Preparation of RNAs with non-canonical 5' ends using novel di- and trinucleotide reagents for co-transcriptional capping

Abstract

Many eukaryotic and some bacterial RNAs are modified at the 5' end by the addition of cap structures. In addition to the classic 7-methylguanosine 5' cap in eukaryotic mRNA, several non-canonical caps have recently been identified, including NAD-linked, FAD-linked, and UDP-glucose-linked RNAs. However, studies of the biochemical properties of these caps are impaired by the limited access to in vitro transcribed RNA probes of high quality, as the typical capping efficiencies with NAD or FAD dinucleotides achieved in the presence of T7 polymerase rarely exceed 50%, and pyrimidine derivatives are not incorporated because of promoter sequence limitations. To address this issue, we developed a series of di- and trinucleotide capping reagents and in vitro transcription conditions to provide straightforward access to unconventionally capped RNAs with improved 5'-end homogeneity. We show that because of the transcription start site flexibility of T7 polymerase, R¹ppApG-type structures (where R¹ is either nicotinamide riboside or riboflavin) are efficiently incorporated into RNA during transcription from dsDNA templates containing both φ 6.5 and φ 2.5 promoters and enable high capping efficiencies (∼90%). Moreover, uridine-initiated RNAs are accessible by transcription from templates containing the φ 6.5 promoter performed in the presence of R²ppUpG-type initiating nucleotides (where R² is a sugar or phosphate moiety). We successfully employed this strategy to obtain several nucleotide-sugar-capped and uncapped RNAs. The capping reagents developed herein provide easy access to chemical probes to elucidate the biological roles of non-canonical RNA 5' capping.

Keywords: FAD; In vitro transcription; NAD; RNA cap; UDP-glucose; dinucleotide; trinucleotide.

FIGURE 1

The set of nucleotide derivatives used for in vitro transcription studies, divided into four categories according to the type of RNA 5′-end they produce.

SCHEME 1

Representative syntheses of the non-canonical caps used in this study.

FIGURE 2

(A) Transcription initiation events expected for the φ 2.5 and φ 6.5 T7 promoters and different initiating nucleotides; For clarity, only the coding (sense) strands of the promoter regions of double-stranded DNA templates are shown. (B) General in vitro transcription protocol for the assessment of nucleotide analog incorporation.

FIGURE 3

High resolution polyacrylamide gel electrophoresis (HRPAGE; 15% PAA, 7 M urea, 1⨯ TBE) analysis of 35 nt IVT RNAs obtained from templates containing either the A φ 2.5 promoter (left gel stained with Ethidium Bromide—EtBr) or the G φ 6.5 promoter (right gel stained with SYBR Gold) in the presence of 0.5 mM NTPs, 0.125 ATP/GTP, and 0.375 mM of the initiating nucleotide analog. The purple labels correspond to the caps that are expected not to incorporate into RNA with the defined promoter.

FIGURE 4

(A) HRPAGE (15% PAA, 7 M urea, 1⨯ TBE, SYBR Gold) analysis of differently capped RNA species with indication of bands used for capping efficiency assessment. (B) Representative HRPAGE (15% PAA, 7 M urea, 1⨯ TBE, SYBR Gold), without APB (upper gel) or with 1% APB (bottom gel), of 25 nt IVT RNA obtained from a template with G φ 6.5 promoter after 3′ end trimming steps. The capping efficiency values given below the gel indicate the mean values from triplicate analyses performed on three independent transcriptions and 6 densitometric measurements (+/- APB). (C) Comparison of the capping efficiencies under different conditions as a function of [nucleotide analog]/[GTP] ratio.

FIGURE 5

(A) Representative HRPAGE (15% PAA, 7 M urea, 1⨯ TBE, SYBR Gold), without APB (upper gel) or with 1% APB (bottom gel), of 25 nt IVT RNA obtained from a template with the φ 2.5 promotor after 3′ end trimming steps. The capping efficiency values given below the gel indicate the mean values from triplicate analyses performed on three independent transcriptions and 6 densitometric measurements (+/- APB) for NAD and NADpG, and from duplicate analyses FAD and FADpG (only APB). (B) Comparison of capping efficiencies under different conditions as a function of nucleotide analog / GTP ratio. (C) HPLC purification profiles of IVT uncapped and FAD-capped RNAs, with corresponding HRPAGE analysis (15% PAA, 7 M urea, 1⨯ TBE, SYBR Gold) of the collected fractions. (D) HRPAGE (15% PAA, 7 M urea, 1⨯ TBE, SYBR Gold) of RNAs after DNAzyme trimming on uncapped, FAD initiated and FADpG initiated IVT RNAs.

FIGURE 6

(A) Capping efficiencies of FAD- and FADpG-RNAs determined using the FAD-capQ method (using 9x excess FADpG). (B) Capping efficiencies obtained for NADpG and FADpG (both quantification methods included) as a function of the T7 promoter.

FIGURE 7

(A) HPLC chromatogram and MS spectrogram issued from LC-MS analyses on FADpG initiated IVT RNA using A φ 2.5 DNA template. (B) HPLC chromatogram and MS spectrogram issued from LC-MS analyses on NAcGlcppUpG initiated IVT RNA using G φ 6.5 DNA template. (C) Capping efficiencies for FAD, FADpG, GlcppUpG and NAcGlcppUpG calculated from LC-MS analyses, as mean values from 3 technical repetitions.