Lipid Impurities in mRNA: Implications and Solutions
Complete the form below to unlock access to ALL audio articles.
The discovery and mass production of life-saving therapeutics and vaccines is a lot more intricate than the initial synthesis. A compound can face various complications throughout the development and manufacturing phases, such as impurities that are hard to detect with traditional analytical methodologies. A recently discovered impurity in mRNA-based products is perhaps the ideal example.
Many pharmaceutical companies rely on lipid nanoparticles (LNPs) for the delivery of mRNA vaccines to the lymph nodes responsible for white blood cell production for immune response. LNPs have been the primary choice of carriers due to their biocompatible nature, ease of scale-up, and chemical properties that improve cellular uptake. Furthermore, studies on the properties of LNPs have been helping to establish them as ideal carriers for gene-based therapies.
Despite promising properties, risks associated with LNP usage surfaced in a recent study conducted by Moderna, where a specific reactive impurity based on one of the lipid components covalently binds the mRNA, altering the mRNA function significantly. In addition, traditional assays failed to detect this concerning modification.
A recent interview with Dr. Adam Crowe, analytical development manager at Precision NanoSystems Inc., sheds light on why this newly discovered impurity is detrimental to mRNA function and how it can be detected more accurately.
Ionizable lipid oxidation as a harmful mRNA impurity source
Kerstin Pohl (KP): How would you describe this new class of mRNA impurity?
Adam Crowe (AC): The impurity we are talking about was first
The lipid adducts are very likely caused by the oxidation of tertiary amine to N-oxide species, which breaks down to form aldehydes that are very reactive toward the mRNA. Basicslly, this adduct is the lipid covalently attached to the mRNA.
KP: How do you anticipate this mRNA impurity becoming an issue for mRNA-LNP developers?
AC: Clients seek nanoparticles tailored to their specific research topics, so there is a vast range of ionizable lipid species which vary significantly in their susceptibility to oxidation. In addition, some types of oxidation are prone to form adducts with mRNA, while others are not. You need to be able to identify not only the presence but also the type of oxidation in the lipid to decipher the extent of damage it can do to your mRNA.
Because these lipid adducts are relatively small compared to the entire mRNA, traditional methods like capillary gel electrophoresis (CGE), which monitors mRNA integrity, cannot recognize them. In smaller oligonucleotide biotherapeutics, such as siRNA, the small percentage of lipid adducts was not significant enough to impact the efficacy of the oligonucleotides. mRNA is much larger in size than siRNA, so the fraction of alkylation per mRNA molecule increases, even though the probability remains the same.
Meredith Packer’s study clearly demonstrated that the presence of a lipid oxidation impurity with a relative abundance of 10-5 (10 ppm) was enough to inhibit the function of the entire mRNA. This particular mRNA impurity needs to be emphasized with cutting-edge analysis.
Integrity analysis vs. fragmentation
KP: What are CGE-based methods, and why are they used as the standard for mRNA impurity analysis?
AC: CGE gives you a robust resolution for profiling the integrity of mRNA. By monitoring the size, mass and integrity throughout manufacturing, we ensure that the mRNA remains intact and determine if there is a manufacturing step causing mRNA degradation. With constantly developing CGE instruments, you can analyze RNA structures of up to 9000 nucleotides.
CGE is used both for the analysis of the raw material mRNA and the mRNA from mRNA-LNP conjugation to ensure that the integrity is retained. This method is often preferred over reversed-phase ion-pair high-performance liquid chromatography (RP-IP-HPLC) because it is better at resolving species with high molecular weight, which makes it ideal for mRNA integrity analysis. Besides the early development phases, we use CGE for quality measurements, because it is a relatively simple and robust assay for releasing a batch for further use.
The problem is that the adduct of the ionizable lipid impurity to the mRNA does not cause a significant enough weight difference. More specifically, one adduct will cause approximately 100-200 Da mass shift, whereas mRNA is in the order of 1-2 MDa. Even with 10 adducts per mRNA, the overall mass shift is still less than a 1000th of the entire mRNA molecular weight. Also, an LNP can undergo oxidation at many sites, but oxidation is deemed harmful only when it is on the tertiary amine. So, detecting oxidation increases is not enough for decision-making.
The best approach to this problem is to analyze your lipid raw material through fragmentation mass spectrometry to see what percentage of oxidation is linked to harmful N-oxide formation. You can only then mitigate the risk of this new class of mRNA impurity from early on.
KP: What are the standard mass spectrometry methods previously used for lipid fragmentation? Why was it not helpful in this particular analysis?
AC: Collision-induced dissociation (CID) has been used very commonly for analyzing biomolecules by cleaving labile bonds, such as ester bonds. The problem is that the derived fragments are often not diagnostic. CID cannot provide comprehensive bond breakage, which is needed to understand oxygen incorporation at different parts of the molecule, such as at an alkene versus a tertiary amine. CID will not give you the range of bond breakage to distinguish between different types of oxidation.
KP: Could nuclear magnetic resonance (NMR) spectroscopy have also been used to analyze oxidation in ionizable lipids?
AC: With NMR, sample preparation and characterization is very time-consuming, not to mention high cost. A lot of nanoparticle manufacturers do not work with NMR. More importantly, the probes used in NMR don’t have the dynamic range to detect lipid oxidation with a relative abundance of approximately 10 ppm.
KP: Can you tell us a bit about the latest liquid chromatography-mass spectrometry (LC-MS) method you used to detect ionizable lipid oxidation more accurately than with the previous methods?
AC: We recently collaborated with SCIEX, who had developed the ZenoTOF 7600 system as a high-resolution mass spectrometry solution. The electron-activated dissociation (EAD) method in the ZenoTOF helped us unlock the level of molecular detail needed for lipid structural analysis. The increased dynamic range enabled us to detect very low abundance impurities, while the EAD gave us much better coverage of bond breakage, allowing us to identify the exact oxidation site — the chain that incorporated the oxygen or the functional group being oxidized.
Thus, we managed to distinguish between different lipid oxidation types with minute relative abundances.
Upon identifying lipid impurities, one can perform additional purification or modify the synthesis process to mitigate the risks associated with N-oxides.
Leveraging lipid fragmentation for structural detail
KP: Do you combine the data from CGE and LC-MS/MS with EAD to get a complete grasp of the mRNA-LNP product at the end of the day?
AC: LNPs are analytically complex materials. You can have anywhere from 5 to 20 different assays on your material depending on the indication and the requirements. Each aspect of quality assurance is non-negotiable in mRNA analysis. So, CGE is still essential to confirm that the mRNA is intact and potent, but you also need structural details from mass spectrometry, namely EAD. You also need to characterize both the raw materials (mRNA and lipids) and their conjugation (LNP), looking at pH, osmolality, the 5’-cap structures and polyadenylation of mRNA. Any one of these could be a production fail point, so the aim is to get a complete picture of the product through diverse data.
KP: What impact will it have on the future of mRNA-LNP therapeutics and vaccines?
AC: Carrying genetic material to their targets via LNPs is a very powerful technique that I think will become much more widely adopted in the near future, not only in COVID-19 research, but also in oncology and personalized medicine. Their functionality has been demonstrated in laboratory settings. The problem is that quality control is not the biggest concern in preliminary studies. As soon as you translate that success to the mass production, you face several risks requiring you to optimize product composition, demonstrate correct lipid behavior and ensure that your downstream processing parameters work correctly.
Today, there is a stronger tendency and need for derisking all the downstream processes as early as possible. Until now, deficient LNP synthesis would have been discovered only halfway through the manufacturing, and the main culprit was usually a reactive species that could have been prevented much earlier. This meant a massive waste of time and resources for the client.
State-of-art technology for raw lipid analysis, such as EAD, allows us to locate problems within the first few weeks – if not days – of the product life cycle. Then, we can notify the client of the potential risks for downstream processes and rectify deficiencies. The ultimate benefit of these advancements is faster access to the manufacturing and the market with much higher confidence in the product.
Dr. Adam Crowe was speaking to Kerstin Pohl, senior global marketing manager, Gene Therapy & Nucleic Acid, SCIEX.
The SCIEX clinical diagnostic portfolio is For In Vitro Diagnostic Use. Rx Only. Product(s) not available in all countries. For information on availability, please contact your local sales representative or refer to www.sciex.com/diagnostics. All other products are For Research Use Only. Not for use in Diagnostic Procedures.
Trademarks and/or registered trademarks mentioned herein, including associated logos, are the property of AB Sciex Pte. Ltd. or their respective owners in the United States and/or certain other countries (see www.sciex.com/trademarks).
