The shock absorber: preventing sRNA transcriptional noise

Abstract

Spacer sequences excised during tRNA (tRNA) maturation were long considered as “junk” RNAs, which are non-functional by-products rapidly degraded by the cell. In our recent report,1 we described a tRNA-derived RNA fragment (tRF) acting as a functional constitutive sRNA sponge that absorbs transcriptional noise from small regulatory RNAs (sRNAs) promoters. In Eukaryotes, natural microRNA (miRNA) sponges or competing endogenous RNAs (ceRNAs) were reported to attenuate specific miRNA-mediated regulation by acting as decoys or traps for miRNAs.2 The first endogenous miRNA sponge was identified in Arabidopsis thaliana. Subsequently, more miRNA sponges have been discovered in mammalian cells or in viruses. Various coding or non-coding RNAs (e.g., pseudogenes, long non-coding RNAs and circular RNAs) can function as miRNA sponges.2 Related examples have been reported in the prokaryotic kingdom attesting to the conservation of this regulatory mechanism. Prokaryotic RNA sponges can indeed originate either from 3′ external transcribed spacer of pre-tRNA,1 3′-untranslated region of mRNA,3 intergenic region of a polycistronic transcript,4 or prophage.5 Computational analyses to identify regulatory RNA targets consider 2 major parameters, namely the complementary of a region of an mRNA to the regulatory RNA and the phylogenetic conservation of this region. As those parameters are also applicable to RNA sponges, some of these computationally identified targets may actually function as competitive inhibitors sequestering regulatory RNAs and preventing them from binding their real mRNA targets. Seitz hypothesized that the action of a miRNA on its real targets will be the result of the integration of information from tens or even hundreds of mRNA transcripts in the cell.6 Similarly, Jost et al. proposed that the role of miRNA sponges is to confer robustness to the regulation of a few principal targets.7 Thus, what we first considered as false positives, or artifacts, could be “real non-targets” so called RNA sponges. Our group recently validated this assumption.1 Using MS2-affinity purification coupled with RNA sequencing technology (MAPS), we observed that the 3′ external transcribed spacer of glyW-cysT-leuZ pre-tRNA transcript (3′ETSleuZ), released during tRNA maturation through RNAse E cleavage, base-pairs with RyhB and RybB sRNAs. Interestingly, mature tRNAs deriving from glyW-cysT-leuZ transcripts are unaffected by overexpression of either RyhB or RybB. We demonstrated that this tRNA-derived RNA fragment (tRF), long considered “junk” RNA, actually prevents any sRNA transcriptional noise to regulate mRNA targets in non-inducing condition. We can perceive 3′ETSleuZ as a buffer that absorbs RyhB and RybB transcriptional noise and sets a concentration threshold that sRNAs have to overcome in order to induce targets regulation (Fig. 1). We also noticed that internal transcribed spacers (ITS) of metZ-metW-metV tRNA transcripts (ITSmetZ-metW and ITSmetW-metV) potentially interact with RybB and MicF sRNAs. In addition to 3′ETSleuZ, ITSmetZ-metW and ITSmetW-metV, we identified others tRF candidates among bacteria by sequence conservation analysis. In this manuscript, we added the notion that sRNA sponges can derive from tRNA transcripts and not only from the mRNA pool. We can also easily imagine that the same observation could be made for rRNA transcripts and assume that sRNA sponges could emerge from every type of RNA in the cell. Overall, these data are blurring the distinction between “junk” RNA and functional RNA. Figure 1. Schematic representation of the sRNA sponge effect. In absence of stress, sRNA transcriptional noise is buffered by the sRNA sponge, preventing regulation of the mRNA target. During stress, sRNA concentration exceeds the threshold set by the sRNA sponge, ... It now appears essential to consider sRNA sponges in comprehensive searches for new sRNA targets, as they can modulate and modify the way a sRNA is able to regulate a pool of targets. Indeed, the loss of these pseudo-targets would increase the sRNA bio-availability and enhance the repression of real targets. By mutating the 3′ETSleuZ to prevent base-pairing with sRNAs, we confirmed that both RyhB and RybB activity increased.1 In this way, sRNA sponges add an additional layer of regulation to the post-transcriptional control of sRNA targets.