Fluorogenic RNA Aptamer-Based Activity Quantification of T7 RNA Polymerase

Abstract

T7 RNA polymerase (T7 RNAP) is a central molecular biology tool that is broadly used for biomedical research and the RNA therapeutic industry. T7 RNAP and its engineered variants provide valuable tools in synthetic biology and gene expression regulation. T7 RNAP activity unit is traditionally quantified by a radioactivity-based polymerization assay, which is often not accessible to many laboratories and raises safety concerns because of exposure to radioactivity. Here, we developed a fluorescence-based in vitro transcription (IVT) assay that enables facile and quantitative determination of T7 RNAP activity units, using the Broccoli RNA aptamer as a transcriptional reporter and commercially available T7 RNAPs as standard references, which are readily accessible materials for most biochemical laboratories. This method will improve the consistency and reproducibility assessment of T7 RNAP prepared from different batches or different laboratories and benefit research communities that rely on the robust performance of RNA polymerases.

Graphical Abstract

Introduction

Since first discovered from T7-infected Escherichia coli cells in 1970,¹ T7 RNA polymerase (RNAP) has been broadly applied for RNA biology research, synthetic biology, and RNA-based therapeutics.^2–7 T7 RNAP functions as a monomeric enzyme that specifically recognizes a double-stranded T7 promoter sequence in the DNA template and catalyzes the 5′−3′ transcription of RNA product in the presence of Mg²⁺ and four ribonucleotide triphosphates (rNTPs).^1,8–12 T7 RNAP exhibits strong promoter specificity^13,14 and generates high yield of RNA products under optimized conditions during in vitro transcription (IVT).^4,6,15 There has been broad and rapidly growing interest in the in vitro applications of T7 RNAP and its engineered variants, including synthesis of RNAs for structural biology and molecular biology research,^2–4,16 synthesis of functional small RNAs such as guide RNAs used in CRISPR technology,¹⁷ large-scale production of therapeutic mRNAs, “SELEX” platform for the rapid selection of RNA aptamers,^18,19 biosensors,^20,21 cell-free protein synthesis,²² and several protein-directed evolution platforms.^23–25 Protein engineering and directed evolution efforts on T7 RNAP have further diversified its biophysical and biochemical properties such as enhanced processivity,^26,27 elevated thermostability,^28,29 orthogonal promoter recognition,³⁰ minimized dsRNA byproducts,⁷ and the incorporation of chemically modified rNTPs,³¹ which further broaden the application scope of T7 RNAP-based enzymes.

Despite there being multiple suppliers for T7 RNAP, commercially available T7 RNAPs often do not meet the quantity and activity demand or are sometimes cost-prohibitive for the desired application in both academic and industrial laboratories. Over the decades, T7 RNAP and its variants have been widely prepared in laboratories to support diverse research applications.^6,23,32–35 T7 RNAP with an N-terminal His-tag can be overexpressed in BL21 (DE3) cells and further purified via Ni-NTA affinity chromatography; additional purifications by ion-exchange and/or size-exclusion chromatography can further help increase the purity or activity of the purified enzyme.^6,34,36 The activity of the purified T7 RNAP was then tested by IVT assays. The activity unit of T7 RNAP was originally defined as the amount of enzyme catalyzing the incorporation of 1 nmol of labeled nucleotide in 1 h at 37 °C.¹ To quantify the activity unit of T7 RNAP, 360 μM T7 phage DNA was used as a template to transcribe RNA by RNAP using 400 μM rATP, rGTP, rCTP, and rUTP in a 40 mM Tris-HCl pH 8.0 buffer with 20 mM MgCl₂, 0.01 mM 2-mercaptoethanol (β-ME), and 1 mg/mL BSA for 10 min at 37 °C, where one nucleotide of the four is labeled with ³H or ³²P (2000−20,000 c.p.m. per nmol). The amount of produced RNA is proportional to the amount of enzyme used. The RNA product is then precipitated under acidic conditions and quantified by radioactivity.¹ Currently, the commercially available T7 RNAPs such as the one from the New England Biolabs (M0251L) are supplied in activity units (U/μL), the activities of which are determined using a radioactivity-based IVT assay with slight differences in DNA templates and reaction conditions.

However, the classic method for polymerase activity quantification utilizes radioactively labeled materials, which are not easily accessible to many laboratories and present health threats to the exposed environment and personnel.^14,37–39 Notably, T7 RNAPs purified from laboratories are often reported in molecular concentrations of purified T7 RNAPs (molar or mg/mL) rather than activity units.^4,6 Lack of quantified activities pose challenges in assessing consistency of T7 RNAP purified from different protocols or batches. To circumvent the major limitation posed by radioactivity, strategies have been developed to report T7 RNAP activity independent from the radiolabeled materials. For example, Vassiliou et al. reported a colorimetric RNAP assay that monitors the transcription reaction using p-nitrophenyl caged GTP and ATP at the γ-phosphate position as substrates, and the released p-nitrophenylpyrophosphate is instantaneously converted to the chromogenic p-nitrophenylate by alkaline phosphatase when a caged NTP gets incorporated into the nascent RNA by RNAP.⁴⁰ Blair et al. reported a “broken-beacon” platform to quantify the RNA synthesis rate by RNAP, where a fluorescent oligonucleotide probe is first annealed with a quencher oligonucleotide with minimal fluorescence; the fluorescent oligonucleotide probe binds with the RNA product, which contains a longer complementary sequence compared to the quencher probe and thus generates fluorescence as RNA is transcribed.⁴¹ Recently, a transcription-translation system was reported to quantify T7 RNAP, where T7 RNAP activity was linearly correlated with the translated green fluorescent protein (GFP) from the transcription-translation system that contains bacterial cell extract, NTPs, amino acids, an energy source, and a T7 promoter-driven expression plasmid for GFP.⁴² While these strategies provide nonradioactive options for RNAP activity measurement, they require the use of non-natural substrates for RNAP,⁴⁰ expensive materials, such as fluorophore-labeled oligonucleotides,⁴¹ multiple-step reactions, or assay components with high variability such as cellular extracts.⁴²

Here, we report a single-step, nonradioactive, and cost-effective assay to quantify the activity units of in-house purified T7 RNAP, which does not require any modified oligonucleotides and is highly accessible to any biochemistry laboratory. Briefly, the IVT reactions were performed based on a short double-stranded DNA template containing a T7 promoter sequence¹⁴ followed by the Broccoli aptamer DNA template by RNAP in the presence of a small-molecule dye, DFHBI1T.⁴³ The transcribed Broccoli RNA aptamer binds the fluorophore DFHBI-1T and generates fluorescence.^43,44 This Broccoli IVT assay was used to monitor and optimize RNA yields for IVT by T7 RNAP¹⁵ and to serve as a critical reporter assay for protein-directed evolution platforms.^23–25,45We found that the observed rate of fluorescence enhancement in the early phase of the IVT assay linearly correlates with the number of activity units of the T7 RNAP. We then applied the relationship between the observed rate of fluorescence enhancement and commercial T7 RNAP units to quantify the activity of in-house purified T7 RNAP. Lastly, we applied commercial and in-house-purified T7 RNAPs with normalized units in preparing structurally distinct RNA transcripts under preparative IVT conditions and revealed the consistency of RNA yields produced by T7 RNAPs from difference sources.

Materials and Methods

Dna Oligonucleotides

Single-stranded DNA (ssDNA) templates encoding the Broccoli RNA aptamer and the tRNA^Gly-like sequence, and PCR primers, (Table S1) were purchased from Integrated DNA Technologies (IDT). Both transcripts included a T7 promoter sequence upstream of the RNA coding region. The tRNA^Gly-like DNA template contained an extended 3′ sequence downstream of the tRNA gene.

Commercial Enzymes

Phusion DNA polymerase (F530L) was purchased from Thermo Scientific and used for the PCR amplification of DNA templates. Commercial T7 RNA Polymerase (M0251L) was purchased from New England Biolabs (NEB) and used as the reference enzyme for activity calibration. Thermo Scientific T7 RNA polymerase (EP0111) was purchased from Thermo Scientific.

Expression and Purification of T7 RNAP

T7 RNAP (UniProt: P00573) was overexpressed in E. coli BL21 (DE3) cells carrying a pBb vector encoding N-terminally 6× His-tagged T7 RNAP under the control of the lacUV5 promoter.²⁵ Cultures were grown at 37 °C to an OD600 of 0.6–0.8, induced with 0.5 mM isopropyl β-d-1-thiogalactopyranoside and incubated overnight at 30 °C. Cells were harvested by centrifugation and lysed by sonication in a lysis buffer containing 50 mM Tris-HCl pH 7.5, 1 M NaCl, 10 mM TCEP, and 20% glycerol, supplemented with a protease inhibitor cocktail (Thermo Scientific, A32965). The protein was purified using His60 Ni Superflow Resins (Takara Bio USA, 635660) and eluted with a gradient of 5 mM to 0.5 M imidazole in the lysis buffer. Fractions were examined by the SDS-PAGE gel, and fractions containing high purity T7 RNAPs were pooled, concentrated, and exchanged into a storage buffer containing 50 mM Tris-HCl pH 7.0, 100 mM NaCl, 1 mM EDTA, 10 mM DTT, and 50% glycerol (Figure S1A). The purified enzyme was aliquoted, flash frozen under liquid N₂, and stored at −80 °C until usage. Once an aliquot is thawed for usage, it will be stored at −20 °C if not consumed all at once, which should avoid freeze-and-thaw cycle in the storage buffer containing 50% glycerol. Yet the activity of the enzyme may decrease with increased storage duration. In-house prepared T7 RNAPs tested in this study are all freshly thawed aliquots from −80 °C freezer.

Preparation of DNA Templates for IVT

DNA templates for IVT were prepared by PCR. PCR amplification was performed using 1× Phusion HF buffer, 0.02 μM single-stranded template DNA (Table S1), 0.5 μM forward and reverse primers, 200 μM dNTPs, and 1.0 U of Phusion polymerase in a total volume of 50 μL. PCRs were carried out in thermocyclers for 30 cycles; each cycle included a denaturing step (98 °C for 15 s), an annealing step (61 or 62 °C for 30 s for Broccoli or tRNA^Gly-like templates, respectively), and an elongation step at 72 °C for 22 s. PCR products were resolved on 1% TAE agarose electrophoresis gels and purified using a QIAGEN DNA gel purification kit (QIAGEN 28704) (Figure S1B). Purified DNA templates were eluted in nuclease-free water, and concentrations were determined by absorbance at 260 nm using a NanoDrop spectrophotometer. The purified DNA template was analyzed on a 3% TBE agarose electrophoresis gel to confirm the DNA product and assess its purity (Figure S1C).

Broccoli IVT Assay

The Broccoli IVT reaction was performed to a final volume of 20 μL in a 1× IVT reaction buffer (40 mM Tris-HCl pH 7.9, 6 mM MgCl₂,1 mM DTT, 2 mM spermidine) and contained 2 mM of each rNTP (NEB N0466S), 0.05 μM Broccoli DNA template, 50 μM DFHBI-1T (LuceRNA, Inc. 410–1 mg), and 0.25% DMSO. An additional 25 mM MgCl₂ and 10 mM DTT were added to each reaction. Variable units of commercial or the in-house purified T7 RNAP were added depending on the purpose of the assay. Once prepared, Broccoli IVT reactions were transferred to a black-edge and clear-bottom 384-well plate (Cellvis P384–1.5H-N) and monitored using a Neo2 plate reader (BioTek, Inc.) for one hour at 37 °C at an interval of 1 read per minute, with excitation and emission wavelengths at 472 and 507 nm.^25,43 Broccoli IVT assays in this study were performed by at least three biological replicates and two technical replicates unless otherwise noted. A biological replicate is defined as a reaction in which all reaction components are independently added to the reaction mixture including the enzyme. The same batch of purified or commercial enzyme was used among different biological replicates. After all reaction components were added, each biological replicate reaction mixture was immediately split into two equal-volume reactions and transferred onto two separate wells on the 384-well plate for fluorescence recording. These two split reactions were referred to as two technical replicates. Each biological replicate contains two technical replicates, and three biological replicates were performed for each IVT condition. Therefore, fluorescence response data of n = 6 total replicates were recorded for each Broccoli IVT condition unless otherwise noted.

Observed Initial Rate of the Fluorescence Response of the Broccoli IVT Assay

We define the observed initial rate of the fluorescence enhancement as the slope of the rapidly increasing initial linear phase of the fluorescence response. To derive the slope, we performed simple linear regressions on the fluorescence responses within the “sliding time windows” across the early phase (5−21 min) of response curves. The width of the time window varies from 8 to 12 min. We then plot the correlation coefficient R² of the linear regression across all sliding windows (Figure S2A), and the best linear-fit window was selected based on the highest R². The slope of the fitted line within the best linear-fit time window was used as the observed initial rate. In cases where multiple ranges produced identical R² values, the range with the smaller regression standard error (Sy.x) was selected. The observed initial rate for a specific IVT reaction is calculated by taking the average of the observed initial rates for all biological and technical replicates, with errors calculated by taking the standard deviation of all replicates (Figures S2B and S3). This analysis allows us to systematically determine the observed initial rate guided by the R² of the linear fit.

Protein Concentration Determination for Activity Normalization

Protein concentrations of commercial and in-house T7 RNAP were determined using the Qubit protein assay following the manufacturer’s protocol (Invitrogen Q33211). Three protein standards provided with the kit were measured to generate the calibration curve (Figure S4). 10 μL of commercial T7 RNAP was mixed with 190 μL of Qubit working solution and measured by the Qubit 4 Fluorometer (Invitrogen Q33226). The in-house purified T7 RNAP was in a much higher concentration than the commercial T7 RNAP; therefore, 1 μL of in-house T7 RNAP was mixed with 199 μL of Qubit working solution. The measurement readings for both commercial and in-house T7 RNAPs fell within the linear calibration range of the Qubit protein assay (Figure S4A).

Preparative IVT Assay for tRNA^Gly-like RNA

For applications of the commercial and in-house T7 RNAPs, IVT reactions were assembled to a final volume of 20 μL containing the following final concentrations: 1× IVT reaction buffer, 2 mM of each rNTP (NEB N0466S), 400 ng DNA template, and 5 U/μL of commercial or in-house T7 RNAP. Reactions were incubated at 37 °C for 3 h, followed by 85 °C for 15 min. To assess RNA yields, IVT products were analyzed on 10% denaturing polyacrylamide (PAGE) gels containing 8 M urea. After completion of the reaction, 1 μL of the IVT reaction mixture was diluted with 9 μL of nuclease-free water. Subsequently, 4 μL of the diluted IVT was mixed with 1 μL of either nuclease-free water (control), DNase I (Thermo Scientific EN0521), or 0.5 mM of NaOH stock solution. The mixtures were incubated at 37 °C for 30 min, followed by 95 °C for 15 min. After treatment, 2 μL of each reaction was combined with 14 μL of formamide (Thermo Scientific Chemicals 327235000) and heated at 95 °C for 10 min, immediately chilled on ice for 2 min. 5 μL of the denatured sample was loaded onto the denaturing PAGE gel. 10% denaturing PAGE gels were prerun at 200 V in 1× TBE buffer for 30 min, and run at 200 V for 45 min after sample loading. Following electrophoresis, gels were stained with SYBR Gold nucleic acid gel stain (Invitrogen S11494) and imaged using a Bio-Rad ChemiDoc System. A low range ssRNA ladder (NEB N0364S) was loaded to assess RNA product size during gel electrophoresis.

Results

Calibration of T7 RNAP Activity via a Broccoli IVT Assay

To establish a fluorescence-based assay for quantifying T7 RNAP activity, we employed the Broccoli aptamer (Table S1) as a transcriptional reporter, which binds with DFHBI-1T and generates fluorescence upon binding.⁴³ Transcription of the DNA template containing the T7 promoter and the Broccoli sequence in the presence of DFHBI-1T directly couples RNA synthesis to fluorescence generation, enabling continuous and real-time monitoring of transcriptional output in solution (Figure 1A).^15,23 The double-stranded Broccoli DNA template (74 bp) was prepared by polymerase chain reaction (PCR) and purified using an agarose gel (Figure S1 and Materials and Methods). We hypothesized that the rate of RNA synthesis in IVT will increase with an increasing amount of active T7 RNAP. To test this hypothesis, we used the commercially available T7 RNAP enzyme purchased from NEB with the enzyme provided in 50 U/μL, where one unit is defined as the amount of enzyme that incorporates 1 nmol of ATP into acid-precipitated RNA product in a 50 μL IVT reaction in 1 h at 37 °C. Broccoli IVT reactions were carried out in a specific assay condition that included 0.05 μM of Broccoli DNA template, 2 mM of each rNTP, 40 mM Tris-HCl pH 7.9, 31 mM MgCl₂,11 mM DTT, 2 mM spermidine, 50 μM DFHBI-1T, 0.25% DMSO, and a variable amount of T7 RNAP (Cat. No. M0251L) from NEB. Note that we added additional MgCl₂ compared to the typical amount used for IVT,^1,46 considering the folding of Broccoli depends on Mg²⁺ binding.^15,23,43 A minute amount of DMSO was included, as the stock solution of DFHBI-1T contains DMSO as solvent. We varied the amount of T7 RNAP within this range: 1.25, 2.5, 5, 7.5, and 10 U/μL in different IVT reactions. After the addition of all the above components, the fluorescence signal (λ_excitation = 472 nm; λ_emission = 507 nm) generated from IVT reactions was monitored in the plate reader at 37 °C for 60 min (Materials and Methods). We observed that an increasing enzyme amount resulted in steeper slopes of the fluorescence response curves from the IVT reactions (Figure 1B).

Figure 1.

Broccoli IVT assay for quantifying the activity unit of T7 RNAP. (A) Schematic of Broccoli aptamer transcription from a DNA template. (B) Time-resolved fluorescence response curves for Broccoli IVT assays, with increasing commercial T7 RNAP activity (U/μL). Shown are averaged fluorescence intensities at an emission wavelength of 507 nm at each time point, and error bars represent standard deviations (s.d.) of n = 6 replicates (3 biological replicates and each biological replicate contains 2 technical replicates). The same set of replicates were performed for Broccoli IVT assays unless otherwise noted. (C) Calibration curve for T7 RNAP activity based on observed initial rates of the Broccoli IVT assay against commercial T7 RNAP activity units. Shown are averaged observed initial rates at each enzyme concentration of n = 6 replicates. Note that the 8 U/μL condition was performed with n = 8 replicates (4 biological replicates and each biological replicate contains 2 technical replicates). (D) Fluorescence response curves for the validation of the calibration curve using NEB T7 RNAP. Data are reported as average and s.d. from n = 6 replicates. (E) Calculated T7 RNAP activity units by the calibration curve were plotted versus expected T7 RNAP activity units. Data are reported as average and s.d. from n = 6 replicates. The dashed line (y = x) serves as a reference for ideal agreement between measured and expected activities.

To quantify T7 RNAP activity from the fluorescence response curves, we extracted the slopes of the fluorescence responses against time during the early, near-linear phase of signal increase off the baseline fluorescence, as a proxy for the rate of RNA being synthesized during the early phase of the IVT reaction (Materials and Methods and Figure S2A). We refer to the extracted slope as the “observed initial rate” of each IVT reaction. Observed initial rates were plotted against the corresponding enzyme activities (U/μL), which shows a strong linear correlation (Pearson’s r = 0.97) within the tested range (Figure 1C). This linear relationship provides a calibration curve between the T7 RNAP activity units per volume and the observed initial rate of the fluorescence response and is promising to facilitate quantification of the activities of T7 RNAPs prepared from different batches or sources.

Next, we performed validation tests of the derived calibration curve. We set up three additional groups of IVT assays using the NEB T7 RNAP, with final enzyme concentrations at 2, 5, or 8 U/μL. We first treated these reactions as IVT with “unknown” enzyme concentrations; we recorded fluorescence as a function of time (Figure 1D) and extracted the observed initial rates from the fluorescence response curves (Figure S3A and Materials and Methods). We then plugged the observed initial rates into the calibration curve (Y = 20.26X + 35.50, where X represents the activity units per volume in U/μL and Y is the observed initial rate) to derive the activity of the T7 RNAP in each reaction. We report the derived activities using the calibration curve as “measured activities”. For reactions containing 2, 5, and 8 U/μL of T7 RNAP, the measured activities are quantified as 1.9 ± 0.8, 4.7 ± 0.8, and 7 ± 2 U/μL, respectively, through the calibration curve. The measured activities for the all three concentrations are in good agreement with the expected activities within experimental error (Figures 1E and S3B), suggesting that the Broccoli IVT assay can provide a quantitative framework for assessing T7 RNAP activities.

Activity Assessment of T7 RNAPs from Different Sources

To assess the activities of T7 RNAPs from a different commercial source or purified in-house, we sought to perform the Broccoli IVT assay using different enzymes at comparable molar concentrations to the NEB T7 RNAP used for establishing the calibration curve. However, the molar concentration of the stock NEB T7 RNAP is below the detection limit when measured by the UV absorbance at 280 nm or the Bradford assay. To address this challenge, we adopted a high-sensitivity Qubit protein assay and determined the concentration of the NEB T7 RNAP to be 259 ng/μL or 2.6 μM (Figure S4A). Therefore, the calibration curve derived from the Broccoli IVT assays was characterized with the final concentrations of the commercial T7 RNAP ranging from 0.07 to 0.5 μM (Figure 1C). We acquired a different commercial T7 RNAP from “Thermo Scientific” (Cat. No. EP0111) for testing purpose using the Broccoli IVT assay. The stock solution of this enzyme was measured at a concentration of 242 ng/μL, comparable with the NEB enzyme (Figure S4A).

The in-house T7 RNAPs were purified by nickel-affinity chromatography (Figure S1A), desalted, buffer-exchanged into the storage buffer, and stored as aliquots at −80 °C after flash freezing in liquid N₂ (Materials and Methods). We measured activities of two independently purified batches of in-house T7 RNAPs (i.e., in-house-1 and in-house-2). In-house-1 T7 RNAP was purified in 2023, whereas in-house-2 T7 RNAP was purified in 2024. A fresh aliquot was taken out from the −80 °C freezer for each batch of T7 RNAP and thawed on ice, and the concentrations of the thawed aliquots were quantified by the Qubit protein assay to be 1638 ng/μL for in-house-1 and 1852 ng/μL for in-house-2 (Figure S4A). We then examined purities of the commercial and in-house prepared enzymes by loading 1.3 μg of enzymes onto 8% SDS-PAGE for separation. The bands corresponding to the size of T7 RNAPs are the most prominent bands for all four enzymes, suggesting high purities of all used enzymes (Figure S4B). There are minor impurity bands observed for the two in-house T7 RNAPs, which could be due to that the in-house T7 RNAPs were purified by a single Ni-affinity chromatography step; adding additional purification step(s) such as size-exclusion chromatography could further improve the purity of the enzyme.

Next, we performed the Broccoli IVT assay with the commercial or in-house T7 RNAPs at a final concentration of 0.13 μM (within the molar concentration range of 0.07−0.5 μM used in the calibration curve). The resulting fluorescence response curves yielded observed initial rates of 68 ± 12 arbitrary fluorescence units (RFU)/min for the Thermo Scientific T7 RNAP, 86 ± 16 (RFU)/min for in-house-1, and 184 ± 49 (RFU)/min for in-house-2 (Figures 2A and S5).

Figure 2.

Characterization and comparison of the specific activities of in-house and commercial T7 RNAPs. (A) Fluorescence response curves for the Broccoli IVT assays using 0.13 μM Thermo Scientific and in-house T7 RNAPs. Data are presented as the average and s.d. from n = 6 replicates (3 biological replicates and each biological replicate contains 2 technical replicates). (B) Summary of measured activity (U/μL), enzyme concentration (ng/μL), and calculated specific activity (U/ng) for the Thermo Scientific and in-house T7 RNAPs. Errors for enzyme activities represent the standard deviations of the calculated activities based on the calibration curve for each enzyme from n = 6 replicates. Errors for enzyme specific activities were determined via error propagation.

Using the calibration curve (Figure 1C), we obtained the observed activities of the three T7 RNAPs (Thermo Scientific, in-house-1 or in-house-2) in the IVT reaction at 1.6 ± 0.6, 2.5 ± 0.8, or 7 ± 2 U/μL, respectively. To compare activities for the T7 RNAP prepared in-house versus the commercial source, we calculated the specific activities for each enzyme in “U/ng” (Figure 2B). Among the tested enzymes, in-house-2 T7 RNAP exhibits the highest specific activity; the in-house-1, NEB, and Thermo Scientific T7 RNAPs show comparable specific activities, of which the Thermo Scientific stock shows slightly lower activity than the other two though the difference is not significant considering the associated errors. The results reveal the variability of the specific activities of T7 RNAP prepared from different sources, or different batches within the same lab. Other factors such as storage condition and storage duration of the purified enzymes may also affect the observed activities. The Broccoli IVT assay enables convenient quantification and comparison of the specific activities of T7 RNAPs from different sources or batches.

IVT by In-House and Commercial T7 RNAPs for RNA Preparation

One major application of T7 RNAP is to prepare RNAs via IVT for basic research or industrial applications in RNA therapeutics. The RNA yield per volume of the IVT reaction is known to be dependent upon the concentration of T7 RNAP.⁴⁶ We tested the RNA yield of IVT reactions for two RNA sequences with normalized units of T7 RNAPs: the Broccoli aptamer and a tRNA^Gly-like transcript (Table S1). These templates were selected to assess transcription performance across structurally distinct RNA products. The Broccoli aptamer was transcribed using the same Broccoli IVT assay condition (high Mg²⁺) for ~20 min, while the tRNA^Gly-like transcript was transcribed using standard T7 reaction condition (40 mM Tris-HCl pH 7.9, 6 mM MgCl₂, 1 mM DTT, 2 mM spermidine, 2 mM of each rNTP) for 3 h. Three RNAPs were used for IVT at 5 U/μL, including the Thermo Scientific, in-house-1, and in-house-2 (Figure 3A). To get a faithful examination of T7 RNAP activity, the IVT reactions were performed in the absence of any RNase inhibitor and directly subjected to Urea-PAGE for characterization. We observed RNA product bands consistent with the expected transcript sizes for both DNA templates (Figure 3B), and we further confirmed the identities of the RNA products rather than DNA template using NaOH digestion (Figure S6). With normalized enzymatic units, the IVTs by the Thermo Scientific and in-house-2 T7 RNAPs produced comparable amounts of Broccoli RNA products. The in-house-1 T7 RNAP produces slightly less products, which could be due to potential RNase contamination (Figures 3B and S6A). Under the standard IVT condition, all three RNAPs generated comparable amounts of the tRNA^Gly-like products, consistent with the normalized reaction units (Figures 3B and S6B). Here, the in-house-1 T7 RNAP does not show lower yield, unlike the case in the Broccoli IVT; it could be due to that the Broccoli IVT condition contains higher Mg²⁺ concentration at 31 mM compared to 6 mM Mg²⁺ used in the standard IVT, which could amplify the impact of minor RNase contamination. Together, these findings demonstrate that the quantified units of T7 RNAPs by the Broccoli IVT assay can be used to normalize product yield expectations for preparative IVT reactions.

Figure 3.

Comparison of IVT products from in-house and commercial T7 RNAPs with normalized activity units. (A) Schematic illustration of the preparative IVT assay from DNA templates containing a T7 promoter and a DNA template of interest (Broccoli DNA or a tRNA^Gly-like DNA), catalyzed by either commercial or in-house T7 RNAP to produce RNA transcripts. (B) Visualization of RNA products from DNase I-treated IVT reactions by 10% 8 M Urea-PAGE gel and SYBR Gold staining. A low range ssRNA ladder was loaded on the gel as a reference for sizing the RNA products, where the tRNA^Gly-like transcript is expected to be 99 nucleotides (nt), and the Broccoli RNA is expected to be 57 nt.

Discussion

In this work, we established a fluorescence-based framework for the quantitative assessment of T7 RNAP activity units using the Broccoli RNA aptamer as a real-time transcriptional reporter. We found that the observed initial rates extracted from the fluorescence response curve of the Broccoli IVT assay increased linearly with increasing activity units of T7 RNAP in the reaction. Therefore, we were able to derive a linear calibration curve based on the commercial T7 RNAP with characterized units to quantify the activity units of other purified T7 RNAPs with unknown activities. A prior study by Kartje et al. used the Broccoli IVT assay to evaluate the dependence of RNA yield on different IVT conditions, including concentrations of rNTPs, MgCl₂, and promoter sequence context, where the yield of RNA is estimated by the fluorescence intensity at a certain time point of the fluorescence response curve.¹⁵ Here, we chose to use the initial observed rate (i.e., the initial slope of the curve), rather than a single fluorescence intensity, to relate to the unit activity of T7 RNAP because the initial rates of IVT are more relevant to the enzymatic activity unit where the major components required for IVT maintain more stable concentrations. A major varying factor in the progression of IVT is the concentration of free Mg²⁺. The activity of T7 RNAP requires a certain amount of Mg²⁺;⁴⁷ meanwhile, the substrate rNTPs bind Mg²⁺ to balance the negative charge of the triphosphate; the Broccoli RNA product requires a certain amount of Mg²⁺ for folding;⁴³ additionally, the produced inorganic pyrophosphate binds Mg²⁺ and can be subjected to precipitation at high concentrations.¹⁵ As IVT proceeds, one expects the consumption of rNTPs and an increasing amount of Broccoli RNA and inorganic pyrophosphate, which results in complex changes of free Mg²⁺ in solution. Therefore, we reason that the observed initial rate of the IVT is a more reasonable property to quantify the T7 RNAP activity unit. In fact, we tried to plot the maximum fluorescence intensities (i.e., intensities at 60 min of the fluorescence response curves) against the activity units, which resulted in a saturating curve rather than a linear trend. If we apply a linear fit to this relationship, the linear correlation coefficient (R² = 0.91 in Figure S7) is significantly lower than the R² of the calibration curve derived from the observed initial rate (R² = 0.95 in Figure 1C).

The Broccoli IVT assay provides a facile, low cost, and radioactivity-free approach to quantify activity units of any purified RNAP in any biochemical lab. The Broccoli IVT assay requires the DFHBI-1T fluorophore and commercial T7 RNAP with known activity unit as a reference, which are readily available and low cost compared to other T7 RNAP activity assays including the traditional ³²P-based radioactive IVT assay,¹ the colorimetric assay,⁴⁰ and the “broken-beacon” assay⁴¹ (Table S2). The current fluorescence-based method cannot determine T7 RNAP activity unit de novo but would provide a robust and convenient approach to quickly determine unknown T7 RNAP activity in reference to the commercial T7 RNAP. Therefore, the activity unit determined by the Broccoli IVT assay carries the same definition as the unit defined by the reference T7 RNAP (in this case, the NEB T7 RNAP). The quality of the reference T7 RNAP is crucial for establishing a reliable calibration curve for this method. We compared the activities of the newly acquired NEB T7 RNAP and an “old” tube of NEB enzyme stored in the −20 °C freezer in the lab that has expired for two months using the Broccoli IVT assay (Figure S8). We observed significant reduction of activity for the “old” tube of NEB T7 RNAP compared to the fresh tube. We would recommend performing the calibration curve with fresh commercial T7 RNAP at the same time as characterizing the unknown T7 RNAP activity. We expect that the current platform can be adapted for other RNAPs such as SP6 RNAP if we switch the T7 promoter to the SP6 promoter sequence in the Broccoli DNA sequence.

Conclusion

In summary, the Broccoli IVT assay provides a cost-effective, facile, nonradioactive, and quantitative tool for assessing in-house T7 RNAP activity units, to facilitate rapid assessment and benchmarking of T7 RNAP activity from different batches of purification or sources in any biochemical lab in research and industry.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.biochem.6c00336.

Data for in-house enzyme and DNA template preparation (Figure S1), fluorescence response curves of individual biological replicates for Broccoli IVT assays (Figures S2, S3, and S5), concentration and purity characterization of commercial and in-house T7 RNAPs (Figure S4), full gel images of electrophoresis data on preparative IVTs (Figure S6), comparison with an alternative approach for quantifying RNAP activity (Figure S7), comparison of enzymatic activities between a fresh and an old tube of the NEB T7 RNAP (Figure S8), oligonucleotide sequences used in this study (Table S1), and comparison of costs among different RNAP activity assays (Table S2) (PDF)