Development and Application of Automated Sandwich ELISA for Quantitating Residual dsRNA in mRNA Vaccines

Abstract

The rise of mRNA as a novel vaccination strategy presents new opportunities to confront global disease. Double-stranded RNA (dsRNA) is an impurity byproduct of the in vitro transcription reaction used to manufacture mRNA that may affect the potency and safety of the mRNA vaccine in patients. Careful quantitation of dsRNA during manufacturing is critical to ensure that residual dsRNA is minimized in purified mRNA drug substances. In this work, we describe the development and implementation of a sandwich Enzyme-Linked Immunosorbent Assay (ELISA) to quantitate nanogram quantities of residual dsRNA contaminants in mRNA process intermediates using readily available commercial reagents. This sandwich ELISA developed in this study follows a standard protocol and can be easily adapted to most research laboratory environments. Additionally, a liquid handler coupled with an automated robotics system was utilized to increase assay throughput, improve precision, and reduce the analyst time requirement. The final automated sandwich ELISA was able to measure <10 ng/mL of dsRNA with a specificity for dsRNA over 2000-fold higher than mRNA, a variability of <15%, and a throughput of 72 samples per day.

Keywords: automation; dsRNA; high throughput; mRNA; sandwich ELISA.

Figure 1

Proof-of-concept assay quantitating residual dsRNA in mRNA elution samples (panel (A,B)) and IVT samples (panel (C,D)). Top panels (A,C) display the dose–response curves obtained for reference standard (black), sample 1 (red), and sample 2 (blue). Bottom panels (B,D) display the apparent concentration of dsRNA in sample 1 (red) and sample 2 (blue) which were back-calculated from the fluorescence values using the parameters obtained from a 4-parameter logistic fit of the standard curve. Error bars represent 1 standard deviation of two sample replicates.

Figure 2

Evaluation of assay interference by components within the IVT reaction mixture (A) or the BDS buffer (B). The 142 bp dsRNA reference standard was spiked into either IVT mixture (A) or formulation buffer (B) to prepare a 1000 ng/mL mock IVT or mock BDS sample, respectively. Serial dilutions were performed either in STE buffer alone (red curve) or in STE containing the respective sample buffer (blue). A representative curve measured from 2 IVT samples prepared without the optional enzyme removal step (red and blue) relative to reference standard (black) is demonstrated in panel (C).

Figure 3

dsRNA ELISA curve parameters obtained from reagent optimization. The blank value (A), signal-to-baseline ratio (B), EC-50 value (C), and the slope of the linear region (D) are given for assays performed using either 2 µg/mL, 4 µg/mL, or 8 µg/mL of J2 coating antibody (black, blue, and red, respectively) in either PBS coating buffer (dark color) or NH₄SO₄ coating buffer (light color) and with detection antibody diluted 2-fold, 4-fold, 8-fold, or 16-fold. Values reflect the average of 2 replicates; error bars indicate 1 standard deviation.

Figure 4

(A) Proof-of-concept assay measuring dsRNA concentration in mRNA IVT reaction products performed using either unmodified UTP or N1-Me-pseudo UTP. Values indicate the concentration of dsRNA expressed as the percentage of dsRNA relative to the concentration of mRNA (see Section 2). Average and standard deviation is of 3–4 dilutional replicates. (B) Residual dsRNA measured by dsRNA ELISA in mRNA construct 1–7 BDS samples. mRNA BDS samples were prepared using either process 1 (solid bars) or process 2 (striped bars). Values indicate the concentration of dsRNA expressed as the percentage of dsRNA relative to the concentration of mRNA (see Section 2). Average and standard deviation is of 3–4 dilutional replicates.

Figure 5

Use of dsRNA ELISA to support process development. (A) Evaluation of the effect of two different polymerase buffer conditions on the relative % dsRNA. (B) Effect of manual (solid) vs. liquid handler (striped) pipetting on the relative percentage of dsRNA in IVT reaction products after 1.5 h or 3 h reaction time (C) Removal of dsRNA using two cellulose-based purification strategies for capturing dsRNA. (D) Relative % dsRNA in an IVT sample before and after a commonly used 3-step purification (steps 1–3) including OdT elution, tangential flow filtration, and buffer exchange as measured in constructs 1–5 (grey scale).

Figure 6

Comparison of ELISA dose–response curve of 142 bp dsRNA reference standard (black circles) and positive control (red squares) generated using the semi-automated ELISA protocol (solid lines) or the fully automated ELISA protocol (dashed lines) before (A) and after (B) optimization of the fully automated ELISA, as described in Section 3. Lines indicate the fit of raw data to a 4PL curve. (C,D) EC-50 value measured for the reference standard when using either the semi-automated or fully automated ELISA before optimization (C) and after optimization (D).