mRNA vaccine biochemistry – What are they molecule-wise? How are they made? How do they work? and – in answer to my dad’s question, Why do they need to be kept so much colder than traditional vaccines?

Note: you can find an expanded version of this post here:

You can hardly turn on the news or log onto Twitter without hearing about vaccines for SARS-CoV-2, the novel coronavirus that causes the disease COVID-19. Especially since a couple of leading companies in the quest to develop an mRNA-based vaccine for it, Moderna and Pfizer, have gotten Emergency Use Authorization in the U.S. and are being given to high-priority groups already! Today’s post isn’t going to be like those typical news stories about number of doses and prioritization rankings, etc., that are important, yet easier to find. I have found it is a bit harder to find information about the biochemistry behind them, so that’s what I’m aiming for in this post. 

TLDR version: In response to viral infection, the immune system makes little proteins called antibodies which specifically bind viral parts and call for help to combat the infection. And then, some stick around for a while to keep watch in case the virus tries again. The point of a vaccine is to get people to make such watchmen antibodies without those people having to actually get the disease. Traditional vaccines have done this by using inactivated (“killed”) or severely weakened (live attenuated) versions of the virus – so your body sees “pre-made” viral parts. Another strategy is to introduce lab-grown and purified viral proteins (“recombinant proteins”). The idea with mRNA vaccines is that, instead of introducing premade viral pieces, you introduce genetic instructions for making some viral proteins (which are harmless on their own) and get the injected person’s body to make them. 

Basically, you take messenger RNA (mRNA), which are copies of the recipe for making a protein (in the case of the SARS-CoV-2 coronavirus, the recipes are of the Spike protein) and get those mRNAs into cells (often snuck in by encapsulation in a lipid coat). These mRNAs are then used (over and over again) by the cell’s ribosomes (protein-making complexes) to make lots of the corresponding protein, Spike, which then gets displayed to the immune system which learns to recognize it as foreign and produce an immune response against it. 

SARS-CoV-2 is a single-stranded RNA virus – it holds all of its genetic information, including instructions for making viral proteins, in one strand of RNA. When the virus travels, it coats its RNA in proteins and a lipid (oily) coat. Embedded in that coat are proteins including the Spike protein, S, which is the one that juts out from the surface and docks onto ACE2 receptors on cell surfaces, then shape-shifts to fuse the viral membrane with the cell membrane and dump the viral RNA inside. 

Since the genetic code is universal, our cells are able to read these instructions and make viral proteins. Viruses “know” this, so they’re able to hijack our cells and get us to do their bidding, including making their proteins for them. But, can we beat them at their game? If we could introduce viral protein-making “recipes” (genes) into our body, our body’s cells could make those proteins – but, since we’re only giving them parts of the viral cookbook, no active virus will get made. ⠀

It’s kinda like making a single car part. Enough to recognize as a car part, but not enough to actually drive. This way, since your cells are doing all the hard work, production can be easier, and, unlike vaccine strategies which consist of pre-made, purified, proteins grown “recombinantly,” the proteins your cells make are “perfect” – just like the real thing! Any “post-translational modifications” like glycosylation (sugar addition) get added where they normally would, which could be really important because the S protein is quite sugar-coated! (it’s a glycoprotein)

jargon watch 1: post-translational modifications are bells and whistles that get added onto the protein after it’s “written,” a bit like drawing a heart over an “i.” The writing process for proteins is called translation and it involves protein-making complexes called ribosomes linking together the amino acids (protein letters) specified by 3-RNA-letter “words” called codons present in the mRNA. There are 20 common genetically-encoded amino acids, each with a generic backbone part which allows them to link together and a unique “side chain” (aka R group) that sticks off like a charm on a charm bracelet. Once a protein’s translated, you can’t change the order of these charms, but you can add “extras” such as sugar chains and phosphates onto some of them and this offers an extra level of regulation, etc. 

jargon watch 2: practically speaking “recombinantly” means made in a lab – technically, it means that you’ve inserted the genetic instructions for making the protein into an easy-to-work-with piece of DNA called a vector which has instructions that tell easy-to-work-with cells (such as bacteria or insect cells) to make that protein. The cells will make mRNA copies from the DNA (in a process called transcription) and then make protein from that mRNA in translation. Since you’ve recombined pieces of DNA we call this recombinant DNA and we call the protein that gets made recombinant protein. 

Recombinant expression is a strategy I use all the time to make proteins to purify and study, but the expression cells might add on different post-translational modifications since those modifications aren’t called for in the mRNA sequence and instead depend on the cellular machinery to know what to do. And different cells have different machinery so the insect cells might not add the right sugar chains, etc. which could cause the immune system to learn to recognize the wrong thing. 

Another disadvantage to using recombinant proteins as a vaccine strategy is that you only get as many protein copies as you purify and put in. On the other hand, if you put in the genetic instructions, you can get lots of protein copies from each mRNA copy and you only have to purify the mRNA & not the protein. 

mRNA can get made commercially using a process called in-vitro transcription. Lot’s more on it here: 

But the basic idea is that you use recombinant DNA to make mRNA (transcribe it), but you “stop” before the translation (protein-making) step. “in vitro” just means that we’re doing it basically “in a test tube” (or a vat or something, just not in an actual organism). So, “in vitro transcription” is a way to make lots of RNA from a DNA template, often using the RNA polymerase (RNA copier) of a bacteriophage (bacteria-infecting virus) called T7. 

Like other RNAPs, T7 RNAP “recognizes” (by selectively binding to) specific stretches of DNA called promoter sequences which are located upstream of the start of the gene that they copy. note: for those people like me who always have to think twice about upstream vs. downstream, promoters are in front! If you stick the T7 promoter in front of a complementary DNA version of the RNA you want made (either in double-stranded DNA (dsDNA) such as a linearized plasmid, or single-stranded DNA (ssDNA) with a double-stranded promoter region), you can get T7 RNAP to make lots of that RNA. 

Problem is, RNA is really unstable. So you need a way to get the instructions safely into cells – and keep it from degrading before you even inject it!

This is a major practical difference between mRNA vaccines and traditional vaccines made from inactivated or weakened viruses. In those traditional vaccines, the viral goodies have the benefit of being protected with the virus’ lipid coat. Then, when the virus gets swallowed up by a type of immune cell called antigen-presenting cells, those cells chop up the viral proteins and display pieces of them on their cell surface to serve as antigens (things the immune system can learn to make antibodies against through a trial-and-error approach of seeing what sticks).

With mRNA vaccines, you don’t have that viral coat. And if you were to just inject a bunch of “naked” RNA into someone, it’d quickly get chewed up by non-specific RNA chewers and cutters called RNases. If you’ve worked in a biochemistry or molecular biology lab, you might be all too familiar with RNases because they’re hardy, practically everywhere, and capable of ruining some of your experiments! RNases are so ever-present because they offer an easy, generic, method of offense against viruses. Cells, organisms, etc, can protect their own RNA by keeping it in membrane-bound areas, guarding it with proteins, etc. and then make and secrete RNases to degrade free RNA which could correspond to viral wannabe invaders. 

One defense scientists can use to protect naked RNA is to store it at really really cold temperatures, where RNases are inactive and RNA is less likely to “break” for other reasons (like one of its -OH legs attacking the other). 

In the lab, for example, we keep our RNA at -80°C and this is the temperature that Pfizer’s vaccine requires, which will likely prove to be a real impediment for getting the vaccine distributed in remote and rural locations, where, unlike in our lab, where we have 3 big -80°C freezers, some places will likely have to rely on dry ice. Even for places with -80°C freezers, dry ice will be critical in the shipping process as well. more on dry ice here: 

So, the super-cold-ness can protect the mRNA before it gets into cells. And then, once released inside cells, it’s protected by the same protections as cellular mRNAs (a 5’ cap at one end and 3’ poly-A tail at the other). 

But how do you actually get the mRNA into cells? Cells don’t just let in anything that comes their way, especially not RNA which, thanks to the negatively-charged phosphate (PO₄³⁻) groups in the RNA backbone, isn’t really compatible with the cell’s oily membrane…

Vaccine manufacturers can sneak it in by encasing it in its own oily membranes – in the form of  self-forming sphere things called Lipid NanoParticles (LNPs) which encompass the RNA in a blend of lipids. Now for a bit more jargon… 

We’ve been talking about lipids as “oily things” and they are – mostly… When I’ve talked about “oily” I’ve basically been talking about “hydrophobicity” which is when a molecule is excluded from water. Water is highly polar, meaning that, although it is neutral overall, it has partly positive regions (the hydrogens) and a partly negative region (the oxygen). Opposite charges attract, so water molecules can form extensive networks and they’ll only let other polar or charged things hang out with them (i.e. hydrophilic things). “Oils” are made up of long hydrocarbon chains (carbon and hydrogen based) which are uncharged, nonpolar, and therefore unattractive to water (i.e. hydrophobic). So if you stick oil and water together, they won’t mix.

But the lipids making up our cell membranes are “phospholipids” – they have phosphate-containing head groups on top of their oily (water-excluded) tails. And, as we just talked about, phosphate groups are negatively-charged. Since phospholipids have parts that water likes (the hydrophilic heads) and parts that water doesn’t want to hang out with (the hydrophobic tails), we call them “amphiphilic.”

As a result, if you stick them in water, they’ll assemble into phospholipid bilayers which are like phospholipid sandwiches where the tails (which are water-excluded) hang out together like the peanut butter and the head groups face towards the water, like the (now-soggy…) bread. This way you can make water-based compartments (like cells) inside water-based compartments (like bodies). 

watery outside environment





watery inside environment

You can also get membrane-bound “sub-compartments” inside of cells, and one way these form is through a process called endocytosis. It’s kinda like a biochemical sinkhole -> basically the membrane sucks in a part of itself (including anything bound to it) and pinches it off (inside the cell) so that, although whatever was bound to it is now inside the cell, it’s surrounded by a membrane, like a little membrane-bound packet called an endosome. 

If we get our LNP to bind to the membrane, it can hitch a ride inside and then, once inside, the lipids in the LNP can destabilize the endosomal membrane, allowing the RNA to spill out into the cell (endosomal release). 

But first the LNP needs to bind the membrane and there seem to be multiple ways this can happen. The exact details depend on the various formulations and the cell types, but LNPs are often a mix of neutral, cationic (positively charged) and/or ionizable lipids (lipids which are neutral or near-neutral at some pHs but charged (ionized) at other pHs). The cationic group can help with binding similarly to the reason we often use cationic carriers such as PEI when doing DNA transfections in the lab (transfection is where you get a cell to take in nucleic acids (DNA or RNA) and the term’s typically used when we’re working in the lab but I guess it’s also what these vaccines are doing in bodies!) 

The reason we use cationic carriers in transfection is that DNA & RNA have negatively-charged (anionic) backbones. And cell membranes have those negatively-charged head groups – and we know that like charges repel. So naked RNA, even if it did manage to evade RNases, wouldn’t want to get anywhere near the cell surface. In fact, that RNA wouldn’t even want to get near itself! But cationic molecules can counter-balance the negative charge making it less repulsive – and potentially even attractive! In our bodies, for example, RNA & DNA often hang out with positively-charged magnesium ions Mg²⁺. 

When we do transfection, we often do the charge-neutralizing with molecules called cationic lipids. They’re similar but opposite to phospholipids – they have a hydrophobic tail and a charged head, but cationic lipids have a positively-charged head. This allows them to glob onto the negatively-charged RNA to keep them from repulsing the membrane and helps them stick to the phospholipid groups on the membrane. It works great for transfection in the lab, but cationic lipids can cause problems in the body, such as binding non-specifically to (and getting stuck on) plasma proteins (proteins in the blood). 

So, instead of relying on “always-positive” lipids, LNPs for in-body use have turned towards “sometimes-positive lipids” called ionizable lipids. Whether they’re positive (and/or how positive) they are depends on the pH. Lower pH (more acidic conditions) means there are more protons (H⁺) available. The ionizable lipids have “protonatable groups” which are able to take some, becoming protonated and positively-charged (cationic). But at higher pH, when there are fewer protons available, they won’t be protonated and thus will be neutral and less plasma-protein-sticky. So, you can assemble the LNPs at a low pH, when the lipids are + charged and then raise the pH to and end up with a complex thing where you have ionizable-lipid-coated RNA inside of little pockets inside of a broader, neutral, lipid shell. 

These can dock onto cells in different ways, some of which rely on sticking to lipid-binding receptor proteins. Then, once swallowed inside but trapped in the endosome, the ionizable lipids get to take a starring role again because, as part of the endosomal maturation pathway, the endosome acidifies, so the ionizable lipids becoming protonated and positively-charged (cationic) again. This helps them interact with, squirm into, and destabilize the endosomal membrane.

At least that’s the general idea I think – I’m a little fuzzy on the actual details and it seems to vary depending on the formulation.  DOI:10.1038/mt.2010.85 

But Pfizer’s mRNA vaccine uses an LNP formulation made by a Canadian company called Acuitas.  And, at least in the past paper their website sends you to, they “contain an ionizable cationic lipid (proprietary to Acuitas), phosphatidylcholine, cholesterol and PEG-lipid.”  DOI:10.1038/nature21428 

Phosphatidylcholine & cholesterol are both components of cell membranes, and they can act as “helper lipids” forming that outer shell. PEG-lipids are lipids that have been had a PolyEthyleneGlycol group attached – PEG is a long hydrocarbon with lots of hydrophilic -OHs, and such PEGylation  is often done to add an extra layer of protection, stability, bulk, and immune system avoidance.  

No matter what the formulation, the end result is that you (hopefully) set the mRNA loose in the cell, ribosomes make protein from it (remember you can get lots of protein copies made from each mRNA copy), and then the immune system makes antibodies against it, which I’m not going to get into in the current post. But I did try to in this post 

mRNA vaccines have never been FDA-approved for human use before. But, in their defense, the technology to make them hasn’t been around that long. Now that it is, however, the technique shows great promise – not just for Moderna and Pfizer/BioNTech and their coronavirus vaccines, but potentially for a wide range of viruses. They have a few key things going for them… mRNA is much easier to manufacture than other vaccine types, and it allows for an “early start” –  all you need to know is the genetic sequence of the virus – you just need the recipe and you can go to work. Of course, they have to choose the right protein and maybe tweak it a bit, but it’s still much much easier that having to actually express and purify the protein! Plus, each mRNA copy can be used countless times by the ribosomes, so you can get a much greater return on investment than if you just put in premade protein. 

But (at least these ones) do have that freezer problem. Pfizer’s has to be stored at -70°C (-94°F) which is really really cold, much colder than a “normal” freezer”. Moderna’s just needs -20°C (-4°F) which is just like your home freezer. But still not ideal 

Thankfully, mRNA vaccines are just one of many many types of vaccines being tried, and some of the other methods don’t require such a cold “cold chain”! Vaccine development has been portrayed as a bit of a race (didn’t help that Russia named their vaccine “Sputnik”…) but in reality, the more vaccines the better! Because we have a looooot of people to vaccinate – all over the world – in all sorts of living situations and nearness to high tech low-temp freezers. And hopefully the technology will advance to make mRNA vaccines more shelf stable – and freezers (and health care in general) more accessible to everyone. 

Hope that helped answer your question dad (and hopefully you made it all the way through the post to see this! sorry-not-sorry for all the much-more-than-you-asked-for content!)

2 Thoughts on “mRNA vaccine biochemistry”

  • A nitty gritty question:

    What cells actually imbibe the Pfizer/Moderna lipid packets to be transfectd, or are infected by the J&J/Astrazeneca adenovirus?

    Are they the MUSCLE cells AT the injection site?
    Or vasvular endothelial cells after being spread about the lymphatic-vascular stem?
    Circulating leukocytes?
    Or any organs throughout the body?

    • Hi – this stuff is out of my expertise, but for the mRNA vaccines I believe the vaccine goes into muscle cells and drains into the lymphatic system. The targeting isn’t specific but it is localized around the injection site and the lymphatic system. They talk about this sort of thing in this TWiV podcast:

Leave a Reply

Your email address will not be published.