A Simple and Efficient Method to Produce RNAs with Homogenous 3 Ends

Introduction

Transcription in vitro using T7 RNA polymerase is commonly used to produce RNAs for a wide variety of applications, including structural and biochemical studies and potential therapeutics ⁽¹⁾ ⁽²⁾ ⁽³⁾ . Despite its usefulness, a number of undesired reactions increase the heterogeneity of transcription products and necessitate the careful purification of the desired RNAs. A major undesired reaction is the addition of one or more nucleotides to the 3´ end of an RNA product, commonly called N+1 activity ⁽¹⁾ ⁽²⁾ . N+1 activity can have detrimental effects in a number of applications, such as reducing RNA synthesis yields by viral RNA replicase ⁽⁴⁾ , complicating the NMR data analysis ⁽⁵⁾ and preventing the effective crystallization of the RNA needed for structural studies ⁽⁶⁾ . We report that modification of the last two nucleotides at the 5´ terminus of the DNA template with ribose C2´ methoxy moieties can significantly reduce N+1 activity of T7 RNA polymerase and, in some cases, increase the yields of RNA products.

Transcription from DNA Templates Containing Ribose C2´-Methoxy Analogs

The effectiveness of the ribose C2´-methoxy modification is illustrated by transcription from DNA template, S5, which encodes a 22nt RNA from within the Tetrahymena thermophila group I intron (Figure 1, Panel A). While S5 does produce a 22nt transcript, a 23nt N+1 product is quite prominent (Figure 1, Panel B, lane 1). Modification of the 5´ terminal template nucleotide did not alter the ratio of the N and N+1 products (Figure 1, Panel B, lane 2). However, modification of the two 5´ terminal nucleotides or the penultimate position alone predominantly produced the correctly terminated 22nt transcript (Figure 1, Panel B, lanes 3 and 4). For this transcript, the abundance of the 22nt RNA is reproducibly two-fold higher than that from the unmodified template.

Figure 1. DNA templates containing 5´ guanosine nucleotides modified with ribose C2´-methoxy reduced the N+1 activity of T7 RNA polymerase.

Panel A. Schematic of the DNA templates, S5 (unmodified) and S5** (modified with two guanylates containing ribose C2´-methoxy), used for transcription. An asterisk (*) denotes the presence of a methoxy-modified guanylate. Gm denotes the position of modified guanylates. The sequence in lower case is the bottom strand of the T7 promoter, while the sequence in upper case is the template for transcription. The arrow denotes the position for the initiation of transcription. Panel B. Transcripts generated from various DNA templates. The RNAs were visualized by staining with 0.02% Toluidine blue. The 22nt product and the result of N+1 activity are indicated to the right of the gel image. Panel C: NMR spectra of the RNA produced by DNA template, S5 (upper spectrum) and S5** (lower spectrum). The signals from 5 to 6.2ppm containing the H1´/H5 region of the RNA are shown. NMR spectra were recorded on a Bruker AMX-600 spectrometer. All samples were dissolved in a phosphate buffer containing 90% H₂O and 10% D₂O. Spectra for the RNAs generated from S5 (2.9mM sample) and S5** (2.7mM sample) were taken at 30°C with 16,384 complex points in 64 scans. The proton carrier frequency was set at 4.72ppm, and the maximum for excitation was set at 7.7ppm.

/~/media/Images/Resources/Figures/eNotes/3039CAW08.ashx

To determine whether the modified template would affect the identity of the nucleotide incorporated into the RNA, 22nt RNAs generated from unmodified S5 and modified S5** templates were compared using nuclear magnetic resonance (NMR) spectroscopy. We examined the signals for the protons at ribose H1´, H5, and H6 of the 3´ terminal cytidylate in the S5 RNA. The chemical shifts of these protons will change if the transcripts have different bases at those positions. Transcripts generated from S5 and S5** had proton signals that are identical (Figure 1, Panel C; and data not shown), demonstrating that modifications in S5** did not alter the sequence of the transcription product.

Table 1. Sequences of RNA Transcribed from P1, VΨ, M5-56 and P5-64.

/~/media/Images/Resources/Tables/9200-9299/9237LA.ashx

Several transcriptions of RNAs ranging from 27 to 64nt were examined using the modified and unmodified DNA templates (Figure 2). Four templates that normally generate significant amounts of the respective N+1 products (Figure 2, lanes 1, 4, 5 and 7) were found to have significantly reduced N+1 activity when modified with two 5´ terminal ribose C2´-methoxys (Figure 2, lanes 2, 3, 6 and 8). Two of the templates (P1 and VΨ) also produced prematurely terminated RNAs that were one or two nucleotides less than full-length. The ribose C2´-methoxy modification did not significantly affect the abundance of these truncated RNAs (Figure 2, lanes 1–4).

Figure 2. Effect of ribose C2´-methoxy modifications of the last two nucleotides of four DNA templates on N+1 activity of T7 RNA polymerase.

RNAs in denaturing polyacrylamide gels stained with Toluidine blue. Both gels contain 20% polyacrylamide, but samples in lanes 1–4 and lanes 5–6 were subjected to electrophoresis for 16 and 24 hours, respectively. The DNA templates used to generate the transcripts are listed on top of the gel images, with asterisks (**) denoting modification of the 5´-terminal two nucleotides with ribose C2´-methoxys. The lengths of the transcripts expected for P1 and VΨ are indicated on the left side of the gel image, while transcripts from M5-56 and M5-64 are listed to the right. Positions of the RNA one nucleotide longer than expected are indicated with "N+1". The sequence of the RNA transcribed from P1 is listed in Table 1.

Producing More Homogeneous RiboMAX™ Transcription System Products Using Methoxy-Modified Templates

To demonstrate that the procedure is robust and works as well on a commercially available RNA transcription system as with our own reagents, we tested normal and methoxy-modified templates using the Promega RiboMAX™ Large Scale RNA Production System-T7 (Cat.# P1300). Modified (S5** and P**) and unmodifed (S5 and P) templates were transcribed for 2 hours in parallel, and the products were resolved by denaturing polyacrylamide gel electrophoresis (Figure 3). The S5** and P** templates, modified with two 5´-terminal ribose C2´-methoxys, yielded greatly reduced N+1 products (compare lanes 1 and 2; 3 and 4).

Figure 3. Transcripts produced using the RiboMAX™ Large Scale RNA Production System-T7 (Cat.# P1300).

An asterisk (*) denotes the presence of methoxy-modified templates. The sequence for S5 is shown in Figure 1, Panel A; the sequence for P1 is shown in Table 1.

Transcription from dsDNA Templates

RNA less than 100nt can be transcribed from chemically synthesized DNA templates where only the promoter sequence is double-stranded (4). However, chemical DNA synthesis is impractical for production of longer transcripts, and the double-stranded DNA (dsDNA) templates generated by PCR are more suitable. Therefore, we compared transcriptions from dsDNA templates with or without the ribose C2´-methoxy modifications at the 5´-terminal two nucleotides. The results (data not shown) were identical to those seen in Figure 1. Therefore, dsDNA amplified by PCR using modified primers will also reduce N+1 activity.

Summary

The ability to generate more homogenous transcription products using the template modification of a normal transcription reaction should permit more rapid RNA purification by high-performance liquid chromatography ⁽⁷⁾ . In addition, it will facilitate interpretation of results. This modification can be incorporated during the chemical DNA synthesis and not affect the identity of the nucleotide incorporated into the nascent RNA. This modification of the normal T7 transcription reaction should generate RNA useful for many applications.

Acknowledgments

This work was originally reported in Kao, C. et al. ⁽⁸⁾ . Patent for this application has been submitted by the University of California Technology Licensing Office. We thank Dr. Ignacio Tinoco, Jr., in whose laboratory this work was performed. Funding was provided to I.T. by the NIH (GM10840) and DOE (DE-FG03-86ER60406) and to C.K. by the NSF (MCB9507344) and USDA (9702126).