Physical stability of RNA–lipid nanoparticles under agitation, freeze–thaw and accelerated stability testing

Abstract

BACKGROUND: The choice of excipient is important in enhancing stability of liquid formulations containing lipid nanoparticles (LNPs) used in RNA delivery where the ability to survive unplanned excursions to higher and lower temperature and vibration is an important attribute. In this study, established parenteral excipients were tested for their capacity to protect LNP structures in solution.

RESULTS: LNPs are very susceptible to damage during freeze–thaw cycles. The most promising excipients to reduce ice crystal damage were sucrose and glycerol. Poloxamer 188 (PX188), polysorbate 80 (PS80), bovine serum albumin (BSA) and trehalose did not perform as well as sucrose and glycerol. Accelerated stability testing (35 °C) observed an increase in hydrodynamic diameter of LNPs. BSA and PS80 were deleterious to LNP stability at 35 °C with a large rise in the polydispersity index. Trehalose, sucrose and glycerol performed better at 35 °C than the other treatment groups. Mild agitation had little effect on LNP structures.

CONCLUSIONS: The excipients of choice for liquid formulation based on freeze–thaw and accelerated testing are sucrose and glycerol. Sucrose is known to be effective in formulations of liposomes and LNPs. Glycerol, however, is rarely used for liposomes and LNP liquid formulations. PS80 and BSA were deleterious to LNP stability during freeze–thaw and accelerated testing. PX188 conferred some protection during freezing, and since its structure and mode of action are radically different from those of sucrose and glycerol it could prove to be synergistic.

© 2026 The Author(s). Journal of Chemical Technology and Biotechnology published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry (SCI).

Introduction

Lipid nanoparticles (LNPs) have become an important delivery vehicle for the current generation of RNA therapeutics and vaccines. The encapsulation of the RNA within LNPs provides protection from enzymatic degradation and promotes cellular uptake and endosomal release.1 Commercial LNP composition comprises ionizable lipids, phospholipids, cholesterol and polyethylene glycol (PEG) lipids, and these components and their ratios influence the LNPs’ size, shape and surface, directly affecting the vesicle formation and, therefore, their effectiveness.1 Ideally, RNA–LNP formulations used in RNA therapeutics and vaccines would be stored as a liquid at 5 ± 3 °C for over 18 months. This would be ready-to-use, not requiring reconstitution before use and be stored under standard refrigerated conditions. Ideally, the formulation would be robust enough to survive excursion to higher temperatures, freeze–thaw cycles and vibration during transportation and storage. This may seem ambitious, but it has been achieved for similarly challenging products.

Liposomes are a soft malleable lipid-based delivery system used for drug delivery and are related to LNPs. The first liposomal-delivered drug was Doxil, which was approved for use in 1995. It was rapidly followed by Abelcet, DaunoXome, DepoCyt and AmBisome.2-6 The first four products were liquid formulations that were stored at 2–8 °C for between 1 and 2 years. Light exposure had to be avoided and they did not tolerate freezing. Egress of the drug and liposome aggregation were issues that had to be resolved. The compositions of the cosolutes in the aqueous phase were simple. Abelcet and DepoCyt liquid formulations had only sodium chloride added to control the osmolarity.2, 5 Doxil and DaunoXome were more complex; they had sucrose as the tonicity modifier but also contained ammonium sulfate and histidine buffer, and glycine and calcium chloride, respectively.4, 6 AmBisome was lyophilized and required storage at −25 °C but was stable for up to 4 years.3

The other family of products that enhanced our understanding of the relationship between excipients and product stability were the therapeutic proteins, especially monoclonal antibodies. Currently around 70% of therapeutic monoclonal antibodies are in a liquid form, physically and chemically stabilized by the appropriate use of excipients.7 There are similarities and differences between RNA–LNPs and monoclonal antibodies. Both can be susceptible to oxidation and hydrolysis reactions. The structural stability is different. Monoclonal antibodies are susceptible to unfolding and subsequent aggregation while LNPs may aggregate, fragment, leach material into the aqueous phase and structurally rearrange over time. LNPs are also soft and can distort when force is applied to them. A putative list of causes of RNA–LNP instability in aqueous formulations is presented in Table S1.

The learning from the development of therapeutic proteins is that pH plays an important role in determining the rate of hydrolysis reactions.8 That pH also determines the charge of a particle and its resulting attraction or repulsion. Charged excipients can pair with charges on the particle and reduce repulsion and alter structural stability.9-11 Charged excipients also effect particle solubility by altering the surface tension at the solution–particle interface which could also be described in terms of competition for water between the particle and cosolute hydration layers.9, 12, 13 A single cosolute can also participate in more than one interaction between a cosolute and a particle, as observed with salts and the amino acid arginine.9, 14

LNP products have a mixed track record. The first FDA-approved LNP-based siRNA therapeutic was Onpattro® (patisiran), used for treatment of hereditary transthyretin-mediated amyloidosis.15 Onpattro® is stored as a liquid solution comprising phosphate buffer and sodium chloride, at 2–8 °C and can tolerate up to 25 °C for 14 days but does not tolerate freezing.16 The mRNA–LNP COVID-19 vaccines Comirnaty® (Pfizer–BioNTech) and Spikevax® (Moderna) had more stringent storage conditions. Comirnaty® was stored between −60 and −90 °C and Spikevax® between −15 and −25 °C.17 They were only stable for a few hours after thawing.

The establishment of therapeutic product stability is well established. There are two standard temperatures used for storage of pharmaceutics that require low temperatures: 5 ± 3 and −20 ± 5 °C. The regulatory authorities require long-term stability testing at the recommended storage temperature to be undertaken to establish self-life.18 This is often supplemented with accelerated stability, agitation and freeze–thaw studies to simulate conditions during transport or storage where the recommended storage conditions are not maintained. Accelerated stability studies are short-term studies at elevated temperatures.18-20 These mimic both short-term loss of refrigeration that could occur and accelerated reactions like hydrolysis reactions, which behave following the Arrhenius equation. Agitation studies introduce active mixing to enhance interaction with air–water or solid–water interfaces through to scaled-down systems that mimic vibration-induced shear as well as the enhanced interaction with interfaces.21, 22 Freeze–thaw studies mimic unplanned product freezing, which can occur in refrigerated storage.

Experience with liquid and lyophilized monoclonal antibody formulations has provided a palette of acceptable excipients that can be used in other parenteral products.7 This can provide a framework for LNP formulation. Commercial lyophilized products that undergo a freezing step usually comprise buffer such as histidine, phosphate or succinate, a surfactant (polysorbate 20 or 80), stabilizers such as an amino acid and bulking agent/tonicity modifiers such as sucrose, trehalose, mannitol, sorbitol, glycine or dextran. Experience with lyophilized liposome drug delivery vehicles is also useful. There is a palette of cryopreservatives: alcohols such as glycerol, ethylene glycol, propylene glycol; dimethyl sulfoxide; polymers such as polyvinylpyrrolidones (PVP) and PEG; sugars such as sucrose, lactose, trehalose; and sugar alcohols such as mannitol.23

In the work reported in this paper, we tested RNA–LNPs for resilience using accelerated stability, agitation and freeze–thaw studies. tRNA was used due to its similarity in size to siRNA and it being inexpensive. In the agitation study, we tested the application of polysorbates (20 and 80), which are widely used excipients in parenteral formulations, to reduce air–water interface-induced structural destabilization.24 Bovine serum albumin (BSA) was also tested due to its prior use to reduce detrimental interactions between drug active ingredients and solid surfaces.25 In the freeze–thaw and accelerated stability studies, the range of excipients was broadened to include some with known stabilizing and cryopreservative properties including glycerol, sucrose and trehalose. A block copolymer with mild surfactant-like properties, poloxamer 188 (PX188), was also tested. Dynamic light scattering (DLS) was the primary analytical technique used to assess stability. It is able to generate a calculated hydrodynamic diameter and polydispersity index (PDI). Changes in the calculated hydrodynamic diameter and rises in the PDI are indicative of LNP structural change and loss of homogeneity, respectively. It is worth noting that, while hydrodynamic diameter and PDI are useful measures of structural stability, they do not necessarily detect changes in morphology or leaching of RNA into solution; and that physical stability alone is not a guarantee of transfection efficiency or behavior in vivo.

Download the full article as PDF here Physical stability of RNA–lipid nanoparticles under agitation, freeze–thaw and accelerated stability testing

or continue reading here

Materials

Heptadecan-9-yl 8-((2-hydroxyethyl)(6-oxo-6-(undecyloxy)hexyl) amino) octanoate (SM-102) was purchased from Sinopeg (Xiamen, China). 1,2-Distearoyl-snglycero-3-phosphocholine (DSPC), 1,2-dimyristoylracglycero-3-methoxypolyethylene glycol2000 (PEG-DMG), cholesterol, tRNA, BSA, polysorbate 20 (PS20), polysorbate 80 (PS80), glycerol, sucrose, trehalose dihydrate, sodium acetate, glacial acetic acid, sodium chloride and ethanol (EtOH) were sourced from Sigma-Aldrich (Bayswater, VIC, Australia). Potassium chloride, potassium dihydrogen phosphate and sodium hydrogen phosphate were sourced from ChemSupply (Gillman, SA, Australia). Snakeskin 10 kDa MWCO dialysis tubing and PX188 were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Agitation stability study used 3 mL borosilicate glass V-vials with screw caps (Sigma-Aldrich, St Louis, MO, USA) and the accelerated stability and freeze–thaw studies used polypropylene 1.5 mL SPINWIN microcentrifuge tube (Tarsons Products Ltd, West Bengal, India).

Laurett Veras, F.W., Wright, L. and Falconer, R.J. (2026), Physical stability of RNA–lipid nanoparticles under agitation, freeze–thaw and accelerated stability testing. J Chem Technol Biotechnol. https://doi.org/10.1002/jctb.70235