Skip to main content
NCBI home page
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2025 Nov 20;53(21):gkaf1124. doi: 10.1093/nar/gkaf1124

Engineered chimeric T7 RNA polymerase improves salt tolerance and reduces dsRNA impurity generation during in vitro transcription of mRNA

Athul Sanjeev 1,2, Caroline Reiss Summers 3,✉, Jason Politi 4, Penny J Beuning 5,✉, Christopher J Cheng 6,✉
PMCID: PMC12630135  PMID: 41261859

Abstract

Messenger RNA (mRNA) is typically produced enzymatically through in vitro transcription using wild-type T7 RNA polymerase; however, this enzyme is also known to generate double-stranded RNA (dsRNA) impurities during transcription. These impurities may evoke an immune response, potentially reducing the therapeutic index of the drug product. Upstream and downstream approaches may be used to mitigate the formation or removal of such dsRNA impurities. However, these processes can be costly, reduce yield, and be challenging to scale for clinical manufacturing. In this work, we engineered a chimeric T7 RNA polymerase by tethering a DNA-binding domain to increase the selectivity of the polymerase for DNA templates and found that it was capable of reducing dsRNA formation. The chimeric T7 RNA polymerase reduced dsRNA levels by three- to four-fold relative to wild-type T7 RNA polymerase with commensurate reduction in immune stimulation in vitro. Additionally, the chimeric T7 RNA polymerase showed improved salt tolerance and was active at NaCl concentrations up to 150 mM, which is otherwise a restrictive condition for wild-type T7 polymerase. These features make this novel enzyme an attractive option for addressing various challenges facing the field of mRNA manufacturing.

Graphical Abstract

Graphical Abstract.

Graphical Abstract

Open in a new tab

Introduction

The messenger RNA (mRNA) therapeutics field has gained considerable traction in recent years and is being used in numerous and varied applications such as prophylactic vaccines, cancer vaccines, immunotherapy, cell therapy, protein replacement therapy, and gene editing [1, 2]. Certain features make mRNA an attractive therapeutic tool. This technology utilizes endogenous translation machinery to facilitate the translation of any protein of interest in a cell, including proteins that would be challenging to synthesize in vitro as protein biologics [1–4]. Moreover, mRNA can be safely administered in vivo through non-viral delivery systems such as lipid nanoparticles, eliminating the risk of genomic integration [1, 4, 5]. It also supports large-scale kilogram-level production and offers a versatile, platform-based approach by enabling a standardized way of manufacturing mRNA molecules with different encoded proteins [6, 7].

These compelling attributes provided the foundation for the rapid expansion of the mRNA manufacturing infrastructure over the last decade, which was also largely driven by the global vaccine response during the COVID-19 pandemic [8, 9]. However, opportunities for innovation in more efficient mRNA process scale-up [10] and control of impurities [2–4] remain.

mRNA is synthesized enzymatically by an in vitro transcription (IVT) reaction utilizing DNA-dependent RNA polymerases. Wild-type T7 RNA polymerase (WT T7 RNAP) is commonly used due to its efficient kinetics, processivity, single-subunit nature, and fidelity [11, 12]. Although IVT using T7 RNAP is a one-pot reaction, IVT requires optimization to efficiently generate high-quality mRNA products. Process development challenges include producing mRNA of high full-length purity and high 5′ capping efficiency, maximizing yield, and reducing formation of T7 RNAP-related impurities that must be removed during downstream processing [10, 13].

A key impurity generated during IVT of mRNA by T7 RNAP is double-stranded RNA (dsRNA) [14–17]. Control and reduction of the dsRNA impurity is particularly important during mRNA manufacturing, as dsRNA may cause innate immune activation [18]. Innate immune receptors like RIG-I, MDA5, PKR, and Toll-like receptors sense this dsRNA impurity and trigger the production of interferons, tumor necrosis factor-α, and cytokines, which could potentially lead to translation inhibition and cell toxicity [18–20]. Significant efforts have been made in the field to understand how dsRNA is generated during IVT and how to reduce dsRNA levels via process operations [21–26].

Factors affecting dsRNA generation during IVT are still unclear; multiple researchers have proposed different mechanisms. These include the T7 RNA polymerase switching to a non-template strand [15, 17], promoter-independent transcription by using RNA as a substrate [27], and RNA forming stem-loop structures to promote T7 transcription [28]. Additionally, dsRNA generation could occur via multiple of these possible mechanisms, and possibly other undescribed mechanisms, in the same IVT reaction. Despite the uncertainty around how dsRNA is generated during IVT, various downstream purification approaches can significantly reduce dsRNA levels in the final mRNA product. These approaches include using cellulose columns [29], ion pair reverse phase chromatography [21], mesoporous silica [30], steric exclusion chromatography [31], and hydrophobic interaction chromatography [32]. These chromatography steps add time and expense and reduce yield, and some approaches use organic solvents, all of which can present challenges for manufacturing mRNA. An alternative approach, which is the subject of the work presented in this paper, is to reduce the production of dsRNA species during the IVT reaction itself.

In this work, we engineered a chimeric T7 RNA polymerase with the goal of improving selectivity for DNA templates and reducing the generation of RNA-templated dsRNA species during transcription. This chimeric T7 RNA polymerase, generated by tethering a DNA-binding domain to the N terminus of T7 RNAP, produced less dsRNA than WT T7 RNAP during IVT. Interestingly, we also observed improved tolerance for high salt conditions, which further reduced dsRNA and might support continuous mRNA manufacturing. mRNA generated by engineered chimeric T7 RNA polymerase was effectively translated and showed lower immune activation relative to mRNAs produced by a WT T7 RNAP when tested in cells. This novel RNA polymerase has several beneficial features that can enable more efficient large-scale manufacturing of therapeutic mRNA.

Materials and methods

Protein expression and purification

Plasmid vector generation, protein expression, and purification were performed commercially (GenScript). Briefly, a pET-28a-T5 vector was created by replacing the T7 promoter in pET-28a with the T5 promoter. All protein-coding sequences were optimized for expression in Escherichia coli and were cloned into the pET-28a-T5 vector. Escherichia coli BL21(DE3) cells were transformed with recombinant plasmids, and colonies were picked for cell bank preparation. The bacteria were grown overnight for seed culture, which was diluted into fresh LB media with 50 μg/ml kanamycin for protein production. Cultures were incubated at 37°C. Once cell density reached OD600 of 0.6, 0.5 mM IPTG was introduced for induction. The cells were harvested and lysed (lysis buffer: 50 mM Tris–HCl, 500 mM NaCl, 1 mM Tris(2-carboxyethyl)phosphine (TCEP), pH 8.0) after 4 h of induction. His-tagged proteins were purified from the lysate using Ni-NTA columns (wash buffer: 50 mM Tris–HCl, 1 M NaCl, 0.5% Triton X-100, 1% Triton X-114, 1 mM TCEP, pH 8.0; elution buffer: 50 mM Tris–HCl, 500 mM NaCl, 1 mM TCEP, 20/50/500 mM imidazole, pH 8.0). The tag was removed using TEV protease cleavage and nickel column purification, followed by size-exclusion chromatography using a Superdex 200 column (SEC buffer: 50 mM Tris–HCl, 500 mM NaCl, 10% glycerol, 1 mM TCEP, pH 8.0). Protein purity was analyzed using sodium dodecyl sulfate–polyacrylamide gel electrophoresis.

In vitro transcription of mRNA

mRNA was generated using an IVT reaction. The transcription reaction contained 25 ng/µl linearized DNA template, 1× IVT buffer (50 mM Tris–HCL, 19 mM MgCl2), a custom ratio of NTPs (N1-methyl pseudouridine triphosphate was used instead of UTP to reduce immunogenicity [33]), 4 mM cap analog, 10 mM dithiothreitol (DTT), 500 U/ml RNase inhibitor, 20 U/ml inorganic pyrophosphatase, and 0.8 µM of WT T7 RNAP or chimeric T7 RNA polymerase variants. One molar sodium chloride was used to adjust salt concentration according to test conditions. The transcription was performed at 38°C for 90 min. The IVT product was subjected to DNase digestion (25 U/ml) for another 40 min at 37°C, and the reaction was quenched using 10 mM ethylenediaminetetraacetic acid. All firefly luciferase mRNA was purified using RNeasy silica purification (Qiagen) following the manufacturer’s instructions. All GFP and PhiLOV3 mRNAs generated for cell study were purified using oligo-dT affinity purification [34], followed by ion-paired reverse phase chromatography if “RP” was mentioned. The purified GFP and PhiLOV3 mRNAs were buffer exchanged into sodium citrate buffer (pH 6.4) using Amicon filters (EMD Millipore).

mRNA yield

mRNA yield after IVT reaction was measured using HPLC with 0.1 ml CIMac Oligo dT18 column (Sartorius) [34]. The mRNA was bound to the column using high-salt sodium phosphate buffer and eluted using 10 mM Tris–HCl buffer and UV absorption was monitored at 260 nm. Purified mRNA of known concentration was used to create a standard curve of elution peak area. Material from IVT reactions was directly subjected to HPLC, and the elution peak area was used to determine yield. The kinetic study mRNA yield was measured using the Quant-iT RNA BR (Broad Range) Assay Kit (Thermo Fisher) following manufacturer’s instruction. Purified mRNA concentration was measured by UV spectrophotometry at 260 nm using a nanodrop spectrophotometer.

Capillary gel electrophoresis for full-length product

The Agilent Fragment Analyzer 5400 system was used to analyze RNA integrity. The mRNA samples were diluted to 0.1 mg/ml, and an Agilent DNF-471 RNA Kit (15 nt) was used according to the manufacturer’s instructions. ProSize data analysis software (Agilent) was used to generate a full-length product (FLP) profile.

dsRNA ELISA

dsRNA was quantified using a sandwich ELISA. EIA plates (96-well, Corning) were coated overnight at 4°C with J2 anti-dsRNA monoclonal antibody (Jena Bioscience) [35]. Plates were blocked with 1× PBST for 1 h, washed, and incubated with mRNA samples for 1 h at 37°C. Plates were rewashed and incubated overnight at 4°C with the anti-dsRNA monoclonal antibody K2 (Jena Bioscience) [35]. Plates were then washed and incubated with alkaline phosphatase-conjugated goat anti-mouse IgM (Jackson ImmunoResearch Laboratories) for 1 h at 37°C. After washing, the plates were developed using p-nitrophenyl phosphate substrate and read at 405 nm using a SpectraMax i3x plate reader. dsRNA concentration was calculated using a standard curve generated with a 142-bp dsRNA positive control (Jena Bioscience).

Transcription fidelity

mRNA samples were reverse transcribed using the Maxima H Minus with oligo dT/TSO RT workflow, following the manufacturer’s instructions (Thermo Fisher). Library preparation from the complementary DNA (cDNA) samples was performed by using the Nextera XT DNA Library Prep kit, following the manufacturer’s recommendations (Illumina). Briefly, cDNAs were tagmented by transposons, and Illumina adapters were added to cDNA fragments by limited polymerase chain reaction cycles. The sequencing libraries were multiplexed and clustered onto an Illumina flow cell. After clustering, the flow cell was loaded onto the Illumina NovaSeq instrument according to the manufacturer’s instructions. The samples were sequenced using a 2 × 150 bp paired-end configuration. Image analysis and base calling were conducted by the Illumina Control Software. Raw sequence data (.bcl files) generated from Illumina were converted into FASTQ files and demultiplexed using Illumina bcl2fastq 2.20 software. After confirming the quality of the reads, sequencing data were trimmed to remove possible adapter sequences at the 3′ end and nucleotides with poor quality before analysis using CLC Genomics Server. Trimmed sequence reads were aligned to the provided reference sequences. SNP/INDEL detection was performed using the probabilistic model with the CLC Genomics Server program. The sequencing was performed commercially by Azentas.

In vitro activity assay

In vitro activity of GFP mRNA was measured using an Agilent NovoCyte flow cytometer. Primary human hepatocytes were seeded in 48-well plates at a density of 1.25 × 105 cells per well. After 5 h, the cells were transfected with 325 ng mRNA per well using lipofectamine, following the manufacturer’s instructions (Thermo Fisher). The cells were dissociated 20 h after transfection using the TrypLE Express Enzyme (Thermo Fisher), resuspended in FACS buffer (5% fetal bovine serum in PBS) containing SYTOX Orange, and then assessed using the flow cytometer. Live cells were gated using forward scatter, side scatter, and SYTOX Orange parameters. Protein expression was quantified using a 488 nm excitation laser and a 525/45 bandpass filter in the B525 channel.

Immunogenicity assay

HEK-Lucia RIG-I cells were purchased from InvivoGen. The cells were cultured in DMEM medium supplemented with 10% HI-FBS (Thermo Fisher) with Normocin (100 µg/ml), penicillin–streptomycin (100 U/ml), blasticidin (30 µg/ml), and zeocin (100 µg/ml). The cells were seeded in 96-well plates at a density of 50 000 cells per well. The test media did not contain blasticidin and zeocin. After 5 h, the cells were transfected with 200 ng of mRNA per well using lipofectamine (Thermo Fisher) and incubated for 24 h. Following the manufacturer’s instructions, the fluorescence signal was quantified using QUANTI-Luc 4 reagent (InvivoGen), and plates were read on a SpectraMax i3x plate reader. dsRNA (50 ng,142-bp, Jena Bioscience) was used as the positive control, and lipofectamine was used as the negative control.

Results and discussion

Design of chimeric T7 RNA polymerase constructs to reduce dsRNA generation

Though multiple mechanisms are proposed for dsRNA generation, one popular putative mechanism is that T7 RNAP accepts RNA as a substrate instead of DNA, leading to dsRNA generation [15, 16, 28]. Under high-yield IVT conditions, RNA products are made in quantities that greatly outnumber the DNA template, which may increase the probability that T7 will initiate using an RNA template in this mixture. We hypothesized that further biasing the T7 RNAP toward binding DNA substrates would reduce RNA-dependent RNA polymerase activity and create a stronger dependence on T7 promoter-based transcription, resulting in reduced dsRNA generation (Fig. 1). Recently published works support this hypothesis [23, 36]. To test this hypothesis, a DNA-binding domain was tethered to the N-terminus of T7 RNAP using two linkers of varying lengths (Linker 1 of 33 amino acids and Linker 2 of 7 amino acids) (Fig. 2A). Modification of the C-terminus of the WT T7 RNA polymerase is generally not effective [37, 38]. Therefore, we initially did not plan to design a C-terminus chimeric T7 RNAP. However, recent work by Pu et al. has reported a novel T7 RNAP variant (G538E, V685A, A724S) that can accommodate C-terminal modifications [37]. As a result, we decided to design a chimeric T7 RNAP by tethering the DNA-binding domain to the C-terminus of the T7 RNAP variant using Linker 1. Sso7d from archaebacteria Sulfolobus solfataricus was selected as the DNA-binding candidate due to its high stability, small size, and ability to bind dsDNA without sequence preference [39, 40]. Since ribonuclease activity was reported for wild-type Sso7d, we used the Sso7d K13L variant, reported to eliminate this RNase activity [41]. In this work, we refer to the Sso7d K13L variant as Sso7d, as it is the only version of Sso7d used throughout this study.

Figure 1.

Figure 1.

Open in a new tab

Engineering chimeric T7 RNA polymerase to reduce dsRNA impurity. Wild-type T7 RNA polymerase accepts RNA as a substrate during high-yield IVT to generate dsRNA impurity [20, 21, 28]. We hypothesize that this mechanism can be perturbed by tethering a DNA-binding domain to the N-terminus of T7 RNA polymerase.

Figure 2.

Figure 2.

Open in a new tab

Sso7d tethered to the N-terminus of T7 RNA polymerase is functional. (A) Three chimeric variants were designed: two with Sso7d linked to the N-terminus of T7 RNAP with linkers of varying length and one C-terminus-linked variant. These variants were evaluated for (B) mRNA yield, (C) FLP, and (D) dsRNA generation at three different concentrations. Each condition was assessed using three replicates. Data are presented as mean values, with error bars representing the standard deviation.

Chimeric T7 RNAPs with an N-terminal Sso7d DNA binding domain are functional

To assess whether the chimeric constructs were functional, IVT reactions to generate mRNA encoding firefly luciferase protein were performed. The polymerase was evaluated at varying concentrations of 0.4, 0.8, and 1.6 µM, where 0.8 µM is the standard reaction condition. The IVT reaction product was subjected to DNase treatment and then purified using silica column purification. The mRNA generated was evaluated for yield, FLP profile, and dsRNA impurity content. The yield data show polymerase productivity at the end of a 90-minute IVT reaction, representing an endpoint measurement for WT T7 RNAP, which provides insight into whether these variants exhibit activity deficits at a time point relevant for maximal mRNA production. A kinetic study was conducted to ensure that the 90-minute reaction time point was not within the linear range (Supplementary Fig. S1A). This builds confidence that if a yield decrease is observed, it is due to an enzyme activity deficit significant enough to impact final yield at a typical manufacturing endpoint. The FLP profile provides an idea of the polymerase’s processivity and an estimate of the percentage of full-length mRNA product of the desired size. Both N-terminal chimeras (Sso7d Linker1 T7 RNAP, Sso7d Linker2 T7 RNAP) were active and obtained similar yield of mRNA and FLP profile as that of WT T7 RNAP, while the C-terminal chimeric polymerase (T7var Linker1 Sso7d) resulted in low yield of mRNA (Fig. 2B and C). This is consistent with previous findings that C-terminal modifications are not well tolerated by T7 RNAP [37, 38]. Since the C-terminal chimeric T7 RNAP resulted in low mRNA yield, it was not further studied.

We observed a negative correlation between increasing polymerase concentration and the level of dsRNA generated by the N-terminal chimeric T7 RNAP constructs. This trend was not observed for the WT T7 RNAP, which produced similar levels of dsRNA across all polymerase concentrations tested (Fig. 2D). The dsRNA impurity generated by Sso7d Linker1 T7 RNAP was lower than WT T7 RNAP only at 1.6 µM polymerase. In all other conditions tested, the chimeras generated dsRNA higher than WT T7 RNAP. This observation was in contrast to our initial hypothesis. Sso7d exhibits salt-dependent binding to DNA [39]. At higher polymerase concentrations, more volume of polymerase was added, and more salt was introduced to the IVT reaction via the enzyme storage buffer solution. We reasoned that the extra salt might have contributed to reduced dsRNA levels, and that chimeric T7 RNAP may require salts for optimal function. Next, to investigate this hypothesis, we tested the effects of different salt conditions on chimeric T7 RNAP performance.

Tethering a DNA-binding domain to T7 RNAP results in improved salt tolerance and dsRNA reduction

To understand how salt affects the performance of both WT and chimeric T7 RNAPs, the polymerases were evaluated at 75, 100, 125, 150, 175, and 200 mM NaCl conditions at 0.8 µM polymerase concentration. The salts present in the polymerase storage buffer were taken into account and balanced across all tested conditions. Aside from the varying NaCl concentrations, all other salt components in the reaction (Tris–HCl and MgCl2) were maintained at a constant level throughout the experiment. The WT T7 RNAP tolerated salts up to 150 mM NaCl, and the yield of mRNA decreased at 175 mM NaCl and higher. Linker 1 and Linker 2 tethered chimeras showed improved salt tolerance over WT T7 RNAP and yielded the same amount of mRNA up to 175 mM NaCl (Fig. 3A). The longer Linker 1 tethered chimera tolerated salt better than the shorter Linker 2 tethered chimera, which is evident from the higher yield for the Linker 1 construct at 200 mM NaCl. Overall, the FLP profile was similar for all constructs across all conditions, and the chimeras showed a small increase in the FLP profile at higher salt concentrations (Fig. 3B).

Figure 3.

Figure 3.

Open in a new tab

Chimeric T7 RNA polymerase improves salt tolerance and reduces dsRNA generation. WT T7 RNA polymerase and two chimeric T7 RNAP variants were tested at varying salt concentrations in IVT reactions. The (A) mRNA yield, (B) FLP, and (C) dsRNA impurity generation by chimera variants were compared to WT T7 RNAP. 0.02% dsRNA is the lower limit of detection for dsRNA ELISA. Each condition was assessed using two replicates. Data are presented as mean values, with error bars representing the range.

With increasing salt concentration, we observed a reduction in the dsRNA impurity up to 150 mM NaCl for WT T7 RNAP, whereas at higher salt concentrations, there was an increase in dsRNA impurity content (Fig. 3C). This increase in dsRNA production might be due to enhanced nonspecific interactions at higher salt concentrations [42, 43]. Both chimeric T7 RNAP constructs outperformed WT T7 RNAP in reducing the dsRNA impurity at all salt conditions tested (Fig. 3C). Sso7d Linker1 T7 RNAP performed better than Sso7d Linker2 T7 RNAP in reducing dsRNA and even decreased the impurity level to below the detection limit [lower limit of detection (LLOD) for the assay is 0.02%] at 150 mM NaCl. We observed an increase in dsRNA by Sso7d Linker2 T7 RNA polymerase at 200 mM NaCl. The Sso7d Linker1 T7 RNAP reduced dsRNA by approximately three-fold (to below the detection limit at 150 mM NaCl) compared to WT T7 RNAP (0.07% at 150 mM NaCl), while Sso7d Linker2 T7 RNAP reduced dsRNA by around two-fold to 0.035% at 150 mM NaCl. Thus, the dsRNA impurity reduction was not solely due to added salt. We observed better dsRNA reduction by chimeras than by WT T7 RNAP at high salt conditions (Fig. 3C).

The tethered DNA-binding domain is responsible for dsRNA reduction and polymerase processivity

To assess whether the tethering of the DNA-binding domain is essential for the observed salt tolerance and dsRNA reduction, we tested two additional T7 RNAP design controls. In the first control, the polymerase had an N-terminal extension of Linker 1 with no DNA-binding domain (NilSso7d-Linker1 T7). In the second control, an untethered Sso7d protein was added in trans to an IVT reaction with WT T7 RNAP (Sso7d trans; Fig. 4A). The Sso7d trans condition behaved similarly to WT T7 RNAP. The yield decreased at and above 175 mM NaCl (Fig. 4B). dsRNA was reduced to ∼0.06% at 150 mM NaCl but increased at and above 175 mM NaCl (Fig. 4D). Thus, this confirmed that supplying Sso7d in trans will not provide the salt tolerance or reduce dsRNA generation as observed with the chimeric polymerase. The NilSso7d-Linker1 T7 had similar dsRNA reduction trends as were observed for Sso7d Linker2 T7 RNAP earlier and reduced dsRNA to ∼0.045% at 150 mM NaCl (Fig. 4D). NilSso7d-Linker1 T7 did not further reduce dsRNA to below the detection limit at higher salt concentrations, but it prevented the increase in dsRNA we observed for WT T7 RNAP at higher salt concentrations. Though NilSso7d-Linker1 T7 had a minor advantage over WT T7 RNAP in dsRNA reduction, the FLP profile was lower compared to WT T7 RNAP and the other chimeric polymerases in all salt conditions we tested (Fig. 4C). NilSso7d-Linker1 T7 also had slightly decreased salt tolerance compared to WT T7 RNAP, as the yield decreased at 150 mM NaCl (Fig. 4B).

Figure 4.

Figure 4.

Open in a new tab

Tethered DNA-binding domain provides salt tolerance and dsRNA reduction. To evaluate whether the tethered Sso7d domain conferred the observed salt tolerance and dsRNA reduction, we evaluated (A) Sso7d Linker1 chimeric T7 RNA polymerase, chimeric T7 RNA polymerase with linker and no DNA-binding domain (NilSso7d-Linker1 T7), and WT T7 RNAP with Sso7d DNA-binding domain in trans. The (B) mRNA yield, (C) FLP, and (D) dsRNA reduction were compared against WT T7 RNA polymerase. Each condition was assessed using two replicates. Data are presented as mean values, with error bars representing the range.

During transcription, the N-terminal domain of T7 RNAP undergoes significant conformational changes during the transition from an initiation complex to an elongation complex [44, 45] and goes through multiple rounds of abortive transcription. The N-terminal extension might be adding resistance to the transition from the initiation phase to the elongation phase, and the T7 RNA polymerase needs much stronger template affinity to compensate for this deficit. This might help further bias DNA over RNA as a template and reduce RNA-dependent RNA polymerase activity. From these results, we propose that N-terminal extensions alone impact T7 RNAP kinetics, causing impacts to both yield and dsRNA generation, but the DNA-binding domain can rescue and even improve the DNA-dependent RNA polymerase activity under these conditions. Thus, a chimeric polymerase with a DNA-binding domain tethered to T7 RNA polymerase is essential to provide both dsRNA reduction and salt tolerance without affecting the FLP profile.

Effect of linker length and composition in chimeric T7 RNA polymerase performance

Since we observed earlier that the chimeric polymerases Sso7d Linker1 and Sso7d Linker2 performed differently, we wanted to understand the effect of linker length and composition on chimeric T7 RNA polymerase performance. We designed two sets of chimeric polymerases for these experiments: (i) chimeras with GS linkers of varying length but the same composition and (ii) chimeras with different linker compositions but similar length. The linkers of varying length evaluated were L_3, L_7 (Sso7d Linker2 T7 RNAP), L_15, L_33 (similar length as Sso7d Linker1 T7 RNAP), and L_51, where the number represents the amino acid length of the linker (Supplementary Table S1). We observed that with increasing linker length, the salt tolerance increased proportionally (Fig. 5A). Whereas all chimeras maintained the same yield at salt concentrations up to 175 mM NaCl, the yield decreased proportionally to linker length at 200 mM NaCl. L_33 and L_51 performed similarly and provided the best salt tolerance. Thus, a linker length of ∼30 amino acids was identified as providing the optimal salt tolerance by the Sso7d T7 RNA polymerase chimera. The same trend was observed for dsRNA reduction (Fig. 5B), for which we observed that only L_33 and L_51 decreased dsRNA below the detection limit at 150 mM NaCl (the same trend we observed for Sso7d Linker1 T7 in earlier experiments), whereas L_3, L_7, and L_15 performed similarly to WT T7 RNAP. At 175 mM NaCl, an increase in dsRNA was observed for WT T7 RNAP, while all chimeras lowered dsRNA proportional to increasing linker length. At 200 mM NaCl, we observed an increase in dsRNA for L_3 and L_7, while L_15, L_33, and L_51 maintained dsRNA below the detection limit.

Figure 5.

Figure 5.

Open in a new tab

Effect of linker length and linker composition. Chimeric variants with varying linker length (A, B) and linker composition (C, D) were evaluated for (A, C) mRNA yield and (B, D) dsRNA impurity generation. Each condition was assessed using two replicates. Data are presented as mean values, with error bars representing the range.

Linker length does affect the extent of dsRNA reduction by chimeric RNA polymerases. If the DNA-binding domain is too close to T7 RNA polymerase, it might provide too much hindrance for T7 RNA polymerase to initiate and obtain optimal processivity and thus the DNA-binding domain should not be too proximal to T7 RNA polymerase. Longer linkers are preferred with a linker length of a minimum of ∼30 amino acids, providing the best dsRNA reduction. All of the chimeric polymerases with different linker lengths tested produced mRNA with similar FLP profiles (Supplementary Fig. S2A).

To study the effect of linker composition, we used linkers of varying composition while maintaining an optimal length of ∼30 amino acids. Sso7d Linker1 T7 RNA polymerase served as a control (Linker 1 is a flexible XTEN family linker), and we also evaluated a GS linker of the same size (Linker_comp_1). Two additional linkers reported by Gräwe et al. [46] were used (Supplementary Table S1) and named Linker_comp_2 and Linker_comp_3. GS sequences were added at both ends to make them the same size as Linker 1. We did not observe a significant difference between the linkers in salt tolerance, as they all provided similar yields at all salt concentrations tested (Fig. 5C). Both Linker_comp_2 and Linker_comp_3 produced significantly higher dsRNA at low salt conditions, but at higher salt conditions, they performed similarly to that of Sso7d Linker1 T7 RNAP and the Linker_comp_1 (Fig. 5D). It is unclear why both Linker_comp_2 and Linker_comp_3 had higher dsRNA impurity than WT T7 RNA polymerase at the lower salt concentrations. Gräwe et al. reported Linker_comp_2 and Linker_comp_3 as semiflexible and rigid linkers, respectively [46]. We hypothesize that this might have caused the observed higher dsRNA at low salt conditions. At higher salt concentrations, the secondary structures of linkers might be disrupted [47, 48], resulting in them behaving like flexible linkers and thus leading to lower dsRNA generation. Further biochemical tests are required to confirm this hypothesis. Apart from this trend of higher dsRNA at low salt concentration for Linker_comp_2 and Linker_comp_3 linkers, at higher salt concentrations, all linker compositions performed similarly in dsRNA reduction. Sso7d Linker1 T7 RNAP and the Linker_comp_1 had marginally better FLP profiles compared to Linker_comp_2 and Linker_comp_3 linkers (Supplementary Fig. S2B).

Effect of different DNA-binding domains in chimeric T7 RNAP performance

We designed our chimeras with the Sso7d K13L variant as the DNA-binding domain and observed improved salt tolerance and reduced dsRNA generation relative to WT T7 RNAP. We wanted to determine if we could reproduce or improve the results with other DNA-binding domains. For this, we evaluated four different DNA-binding domains reported in recent studies to improve processivity and salt tolerance (Supplementary Table S2). Sto7d_K13L was used to improve processivity of reverse transcriptase [49], engrailed homeodomain from Drosophila melanogaster was used to improve catalytic activity of RNA ligase [50], NEQ150 DNA-binding protein from Nanoarchaeum equitans improved processivity and inhibitor tolerance of Taq DNA polymerase [51], and helix–hairpin–helix (HhH) domains of topoisomerase V from Methanopyrus kandleri were used to improve the replication efficiency of Phi29 DNA polymerase [52]. Though Gao et al. ended up using eight domain repeats to obtain improvement in activity [52], we only used the HhH motif 1 to keep the size of chimeric polymerase comparable to that of other chimeras. We used Linker 1 to tether all DNA-binding domains, as it was our best-performing linker in the Sso7d context. We observed that none of the DNA-binding domains improved salt tolerance for T7 RNAP and even performed worse than WT T7 RNAP at high salt concentrations (Fig. 6A). The FLP profile was slightly lower for the engrailed homeodomain construct at low salt concentrations but increased with increasing salt concentration (Fig. 6B). All other domains had a similar FLP profile as WT T7 RNA polymerase. All chimeric T7 RNA polymerase constructs reduced dsRNA compared to WT T7 RNA polymerase, but only Sso7d and Sto7d K13L prevented the increase in dsRNA at higher salt concentrations (Fig. 6C). Sso7d and Sto7d are from the same family [37], and both showed a similar trend in dsRNA reduction, but Sto7d K13L did not improve the salt tolerance.

Figure 6.

Figure 6.

Open in a new tab

Effect of different DNA-binding domains. Chimeric variants with different DNA-binding domains were evaluated for (A) mRNA yield, (B) FLP, and (C) dsRNA impurity generation. Each condition was assessed using two replicates. Data are presented as mean values, with error bars representing the range.

Though all chimeras tested had benefits over WT T7 RNAP in reducing dsRNA generation, they failed to provide the salt tolerance we observed with Sso7d. This was despite the fact that Sto7d reportedly has a higher affinity for DNA than Sso7d [39]. This again raises the question of optimal DNA binding strength. High salt conditions improved the yield of mRNA from the Sso7d chimeras (Figs 3A, 5A, and 6A); thus, stronger DNA binding might not be optimal. Future studies are required to understand the role of DNA binding strength of various domains in dsRNA generation. It should be noted that we used an optimal linker in the context of Sso7d, whereas the chimeras with other DNA-binding domains may need additional linker optimization. Gao et al. used multiple HhH domains to obtain improved processivity for DNA polymerase [52]. Thus, all these proteins require further optimization with linkers and number of DNA-binding domains to obtain optimal performance. Under the current conditions evaluated, Sso7d outperformed all other domains evaluated.

mRNA generated using chimeric T7 RNAP was effective at expressing protein and lowered immune response

We have shown that chimeric T7 RNA polymerase generates mRNA with reduced dsRNA impurity at high salt conditions. Next, we evaluated whether the mRNA generated by this chimeric polymerase is functional and its impact on immune stimulation. Since luciferase mRNA interferes with the immunogenicity assay, mRNA encoding GFP (∼1000 bp) and the phiLOV3 protein (∼600 bp) was used in this study. mRNA generated by Sso7d Linker1 chimeric T7 RNAP at a high salt condition (150 mM NaCl) was compared against mRNA generated by WT T7 RNAP under the standard IVT condition. WT T7 RNA polymerase yield decreased at 150 mM NaCl, which is why this condition was designated as the high salt condition (Supplementary Fig. S3).

We observed that chimeric T7 RNAP had similar yield to WT T7 RNAP for both GFP and phiLOV3 mRNA (Fig. 7A and D). mRNAs generated from all conditions were purified using oligo-dT affinity purification, and in some instances, the mRNA generated by WT T7 RNAP was further purified by reverse phase (designated as WT T7-RP) purification to remove dsRNA. WT T7 RNAP generated GFP mRNA with 0.31% dsRNA and phiLOV3 mRNA with 0.54% dsRNA, while chimeric T7 RNA polymerase reduced production of dsRNA for both mRNAs to below the LLOD and ∼0.05%, respectively (Fig. 7B and E). We did evaluate the performance of WT T7 RNAP at high salt concentration, and akin to our previous finding, WT T7 RNAP failed to reduce dsRNA below detection limit (Supplementary Fig. S4). Reverse phase purification reduced dsRNA levels for both mRNAs below the detection limit, as expected.

Figure 7.

Figure 7.

Open in a new tab

Evaluating mRNA generated by chimeric T7 RNAP at high salt IVT for potency and immune stimulation. GFP and PhiLOV3 mRNA were generated with WT T7 RNAP (with and without RP purification) and chimeric T7 RNAP at 150 mM NaCl. These mRNAs were evaluated for (A, D) yield, (B, E) dsRNA, (C, F) potency, and (G, H) immune stimulation. 0.02% is the LLOD for dsRNA ELISA. One hundred forty-two base pairs (bp) of dsRNA was used as a positive control, and lipofectamine was used as a negative control for immune stimulation (G, H). A single IVT reaction was performed for each condition. Yield and dsRNA levels were assessed with two technical replicates. For the cell-based assay, three independent wells were transfected per condition. Data are presented as mean values, with error bars indicating the range for yield and dsRNA, while error bars indicate the standard deviation for potency and immune stimulation.

To evaluate whether the mRNAs were functional, they were transfected into primary human hepatocyte cells and the expression was evaluated using flow cytometry after 20 h. For both GFP and PhiLOV3, all conditions tested had similar expression, with the GFP mRNA produced by the chimeric T7 RNAP having slightly higher GFP expression (Fig. 7C and F). Thus, mRNAs generated by chimeric T7 RNA polymerase at high salt concentration are functional and have similar or better expression than that produced by WT T7 RNA polymerase.

Immune stimulation was studied using HEK-Lucia RIG-I cells. These cells express human RIG-I receptors that detect dsRNA impurities and induce type 1 interferon (IFN) response. The cells contain a Lucia luciferase reporter gene under the control of IFN-inducible promoter, and IFN response can be studied by measuring secreted luciferase. We observed that the mRNA generated by chimeric T7 RNAP produced a significantly lower immune response than that generated by WT T7 RNAP and an IFN response similar to mRNA purified by reverse phase purification (Fig. 7G and H). Though high salt IVT conditions helped WT T7 RNAP to generate mRNA with reduced immunogenicity, it failed to eliminate immune stimulation (Supplementary Fig. S4).

For phiLOV3 mRNA, RP material gave a modestly higher IFN signal compared to chimeric T7 RNAP (Fig. 7H). Since there was significant dsRNA clearance for RP-purified phiLOV3 mRNA, we expected a lower signal from the RP-purified material. Piao et al. reported that RP purification can result in mRNA degradation and thus result in lower potency [26]. Dousis et al. also reported a similar trend in which the RP-purified material with lower dsRNA content provided a similar IFN response to that produced by WT T7 RNA polymerase [53]. Though we did not observe a significant difference in the FLP profile (Supplementary Fig. S5), the extra RP purification step might result in mRNA fragmentation, which might have triggered the higher IFN response [18, 54].

Chimeric T7 RNAP has the potential to support continuous mRNA manufacturing

We have shown that Sso7d Linker1 chimeric T7 RNAP has higher salt tolerance and generates mRNA with reduced immunogenicity at high salt conditions. The ability of chimeric T7 RNAP to perform IVT at high salt conditions might provide an added advantage to support the concept of continuous mRNA manufacturing. Continuous IVT involves the idea of reusing IVT components by recycling them back to the reaction vessel, which is otherwise lost as waste after the IVT reaction. The concept of continuous IVT is gaining popularity in the field as it can result in higher yield and make mRNA manufacturing cost-effective.

Oligo-deoxythymidylic acid (OdT) affinity column chromatography has become an industry standard in purifying poly-adenylated mRNA [55]. At high salt concentrations, the polyA tail of mRNA hybridizes with the OdT column, while the other IVT components are in the flow-through, and the bound mRNA is later eluted using a low conductivity buffer. Standard IVT is performed under low salt conditions and thus does not support direct binding of RNA to the OdT column [55, 56]. Therefore, a sample preparation step is required for OdT purification. Since chimeric Sso7d Linker1 T7 RNA polymerase improved salt tolerance, we compared the direct column binding of mRNA produced by chimeric T7 RNAP at 150 mM NaCl to the standard OdT WT T7 RNAP sample preparation for firefly luciferase mRNA. We observed that the chromatogram from direct binding of the mRNA produced by the chimeric polymerase at high salt conditions was the same as that from the standard OdT protocol [34] (Fig. 8A). The eluted material was evaluated for fidelity, dsRNA clearance, and FLP profile, and we observed a similar product profile and expected dsRNA clearance for chimeric T7 RNA polymerase (Fig. 8B). The sequence identity of mRNA transcribed by WT T7 RNAP was compared against mRNA transcribed by chimeric T7 RNA polymerase at high salt conditions using next-generation sequencing and no variance was detected (Supplementary Table S3). Thus, the increased salt tolerance of chimeric polymerase not only helps reduce dsRNA but can also help prevent the extra sample preparation step for OdT affinity chromatography and can make downstream processing seamless. This process has potential for further optimization to support continuous IVT by reusing the flowthrough from OdT step, which contains the enzymes and DNA template for IVT. Thus, chimeric T7 RNA polymerase may have utility for continuous mRNA manufacturing.

Figure 8.

Figure 8.

Open in a new tab

mRNA generated by high salt transcription promotes direct binding to OdT columns. Direct OdT purification of mRNA generated using Chimeric T7 RNAP (magenta) at high salt conditions (150 mM NaCl) was compared against OdT purification following standard sample preparation (black). The chromatogram at A260 was plotted for the two runs (A). The eluted material was tested for dsRNA content and FLP (B). A single chromatography experiment was conducted for each condition, and the resulting chromatograms were plotted. FLP and dsRNA levels were quantified using two technical replicates. Data are presented as mean values, with error bars representing the range.

Conclusion

In this work, we engineered a chimeric T7 RNA polymerase by tethering a DNA-binding domain to the N-terminus of T7 RNA polymerase. Chimeric T7 RNAP design optimization showed that a flexible linker of ∼30 amino acid length and Sso7d DNA-binding domain are optimal for chimeric T7 RNAP performance. The engineered chimeric polymerase exhibited improved salt tolerance and low dsRNA impurity generation over WT T7 RNAP. Salt played an important role in chimeric T7 RNAP performance, and at high salt concentrations, the chimeric T7 RNAP reduced dsRNA impurity to below the detection limit. Increasing salt concentration helped WT T7 RNAP reduce dsRNA to some degree, but a further increase in salt resulted in enriched dsRNA impurity and decreased yield. This observation of an increase in dsRNA with a decrease in yield at higher salt concentrations is reported for the first time and may be due to an enhanced nonspecific RNA interaction by salt-disrupted T7 RNAP conformation. mRNA generated using chimeric T7 RNA polymerase at high salt concentration displayed good translation and low immunogenicity, similar to mRNA purified by reverse phase chromatography. Thus, the innovation presented here eliminates the need for additional downstream efforts to reduce dsRNA impurities and helps scale up mRNA therapeutics production in a simple, cost-effective way at a raw material level. Thus, pairing chimeric T7 RNA polymerase with high salt concentrations during IVT ends up being the ideal solution for reducing dsRNA generation.

Supplementary Material

gkaf1124_Supplemental_File
gkaf1124_supplemental_file.pdf (586KB, pdf)

Acknowledgements

We would like to thank Shipra Malik for her valuable guidance in planning and analyzing the flow cytometry experiments. We would like to thank the members of the Process Development and RNA Technology teams at Verve Therapeutics and Beuning laboratory members for insightful discussions.

Author contributions: A.S. developed the concept with input from C.R.S. and C.J.C. A.S. and C.R.S. designed experiments. J.P., C.J.C., and P.J.B. supervised the study. A.S. performed all experiments. A.S., C.R.S., C.J.C., and P.J.B. wrote the manuscript, and all authors reviewed and approved the manuscript.

Contributor Information

Athul Sanjeev, Verve Therapeutics, Boston, MA 02215, United States; Department of Chemistry and Chemical Biology, Northeastern University, Boston, MA 02115, United States.

Caroline Reiss Summers, Verve Therapeutics, Boston, MA 02215, United States.

Jason Politi, Verve Therapeutics, Boston, MA 02215, United States.

Penny J Beuning, Department of Chemistry and Chemical Biology, Northeastern University, Boston, MA 02115, United States.

Christopher J Cheng, Verve Therapeutics, Boston, MA 02215, United States.

Supplementary data

Supplementary data is available at NAR online.

Conflict of interest

A.S., C.R.S., J.P., and C.J.C. are employees of Verve Therapeutics and hold equity in Verve Therapeutics. Verve Therapeutics has filed for patent protection related to various aspects of chimeric polymerase use, with A.S., C.R.S., and C.J.C. as the inventors.

Funding

Funding for this work and funding to pay the Open Access publication charges for this article were provided by Verve Therapeutics.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

References

  • 1. Parhiz  H, Atochina-Vasserman  EN, Weissman  D. mRNA-based therapeutics: looking beyond COVID-19 vaccines. Lancet. 2024;403:1192–204. 10.1016/S0140-6736(23)02444-3. [DOI] [PubMed] [Google Scholar]
  • 2. Rohner  E, Yang  R, Foo  KS  et al.  Unlocking the promise of mRNA therapeutics. Nat Biotechnol. 2022;40:1586–600. 10.1038/s41587-022-01491-z. [DOI] [PubMed] [Google Scholar]
  • 3. Pardi  N, Hogan  MJ, Porter  FW  et al.  mRNA vaccines—a new era in vaccinology. Nat Rev Drug Discov. 2018;17:261–79. 10.1038/nrd.2017.243. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Wadhwa  A, Aljabbari  A, Lokras  A  et al.  Opportunities and challenges in the delivery of mRNA-based vaccines. Pharmaceutics. 2020;12:102. 10.3390/pharmaceutics12020102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Chaudhary  N, Weissman  D, Whitehead  K. mRNA vaccines for infectious diseases: principles, delivery and clinical translation. Nat Rev Drug Discov. 2021;20:817–38. 10.1038/s41573-021-00283-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Zeng  C, Zhang  C, Walker  PG  et al.  Formulation and delivery technologies for mRNA vaccines. Curr Top Microbiol Immunol. 2022;440:71–110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Skerritt  JH, Tucek-Szabo  C, Sutton  B  et al.  The platform technology approach to mRNA product development and regulation. Vaccines. 2024;12:455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Baden  LR, El Sahly  HM, Essink  B  et al.  Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med. 2021;384:403–16. 10.1056/NEJMoa2035389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Polack  FP, Thomas  SJ, Kitchin  N  et al.  Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med. 2020;383:2603–15. 10.1056/NEJMoa2034577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Rosa  SS, Prazeres  DMF, Azevedo  AM  et al.  mRNA vaccines manufacturing: challenges and bottlenecks. Vaccine. 2021;39:2190–200. 10.1016/j.vaccine.2021.03.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Tabor  S. Expression using the T7 RNA polymerase/promoter system. Curr Protoc Mol Biol. 2001;Chapter 16:Unit 16.2. 10.1002/0471142727.mb1602s11. [DOI] [PubMed] [Google Scholar]
  • 12. Kartje  ZJ, Janis  HI, Mukhopadhyay  S  et al.  Revisiting T7 RNA polymerase transcription in vitro with the Broccoli RNA aptamer as a simplified real-time fluorescent reporter. J Biol Chem. 2021;296:100175. 10.1074/jbc.RA120.014553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Sahin  U, Karikó  K, Türeci  Ö. mRNA-based therapeutics—developing a new class of drugs. Nat Rev Drug Discov. 2014;13:759–80. 10.1038/nrd4278. [DOI] [PubMed] [Google Scholar]
  • 14. Triana-Alonso  FJ, Dabrowski  M, Wadzack  J  et al.  Self-coded 3'-extension of run-off transcripts produces aberrant products during in vitro transcription with T7 RNA polymerase. J Biol Chem|. 1995;270:6298–307. 10.1074/jbc.270.11.6298. [DOI] [PubMed] [Google Scholar]
  • 15. Konarska  MM, Sharp  PA. Replication of RNA by the DNA-dependent RNA polymerase of phage T7. Cell. 1989;57:423–31. 10.1016/0092-8674(89)90917-3. [DOI] [PubMed] [Google Scholar]
  • 16. Cazenave  C, Uhlenbeck  OC. RNA template-directed RNA synthesis by T7 RNA polymerase. Proc Natl Acad Sci USA. 1994;91:6972–6. 10.1073/pnas.91.15.6972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Rong  M, Durbin  RK, McAllister  WT. Template strand switching by T7 RNA polymerase. J Biol Chem. 1998;273:10253–60. 10.1074/jbc.273.17.10253. [DOI] [PubMed] [Google Scholar]
  • 18. Hartmann  G. Nucleic acid immunity. Adv Immunol. 2017;133:121–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Saito  T, Gale  M. Differential recognition of double-stranded RNA by RIG-I-like receptors in antiviral immunity. J Exp Med. 2008;205:1523–7. 10.1084/jem.20081210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Chen  Y-G, Hur  S. Cellular origins of dsRNA, their recognition and consequences. Nat Rev Mol Cell Biol. 2022;23:286–301. 10.1038/s41580-021-00430-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Karikó  K, Muramatsu  H, Ludwig  J  et al.  Generating the optimal mRNA for therapy: HPLC purification eliminates immune activation and improves translation of nucleoside-modified, protein-encoding mRNA. Nucleic Acids Res. 2011;39:e142. 10.1093/nar/gkr695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Mu  X, Greenwald  E, Ahmad  S  et al.  An origin of the immunogenicity of in vitro transcribed RNA. Nucleic Acids Res. 2018;46:5239–49. 10.1093/nar/gky177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Cavac  E, Ramírez-Tapia  LE, Martin  CT. High-salt transcription of DNA cotethered with T7 RNA polymerase to beads generates increased yields of highly pure RNA. J Biol Chem. 2021;297:100999. 10.1016/j.jbc.2021.100999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Ziegenhals  T, Frieling  R, Wolf  P  et al.  Formation of dsRNA by-products during in vitro transcription can be reduced by using low steady-state levels of UTP. Front Mol Biosci. 2023;10:1291045. 10.3389/fmolb.2023.1291045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Martínez  J, Lampaya  V, Larraga  A  et al.  Purification of linearized template plasmid DNA decreases double-stranded RNA formation during IVT reaction. Front Mol Biosci. 2023;10:1248511. 10.3389/fmolb.2023.1248511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Piao  X, Yadav  V, Wang  E  et al.  Double-stranded RNA reduction by chaotropic agents during. Mol Ther Nucleic Acids. 2022;29:618–24. 10.1016/j.omtn.2022.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Sarcar  SN, Miller  DL. A specific, promoter-independent activity of T7 RNA polymerase suggests a general model for DNA/RNA editing in single subunit RNA polymerases. Sci Rep. 2018;8:13885. 10.1038/s41598-018-32231-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Gholamalipour  Y, Karunanayake Mudiyanselage  A, Martin  CT. 3' end additions by T7 RNA polymerase are RNA self-templated, distributive and diverse in character—RNA-Seq analyses. Nucleic Acids Res. 2018;46:9253–63. 10.1093/nar/gky796. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Baiersdörfer  M, Boros  G, Muramatsu  H  et al.  A facile method for the removal of dsRNA contaminant from in vitro-transcribed mRNA. Mol Ther Nucleic Acids. 2019;15:26–35. 10.1016/j.omtn.2019.02.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Cho  E, Namgung  J, Lee  JS  et al.  Mesoporous silica particle as an RNA adsorbent for facile purification of in vitro-transcribed RNA. Int J Mol Sci. 2023;24:12408. 10.3390/ijms241512408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Levanova  A, Poranen  MM. Application of steric exclusion chromatography on monoliths for separation and purification of RNA molecules. J Chromatogr A. 2018;1574:50–9. 10.1016/j.chroma.2018.08.063. [DOI] [PubMed] [Google Scholar]
  • 32. Gagnon  P, Goričar  B, Peršič  Š  et al.  Two new capture options for improved purification of large mRNA. Cell Gene Ther Insights. 2020;6:1035–46. 10.18609/cgti.2020.114. [DOI] [Google Scholar]
  • 33. Andries  O, Mc Cafferty  S, De Smedt  SC  et al.  N¹-Methylpseudouridine-incorporated mRNA outperforms pseudouridine-incorporated mRNA by providing enhanced protein expression and reduced immunogenicity in mammalian cell lines and mice. J Control Release. 2015;217:337–44. 10.1016/j.jconrel.2015.08.051. [DOI] [PubMed] [Google Scholar]
  • 34. Mencin  N, Štepec  D, Margon  A  et al.  Development and scale–up of oligo–dT monolithic chromatographic column for mRNA capture through understanding of base–pairing interactions. Sep Purif Technol. 2023;304:122320. 10.1016/j.seppur.2022.122320. [DOI] [Google Scholar]
  • 35. Schönborn  J, Oberstrass  J, Breyel  E  et al.  Monoclonal antibodies to double-stranded RNA as probes of RNA structure in crude nucleic acid extracts. Nucleic Acids Res. 1991;19:2993–3000. 10.1093/nar/19.11.2993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. MalagodaPathiranage  K, Banerjee  R, Martin  CT. A new approach to RNA synthesis: immobilization of stably and functionally co-tethered promoter DNA and T7 RNA polymerase. Nucleic Acids Res. 2024;52:10607–18. 10.1093/nar/gkae599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Pu  J, Disare  M, Dickinson  BC. Evolution of C-terminal modification tolerance in full-length and split T7 RNA polymerase biosensors. ChemBioChem. 2019;20:1547–53. 10.1002/cbic.201800707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Gardner  LP, Mookhtiar  KA, Coleman  JE. Initiation, elongation, and processivity of carboxyl-terminal mutants of T7 RNA polymerase. Biochemistry. 1997;36:2908–18. 10.1021/bi962397i. [DOI] [PubMed] [Google Scholar]
  • 39. Kalichuk  V, Béhar  G, Renodon-Cornière  A  et al.  The archaeal “7 kDa DNA-binding” proteins: extended characterization of an old gifted family. Sci Rep. 2016;6:37274. 10.1038/srep37274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Wang  Y, Prosen  DE, Mei  L  et al.  A novel strategy to engineer DNA polymerases for enhanced processivity and improved performance in vitro. Nucleic Acids Research. 2004;32:1197–207. 10.1093/nar/gkh271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Shehi  E, Serina  S, Fumagalli  G  et al.  The Sso7d DNA-binding protein from Sulfolobus solfataricus has ribonuclease activity. FEBS Letters. 2001;497:131–6. 10.1016/S0014-5793(01)02455-3. [DOI] [PubMed] [Google Scholar]
  • 42. Athavale  SS, Ouyang  W, McPike  MP  et al.  Effects of the nature and concentration of salt on the interaction of the HIV-1 nucleocapsid protein with SL3 RNA. Biochemistry. 2010;49:3525–33. 10.1021/bi901279e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Spencer  DS, Xu  K, Logan  TM  et al.  Effects of pH, salt, and macromolecular crowding on the stability of FK506-binding protein: an integrated experimental and theoretical study. J Mol Biol. 2005;351:219–32. 10.1016/j.jmb.2005.05.029. [DOI] [PubMed] [Google Scholar]
  • 44. Yin  YW, Steitz  TA. Structural basis for the transition from initiation to elongation transcription in T7 RNA polymerase. Science. 2002;298:1387–95. 10.1126/science.1077464. [DOI] [PubMed] [Google Scholar]
  • 45. Steitz  TA. The structural changes of T7 RNA polymerase from transcription initiation to elongation. Curr Opin Struct Biol. 2009;19:683–90. 10.1016/j.sbi.2009.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Gräwe  A, Ranglack  J, Weyrich  A  et al.  iFLinkC: an iterative functional linker cloning strategy for the combinatorial assembly and recombination of linker peptides with functional domains. Nucleic Acids Res. 2020;48:e24. 10.1093/nar/gkz1210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Kohn  WD, Kay  CM, Hodges  RS. Salt effects on protein stability: two-stranded alpha-helical coiled-coils containing inter- or intrahelical ion pairs. J Mol Biol. 1997;267:1039–52. 10.1006/jmbi.1997.0930. [DOI] [PubMed] [Google Scholar]
  • 48. Basu  T, Das  S, Majumdar  S. Elucidating the influence of electrostatic force on the re-arrangement of H-bonds of protein polymers in the presence of salts. Soft Matter. 2024;20:2361–73. 10.1039/D3SM01440A. [DOI] [PubMed] [Google Scholar]
  • 49. Oscorbin  IP, Wong  PF, Boyarskikh  UA  et al.  The attachment of a DNA-binding Sso7d-like protein improves processivity and resistance to inhibitors of M-MuLV reverse transcriptase. FEBS Letters. 2020;594:4338–56. 10.1002/1873-3468.13934. [DOI] [PubMed] [Google Scholar]
  • 50. Tong  CL, Kanwar  N, Morrone  DJ  et al.  Nature-inspired engineering of an artificial ligase enzyme by domain fusion. Nucleic Acids Res. 2022;50:11175–85. 10.1093/nar/gkac858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Olszewski  M, Śpibida  M, Bilek  M  et al.  Fusion of Taq DNA polymerase with single-stranded DNA binding-like protein of Nanoarchaeum equitans—expression and characterization. PLoS One. 2017;12:e0184162. 10.1371/journal.pone.0184162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Gao  Y, He  Y, Chen  L  et al.  Chimeric Phi29 DNA polymerase with helix–hairpin–helix motifs shows enhanced salt tolerance and replication performance. Microb Biotechnol. 2021;14:1642–56. 10.1111/1751-7915.13830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Dousis  A, Ravichandran  K, Hobert  EM  et al.  An engineered T7 RNA polymerase that produces mRNA free of immunostimulatory byproducts. Nat Biotechnol. 2023;41:560–8. 10.1038/s41587-022-01525-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Jung  S, von Thülen  T, Yang  I  et al.  A ribosomal RNA fragment with 2',3'-cyclic phosphate and GTP-binding activity acts as RIG-I ligand. Nucleic Acids Res. 2020;48:10397–412. 10.1093/nar/gkaa739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Feng  X, Su  Z, Cheng  Y  et al.  Messenger RNA chromatographic purification: advances and challenges. J Chromatogr A. 2023;1707:464321. 10.1016/j.chroma.2023.464321. [DOI] [PubMed] [Google Scholar]
  • 56. Green  MR, Sambrook  J. Isolation of Poly(A)+ messenger RNA using magnetic oligo(dT) beads. Cold Spring Harb Protoc. 2019;2019:711–15. https://cshprotocols.cshlp.org/content/2019/10/pdb.prot101733. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

gkaf1124_Supplemental_File
gkaf1124_supplemental_file.pdf (586KB, pdf)

Data Availability Statement

The data underlying this article are available in the article and in its online supplementary material.

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES