The Chemistry of mRNA Vaccines

In December 2020, the first mRNA vaccines were approved for use against COVID-19, less than a year after the virus’s genetic sequence was published. Their speed looked like a miracle, but it rested on decades of chemistry: learning how to make RNA that the body tolerates, and how to wrap a fragile, negatively charged molecule in fat so it can get inside cells. This article explains how mRNA vaccines work from a chemist’s point of view.

The basic idea

A traditional vaccine gives the immune system a weakened or inactivated virus, or a piece of a viral protein, so it can learn to recognise the real thing. An mRNA vaccine instead delivers the instructions for making one viral protein. Your own cells read the instructions, make the protein, and display it, and your immune system learns to recognise it.

For the COVID-19 vaccines, the protein was the coronavirus spike protein, which the virus uses to enter cells. The vaccine contains no virus, and the mRNA can’t make a virus: it codes for just one protein.

Part 1: the mRNA

Messenger RNA is a single-stranded nucleic acid: a chain of nucleotides, each with a ribose sugar, a phosphate group and a base (A, U, G or C) (see nucleic acids and DNA vs RNA).

The vaccine mRNA is made in the lab, not in cells, by in vitro transcription: an enzyme (RNA polymerase from a virus that infects bacteria) copies a DNA template into RNA using ribonucleoside triphosphates. It’s then carefully purified.

A therapeutic mRNA has several designed parts:

1. A 5′ cap: a modified guanine nucleotide attached back-to-front, which protects the RNA from being chewed up from that end and helps ribosomes recognise it.
2. Untranslated regions (UTRs): sequences before and after the coding region, chosen to make the mRNA stable and efficiently translated.
3. The coding sequence for the spike protein. The sequence was tweaked with two amino acid changes (two prolines) that lock the spike in the shape it has before it fuses with a cell, the shape antibodies need to recognise. The codons were also optimised for efficient translation (see transcription and translation).
4. A poly(A) tail: a long stretch of adenine nucleotides at the 3′ end, which protects the RNA and boosts translation.

The key chemical trick: modified nucleosides

Early attempts to use mRNA as a medicine hit a problem: injected RNA triggered strong inflammation. The immune system has sensors (such as Toll-like receptors) that detect foreign RNA, because many viruses use RNA. Detection caused inflammation and shut down translation, so little protein was made.

In 2005, Katalin Karikó and Drew Weissman found that replacing uridine in the mRNA with a modified version, pseudouridine, largely avoided these sensors. Later, N1-methylpseudouridine worked even better. In pseudouridine, the uracil base is attached to the ribose sugar through a carbon atom instead of a nitrogen atom, subtly changing its shape and hydrogen-bonding properties, while still pairing with adenine. The modified mRNA is translated efficiently and provokes much less unwanted inflammation. Karikó and Weissman received the 2023 Nobel Prize in Physiology or Medicine.

Part 2: the lipid nanoparticle

Naked mRNA can’t do much on its own:

• it’s a large, negatively charged molecule (every phosphate carries a negative charge), so it can’t cross the non-polar core of cell membranes (see phospholipids and cell membranes);
• enzymes called ribonucleases, present everywhere in the body, break it down within minutes.

The solution is to package it in lipid nanoparticles (LNPs): tiny fat spheres about 80–100 nm across. Each particle contains four kinds of lipid:

Lipid	Job
Ionisable lipid	The key component. It carries a positive charge at acidic pH, binding the negative mRNA during manufacture, but is nearly neutral at blood pH, which makes it less toxic.
Helper phospholipid (such as DSPC)	Forms part of the particle’s structure, like a membrane lipid.
Cholesterol	Fills gaps and stabilises the particle, as it does in cell membranes (see cholesterol chemistry).
PEG-lipid	A lipid attached to a polyethylene glycol chain; forms a water-loving outer coating that controls particle size and stops particles clumping.

How the particles are made

The lipids are dissolved in ethanol, and the mRNA in an acidic buffer. The two liquids are mixed rapidly in precise ratios, often in microfluidic devices. In the acidic conditions, the ionisable lipids become protonated (positively charged) and bind the negatively charged mRNA. As the ethanol is diluted away, the lipids assemble around the mRNA into particles. This depends on acid–base chemistry: the ionisable lipids are designed with a pKa around 6 to 7.

How the particles deliver mRNA

1. After injection into muscle, the particles are taken into cells by endocytosis: the cell membrane wraps around a particle and pinches off into a small internal bubble called an endosome.
2. The endosome becomes more acidic as the cell processes it.
3. At the lower pH, the ionisable lipids become positively charged again. They interact with negatively charged lipids in the endosome membrane, disrupting it.
4. The mRNA escapes into the cytoplasm, where ribosomes translate it into spike protein.

Only a small fraction of mRNA escapes this way, but enough protein is made to train the immune system.

Part 3: what happens next

• Cells display pieces of the spike protein on their surfaces, and some protein is released.
• The immune system makes antibodies and T cells that recognise the spike.
• The mRNA doesn’t enter the nucleus and can’t change your DNA. Human cells have no mechanism to copy ordinary mRNA into DNA and insert it into chromosomes, and the mRNA is broken down by normal cellular enzymes, typically within days.
• The lipids are broken down and cleared.

Why they needed such cold storage

The early mRNA vaccines required freezer storage, one at about −70 °C, which made distribution difficult. The reasons are chemical:

• RNA is chemically fragile: its 2′-OH group can attack the neighbouring phosphodiester bond, breaking the chain (hydrolysis), and this happens faster at higher temperatures (see reaction rates and catalysts);
• lipid nanoparticles can fuse or change structure over time, especially when warm.

Low temperatures slow both processes. Improved formulations and freeze-drying have since made mRNA products easier to store.

Why so fast?

Once scientists know a pathogen’s genetic sequence, designing an mRNA vaccine is mostly a matter of writing the new coding sequence; the cap, tail, modified nucleosides and lipid nanoparticles stay the same. That “plug-and-play” chemistry is why development was so quick, although safety testing in large clinical trials still took many months. The same technology is being developed for influenza, other infections and cancer treatments.

Key takeaways

• mRNA vaccines deliver instructions for one viral protein; your cells make the protein and your immune system learns from it.
• The mRNA has a 5′ cap, UTRs, an optimised coding sequence and a poly(A) tail.
• Replacing uridine with N1-methylpseudouridine reduces unwanted inflammation and boosts protein production (Karikó and Weissman).
• Lipid nanoparticles made of an ionisable lipid, a phospholipid, cholesterol and a PEG-lipid protect the RNA and use pH changes to release it inside cells.
• The mRNA is broken down within days and doesn’t enter the nucleus; cold storage slows RNA hydrolysis and particle changes. For how cells normally use mRNA, see transcription and translation.