(W1230-11-74) Predicting Stability of Lyophilized mRNA/LNP Formulations Using Solid State NMR

Purpose: Messenger ribonucleic acid/lipid nanoparticle (mRNA/LNP) vaccines have recently emerged as a popular modality for curing many diseases. mRNA/LNP formulations must be stored at frozen temperatures (typically -80 ºC) to maintain stability. It is of great interest to develop mRNA/LNP formulations that are stable at room temperature. Lyophilization of therapeutic proteins with a disaccharide has been used for decades to prepare shelf stable formulations. The disaccharide stabilizes the protein both during the lyophilization process (prevents buffer-induced pH shift, protein self-association, and cold denaturation) and during storage. The leading hypothesis to explain how the disaccharide stabilizes the protein during storage is the vitrification theory, which states that by dispersing the protein throughout a relatively immobile glass (the amorphous disaccharide), the mobility of the protein will be lowered, decreasing the propensity for aggregation. For maximum stability, the mRNA/LNP and the sucrose must be homogeneously mixed. Solid state nuclear magnetic resonance spectroscopy (SSNMR) will be used to measure both mobility and homogeneity.

Methods: mRNA/LNP formulations were prepared in a Precision NanoSystems IGNITE NanoAssemblr using TriLink CleanCap EFGP mRNA and a cocktail of lipids including an ionizable lipid, a poly(ethylene glycol) lipid, cholesterol, and a zwitterionic phospholipid in a flow rate ratio of 3:1. Sucrose or trehalose was added to the solution at concentrations of 2.5, 5, 10, and 20% (w/v). Solutions were lyophilized with a SP Scientific LyoStar 3. Moisture content (Karl Fischer titration), particle size (dynamic light scattering), encapsulation efficiency (RiboGreen Assay), and mRNA Integrity (capillary electrophoresis) were performed after mRNA/LNP fabrication, as well as after lyophilization. All SSNMR data were acquired using a 400 MHz Bruker Avance (Billerica, MA) spectrometer and Phoenix NMR HX MAS probe. Samples were packed into 6mm zirconia rotors in a dry box with < 1% RH. All spectra were acquired using cross-polarization, magic angle spinning at 4 kHz, and total sideband suppression. ¹³C detected or ³¹P detected saturation recovery experiments were used to measure proton spin–lattice relaxation times in the laboratory frame (¹H T₁). ¹H T₁ relaxation times from each component in the system(s) were compared, to assess homogeneity and mobility. Additional experiments were performed by freezing mRNA/LNP + disaccharide formulations in-situ. During freezing, NMR spectra were collected periodically.

Results: Before lyophilization, all samples had particle sizes of approximately 90 nm. This is in the acceptable range from approximately 80-110 nm. Before lyophilization, encapsulation efficiency of all samples was approximately 95%. After lyophilization, particle size ranged from 136 -170 nm, depending on the stabilizer used. After lyophilization, encapsulation efficiency ranged from 50-80%. In all cases, before and after lyophilization, at time zero, mRNA integrity was acceptable (one peak, electrophoresis). SSNMR was used to assess homogeneity and mobility of lyophilized systems. Adding more sucrose (from 2.5% (w/v) to 20% (w/v)) resulted in a decrease in system mobility, as measured by ¹H T₁ relaxation times. ¹H T₁ will be linked to mRNA/LNP stability (particle size, encapsulation efficiency, mRNA integrity) over 1 year. In mRNA spiked samples (mRNA is outside of the LNP, not encapsulated) a biexponential fit was observed in ³¹P SSNMR ¹H T₁ measurements. This suggests that the encapsulated mRNA has slower mobility than the mRNA that is outside of the LNP.

Conclusion: Stabilizer type and concentration effect mRNA/LNP critical quality attributes. At extreme low or high sucrose concentration, particle size is high and encapsulation efficiency is low. Stabilizer concentration has an effect on the mobility and homogeneity of the sample. Freezing behavior of mRNA/LNP/Sucrose solutions was studied by SSNMR, and freezing point depression was observed.

Acknowledgements: This work was performed under a Project Award Agreement from the National Institute for Innovation in Manufacturing Biopharmaceuticals (NIIMBL) and financial assistance award 70NANB21H085 from the U.S. Department of Commerce, National Institute of Standards and Technology.