HPLC in IVT mRNA Process Monitoring

The success of mRNA-based COVID immunizations has opened new avenues to gene-based vaccines and treatments. In addition to obvious opportunities for preventing infectious diseases, mRNA is under investigation for treating genetic disorders, cancer, heart disease, and to replace diseased, damaged, or missing body parts through regenerative medicine. In contrast to therapies that act directly on molecular targets, or that supply a missing protein (e.g., insulin in diabetes), mRNA harnesses the innate protein-production capabilities of cells to produce antigens or therapeutic proteins in vivo.

In vitro transcribed (IVT) mRNAs are synthetic messenger RNAs, designed and produced in vitro through cell-free manufacturing, to express therapeutic or immunogenic proteins in vivo. Because they are synthetic, IVT mRNAs may be tailored to overcome issues with conventional gene therapies, for example low cellular uptake, poor translational efficiency, and immunogenicity.

As the current gold standard for mRNA manufacturing, IVT mRNA uses bacteriophage-derived modified RNA polymerases to synthesize the gene from synthetic or chemically modified ribonucleoside triphosphate (rNTPs) building blocks. By sidestepping cell culture, IVT is efficient, robust, and versatile but it still requires sequence-specific promoters.

Like DNA-based therapies, mRNA exploits innate protein manufacturing capabilities to generate therapeutic proteins directly in cells. But unlike DNA, which must enter the nucleus and whose effects are of varying duration, mRNA is processed into protein in the cytoplasm and degraded immediately thereafter, with little to no opportunity to enter the nucleus and cause genetic mutations. The potential to generate proteins transiently and predictably opens opportunities for novel, sequential dosing strategies within the parameters of mRNA degradation, protein expression, and according to the patient’s response.

The role of analytics

mRNA treatments are biologicals manufactured through complex unit operations. A “block diagram” view of the process, from concept to patient, includes identification of the therapeutic protein or antigen, mRNA design, in vitro transcription (usually done through polymerase enzymes), formulation, administration, and monitoring. Analytics are involved at every step, from quality checking all ingredients, quantitation and quality monitoring of reaction parameters, to post-treatment studies of the drugs’ activity in patients (potential genetic changes, protein expression levels, in vivo degradation, etc.).

The potential analytical targets are therefore substantial in number, chemically diverse, and require different levels of care in their utilization. Consider the differing requirements for characterizing ingredients and reagents, to checking enzymes and other active reagents, to characterizing the therapeutic gene construct, to process monitoring, to measuring its effects in patients, and the criticality of each.

The mRNA alone incorporates several structural features representing critical quality attributes, for example:

• The 5’ cap, located at the 5’ mRNA terminus, may carry different degrees of methylation and governs such diverse activities as resistance to degradation and translation efficiency. The 5’ cap may be added during or after IVT.
• Consisting of up to 250 adenine ribonucleotides, poly-adenylated (poly(A)) tails regulate mRNA translation efficacy and therefore therapeutic protein expression. Poly(A) may be incorporated through the DNA template or added enzymatically after IVT
• Untranslated regions (UTRs) at the 3′ and 5′ terminals that regulate protein expression, assist in mRNA subcellular localization, and regulate mRNA stability.
• Biologically relevant (“open”) reading frames, or regions translated to protein through combinations of consecutive oligonucleotides (codons)

Optimizing IVT rests on understanding the effects of reagents and reaction conditions on mRNA production kinetics and the consumption of oligonucleotide building blocks, both of which are problematic due to slow, low-throughput, end-point analytics.

These non-genetic components of IVT are also sources of impurities including, but not limited to, unused or breakdown products from:

• Plasmid DNA, enzymes, and reagents
• Process impurities from leftover DNA templates, enzymes used for IVT or post-translational modifications, capping reagents
• IVT-produced RNA variants such as fragments, dsRNA, aggregates, and truncated RNA

In response to these needs, a group at Bia Separations (a Sartorius company) have developed an analytical platform based on high-performance liquid chromatography (HPLC), which (among other things) allows the adjustment of IVT reaction conditions in real time in response to the depletion of key nucleoside triphosphate (NTP) building blocks. This allowed investigators to convert the production process from batch to fed-batch, thus extending production time and doubling yield.

Given its suitability to a wide range of analytes, HPLC is perfect for this particular job. “Historically, HPLC is one of the most selective and scalable tools for analyzing biologics,” says Rok Sekirnik, Ph.D., Head of Process Development for mRNA/pDNA at Bia Separations. “Yet, given the physical size of mRNA, substantial modifications are required for IVT mRNA processing.”

Conventional porous HPLC media are sized to facilitate mass transfer of small molecules or proteins into and out of the pores. Because of its size, mRNA diffusion into these pores is slow or nonexistent. Pores are also a source of laminar forces that could disrupt or inactivate mRNA.

mRNA is much larger than a typical biologic, with molecular weights up to eight times larger than for antibodies (1.3 MDa vs. 150 kDa, respectively). The molecule is negatively charged, with a natural affinity to anion exchangers, but hydrophobic in the presence of high salt and frequently poly-adenylated during IVT processing.

“This is where chromatography resins based on wide, interconnected flow channels, modified with suitable ligands, come into play,” Sekirnik continues. The Bia effort finally settled on affinity ligands and multimode ligands that combine anion exchange with hydrogen bonding. “The stationary phases on these monolithic columns purify mRNA based on convective vs. diffusive mass transport, and minimize turbulent and shear forces, which can damage mRNA.”

Additionally, by coupling HPLC to exonuclease digestion, for example, the authors extended the utility of HPLC analytics to monitoring the capping efficiency under different reaction conditions.

“At-line analytics provides insights into the kinetics of mRNA production and NTP consumption, which allows us to optimize individual IVT reaction components,” Sekirnik tells Biocompare. Critical concentration parameters include levels and depletion of magnesium, plasmid, and nucleoside triphosphate. “We showed that optimal magnesium ion concentrations may be construct-dependent, and high concentrations can be inhibitory. We also demonstrated that NTP concentration optimization must simultaneously include considerations for magnesium levels.”

And from the perspective of cost of goods, the Bia team found that plasmid concentration does not correlate with mRNA yield, which, according to Sekirnik, minimizes production costs.