Read Time: 7 minutes

Interviewee: Jane Luo (PhD), Senior Scientist, SCIEX

Lipid nanoparticles (LNPs) are widely used vehicles for mRNA-based therapeutics and vaccines. However, ionizable lipids used in LNPs can be susceptible to N-oxide impurities that can cause functional loss of the mRNA cargo.

In a recent webinar, available on demand, Jane Luo, Senior Scientist (SCIEX), demonstrated how capillary electrophoresis (CE) was used to understand mRNA integrity, purity and encapsulation efficiency. Moreover, Jane showed how liquid chromatography coupled to mass spectrometry (LC-MS) and electron activated dissociation (EAD) enabled the structural elucidation of ionizable lipid raw material, including localizing double bonds and saturated impurities and pinpointing oxygen incorporations.

In this blog, I summarized Jane’s answers to the most pressing questions to help you improve your mRNA-LNP analysis.

Published information

Can you share any of your published information on mRNA-LNPs?

Jane Luo: Yes, I’m very happy to share further details. Please find the two most recent studies on mRNA integrity and encapsulation efficiency with CE here and here. The investigation of different ionizable and cationic lipids with EAD can be found here: ALC-0315, MC3, DODAP. Please do not hesitate to reach out to us if you have any questions.

N-oxides

Is the oxidation of the tertiary amines happening during neat compound storage or after formulation? Is it pH dependent?

Jane Luo: The tertiary amines can undergo oxidation prior to LNP formulation and thereafter. To my knowledge, the process is dependent on oxygen exposure, temperature and pH. Even the filling volume, meaning how much air is in the vial or container, will have an impact on the formation of N-oxides.

Can you comment on which levels and oxides are affecting mRNA efficacy?

Jane Luo: It is difficult to determine an exact level of concern for N-oxides, as the N-oxides are not the final product. These impurities are believed to degrade further to aldehydes that then react with the mRNA, causing activity loss, as published by Packer et al. in 2021. What we do know is that even a very low abundance of N-oxides affects the mRNA efficacy. We discussed this topic with scientists at Precision NanoSystems ULC (part of Cytiva), and the conclusion was that N-oxides levels >0.1% are concerning and may cause lipidation and, therefore, a loss of function of mRNA.

Sample preparation for CE

Can you explain why the mRNA solution was heated prior to CE analysis? Can the same sample preparation be used for different mRNAs?

Jane Luo: Heating the mRNA helps to unfold secondary structures and break up any aggregates. If mRNA samples are not heated prior to CE analysis, multiple peaks and shoulder peaks may be detected. Different mRNAs may fold differently based on their base compositions. In addition, they may have different chemical modifications, which can impact their stability toward sample pre-treatment. I highly recommend doing an optimization experiment to identify the best treatment condition for your mRNA.

Why was the mRNA-LNP sample not heated for determining the free mRNA amount?

Jane Luo: When we want to determine the free mRNA outside the LNP, we need to be as gentle as possible to avoid disruption of the LNPs and prevent the accidental release of the mRNA from the LNPs. We therefore did not heat the samples for determining the amount of free mRNA.

Which sample preparation did you use for your mRNA-LNP sample?

Jane Luo: We used an mRNA encoding for luciferase, a common reporter gene. When we tested different temperatures and incubation times, it turned out this mRNA sample is quite stable. We determined that heating to 70⁰C for five minutes in the presence of formamide was eliminating secondary structures, providing best results for our CE measurements.

Is the use of formamide before CE analysis required, or would heating alone be sufficient?

Jane Luo: mRNA can spontaneously form secondary structures. Even though we used heat treatment to unfold the mRNA followed by quick chilling of the sample, we found that for samples without formamide there were additional peaks in the electropherograms, linked to secondary structures. Using formamide prevented these additional peaks. We think that the formamide helps keep the mRNA in its unfolded state after heat treatment and achieve more consistent data.

Do the optimized LNP extraction conditions depend on the lipid formulation?

Jane Luo: I think that the extraction conditions should be optimized for each lipid formulation used. In the published literature, Triton X-100 is the most used detergent to deformulate LNPs, with concentrations ranging anywhere from 0.1% to about 5%. Some scientists use another detergent named Brij 58. Generally, it is a good idea to optimize the extraction condition with a titration experiment, as I presented.

Do you perform any kind of cleanup on the LNP samples prior to CE analysis, or are the lipids still present in the sample?

Jane Luo: We do not perform any cleanup after the deformulation with Triton X-100. With the sample preparation presented, we have not experienced any issues with subsequent CE measurements. When using a different sample preparation with other detergents, you may want to verify that the buffers do not damage the capillary. Lipids present in a sample do not interfere with the dye-based detection of the mRNA.

Did you try phenol/chloroform extraction for mRNA extraction from LNPs? If yes, did you see any differences in percentage purity for the main peak when comparing different sample preparation methods?

Jane Luo: We did not try phenol/chloroform for deformulating LNPs. For our samples, the mRNA was efficiently released from the LNPs using Triton X-100. Since the phenol/chloroform extraction presents a harsh condition, I suggest using caution for subsequent CE analysis. The residual amount of chloroform could damage the capillary walls. Therefore you may need to do further sample cleanup to ensure removal of residual phenol and chloroform prior to CE analysis.

CE analysis

Which CE system did you use for determining LNP encapsulation?

Jane Luo: We used the BioPhase 8800 system, a multi-capillary electrophoresis system, that allows for running eight samples in parallel. We ran this system with the RNA 9000 Purity & Integrity kit and bare-fused silica capillaries that enable excellent data quality of single-stranded nucleic acid samples. This allows for separating truncated species with high resolution and determining the integrity of your mRNA samples.

Is the capillary temperature controlled and consistent between untreated LNP and deformulated LNP measurements?

Jane Luo: Yes, we control the capillary temperature and use the same temperature for separating the samples from the untreated LNP for the determination of free mRNA and total mRNA after deformulation of the LNP samples.

LC-MS analysis

Have you used the ZenoTOF 7600 system to investigate other lipids in LNPs? Did you find any other impurities of concern?

Jane Luo: Yes, in addition to ALC-315, we have used the ZenoTOF 7600 system to analyze MC3 and DODAP and a few more proprietary lipids. We detected N-oxides at various levels in all of them and were able to differentiate between N-oxides and epoxides. We also found saturations and desaturation of the fatty acid chains as well as methylation events, loss of headgroups and loss of alkyl or acyl chains to name a few more. EAD can help with pinpointing the exact location of these events and determining the potential liabilities of different lipids. The changes in the structure of lipids could potentially affect the efficacy of LNPs. Understanding the lipid structures and their impurities better and determining the quality of a given raw material batch can help with improving the quality of LNPs.

Can you elaborate how much method optimization is needed for lipid analysis with EAD and how much time is required to process the data? Can you efficiently transfer methods to new lipids?

Jane Luo: The amount of optimization with EAD is comparable to CID experiments. While EAD is highly tunable, we already know that efficient fragmentation for singly charged ionizable lipids can be achieved using high electron kinetic energy settings of approximately 10 electron volts (eV). Data-dependent acquisition (DDA) is compatible with EAD, which ensures you acquire fragmentation information on the most abundant impurities in your sample without prior knowledge of the impurities. Optionally, a user can specify an inclusion list to check for expected impurities, such as N-oxides for instance. In terms of processing, the fragment-rich data could take some time to manually investigate. Therefore, SCIEX released lipid processing features in the Molecule Profiler software that can reduce the data processing time from days to hours or minutes.

Can you explain how EAD compares to electron capture dissociation (ECD) and electron transfer dissociation (ETD)?

Jane Luo: Basically, ECD is an electron fragmentation with a very low electron kinetic energy of around 0–1 eV. It can be used to fragment molecules with multiple charges, such as proteins and peptides, but struggles with fragmenting singly charged molecules efficiently. Electron transfer dissociation, or ETD, is a fragmentation method that transfers electrons from a gas to positively charged analytes. Apart from the method requiring long reaction times, it also requires multiply charged analytes. Both types—ECD and ETD—are, therefore, not great for analyzing singly charged lipids on an LC timescale. EAD, on the other hand, can be tuned from 0 to 25 eV. The higher energies of 10 eV and above are comparable to electron impact dissociation (EI). The energy can even break C-C bonds, which enables the detailed structural elucidation of lipids as shown in my presentation. In addition, the fragmentation with the EAD is very efficient since it occurs in a fragmentation cell in which the ions are trapped, not a flow-through cell.