Digital Twin Fundamentals of mRNA In Vitro Transcription in Variable Scale Toward Autonomous Operation

Abstract

The COVID-19 pandemic caused the rapid development of mRNA (messenger ribonucleic acid) vaccines and new RNA-based therapeutic methods. However, the approval rate for candidates has the potential to be increased, with a significant number failing so far due to efficacy, safety, and manufacturing deficiencies, hindering equitable vaccine distribution during pandemics. This study focuses on optimizing the production of mRNA, a critical component of mRNA-based vaccines, using a scalable machine by investigating the key mechanisms of mRNA in vitro transcription. First, kinetic parameters for the mRNA production process were determined. The validity of the determination and the robustness of the model are demonstrated by predicting different reactions with and without substrate limitations as well as different transcripts. The optimized reaction conditions, including temperature, urea concentration, and concentration of reaction-enhancing additives, resulted in a 55% increase in mRNA yield with a 33% reduction in truncated mRNA. Additionally, the feasibility of a segmented flow approach allowed for high-throughput screening (HTS), enabling the production of 20 vaccine candidates within a short time frame, representing a 10-fold increase in productivity, compared to nonsegmented reactions limited by the residence time in the plug flow reactor. The findings presented for the first time here contribute to the development of a fully continuous and efficient manufacturing process for mRNA and other cell and gene therapy drugs/vaccine candidates as presented in our previous work, which discussed the integration of process analytical technologies and predictive process models in a Biopharma 4.0 facility to enable the production of clinical and large-scale doses, ensuring a rapid and resilient supply of critical therapeutics. The results in this study especially highlight that the same machine and equipment can be used for screening and manufacturing different drug candidates in continuous operation. By streamlining production and adhering to quality standards, this approach enhances the industry's ability to respond swiftly to pandemics and public health emergencies, addressing the urgent need for accessible and effective vaccines.

Figure 1

Ion-pair reversed-phase chromatogram of the mRNA in vitro transcription product with smaller transcripts corresponding to the first peak and the full-length transcript corresponding to the second peak.

Figure 2

Setup of the test unit in which the segmented slugs are generated, detected, and fractionated by the resulting detection signal.

Figure 3

Calibration of the mRNA concentration measured by ion-pair reversed-phase chromatography. Chromatogram with increasing injection mass (a) and calibration line (b).

Figure 4

Calibration of the mRNA concentration measured by anion exchange chromatography. Chromatogram with increasing injection mass (a) and calibration line (b).

Figure 5

Calibration of the nucleotide concentration by anion exchange chromatography. Chromatogram showing baseline separation of ATP, CTP, GTP, and UTP (a), calibration line for ATP (b), calibration line for CTP (c), calibration line for UTP (d), and calibration line for GTP (e).

Figure 6

Experimental determination of Michaelis–Menten constants and Monte Carlo simulations for ATP (a), CTP (c), UTP (e), and GTP (g) and corresponding Lineweaver–Burk plots (b, d, f, and h). Validation of the model on an open-access independent data set from Rosa et al. with three alternative transcripts (i).

Figure 7

Progression of in vitro transcription resolved on agarose gel electrophoresis; NTP concentration of 10 mM, 6 time points, each with two dilutions indicated by the text.

Figure 8

Input (a) and output (b) signals (conductivity) for six experimental points of the DoE.

Figure 9

Statistical evaluation of the experiments. Actual vs predicted plot for mRNA concentration (a), profile plot (b), normal quantile plot (c), and residual plot (d).

Figure 10

Statistical evaluation of the experiments. Actual vs predicted plot for truncated mRNA percentage (a), profile plot (b), normal quantile plot (c), and residual plot (d).

Figure 11

Contour plots showing the effect of the reaction-enhancing additive concentration and temperature on mRNA concentration (a) as well as for percent truncated mRNA (b). Effect of the urea concentration and reaction-enhancing additive concentration of percent truncated mRNA is shown on the right (c).

Figure 12

Agarose gel electrophoresis image (left), anion exchange (mid), and ion-pair reversed-phase (right) chromatogram of the optimal operating point.