Formulation screening of lyophilized mRNA-lipid nanoparticles

Abstract

Lipid nanoparticles (LNPs) have demonstrated their therapeutic potential as safe and effective drug delivery systems for nucleic acids during the COVID-19 pandemic. However, one of the main challenges during technical CMC (Chemistry, Manufacturing, and Controls) development is their long-term stability at temperatures of 2–8 °C or higher, which may be improved by the removal of water by lyophilization. In this study, we identified lyo-/cryo-protectants for freeze-dried mRNA-LNP formulations beyond conventional excipients such as sucrose and trehalose as T_g-modifiers using polyA as a surrogate. Hydroxypropyl-beta-cyclodextrin, Kollidon® 12 PF (PVP), and dextran 40 kDa were tested in combinations to best stabilize the mRNA-LNPs during the lyophilization process as well as during storage for up to 6 months at 2–8 °C, 25 °C/60 % r.h., and 40 °C/75 % r.h.. We also tested the formulation principle including protectants in- and outside of the LNPs. Formulations were assessed for size, PDI, encapsulation efficiency, and properties related to the lyophilized dosage form. While 10 % (w/V) sucrose formulations successfully stabilized LNPs during the lyophilization process, they were not suitable for storage at temperatures beyond 2–8 °C. The most promising formulations for storage at higher temperatures were identified as 9 % (w/V) trehalose + 1 % (w/V) PVP with only a small increase in size over 6 months at 25 °C maintaining PDI and encapsulation efficiency. Results were verified with eGFP-mRNA-LNPs and tested in cell culture experiments. This study may serve as guidance for formulation scientists to further optimize freeze-dried mRNA-LNP formulations and eventually eliminate the cold chain for mRNA-LNP products.

Introduction

The messenger RNA (mRNA) vaccines Comirnaty® and Spikevax® showed the full potential of lipid nanoparticles (LNPs) as new delivery platform technology during the COVID-19 pandemic (Crommelin et al., 2021). In vitro production allowed the cost-conscious development of effective and safe vaccines in no time (Gote, 2023). They were produced in large amounts, e.g., three billion delivered Comirnaty® doses in 2021 (Warne, 2023); and the mRNA encoding for the spike protein was adjusted in only a few months after the Omicron variant emerged (Webb, 2022). However, one of the main challenges associated with technical CMC development of mRNA-LNPs is their stability (Schoenmaker, 2021). mRNA is sensitive, among others, to thermal and enzymatic degradation (Oude Blenke, 2022). Even a single alteration in the sequence (e.g. strand break, oxidation) in the long mRNA strand (1000 – 5000 nucleotides) can stop translation and lead to loss of efficacy. Therefore, mRNA COVID-19 vaccines have to be stored frozen at temperatures between −90 °C and −60 °C (Comirnaty® from BioNTech) and between −50 °C and −15 °C (Spikevax® from Moderna) respectively, which is challenging from a supply chain perspective including storage logistics and global distribution.

There are multiple approaches to improve the stability of mRNA-LNPs including the engineering of the mRNA on a molecular basis, optimization of the lipid composition of the delivery system, as well as optimization of the surrounding formulation of the LNP. Among others, the removal of water from the drug delivery system has been previously described as a promising formulation strategy to improve long-term stability and enable storage at temperatures of 2–8 °C or higher (Oude Blenke, 2022, Muramatsu, 2022). Since spray-drying may be problematic due to the use of high temperatures (Friis, 2023), lyophilization has been described as a promising approach for drying the formulations (Gote, 2023). In addition, it has been previously discussed that the secondary structure of the mRNA might play a decisive role for long-term stabilization (Oude Blenke, 2022). Moreover, the structural properties of LNPs also appear to influence the functionality and stability of mRNA-LNPs. In particular, the location of the mRNA – whether segregated into aqueous bleb structures or lipid-associated in the core – plays a crucial role (Simonsen, 2024).

Comirnaty® and Spikevax® are stored frozen and both vaccines contain sucrose as a cryoprotectant. While Spikevax® uses a Tris buffer, Comirnaty® used a phosphate buffer with potassium chloride and sodium chloride in the first generation vaccine, which was changed to Tris buffer for the second generation (EMA. Comirnaty Product Information. https://www.ema.europa.eu/en/documents/product-information/comirnaty-epar-product-information_en.pdf, 2023, EMA. Spikevax Product Information. https://www.ema.europa.eu/en/documents/product-information/spikevax-previously-covid-19-vaccine-moderna-epar-product-information_en.pdf, 2024). Tris buffer has the advantage of acting as an “aldehyde sink” reducing the formation of mRNA-lipid adducts (Oude Blenke, 2022). Additionally, the switch to Tris buffer may eliminate the pH shift during the freezing step (Thorat and Suryanarayanan, 2019). As both vaccines contain sucrose and Tris buffer, they are promising formulations for lyophilization. In a recent study, Fan et al. proposed an acetate buffer at a pH of 5 yielding a superior biological read-out after lyophilization compared to other buffer systems such as PBS; and also reported promising results for a formulation containing Tris buffer (Fan, 2024).

Recent publications demonstrate that lyophilized mRNA-LNPs can be stored for several months at 2–8 °C and for a few weeks at 25 °C. Suzuki et al. used LNPs consisting of the newly synthetized, ionizable lipid L202, DSPC, cholesterol, and PEG-DMG. The encapsulated mRNA coded for the SARS-CoV-2 spike protein and the lyophilized LNPs were formulated in 20  mM Tris buffer with different sucrose concentrations. Samples with 16 % (w/V) sucrose showed consistent immunogenicity in mice after 1 month of storage at 5 °C and 25 °C (Suzuki, 2022). Muramatsu et al. manufactured LNPs of the ionizable lipid Lipid 10, DSPC, cholesterol, and PEG2000-c-DMA. They used mRNA encoding for firefly luciferase (fLuc) and influenza virus hemagglutinin. With a formulation of 5  mM Tris containing 10 % sucrose and 10 % maltose (w/V), they reached 24 weeks of stability at 4 °C and 12 weeks of stability at room temperature (Muramatsu, 2022). Meulewaeter et al. prepared LNPs using the ionizable lipid C12-200, DSPC, cholesterol, and DMG-PEG2000 and encapsulated enhanced green fluorescent protein (eGFP) or fLuc-mRNA. They tested Tris, phosphate, and PBS buffer in buffer capacities between 10  mM and 40  mM with 12.5 % (w/V) sucrose or trehalose. The samples were exposed to a continuous freeze-drying process. A formulation with 20  mM Tris and 12.5 % sucrose maintained the transfection properties after 12 weeks at 4 °C, 22 °C, and even at 37 °C. However, the formulation with 10  mM phosphate buffer and 12.5 % sucrose showed good results (Meulewaeter, 2023). Shirane et al. prepared LNPs with ssPalmO-Phe-P4C2, DOPC, cholesterol, and DMG-PEG2000 encapsulating luciferase or hEPO mRNA. The lipid solution was prepared in 90 % t-BuOH because the organic phase was not removed by dialysis but directly during the freeze-drying process. Sucrose was added to the LNPs before they were lyophilized. They tested different buffers (pH values and sodium chloride concentrations) and different sucrose concentrations and concluded that for their alcohol dilution-lyophilization method, acidic pH and low salt concentration were beneficial. Lyophilized LNPs protected with 16 % (w/V) sucrose preserved the gene expression efficiency for at least 4 months at 4 °C (Shirane, 2023). Highest stability was achieved by Ai et al., who lyophilized SARS-CoV-2 vaccines that showed no change in physiochemical properties and bioactivities after 6 months at 25 °C (Ai, 2023). However, they did not disclose the ionizable lipid, buffer system, or protectant used.

In sum, most of these studies use conventional sucrose or trehalose as a cryo- and lyoprotectant to stabilize the LNPs against stress during both the freezing drying process. While trehalose exhibits higher glass transition temperatures (Tg’/Tg), which are beneficial for more aggressive lyophilization cycle conditions and enable higher storage temperatures, trehalose crystallization during frozen storage needs to be investigated (Singh, 2011). In contrast, sucrose has a slightly lower Tg’/Tg and may crystallize and hydrolyze at elevated storage temperatures. Further tested sugars reported in literature are mannitol, lactose, and glucose, which were not successful in sufficiently stabilizing mRNA-LNPs during lyophilization or storage (Zhao, 2020, Li, 2023).

Tg’/Tg-modifiers have been described in the literature as promising excipients to increase the glass transition temperature of the freeze-concentrated solution (Tg’), as well as of the lyophilized cake (Tg). As a result, lyophilization cycles may be significantly shortened due to more aggressive primary drying, saving time and energy, and lyophilized products can be potentially stored at higher temperatures without a collapse or impairment of the lyophilized cake. Häuser et al. described the use of sucrose (Suc), hydroxypropyl-beta-cyclodextrin (CD), dextran 40 kDa (Dex), and PVP for lyophilization of monoclonal antibodies. Formulations with CD/Suc and HP-beta-CD/PVP/Suc performed better for the amorphous protein matrix than pure sucrose for 9 months of storage at 40 °C, potentially enabling storage at room temperature (Haeuser et al., 2020). In another publication, the same authors showed that lyo cycle time may be reduced by 50 % using a CD/Suc formulation (Haeuser et al., 2019). In a very recent study, Fan et al. evaluated the addition of various cryo/lyo-protectants including sucrose, maltose, PEG-1500 and PVP-K-12 as well as lysine for their ability to stabilize eGFP encoding mRNA during lyophilization. They concluded that PVP in combination with an acetate or Tris buffer might be promising excipients based on various analytical read-outs including cell culture experiments. However, they did not provide stability data on their formulations offering mechanistic insights about the influence of physiochemical and structural properties of lyophilized mRNA-LNPs on functionality (Fan, 2024). The formulation principle of using Tg’/Tg-modifiers to stabilize lyophilized mRNA-LNPs on long term storage and above 2–8 °C has not been tested. For the sake of clarity, Tg’/Tg-modifiers are only called Tg-modifiers later in this article.

The objective of this study was to identify Tg-modifiers that best stabilize mRNA-LNPs during the lyophilization process as well as during long-term storage, to guide future formulation development of lyophilized mRNA-LNPs. For this purpose, we included hydroxypropyl-beta-cyclodextrin, Kollidon® 12 PF (PVP), and dextran 40 kDa as Tg-modifying excipients into the study. While the optimization of pH and buffer system was not the focus of this study, we further studied the combination of conventional cryo-/lyo-protectants such as sucrose and trehalose with the Tg-modifiers, as well as the presence of sugars and Tg-modifiers inside the LNPs. This was achieved by adding protectants into the aqueous phase before the mixing step.

PolyA LNPs with a lipid composition equivalent to the Moderna vaccine Spikevax® were used as a surrogate for mRNA-LNPs consisting of SM-102, cholesterol, DSPC, and DMG-PEG2000. Results were confirmed with eGFP-mRNA-LNPs. We determined several product attributes for the lyophilized product including Tg, cake appearance, residual moisture, and reconstitution time, as well as formulation attributes of the reconstituted solution including pH, osmolality, and Tg’. Critical quality attributes (CQAs) of the LNPs such as size/PDI and encapsulation efficiency were assessed before and after lyophilization and after storage up to 6 months at 2–8 °C, 25 °C/60 % r.h., and 40 °C/75 % r.h.. eGFP-mRNA protein expression was tested in cell culture experiments using HeLa cells.

Download the full article as PDF here Formulation screening of lyophilized mRNA-lipid nanoparticles

or read it here

Materials

Nucleic acid solutions were prepared from either polyadenylic acid (polyA) (ABP Biosciences, Virginia, USA) or CleanCap EGFP-encoding mRNA (TriLink, San Diego, USA) in 50  mM citrate buffer pH 4.0 (citric acid, Thermo Fisher Scientific; sodium citrate tribasic, Sigma-Aldrich, St. Louis, USA). Stock solutions of the ionizable lipid SM-102 (BroadPharm, San Diego, USA), cholesterol (Sigma-Aldrich), the helper lipid DSPC (Lipoid, Ludwigshafen, Germany), and DMG-PEG2000 (Avanti Polar Lipids, Birmingham, USA) were prepared in ethanol (Carl Roth, Karlsruhe, Germany) and mixed at a molar ratio of 50: 38.5: 10: 1.5.

Anna Ruppl, Denis Kiesewetter, Monika Koell-Weber, Thomas Lemazurier, Regine Süss, Andrea Allmendinger, Formulation screening of lyophilized mRNA-lipid nanoparticles, International Journal of Pharmaceutics, 2025, 125272, ISSN 0378-5173, https://doi.org/10.1016/j.ijpharm.2025.125272.