodies (actually we raise antibodies for testing from animals like mice, rabbits, and goats - humans raise antibodies for their own personal use). But antibody-based tests can be a little finicky and are notorious for a high rate of false positives – meaning the test says you have the virus (positive) but the test is wrong (false).
I have some personal experience with antibody tests as I am sure many of you do. I have had Lyme disease (caused by a bacterium called Borrelia burgdorferi transmitted by deer ticks), and the usual test is a Western blot whose goal is to detect burgdorferi proteins in your blood. Here’s how the Western blot goes. First, the proteins from your blood are run on a gel which sorts the proteins in a size-ordered array (big proteins at the top, small proteins at the bottom). We know how big the burgdorferi proteins are – but there are many proteins of similar size so we need something more specific than size alone. The proteins are transferred from the floppy, fragile gel to a more robust plastic membrane and then an antibody against the burgdorferi proteins can be applied to the membrane. If the antibody attaches to the burgdorferi proteins affixed to the membrane, that antibody in turn can be detected by a second antibody which binds to and amplifies the first antibody. A light-emitting marker attached to the 2nd can be detected by a camera. No light means no Lyme disease. A light band at the correct size might mean you have Lyme. Since these tests are prone to false positives, the CDC has recommended a two-step diagnosis starting with a 1st screening test in which a negative result is reliable and no further test is required. IF the 1st test is positive then it is followed by a second test in which multiple positives are required for a positive diagnosis.
Whew! A mess, right? These antibody assays are not quick and reliable tests.
What else can we look at in the virus as a target for a test?
The coronavirus envelope (the hollow red spherical part in the above figure) is actually derived from part of the host cell’s membrane. When the coronavirus blebs out of the human cell, the human cell membrane wraps around the viral RNA creating the finished viral particle. Therefore the envelope is not at all diagnostic. The proteins embedded in the viral envelope of course could be the target of a diagnostic test, but those would have the same problems as the other proteins we discussed above.
The last thing we can really focus on as a target for a good diagnostic test is actually the best one – and that is the virus’s RNA. The RNA encodes all the protein needed by the virus, and embodies all the information which makes the virus unique. Furthermore, test methods for detecting specific RNA sequences are routine and fairly fast. Although RNA is the best diagnostic target it is not perfect.
There are a few bad things (when I say bad I mean unfortunate, not morally bad) about RNA… first, look at where the RNA is located. It is protected by the envelope and a lot of protein. Another bad thing is that RNA is very fragile and degrades quickly by natural chemical degradation. Even worse, since RNA viruses have been with us since the beginning of time, evolution has ensured that every single cell and organism has designed innumerable ways to detect and destroy RNA. We have a whole suite of proteins called ribonucleases (or RNases for short) whose job is to destroy RNA. RNases exist on our skin, in our saliva, every cell in our body makes them, and they are among the toughest proteins to destroy. While these issues are “bad”, they are manageable.
We’ll see that available test methods are able to address and manage these various issues.
I believe the reagent in the tube that the nurse puts the swab into after mopping out my nose is similar to what I use in my research to extract RNA from the yeast cells I work with. The trade name is Trizol, and it contains the key ingredients guanidinium isothiocyanate and acid phenol.
The core function of the guanidinium isothiocyanate is to denature or melt proteins in an RNA preparation. There are a couple important effects of guanidinium isothiocyanate. One effect of protein denaturing is to break open cells. Mammalian cells are relatively fragile things, unlike the yeast in my lab which requires shaking with ceramic beads to beat the living bejeebubs out of them to release their contents. In contrast our cells are soft watery bags of stuff held together by protein scaffolding, and so a chemical like guanidinium that denatures proteins is sufficient to break open cells (and similarly coronaviruses). A second effect of protein denaturation is that, since RNases are proteins, this guanidinium isothiocyanate effectively protects the RNA from biological degradation. See? Right away we are managing the issues we raised about RNAs.
Acid phenol is used to separate RNA from DNA based on the differences in their chemistry. RNA at low pH prefers to associate with water, while DNA at low pH prefers to associate with an organic solvent. If we add chloroform to the mix and spin the solution in a centrifuge, the aqueous and organic phases separate: RNA remains in the clear aqueous solution which sits atop a red organic solution containing the proteins and DNA.
We can now carefully extract the aqueous solution which has purified RNA leaving the DNA and proteins behind. We have purified intact RNA.
Now we’re ready to test.
The first thing we need to do is to convert RNA into DNA so we have a stable molecule to work with, and that we can easily analyze. Although we removed the RNases that enzymatically digest our RNA molecules, RNA is still liable to chemical degradation as well.
Here’s the fun thing. In order to convert RNA to DNA, we use a viral enzyme called reverse transcriptase. It’s kind of like spy versus spy… using a virus to catch a virus.
There are some fascinating viruses called retroviruses, which are RNA viruses that are able to make DNA from their RNA, and insert that newly made viral DNA into the genome of their host. An example of a retrovirus is HIV. About 5-8% of our genome is comprised of “fossilized” retrovirus sequences that in the past have become trapped in our DNA.
Normally transcription goes from DNA to RNA. Messenger RNA (mRNA) is made by an RNA polymerase which uses a DNA as a template to make the RNA.
That is why we call the retrovirus’s enzyme a reverse transcriptase, because it uses RNA as a template, and makes DNA from it.
So in our assay, we take the retroviral reverse transcriptase enzyme and make DNA from the coronavirus RNA (and all the other human and bacterial and other RNAs that came along with the sample).
We often abbreviate the reverse transcription reaction as “RT”.
And to distinguish the DNA we made from RNA from native DNA from our cells, we call this complementary DNA or cDNA. We will see an example of a complementary pair next when we talk about primers binding to a DNA sequence that is complementary to the primer sequence.
Now that we have a bunch of cDNA from human, bacteria, viruses, and who knows what else was lurking in my nose goop, how do I pick out cDNA specific to possible SARS-CoV-2?
Here we use a classic laboratory method and this time I’ll tell you the abbreviation up front: PCR.
Many of you have heard of PCR, or polymerase chain reaction. But let’s break it down some.
Polymerase refers to an enzyme all living things have which we need for DNA to replicate. We need to replicate DNA so cells can divide, so we can pass our DNA to our kids, to repair DNA damage, etc. So we are talking about DNA polymerase which makes new DNA molecules from a DNA template (previously we talked about a viral reverse transcriptase which makes DNA from an RNA template).
The chain reaction is conceptually straight forward. Right? When we think of a nuclear chain reaction, for example, we think of a single uranium-235 atom which splits (fission) and gives off energy plus 3 high-energy neutrons. Each released neutron causes another U235 atom to split and which releases 3 more neutrons for a total of 9. Those 9 neutrons each causes fission in another U235 so we have 27 neutrons released, and so on and on for an exponential and ultimately explosive reaction.
Similarly for PCR, the polymerase enzyme makes a DNA copy of the original cDNA. Then the next round, a copy is made of each of those for 4 copies. In the 3rd round a copy is made of each of those 4 copies, for a total of 8 copies… and we have another exponential growth of copies.
But for a polymerase to make a DNA copy, it needs something called a primer. A primer is a short DNA sequence that is complementary to the sequence of the target DNA. And that primer being of a sequence that matches a particular target’s sequence is what makes PCR so specific and able to detect viral cDNA (i.,e., RNA) among so much other contaminating cDNA.
Let’s take a look at that complementary sequence in a little more detail.
Since we know that the RNA sequence contains information that distinguishes the SARS-CoV-2 coronavirus from other viruses, as well as from every living thing, we want a test (PCR) that takes advantage of that informational specificity. RNA is composed of a four-letter code arranged in a particular sequence. Let me show you part of the RNA sequence that encodes the spike S protein as an example (the full sequence is in the link below):
ATGTTTGTTTTTCTTGTTTTATTGCCACTAGTCTCTAGTCAGTGTGTTAATCTTACAACCAGAAC
Each three letters of the RNA or DNA sequence encodes a specific amino acid. For example, the first three letters of the sequence above, ATG, encodes the amino acid methionine (abbreviated M). When you see a sequence of a protein, you will see a long sequence of twenty different letters, not just four like our RNA code above.
The RNA sequence for the same S protein but in the SARS-CoV-1 virus is listed in this link:
The first line of the SARS-CoV-1 S protein RNA sequence is as follows:
ATGTTTATTTTCTTATTATTTCTTACTCTCACTAGTGGTAGTGACCTTGACCGGTGCACCAC
Let’s line up the first line of both viral S proteins next to each other
ATGTTTATTTTCTTATTATTTCTTACTCTCACTAGTGGTAGTGACCTTGACCGGTGCACCAC
ATGTTTGTTTTTCTTGTTTTATTGCCACTAGTCTCTAGTCAGTGTGTTAATCTTACAACCAGAAC
| | | | | | | | | | | | | | | | | | | | | ...etc
I put a vertical line under the letters that differ between these two closely related viral species.
Recall that for DNA, the letters of the DNA sequence bind in pairs: T:A and C:G.
So if we have the first viral sequence for the S protein (1st long line below) aligned with its short complementary primer (2nd short line), we would have:
ATGTTTATTTTCTTATTATTTCTTACTCTCACTAGTGGTAGTGACCTTGAC (cDNA sequence)
TACAAATAAAAG (a pretend primer)
Note in our pretend primer, a T sits opposite an A, and a C sits opposite a G. When the sequence of the primer matches the target sequence in the RNA or cDNA, it will zip up and bind strongly and specifically to that site. If there are differences in the sequence of the primer to the target, the binding will be weak.
And if we do the same for the second viral sequence for the S protein aligned with its own short primer, it would look as follows:
ATGTTTGTTTTTCTTGTTTTATTGCCACTAGTCTCTAGTCAGTGTGTTAAT (cDNA sequence)
TACAAACAAAAAG (another pretend primer)
You can see easily that the short primer for the 2nd virus does not pair up with the sequence for the 1st virus (and vise versa).
That sequence specificity is what enables PCR to precisely pick out a viral cDNA from all the other cDNA from whatever sources happened to be up my nose, including my own.
If there are no viral RNA (or cDNA) sequences, during the PCR reaction the primers have nothing to bind to, and no DNA copies are made.
But if there are even a few viral RNA copies in the nasal swab sample, the primers will bind to them and become amplified, and we will be able to detect the presence of so many copies of DNA.
There we have it in a nutshell... the nasal swab sample goes through an RNA extraction step where we purify RNA from all the other nose goop, followed by reverse transcription (RT) where we make a stable analyzable cDNA from all RNA in the purified sample, and last by polymerase chain reaction (PCR) which amplifies a specific cDNA depending on the primers used. This is why you see the RT-PCR acronym popping up everywhere. Hopefully that explanation sheds a little light on what happens to that nasal swab – and will make an unpleasant experience a little less traumatic.
If this was helpful, please email it to friends and family, and also subscribe and comment. I appreciate corrections, suggestions, and a howdy duuude!