Understanding mRNA Vaccines: Structure, Delivery, and Immune Response

mRNA vaccines work by delivering a small set of genetic instructions into your cells, which then produce a protein that trains your immune system to recognize a specific pathogen. The concept sounds simple, but the engineering behind each component of these vaccines is remarkably precise. Every piece of the mRNA molecule, from its chemically modified building blocks to the fatty shell that carries it into cells, has been designed to solve a specific biological problem. How all of these parts work together to generate a strong, lasting immune response is worth understanding in detail.

What the mRNA Molecule Actually Looks Like

The mRNA inside a vaccine is not identical to the mRNA your cells naturally produce. It has been chemically modified in several ways to make it more stable, more translatable, and less likely to trigger the wrong kind of immune alarm. The most important modification involves swapping out one of the standard RNA building blocks, uridine, for a synthetic alternative called N1-methylpseudouridine (m1Ψ). This single substitution is a cornerstone of both the Pfizer-BioNTech and Moderna COVID-19 vaccines, and it dramatically increases the amount of protein the mRNA can produce once it gets inside a cell.¹ The modification also helps the mRNA slip past certain innate immune sensors that would otherwise recognize foreign RNA and destroy it before it could do its job.²

Beyond this chemical swap, the regions flanking the protein-coding sequence matter enormously. The 5′ untranslated region (the stretch of RNA before the instructions begin) affects how efficiently a ribosome can latch on and start reading. Its length and internal folding patterns influence how much protein ultimately gets made. Researchers have found that even the position and stability of small hairpin-like folds near the cap structure at the start of the mRNA can have a large effect on translation efficiency.³ At the opposite end, the poly(A) tail, a string of adenine bases appended to the mRNA’s tail, protects the molecule from degradation and keeps it functional longer. Recent work has shown that adding a loop structure within this poly(A) tail can further boost both stability and protein output compared to a standard straight tail.⁴

The Lipid Nanoparticle Shell

Naked mRNA injected into the body would be chewed up by enzymes within minutes and could never cross a cell membrane on its own. The solution is to wrap the mRNA in a tiny sphere of lipids, roughly 80 to 100 nanometers across, called a lipid nanoparticle (LNP). These particles are not just passive containers. Each of the four main lipid components serves a distinct function.

The most critical ingredient is the ionizable lipid. At the neutral pH of your bloodstream, it carries no electrical charge, which helps the particle avoid rapid clearance. But once a cell swallows the LNP into an acidic compartment called an endosome, the ionizable lipid picks up a positive charge. That charge shift lets it interact with negatively charged lipids in the endosomal membrane, destabilizing the compartment and allowing the mRNA cargo to spill into the cell’s interior where ribosomes can read it.⁵ Exactly how this destabilization works at the molecular level is still debated. One hypothesis holds that the ionizable lipid and endosomal lipids form cone-shaped complexes that punch through the membrane, though molecular simulations suggest these complexes are transient rather than stable, appearing only during a brief phase transition that accelerates escape.⁶

The remaining lipid components handle structural and logistical tasks. Helper lipids like phosphatidylcholine stabilize the particle’s bilayer structure so it holds up in the bloodstream. Cholesterol improves both structural integrity and intracellular delivery. A PEGylated lipid, coated with polyethylene glycol chains, prevents particles from clumping together in the vial and extends their circulation time, though too much PEG can actually slow cellular uptake and hinder endosomal escape.⁷

From Injection Site to Protein Factory

After an intramuscular injection, the LNPs do not stay put in your deltoid. Biodistribution studies show that the mRNA concentrations peak quickly, within the first eight hours or so, and then decline. The highest mRNA exposures occur at the injection site itself, the draining lymph nodes, and the spleen. Smaller amounts reach the liver and circulate briefly in the blood. Levels at the injection site and lymph nodes drop below detection within about 24 hours, while trace amounts in the spleen and blood can be detected for up to a week before falling to negligible levels.⁸ Broader studies in rats confirm this pattern: the injection site, draining lymph nodes, and spleen see the highest concentrations, with minimal distribution to other organs and overall clearance projected within roughly two weeks.⁹

Once inside a cell, the mRNA is read by the cell’s ribosomes, which assemble the encoded protein, in the case of COVID-19 vaccines, the spike protein of SARS-CoV-2.¹⁰ That protein then gets displayed on the cell’s surface, alerting the immune system to something foreign. The mRNA itself is degraded by normal cellular processes and does not integrate into your DNA.

How the Innate Immune System Responds

Before the adaptive arm of the immune system (antibodies, T cells) even gets involved, the innate immune system reacts to both the mRNA and the LNP shell. This initial reaction is critical because it shapes the strength and quality of the longer-term immune response. The LNP itself acts as an adjuvant, a substance that amplifies the immune response to whatever antigen accompanies it. Research has made clear that the lipid carrier is not immunologically inert; it actively drives inflammation that helps the vaccine work.¹¹

This is where the m1Ψ modification plays a second role. Unmodified RNA would strongly activate innate sensors like TLR3, TLR7, and TLR8, toll-like receptors that evolved to detect viral RNA. Incorporating m1Ψ largely suppresses this detection. The modified RNA resists breakdown into the fragments these receptors recognize, and it is poorly bound by TLR7 and TLR8 binding pockets. In the cytoplasm, the modification also reduces signaling through another sensor, RIG-I.¹² The result is a carefully calibrated innate response: enough inflammation to recruit immune cells and jumpstart adaptive immunity, but not so much that the mRNA is destroyed before it can produce adequate protein.

That said, some innate sensing does still occur, and it appears to work through different pathways than initially expected. Rather than TLR7, the primary cytoplasmic sensor responding to mRNA vaccines may be MDA5, which triggers production of type I interferons and inflammatory cytokines like IL-1β and IL-6. These signals serve double duty: they improve vaccine effectiveness but also contribute to side effects.¹³

Why You Feel Lousy After the Shot

The sore arm, fatigue, fever, and body aches many people experience after an mRNA vaccine are direct consequences of the innate inflammatory cascade. Studies in mice have pinpointed the key players. IL-6 is a major driver of fever; blocking it with a neutralizing antibody significantly reduced febrile responses. TNF-α, type I interferon, and IL-1α contributed to weight loss and systemic malaise, though blocking any single one of these alone only partially reduced symptoms.¹⁴ Other research has identified HMGB1, a damage-associated molecular signal, as an early trigger. One reactogenic LNP formulation (using the ionizable lipid SM-102, the same one in the Moderna vaccine) caused a more than seven-fold spike in HMGB1 within two hours, which in turn drove TNF-α secretion from immune cells. A different, less reactogenic lipid produced far less HMGB1. Blocking the IL-1 receptor had the strongest effect on reducing fever in that study.¹⁵

The practical takeaway is that the specific ionizable lipid used in the LNP formulation has a direct and measurable influence on how intense side effects are. This is one reason next-generation vaccine development is focused so heavily on finding lipid formulations that maintain strong immune responses while dialing down reactogenicity.

Building Adaptive Immunity

The real prize of vaccination is the adaptive response: targeted antibodies that neutralize the pathogen and memory cells that stand ready for future encounters. mRNA vaccines have proven particularly strong at activating a type of immune cell called a T follicular helper cell (Tfh), which is essential for setting up germinal centers in the lymph nodes. Germinal centers are where B cells undergo a Darwinian selection process, competing to produce ever-more-precise antibodies. In both mice and primates, nucleoside-modified mRNA-LNP vaccines produced strong Tfh responses, high numbers of germinal center B cells and plasma cells, and durable high-affinity neutralizing antibodies.¹⁶

An interesting wrinkle in how mRNA vaccines activate helper T cells (CD4+ T cells) has recently emerged. Traditionally, these cells respond to protein fragments that immune cells scavenge from outside the cell and display on their surface. But with mRNA-LNP vaccines, the most efficient CD4+ T cell activation appears to come from antigen that is produced directly inside the immune cell that took up the LNP, without the protein ever leaving the cell first.¹⁷ This means the vaccine leverages a somewhat non-standard pathway to prime helper T cells, which may partly explain why mRNA vaccines generate such robust T cell responses.

The adaptive response extends beyond just making antibodies. Functional memory B cells, the long-lived cells that can rapidly ramp up antibody production upon re-exposure, actually increased from three to six months after vaccination with COVID-19 mRNA vaccines. The majority of these memory cells could cross-recognize several SARS-CoV-2 variants, including Alpha, Beta, and Delta.¹⁸

How Long Does Immunity Last

This is probably the question most people care about, and the answer is nuanced. On the antibody side, levels do decline over time, with an estimated half-life of roughly five months.¹⁹ Whether the bone marrow establishes long-lived plasma cells, the factories that continuously secrete antibodies for years or even decades, has been an area of active investigation. Early data raised concern that mRNA vaccines might not produce these cells in people who had never been infected with SARS-CoV-2.²⁰ More recent findings, however, have detected SARS-CoV-2-specific plasma cells in human bone marrow after a basic mRNA vaccination series, including cells with phenotypic features associated with long-lived memory.²¹ The picture is still developing, but the overall evidence points toward durable humoral memory even as circulating antibody levels wane.

For people with weakened immune systems, the story is different. Solid organ transplant recipients vaccinated with an mRNA COVID-19 vaccine did mount CD4+ T cell responses with the right profile (biased toward a Th1 pattern, which is better for fighting viruses), but the magnitude and rate of response were lower than in healthy individuals. CD8+ T cell responses were also weaker.²² This underscores why immunocompromised people often need additional vaccine doses and remain at higher risk even after vaccination.

mRNA Vaccines Compared to Other Platforms

A natural question is how mRNA vaccines stack up against alternatives like viral vector vaccines (which use a harmless virus to deliver genetic instructions) or protein subunit vaccines (which deliver the antigen protein directly). Comparative studies with COVID-19 vaccines provide some answers. mRNA vaccines produced higher binding and neutralizing antibody responses than the Sputnik V viral vector vaccine, and this advantage held through at least six months.¹⁹ Interestingly, viral vector vaccines can produce a strong CD8+ T cell and antibody response after a single dose, but a second dose of the same vector often fails to boost the response much further, likely because the immune system starts targeting the vector itself.²³ mRNA vaccines do not have this problem, since there is no viral carrier to generate anti-vector immunity, making them well suited to multi-dose regimens and updated boosters.

The Anti-PEG Antibody Question

PEG (polyethylene glycol) is found in many consumer products and medications, and some people have developed antibodies against it. Since LNPs contain a PEGylated lipid, researchers have investigated whether pre-existing anti-PEG antibodies could undermine vaccine effectiveness. In animal models, mice with high levels of anti-PEG antibodies showed significantly reduced spike protein production after vaccination, on the order of 30- to 50-fold lower than naive mice, along with lower anti-spike antibody responses and signs of complement activation.²⁴

However, the story may differ depending on the route of administration and the type of anti-PEG antibody involved. A separate study focused on anti-PEG IgM (a different antibody class) found that these antibodies had only a minor effect on mRNA translation at the intramuscular injection site. While they did redirect LNPs in the liver to the wrong cell type (immune scavenger cells instead of liver cells), repeated intramuscular injections still elicited a robust immune response.²⁵ The relevance to human vaccination likely depends on how high someone’s anti-PEG antibody levels are and which class of antibody predominates. For most people, this is not a practical concern, but it is an active area of research, especially as LNP-based therapies expand beyond vaccines.

Targeting Organs Beyond the Liver

Standard LNPs have a strong tendency to accumulate in the liver after intravenous administration. This is useful for liver-targeted therapies but a significant hurdle for treating diseases elsewhere in the body.²⁶ A strategy called SORT (selective organ targeting) has shown that adding a specific supplemental lipid molecule to the LNP formulation can redirect delivery to the lungs, spleen, or liver with high selectivity. In animal studies, SORT nanoparticles were engineered to edit specific cell types in each organ, including lung epithelial cells, splenic B cells and T cells, and liver hepatocytes.²⁷ While this work is not yet in clinical use for vaccines, it points toward a future where mRNA therapies can be directed to specific tissues for applications ranging from gene editing to cancer immunotherapy.

Self-Amplifying RNA

One limitation of current mRNA vaccines is that each molecule of mRNA can only be read a limited number of times before it degrades. Self-amplifying RNA (saRNA) is designed to overcome this. Based on alphavirus genomes, saRNA retains the molecular machinery the virus uses to copy its own RNA, but replaces the virus’s structural genes with the antigen of interest. Once inside a cell, the saRNA replicates itself, generating many more copies of the instructions and producing significantly more antigen from a much smaller initial dose.²⁸ The first approved saRNA vaccine, ARCT-154 for COVID-19, was authorized in Japan in late 2023. The trade-off is that saRNA molecules are larger than standard mRNA, making them more challenging to formulate and potentially more reactogenic. But the dose-sparing advantage could lower manufacturing costs and make vaccines more accessible, particularly in resource-limited settings.

Manufacturing Challenges

Producing mRNA vaccines at scale introduces its own set of problems. The mRNA is synthesized through an enzymatic reaction called in vitro transcription, and the process is sensitive to conditions. Even at optimized operating points, a substantial fraction of the RNA produced can be truncated, meaning incomplete and non-functional. One process optimization study reported that at its best configuration, about 55% of the mRNA was truncated, though this represented a 33% reduction from the starting conditions alongside a 55% improvement in overall yield.²⁹ Purification to remove these truncated fragments, along with residual enzymes and DNA template, adds cost and complexity. The cold-chain requirements for shipping and storing the final product, particularly the ultra-cold temperatures initially required for the Pfizer vaccine, have been another logistical bottleneck, though reformulation efforts have progressively relaxed storage conditions.

A separate challenge is the endosomal escape efficiency of LNPs, which remains surprisingly low. A recent re-evaluation of existing data argues that LNPs may escape endosomes through a mechanism called vesicle budding-and-collapse, and that even after escape, the mRNA forms an insoluble aggregate with lipids in the cytoplasm. The slow dissolution of this aggregate could be a rate-limiting step that constrains how much mRNA actually reaches ribosomes.³⁰ Improving this bottleneck is one of the most active frontiers in LNP engineering, since even modest gains in cytoplasmic delivery could allow lower doses, fewer side effects, and stronger immune responses.