Thread by @taaltree

You're home. What else do you and/or your kids have to do but learn something about DNA, RNA, and RT-PCR? Never a more important time to understand some of the details. A thread on how we detect the novel coronavirus starting with the basics of DNA and RNA.

The first thing to remember is that many of the most important ingredients for life are chains of molecules called polymers. Starch and fiber are polymers of sugar molecules. Proteins are polymers of amino acids. DNA and RNA are polymers of nucleic acids.

The polymer is connected by a sugar and phosphate chain, but the information is stored in one of 5 nitrogen base that they carry (T, G, C, A, and U). DNA uses the first 4 of those bases. RNA the last 4. Each base can loosely bind to a complementary base with a hydrogen bond.

A hydrogen bond is like a magnet. H gives its electron to N, so the H is positively charged. This + is attracted to the - on an O or N because they have pulled electrons from whoever they're bound to. This sound crazy, but hydrogen bonds are the key to everything we do with DNA.

So H-bonds connect C to G and A to T. This is how two strands of DNA come together to form a double helix. We call them the "sense strand" and "antisense strand". A small part of our DNA has "genes" that make RNA that matches the sense strand except that T is replaced with U.

This is called "transcription". This RNA is a single strand unlike DNA. Some splicing happens to the RNA and it's then made into a protein in a process called "translation". Now, this virus is mostly a strand of RNA and proteins.

Once it gets into your cells, your body sees its RNA and thinks it knows what to do. The virus hijacks your cell and you help it replicate given the genes it carries. Now THIS RNA strand is what a coronavirus test is trying to detect. How do we do it?

First we extract RNA. RNA degrades quickly and most things we do in the lab use DNA, so the next thing is to convert RNA into complementary DNA (cDNA). Now we have something double stranded but with the U from RNA instead of the T from DNA. OK, how do we detect this cDNA?

Our tool to detect DNA is to see if we can replicate it many times using "polymerase chain reaction". This is where the H-bonds come in. This process happens in a "thermocycler". We heat the cDNA so it denatures into two single strands. Then we rebuild the complementary strand.

To rebuild the strand we introduce primers, which are just a short stretch of nucleotides that match *only* the thing we're trying to detect. So the coronavirus test has two primers that amplify this cDNA but not other things. The second DNA strand rebuilds using an enzyme.

This enzyme called "Taq polymerase" has a VERY interesting story. Remember that enzymes are supposed to denature at high temperature. "Taq" stands for Thermus aquaticus, which is a species of bacteria isolated in 1969 from geysers in Yellowstone. This enzyme likes it hot!

Now, the thermocycler alternates temperature so that, just like COVID19, DNA grows exponentially as one molecule becomes 2, 2 become 4, 4 become 8, and 8 become 16. Pretty soon you have billions of DNA strands. We can see if the virus is there by seeing if anything amplified.

In the past, you would detect whether anything amplified using gel electrophoresis, which is basically using electricity to pulling the DNA you made through some jelly. If you see a dark band at the right place (how much it moves depends on size of the DNA) then you test positive

Now tests use a more sensitive measure called real time PCR. In addition to the two primers, we add a probe that is a short string of nucleotides that binds between the primers. It includes something fluoresces (F) and something that stops it from fluorescing (quencher).

The probe doesn't fluoresce because of the quencher, but as the Taq polymerase makes the complementary strand, it kicks off the thing that fluoresces first. In real-time PCR, a much more expensive thermocycler monitors the fluorescence in "real time".

Our lab primarily uses this to detect animals or pathogens in water samples, but also pathogens like the fungus that causes white-nose syndrome in bats, the protozoa that causes Leishmaniasis, and some mosquito-borne viruses. I digress - now, how do we interpret RT-PCR output?

Just like COVID19 cases are linear on a log-scale until the pool of susceptible hosts runs out, so is the exponential process of DNA replication until reagents run out. If your sample has more of your target DNA that you made from RNA, then it amplifies earlier.

Each sample produces one curve. The number of cycles it took to reach a threshold is called the Ct score. The green samples had a lot more DNA than the red and way more than the other samples. We quantify the amount of DNA by comparing with a "standard curve" of known DNA amount

The coronavirus tests look for fluorescence rising above the threshold. That's what a positive test result means. The process that I've described can be easily done by thousands of labs all over the country. So why weren't we rapidly mobilized to do testing in our communities?

The answer is that as a research lab, we're allowed to test for research purposes but not for diagnostic purposes. Congress passed the Clinical Laboratory Improvement Amendments (CLIA) in 1988 which requires certification to do diagnostics.

Helen Chu from UW was doing research on influenza in Seattle when she "illegally" added the coronavirus primers. Her lab identified the epidemic in Seattle early but was not CLIA certified. Her request to continue testing was denied by the federal gov't.
mynorthwest.com/1758762/corona…?

If we want to test for research, a different bureaucracy stands in the way called the Institutional Review Board. We need to go through this, for good reason, whenever working with human subjects. There were weeks when all the labs at universities could have been testing.

Instead people like Helen Chu were asked not to test. My university formally asked for all R1 research universities to be allowed to test with a letter to the Whitehouse in early March. Nope.

OK. Hopefully this has been somewhat interesting and informative.