Transcription Regulation Through Nascent RNA Folding

Abstract

Folding of RNA into secondary structures through intramolecular base pairing determines an RNA's three-dimensional architecture and associated function. Simple RNA structures like stem loops can provide specialized functions independent of coding capacity, such as protein binding, regulation of RNA processing and stability, stimulation or inhibition of translation. RNA catalysis is dependent on tertiary structures found in the ribosome, tRNAs and group I and II introns. While the extent to which non-coding RNAs contribute to cellular maintenance is generally appreciated, the fact that both non-coding and coding RNA can assume relevant structural states has only recently gained attention. In particular, the co-transcriptional folding of nascent RNA of all classes has the potential to regulate co-transcriptional processing, RNP (ribonucleoprotein particle) formation, and transcription itself. Riboswitches are established examples of co-transcriptionally folded coding RNAs that directly regulate transcription, mainly in prokaryotes. Here we discuss recent studies in both prokaryotes and eukaryotes showing that structure formation may carry a more widespread regulatory logic during RNA synthesis. Local structures forming close to the catalytic center of RNA polymerases have the potential to regulate transcription by reducing backtracking. In addition, stem loops or more complex structures may alter co-transcriptional RNA processing or its efficiency. Several examples of functional structures have been identified to date, and this review provides an overview of physiologically distinct processes where co-transcriptionally folded RNA plays a role. Experimental approaches such as single-molecule FRET and in vivo structural probing to further advance our insight into the significance of co-transcriptional structure formation are discussed.

Keywords: RNA folding; RNA polymerase; RNA splicing; nascent RNA; transcription.

Fig. 1 :. Co-transcriptional RNA processing events are regulated by nascent RNA structure.

(a) Eukaryotic RNA processing involves protein and RNA/protein complexes binding sequence or structure features on the elongating transcript. The RNA 5’-end is co-transcriptionally capped and occupied by the Cap-binding complex. Components of the spliceosome (snRNPs) recognize intron boundaries, the 5’ splice site (5’SS) and branch site (BS) to facilitate co-transcriptional splicing. The exon junction complex (EJC) marks successfully spliced exon-exon boundaries. RNA editing factors such as adenosine deaminases (ADAR) bind to and act on double-stranded RNA. Small proteins specifically recognize stem loops and regulate e.g. termination of histone gene transcription. The signal for polyadenylation is read by the cleavage and polyadenylation specificity factor (CPSF), recruiting additional proteins such as cleavage stimulation factor (CstF) to release and polyadenylate the nascent transcript. Not to scale. (b) E. coli chromosome spread [16] showing ribosomes covering the nascent RNA (top); chromatin spread of a Drosophila chorion gene [15] showing spliceosome complexes assembled on nascent RNA (bottom).

Fig. 2:. Interplay of nascent RNA structure close to the polymerase and transcription speed.

(a) Single-molecule transcription assay with optical tweezers, measuring force generated by transcribing Pol II over time. Force is proportional to polymerase position along the template. The inset shows a Pol II backtracking event. Without TFIIS, the polymerase recovers from backtracked states less efficiently. Data from [30]. (b) Same transcription assay as in (a), but DNA templates with different base compositions are used. Data from [38]. (c) Metaplot for 400 nt region around Pol I occupancy peaks in the 5’ external transcribed spacer region of rDNA repeats. Left axis: Energy of folding, where each data point corresponds to a 65 nt window ending 14 nt upstream of the respective Pol I position. ΔG at 30°C was calculated with ViennaRNA [39]. Right axis: RNA crosslinking and Pol I immunoprecipitation profile (fraction of reads ×10³). Data from [40]. (d) Model for inhibition of backtracking by nascent RNA structure forming close to the polymerase. This leads to net acceleration of transcription across regions with high structure formation potential.

Fig. 3:. Methods to study co-transcriptional RNA folding.

(a) Principle of RNA chemical probing using SHAPE reagents or DMS (dimethyl sulfate) combined with mutational profiling. (b) Co-transcriptional SHAPEseq shows structural transitions of the crcB riboswitch during transcription. Raw data from [49]. (c) Co-transcriptional folding pathway of E. coli 23S rRNA helix 23 determined by in vivo DMS probing of nascent RNA [52]. (d) Monitoring co-transcriptional RNA folding using single-molecule FRET. Rep-X or the 10-fold slower PcrA-X helicase are loaded onto an immobilized RNA/DNA hybrid. Upon addition of missing buffer components, the helicase unwinds the duplex and releases RNA in analogy to an elongating 3’-end during transcription. Folding is monitored in real time using FRET [59].

Fig. 4:. Co-transcriptionally formed RNA structures regulate gene expression.

(a) Riboswitches fold in conjunction with small molecule ligands. If the ligand (e.g. Fluoride) is present, formation of a terminator hairpin is prevented. In the absence of fluoride the hairpin folds, recruits NusA and the transcription complex is destabilized, leading to termination. (b) A stem loop within the intron of yeast APE2 occludes the cryptic splice site AAG under normal conditions. Under heat shock the stem is melted and alternative splicing is favored [93]. (c) Histone pre-mRNA 3′-end processing complex [11,94,95]. Pol II not to scale. (d) Alternative splicing of the HIV-1 primary transcript at the A3 splice site is regulated by coexistence of two alternative conformations involving the 3’ splice site [96], leading to regulation of viral Tat protein production. The stretch pairing with A3 is in red. D: 5’ splice sites, A: 3’ splice sites.