The stability problem is not just storage — it's encapsulation and release
mRNA stability is often discussed in terms of storage conditions: what temperature, what buffer, what excipient protects the RNA from degradation until the dose is administered. This is a real and important question. But for LNP-encapsulated mRNA, the stability problem has three distinct phases — each governed by different physical chemistry — and formulation decisions affect all three:
- Pre-encapsulation stability: mRNA stability in the aqueous buffer used for ethanol-injection synthesis, prior to LNP formation.
- Encapsulated stability: mRNA integrity inside the formed LNP particle, during storage or in circulation.
- Post-endosomal stability: mRNA integrity after endosomal escape and release into the cytoplasm, where translation must occur before ribosomal degradation of the message.
Each phase has distinct formulation dependencies. A formulation optimized for encapsulation efficiency and transfection may not maximize stored mRNA integrity. A formulation that protects mRNA from nucleases in serum may not maximize cytoplasmic translation efficiency. Understanding these distinctions guides formulation decisions toward the outcome that actually matters — functional protein expression in the target cell — rather than a single proxy metric.
Why mRNA is inherently unstable
The chemical basis of mRNA instability is primarily hydrolytic. The 2'-hydroxyl group on the ribose sugar of each RNA nucleotide is a chemically active site that undergoes 2',3'-cyclic phosphate formation via intramolecular transesterification under basic conditions, and direct hydrolysis under acidic conditions. The net result in either direction is RNA backbone cleavage. This is the fundamental reason RNA is approximately 100,000-fold less stable than double-stranded DNA under aqueous conditions: DNA lacks the 2'-hydroxyl group and cannot undergo this reaction.
In addition to the intrinsic chemical instability of the ribose 2'-OH, mRNA is a substrate for cellular ribonucleases (RNases) — endonucleases and exonucleases with diverse sequence and structure specificities. RNases are present in serum (RNase A superfamily) and within the endosome (lysosomal RNases activated at pH 4.5–5.0). Both enzymatic and non-enzymatic hydrolysis contribute to mRNA degradation in a delivery context.
The standard chemistries that improve mRNA stability are: (1) N1-methylpseudouridine (m1Ψ) substitution of uridines, which reduces TLR7/8 recognition and improves in vivo mRNA half-life by reducing innate immune-mediated clearance; (2) optimized 5' and 3' UTR sequences derived from highly expressed, stable endogenous mRNAs (beta-globin, alpha-globin UTRs are commonly used); (3) poly(A) tail length optimization (longer poly(A) tails, 100–150 nt, slow exonucleolytic degradation from the 3' end); and (4) capping with a cap 1 structure (5'-m7GpppNm-) that is recognized by the translational apparatus and resists decapping enzymes more effectively than cap 0.
These modifications address the mRNA molecule itself. The question for LNP formulation is: how does the particle's internal environment and surface chemistry affect each stability phase?
Internal LNP environment and encapsulated RNA stability
Encapsulated mRNA is not floating free in water inside the LNP core. During LNP formation via microfluidic ethanol injection, ionizable lipid protonated at synthesis pH (typically pH 4.0–4.5 for efficient encapsulation) forms electrostatic interactions with the phosphate backbone of the RNA. The RNA is condensed within the lipid bilayer inverted micelle structure, with the RNA strands threading through an aqueous core bounded by a lipid shell.
The ionizable lipid pKa determines how tightly the RNA is complexed with lipid and what the local environment of the RNA interior is. At synthesis pH (below pKa), ionizable lipids are positively charged and bind RNA tightly, achieving high encapsulation efficiency. In storage at pH 7.4 (above pKa), ionizable lipids are substantially neutral and the electrostatic complexation relaxes — the RNA is less tightly condensed.
Internal pH and RNA stability are linked. Residual proton activity in the LNP interior during storage at neutral pH can drive slow hydrolytic degradation of the encapsulated RNA, even in the absence of nucleases. Formulations that maintain a more neutral internal environment (achieved partly by ionizable lipid pKa selection) reduce this hydrolytic stress on the encapsulated mRNA during storage.
Helper lipid selection also matters for encapsulated RNA integrity. The helper lipid composition determines bilayer fluidity and the "leakiness" of the lipid shell. DSPC (a saturated, high-Tm lipid, Tm ~55°C) provides a more ordered, less permeable bilayer than DOPE or DPPC. For programs where long-term storage stability at elevated temperatures is critical, helper lipid combinations that maximize bilayer order at storage temperature improve RNA retention within the particle and reduce exposure of encapsulated RNA to external nucleases from diffusion across the lipid shell.
Cholesterol fraction modulates membrane fluidity independently of the helper lipid. Cholesterol intercalates between phospholipid tails, reducing lateral diffusion and tightening the bilayer at physiological temperatures. Cholesterol-rich LNPs (40 mol% versus 30 mol%) show improved stability against aggregation and maintained particle size during storage, both of which are correlated with maintained RNA integrity (particle aggregation often coincides with membrane disruption and RNA leakage).
The ionizable lipid pKa and endosomal release kinetics
The same pKa that governs encapsulation efficiency (discussed in our earlier article on ionizable lipid selection) also determines endosomal release kinetics, which affects the stability window during which encapsulated mRNA must survive to be translated.
In the endosomal lumen, pH drops progressively from approximately 6.2 (early endosome) to 5.5 (late endosome) to 4.5 (lysosome). Ionizable lipids with pKa 6.2–6.8 become protonated and positively charged in this pH range, enabling interaction with the negatively charged endosomal membrane phospholipids. This interaction drives the formation of the hexagonal HII phase (highly curved, cone-shaped lipid geometry) that destabilizes the endosomal membrane and facilitates mRNA escape into the cytoplasm.
The timing of endosomal escape matters for mRNA integrity. If mRNA escapes early endosomes before lysosomal acidification (pH 4.5), it has been exposed to relatively mild nuclease conditions (early endosomal RNases are less active than lysosomal RNases). If mRNA is retained in late endosomes approaching the lysosome, it faces the full complement of lysosomal RNases (RNase II, RNase T1, RNase A activity) at pH 4.5–5.0, conditions that are highly destructive to unprotected RNA.
The relationship between pKa and endosomal escape kinetics is therefore a stability consideration as much as a transfection efficiency consideration. Higher-pKa ionizable lipids (6.6–6.8) become protonated earlier in the endosomal maturation pathway — in early endosomes at pH 6.2–6.5 — potentially enabling escape before full lysosomal acidification. Lower-pKa ionizable lipids (<6.2) may require late endosomal acidification to achieve the protonation needed for membrane destabilization, exposing the mRNA to greater lysosomal nuclease activity before escape.
In practice, the optimal pKa for RNA stability and transfection efficiency is in the same range (6.2–6.8) — which suggests that pKa optimization for encapsulation efficiency, transfection efficiency, and RNA protection during endosomal transit are largely aligned objectives rather than competing tradeoffs.
PEG-lipid and protein corona effects on serum stability
In circulation after IV injection, LNPs are exposed to serum nucleases that can degrade RNA if it escapes the particle or the particle's bilayer barrier is compromised. PEG-lipid serves two functions relevant to RNA stability: (1) steric shielding that reduces opsonization and reduces macrophage/Kupffer cell uptake before reaching hepatocytes; and (2) protein repulsion that limits non-specific adsorption of serum nucleases to the particle surface.
LNPs with inadequate PEG-lipid density adsorb a complex protein corona including serum albumin, fibronectin, immunoglobulins, complement proteins, and nucleases. Nuclease adsorption in the protein corona can act as a sustained nuclease source in contact with the particle surface — increasing the degradation rate of any RNA that leaks from the particle or is accessible at the surface.
However, PEG density has the competing effect of limiting ApoE adsorption (as discussed in detail in our article on hepatic targeting). The formulation parameter space for optimizing serum stability, ApoE-mediated hepatic targeting, and endosomal escape simultaneously is constrained: increasing PEG density improves serum stability and reduces nuclease adsorption but reduces ApoE adsorption and hepatic uptake. This tradeoff is one of the reasons LNP optimization is a multi-variable problem that benefits from systematic computational exploration rather than single-parameter tuning.
Assessing RNA integrity in formulation development
Several analytical methods are used to assess RNA integrity during formulation development. Their selection determines which aspects of stability are actually being measured.
Ribogreen assay (EE measurement). Ribogreen is a dye that intercalates into RNA and fluoresces. The standard EE assay measures the fluorescence signal from free RNA (accessible to Ribogreen) versus total RNA (after detergent-mediated LNP disruption). This assay reports on encapsulation but not on RNA integrity — degraded RNA fragments that remain inside the LNP still register as encapsulated, so Ribogreen EE does not detect RNA fragmentation within the particle.
Gel electrophoresis / Bioanalyzer RIN. RNA Integrity Number (RIN) from capillary electrophoresis (Agilent Bioanalyzer or similar) measures the ratio of intact RNA to degraded fragments. RIN scores range 1–10, with >8 considered intact for mRNA. This assay requires RNA extraction from the LNP (using proteinase K digestion followed by standard RNA isolation), which adds potential degradation artifacts. RIN-based integrity measurement is the standard for assessing stored LNP RNA integrity over time.
Functional translation assay. A luciferase reporter mRNA can be formulated in parallel with the therapeutic mRNA of interest and used as a functional stability indicator. Luciferase signal from cell transfection is proportional to translatable mRNA delivered — providing an integrated readout of mRNA integrity, LNP uptake, endosomal escape, and translation efficiency together. This is the most biologically relevant assay but confounds multiple variables.
For formulation optimization programs where the goal is to identify which formulation parameters specifically improve stored RNA integrity versus endosomal release efficiency, combining Bioanalyzer RIN (for stored integrity) with functional luciferase assay (for total delivery) and a separate endosomal escape probe (such as a galectin-8 puncta assay for endosomal membrane damage) provides the mechanistic resolution needed to distinguish which step is rate-limiting for a given formulation candidate.
Practical formulation guidance for mRNA stability
- For storage stability at 2–8°C, prioritize DSPC over DOPE as helper lipid — DSPC's higher phase transition temperature maintains bilayer order and reduces RNA leakage during refrigerated storage.
- Optimize ionizable lipid pKa in the 6.2–6.8 range. Avoid pKa <6.0 (poor encapsulation at acidic synthesis pH, insufficient protonation for early endosomal escape) and pKa >7.0 (protonation at physiological blood pH generates positive surface charge, driving complement activation and nonspecific serum protein adsorption).
- For programs requiring freeze-thaw stability, sucrose cryoprotectant (10% w/v) in the final formulation buffer reduces ice-crystal-mediated particle disruption and RNA leakage during frozen storage cycles.
- Assess RNA integrity at the formulation level (Bioanalyzer after RNA extraction) separately from encapsulation efficiency (Ribogreen). These measure different things and should not be conflated.
- Include mRNA integrity as a stability-indicating method in your IND Module 3 CMC specification, with acceptance criteria for RIN and percent full-length mRNA by capillary electrophoresis at specified timepoints and storage conditions.