- Key takeaways
- In vitro transcription: from DNA template to messenger RNA
- Why do clinical mRNA products use nucleoside-modified transcripts?
- The four-component LNP system: choosing and optimizing each lipid
- What mixing parameters determine LNP particle quality?
- mRNA drug product quality: the attributes that predict clinical performance
- What stability challenges define mRNA LNP drug product storage and shelf life?
mRNA vaccines and therapeutics depend on two tightly coupled manufacturing steps: in vitro transcription, which synthesizes the mRNA from a plasmid DNA template, and lipid nanoparticle formulation, which encapsulates it in the delivery vehicle. Each step generates its own critical quality attributes, and the potency, safety, and stability of the finished drug product depend on getting both right.
Key takeaways |
|
|
|
|
|
The plasmid DNA that serves as the IVT template is the upstream raw material on which the entire mRNA manufacturing process depends. For the fermentation science behind pharmaceutical-grade pDNA production, see optimizing plasmid DNA yields in microbial systems. For the strategic overview of mRNA manufacturing in the broader advanced therapy landscape, see scaling the unscalable: manufacturing cell, gene, and mRNA therapies.
In vitro transcription: from DNA template to messenger RNA
The IVT reaction synthesizes mRNA by a bacteriophage RNA polymerase, almost universally T7 RNA polymerase in pharmaceutical applications, reading a linearized double-stranded DNA template and producing the complementary single-stranded RNA transcript. The template is the plasmid DNA produced by E. coli fermentation, which has been linearized by restriction endonuclease digestion at a specific recognition site downstream of the poly-A sequence encoded in the plasmid, ensuring that the polymerase terminates at the correct position and produces a transcript of the correct length.
The reaction components in a pharmaceutical IVT reaction include the four ribonucleoside triphosphates (ATP, CTP, GTP, and the modified UTP that substitutes for native uridine), T7 RNA polymerase, the linearized pDNA template, magnesium chloride as a cofactor, spermidine to neutralize the template charge, and an RNase inhibitor to protect the transcript from nuclease degradation during synthesis. The reaction runs at 37 degrees Celsius for two to four hours, and mRNA yield from an optimized reaction is typically two to four milligrams per milliliter of reaction volume.
The 5' cap, which protects the mRNA from exonuclease degradation and is essential for ribosome recognition and translational initiation, can be added during the IVT reaction (co-transcriptional capping) or as a separate enzymatic step after IVT (post-transcriptional capping). Co-transcriptional capping using synthetic cap analogs such as CleanCap AG is increasingly preferred for pharmaceutical applications because it eliminates the additional enzymatic step and generates predominantly the Cap1 structure, which is the naturally occurring capped form in mammalian mRNA that avoids innate immune recognition by the cap-sensing receptor IFIT1. The poly-A tail, which is critical for mRNA stability and translational efficiency, is encoded directly in the linearized template and therefore transcribed as part of the IVT reaction rather than added enzymatically.
After IVT, the reaction mixture contains the target mRNA, residual pDNA template, the enzymes and NTP substrates from the reaction, pyrophosphate precipitated from the reaction, and double-stranded RNA byproducts. DNase I treatment removes the pDNA template. Sequential purification by lithium chloride precipitation or cellulose-based chromatography removes dsRNA, which is the most immunostimulatory impurity in the IVT mixture and must be controlled to specification before formulation. Final purification by anion exchange chromatography or HPLC and concentration by tangential flow filtration produce the purified mRNA drug substance ready for LNP encapsulation.
Why do clinical mRNA products use nucleoside-modified transcripts?
Unmodified IVT mRNA is recognized as foreign RNA by the innate immune system through pattern recognition receptors, including endosomal toll-like receptors TLR3, TLR7, and TLR8, and cytosolic sensors RIG-I and MDA5. Recognition of unmodified mRNA by these receptors triggers signaling cascades that elevate interferon and pro-inflammatory cytokine production, which inhibits translation of the mRNA, degrades the RNA, and generates systemic inflammatory responses that limit the therapeutic utility of the product. The first mRNA therapeutic approaches in the early 2000s were largely limited by this immunogenicity problem before the discovery by Karikó and Weissman that incorporation of modified nucleosides could render mRNA invisible to these innate immune sensors. A 2024 review of pseudouridine and N1-methylpseudouridine as nucleotide analogues for RNA therapy and vaccine development confirmed that modified nucleosides are now integral to all clinically successful mRNA therapeutics.
N1-methylpseudouridine (m1Ψ) is the modification used in the two approved mRNA COVID-19 vaccines, with complete uridine substitution with m1Ψ in both mRNA-1273 and BNT162b2 reducing immunogenicity and increasing translational efficiency relative to unmodified IVT transcripts. The mechanism is structural: m1Ψ has a methyl group at the N1 position of the uracil ring that sterically prevents binding to TLR7 and TLR8 recognition sites, while the pseudouridine rearrangement of the N1-C1' glycosidic bond also stabilizes the RNA secondary structure in a way that reduces RIG-I recognition.
The modification level, the fraction of uridines replaced with m1Ψ, is a design variable that must be optimized for each therapeutic application. Research correlating m1Ψ modification level with protein expression, immunogenicity, and stability demonstrated that full m1Ψ substitution (100% replacement of uridines) maximizes translational efficiency and minimizes innate immune stimulation for infectious disease vaccine applications, where minimizing inflammation while maximizing antigen expression is the objective. For cancer vaccines and immunotherapy applications, partial modification ratios may intentionally preserve some innate immune stimulation to support the anti-tumor immune response.
The four-component LNP system: choosing and optimizing each lipid
All clinically approved and advanced-stage mRNA LNP formulations use four lipid components whose functions are distinct and non-interchangeable. The ionizable lipid is the functional core of the system: it drives mRNA encapsulation during particle formation by becoming positively charged at the low pH of the formation buffer, electrostatically condensing the negatively charged mRNA into the forming particle, and then returning to a near-neutral charge at physiological pH to minimize toxicity and non-specific protein binding in circulation. In the acidic environment of the endosome after cellular uptake, the ionizable lipid again becomes positively charged and destabilizes the endosomal membrane, releasing the mRNA into the cytoplasm where ribosomes can translate it.
Ionizable lipid selection is the primary driver of LNP potency, tissue tropism, and tolerability. The clinical benchmark is DLin-MC3-DMA, used in Onpattro, the first approved RNA-LNP product, which established that the ionizable lipid headgroup chemistry, linker, and tail structure together determine endosomal escape efficiency. Subsequent ionizable lipid development for mRNA vaccines produced SM-102 (Moderna) and ALC-0315 (Pfizer-BioNTech), both of which achieve higher transfection efficiency at lower doses than MC3 for intramuscular delivery, reflecting the different optimization requirements for intramuscular injection versus the intravenous delivery route used for Onpattro.
Product | Ionizable lipid | Helper lipid | Cholesterol (mol%) | PEG-lipid | Molar ratio (IL:HL:Chol:PEG) | N/P |
Onpattro/patisiran (Alnylam, siRNA) | DLin-MC3-DMA | DSPC | 38.5 | PEG2000-DMG | 50:10:38.5:1.5 | ~3 |
Spikevax/mRNA-1273 (Moderna, mRNA) | SM-102 | DSPC | 38.5 | PEG2000-DMG | 50:10:38.5:1.5 | ~6 |
Comirnaty/BNT162b2 (Pfizer, mRNA) | ALC-0315 | DSPC | 42.7 | ALC-0159 | 46.3:9.4:42.7:1.6 | ~6 |
The helper phospholipid, DSPC in all three approved formulations, provides structural support to the lipid bilayer, modulates membrane fluidity, and influences the fusogenic properties of the LNP membrane. The PEG-lipid (PEG2000-DMG in Onpattro and mRNA-1273, ALC-0159 in BNT162b2) prevents LNP aggregation in suspension, reduces non-specific protein binding in biological fluids, and modulates particle size. The PEG content of 1.5 to 1.6 mol% represents a balance between adequate surface protection and sufficient exposure of the ionizable lipid surface for endosomal uptake, since too much PEG shielding can inhibit cellular uptake by covering the particle surface.
What mixing parameters determine LNP particle quality?
LNP formation occurs by rapid mixing of an ethanolic lipid solution with an aqueous mRNA solution under conditions where the sudden decrease in ethanol concentration drives spontaneous lipid self-assembly around the mRNA molecules. The quality attributes of the resulting LNPs, specifically particle size, polydispersity index (PDI), and encapsulation efficiency, are sensitive functions of the mixing dynamics at the moment of particle formation.
The flow rate ratio (FRR), the volumetric ratio of aqueous mRNA stream to organic lipid stream, is the primary parameter controlling particle size. Higher FRR results in faster ethanol dilution, which drives faster nucleation and smaller particle size. Clinical mRNA LNP formulations typically use FRRs of 3:1 (aqueous:organic), producing particles in the 80 to 150 nanometer size range. The total flow rate (TFR), which determines the mixing intensity and Reynolds number at the mixing point, affects both particle size and uniformity: higher TFR in the mixing device increases turbulent mixing and generally produces smaller, more monodisperse particles.
The N/P ratio is the formulation parameter most directly linked to encapsulation efficiency and potency. A higher N/P ratio increases the molar excess of ionizable lipid nitrogen atoms over mRNA phosphate groups, driving more complete mRNA condensation into the particle interior. The N/P ratios of approximately 6 used in approved mRNA vaccines reflect optimization of encapsulation efficiency (target greater than 85-90%) against tolerability, since excess ionizable lipid contributes to the local injection site reactogenicity that limits achievable dose levels.
For the engineering design of the microfluidic, impingement jet, and parallelized mixing hardware that implement these parameters at manufacturing scale, including the scale-up challenges of maintaining consistent particle quality from clinical to commercial volumes, see the dedicated article at Separation Science on single-use mixers in formulating lipid nanoparticles. The fluid path engineering and single-use mixing environment are covered there; the formulation science parameters and their drug product consequences are the focus here.
mRNA drug product quality: the attributes that predict clinical performance
mRNA drug product quality is characterized through a set of analytical attributes that together describe the integrity of the mRNA payload, the physical quality of the LNP, and the safety profile of the formulated product. A 2024 protocol for mRNA LNP vaccine development and immunization efficiency characterization confirmed the core analytical workflow for mRNA LNP quality control, including RNA purity assessment, particle sizing, and encapsulation efficiency measurement, as the standard characterization set for mRNA LNP programs entering clinical development.
mRNA integrity, expressed as the percentage of full-length intact RNA molecules in the drug substance, is the primary stability-indicating quality attribute. Intact mRNA produces full-length protein of the correct sequence; fragmented mRNA produces truncated protein that may not fold correctly, may not elicit the intended immune response in vaccine applications, or may not perform the intended therapeutic function. mRNA integrity is measured by automated electrophoresis (Bioanalyzer or Fragment Analyzer) and is the most sensitive indicator of thermal degradation, nuclease activity, or chemical hydrolysis during manufacturing, storage, or in-use.
The dsRNA content of the purified mRNA drug substance is a safety-relevant quality attribute that reflects the completeness of the IVT purification process. dsRNA molecules formed as IVT byproducts are recognized as viral RNA mimics by innate immune sensors, and their presence in the drug substance at levels above specification contributes to reactogenicity, interferon induction, and reduced translational efficiency of the mRNA payload. Cellulose-based purification methods that selectively bind dsRNA can remove the majority of the dsRNA impurity from the IVT product, and the residual dsRNA content is tested by double-strand-specific antibody ELISA or dot blot assay.
Encapsulation efficiency, the fraction of total mRNA molecules that are protected within the LNP interior rather than remaining in the external aqueous phase, is measured by comparing mRNA fluorescence signal before and after detergent treatment to disrupt the LNP membrane. Encapsulation efficiency targets of greater than 85-90% are standard for clinical mRNA LNP products. mRNA present on the outer surface of LNPs or free in solution is not protected from nuclease degradation in biological fluids and does not contribute to the therapeutic dose.
What stability challenges define mRNA LNP drug product storage and shelf life?
mRNA is intrinsically susceptible to two degradation pathways that distinguish its stability profile from conventional biologic drug products: enzymatic hydrolysis by ribonucleases, which are ubiquitous environmental contaminants and must be excluded from every step of the manufacturing process, and non-enzymatic base-catalyzed hydrolysis of the 2'-hydroxyl group of ribose at every nucleotide in the backbone. The second mechanism is thermally activated, with degradation rate increasing approximately 10-fold for every 10 degrees Celsius rise in temperature, which is the fundamental reason why current mRNA LNP formulations require storage at minus 60 to minus 80 degrees Celsius. Model-informed drug development analysis of mRNA-LNP therapeutics published in 2025 identifies the stability and degradation kinetics of mRNA in LNP formulations as one of the key pharmacokinetic parameters requiring characterization in regulatory submissions for this therapeutic class.
The LNP encapsulation itself provides partial protection against mRNA degradation by excluding external ribonucleases from the RNA interior. However, the lipid components themselves undergo hydrolysis and oxidation during storage, and degraded lipid components can in turn damage the mRNA they are protecting. Formulation optimization for mRNA LNP stability therefore includes both the chemical stability of the lipid components and the physical stability of the LNP particle against fusion, aggregation, and size increase during storage.
The cold chain requirement for minus 60 to minus 80 degrees Celsius storage, demonstrated during the COVID-19 vaccine rollout, creates substantial logistical challenges and access limitations for mRNA therapeutics. Next-generation formulation approaches targeting refrigerator-stable (2 to 8 degrees Celsius) mRNA LNP products, through the combination of lyophilization, modified lipid compositions, cryoprotectants, and mRNA chemical modifications that increase backbone stability, are an active area of pharmaceutical development with significant regulatory and commercial implications.
This article was produced under Drug Discovery News' AI Editorial Guidelines.
