Rapid quality analysis of mRNA lipid nanoparticles using micelle formulations

 Pharmaceutical manufacturers developed mRNA lipid nanoparticle (LNP) vaccines as an effective method to combat COVID-19. With multiple COVID-19 variants already in circulation, the requirement of booster vaccines every several months to maintain immunity, and manufacturers’ new focus on using mRNA LNP vaccines against other infectious diseases, the high demand for mRNA LNP vaccines will only increase. This necessitates a high throughput method of rapidly determining vaccine quality to ensure the correct dosage and functionality is maintained in each vaccine. The critical quality attributes (CQA) for mRNA LNP vaccines are: mRNA length (a measure of the identity of mRNA), mRNA titer, and LNP degradation (to ensure LNPs have properly encapsulated the mRNA). All of these CQAs can be determined with a micelle formulation method.

Using micelle capillary electrophoresis, we developed an electrophoretic technique where instead of a sieve-like gel we use an entangled micelle network to perform length-based separations. We aimed to use this technique to determine length, titer, encapsulation efficiency, and lipid quality (as measured by the release kinetics of mRNA from the lipid shell while being lysed by the surfactant during electrophoresis). However, the entangled micelle network rapidly lyses the mRNA LNP, thus it is impossible to measure the release of mRNA while also achieving a length-based separation. Regardless, when using micelle capillary electrophoresis, we can determine the identity and titer of mRNA LNPs. Furthermore, we are capable of assessing degradation effects on mRNA LNPs to due temperature, freeze/thaw cycles, and enzymatic digestion. We discovered that lyophilization with the use of a cryoprotectant prior to long term high temperature storage is protective of mRNA. Without lyophilization, the mRNA LNPs degrade rapidly at room temperature. Even mRNA stored for 3 months at 37°C remains intact if previously lyophilized. Freeze/thaw cycling is not as damaging mRNAs as room temperature storage, and enzymes are capable of permeating the LNP membrane and degrading mRNA. 

To overcome the limitation of micelle capillary electrophoresis, we developed another analytical technique using stopped-flow kinetics and spectrometry. The assay uses a stained mRNA LNP sample and mixes it with a low surfactant buffer. The lysing process is tracked using spectrometry and with it we can determine the encapsulation efficiency of the mRNA LNP and the release kinetics of the mRNA from the lipid shell. We determine that manufacturing conditions and lipid concentrations do not have a significant impact on the release kinetics of the mRNA, however they do dramatically affect the encapsulation efficiency. However, the PEGylated lipid content dramatically affects the release kinetics, with less PEGylated lipid, there is significantly higher release time. We propose that the PEG causes steric hindrance and prevents mRNA LNP aggregation, thus upon removal of the PEGylated lipid, the mRNA LNPs aggregate and thus require more time to be fully lysed. 

Furthermore, we have developed an on-line concentration technique called reverse isotachophoresis with micelles, where a system of different electrophoretic mobility buffers is used to concentrate a sample and then a separation occurs. An application of which is for adventitious agent detection in biological processes, such as determining the presence of mouse hepatitis virus in CHO cells. We show that we are capable of achieving femtomolar level detection of MHV in 5 minutes using this technique. This technique relies upon electro-osmotic flow instead of electrophoretic flow, thus in order to achieve 1 fM detection with a high viscosity system, the time allowed to concentrate our sample must be adjusted. To do this, we had to increase the length of our capillary and increase the concentration time. To maintain a 5-minute runtime, we reversed the flow of the technique, where the sample is concentrated out the outlet and sent back towards the detector, as opposed to being concentrated at the inlet and allowing it to migrate towards the detector. Effectively, this decreased the length of the detector from 90 cm to 10 cm. Finally, the alkane tail on the γPNA probe plays an important role in increasing the LOD by causing a sweeping effect to occur. The alkane tail causes the viral DNA to concentrate at the interface between the micelle-free and micelle zones, thus introducing another mechanism to increase the limit of detection. 

We discussed our development of different types of nucleic acid probes to bind to mRNA. We developed PNA, γPNA, LNA, and DNA probes. We had the most success with binding EPO specific PNA or γPNA, however we would still observe multiple binding peaks. Future work should be focused on improving probe design for high binding affinity and reducing the number of multiple peaks present in the electropherogram.