Discovering riboswitches: the past and the future

Abstract

Riboswitches are structured noncoding RNA domains used by many bacteria to monitor the concentrations of target ligands and regulate gene expression accordingly. In the past 20 years over 55 distinct classes of natural riboswitches have been discovered that selectively sense small molecules or elemental ions, and thousands more are predicted to exist. Evidence suggests that some riboswitches might be direct descendants of the RNA-based sensors and switches that were likely present in ancient organisms before the evolutionary emergence of proteins. We provide an overview of the current state of riboswitch research, focusing primarily on the discovery of riboswitches, and speculate on the major challenges facing researchers in the field.

Keywords: allosteric ribozyme; aptamer; gene regulation; metabolite; noncoding RNA.

Figure 1.. Riboswitch components and mechanisms.

(A) Riboswitches typically use partly overlapping aptamer and expression platform domains to regulate transcription termination (left), translation initiation (center), or alternative splicing (right). Rarer mechanisms include transcriptional interference [205] and regulation of mRNA stability [56, 57, 206]. RBS designates the ribosome binding site and SS designates a splice site. Arrows indicate alternative base-pairing that can form in a manner dictated by ligand (X) binding to the notional aptamer structure depicted. (B) Schematic representation of the differences in RNA structure for a genetic “OFF” riboswitch that suppresses gene expression when ligand is bound. If ligand binds the aptamer domain during transcription (left), a folding pathway is favored that forms an intrinsic terminator stem, which triggers transcription termination within a run of U nucleotides. If the ligand is not quickly bound by the aptamer (right), the RNA folds along a different pathway to form the anti-terminator structure. This blocks formation of the terminator stem and promotes transcription of the full mRNA.

Figure 2.. The abundances of experimentally validated riboswitch classes.

The abundances of riboswitch classes plotted were obtained from previous publications [7, 16] and are derived from computational searches using databases available at the time of these references. The number of undiscovered riboswitch representatives (~28,000) was estimated using power law projections (Box 2) as described elsewhere [7, 15, 17]. Note that some riboswitch classes are too rare to be visible on the graphic.

Figure 3.. The ligands sensed by the known classes of bacterial riboswitches.

(A) Riboswitch classes are assigned to one or more groups based on the roles these ligands serve. The four major categories were chosen to highlight trends relevant to fundamental and ancient biochemical processes. The graphic was adapted from a previous publication [16]. (B) Plot of the number of riboswitch representatives associated with the group assignments as depicted in A. The numbers of representatives for each class were presented elsewhere [16].

Figure I.. Key considerations used when classifying riboswitches.

X and Y represent different ligands.

Figure II.. Predicting the total number of riboswitch classes using a power law model.

Log-log plot of the abundance (count) of each experimentally validated riboswitch class presented in order (rank) based on its abundance. For the power law equation Y = mX^b, Y represents the count, X represents the rank, m is the theoretical number for the most abundant predicted riboswitch class, and b is the exponent (slope). A line with the slope of -1.5 is estimated to best reflect the linear portion of the data points. Right inset: Plots depicting the data for validated riboswitch classes only (first curve) or for both validated riboswitch classes plus orphan riboswitch candidates. Axis labels are the same as for the left plot. Note that the inclusion of the longest-standing orphan riboswitch classes [182] extends the linear portion of the plot because novel riboswitch classes tend to be rare compared to those discovered previously.

Figure III.. The roadblocks to experimentally validating orphan riboswitch candidates.

Former orphan riboswitches are listed in the order in which they were experimentally validated. The barriers overcome to match the ligand with its riboswitch aptamer class are indicated for each orphan. Rare variants of a common riboswitch class that are adapted to bind a different ligand have been called “snugglers” [187]. The figure was expanded from an earlier version published elsewhere [187].

Figure IV.. Riboswitches and elemental ions.

Highlighted are elements whose ions are monitored by known riboswitch classes (light and dark green) or that are obvious potential ligands for riboswitches that have yet to be discovered (red).

Figure V.. Mechanisms of ribozyme regulation by natural riboswitches.

(A) Configuration of a natural allosteric ribozyme wherein a c-di-GMP-II aptamer regulates group I ribozyme access to the 5′ splice site (5′ SS) [23]. (B) Common arrangement of bacterial riboswitches that regulate translation. The aptamer controls folding of the expression platform to hide or display the Shine-Dalgarno sequence (SD), which regulates ribosome binding [9]. (C) Common arrangement of eukaryotic riboswitches that regulate spliceosome access to 5′- or 3′-splice sites, or the branch site to control mRNA processing [94-101].