Many techniques have been developed for studying inducible gene expression, but all of them are multicomponent systems consisting of cis-acting elements at the DNA or RNA level, trans-acting regulator proteins and/or small molecules as inducers. RNA Thermometers are the only known single-component regulators of gene expression. RNA Thermometers are thermosensors that regulate gene expression by temperature-induced changes in RNA conformation. Naturally occurring RNA Thermometers exhibit complex secondary structures which are believed to undergo a series of gradual structural changes in response to temperature shifts. RNA Thermometers often regulate genes required during either a heat shock or cold shock response, but have been implicated in other regulatory roles such as in pathogenicity, starvation, phage l development and expression of virulence genes.
They consist of a temperature-sensitive secondary structure in the 5′ untranslated region of the mRNA, which contains the ribosome-binding site. The ribosome-binding site can be masked or unmasked by a simple temperature shift, thereby repressing or inducing translation.
RNA thermometers are structurally simple and can be made from short RNA sequences; the smallest is just 44 nucleotides and is found in the mRNA of a heat-shock protein, Hsp17, in Synechocystis. Generally these RNA elements range in length from 60-110 nucleotides and they typically contain a hairpin with a small number of mismatched base pairs which reduce the stability of the structure, thereby allowing easier unfolding in response to a temperature increase.
At low temperatures, the mRNA adopts a conformation that masks the Shine-Dalgarno sequence, the sequence that properly aligns the start AUG codon in the P site of the 30S ribosomal unit, within the 50-untranslated region (50-UTR) and, in this way, prevents ribosome binding and translation. At elevated temperatures, the RNA secondary structure melts locally, thereby giving the ribosomes access to the ribosome binding site to initiate translation.
RNA Thermometers do not require binding of a ligand (metabolite) to induce the conformational change, but instead, directly respond to temperature. The highly complex RNA secondary structures into which most naturally occurring RNA Thermometers that can be folded has lead to the belief that RNA thermometers may not function as simple on/off switches, but rather represent dimmers, in that they may go through a series of distinct conformational changes which in turn may gradually change translation levels with temperature.
Many techniques have been developed for studying inducible gene expression, but all of them are multicomponent systems consisting of cis-acting elements at the DNA or RNA level, trans-acting regulator proteins and/or small molecules as inducers. RNA Thermometers are the only known single-component regulators of gene expression. RNA Thermometers are thermosensors that regulate gene expression by temperature-induced changes in RNA conformation. Naturally occurring RNA Thermometers exhibit complex secondary structures which are believed to undergo a series of gradual structural changes in response to temperature shifts. RNA Thermometers often regulate genes required during either a heat shock or cold shock response, but have been implicated in other regulatory roles such as in pathogenicity, starvation, phage l development and expression of virulence genes.
All known RNA thermometers are cis-acting modulators that control translation initiation by masking the Shine-Dalgarno sequence at low temperatures. At increasing temperatures, the thermometer structure melts and thereby permits ribosome access.
The Shine-Dalgarno sequence is a ribosomal binding site in the mRNA, generally located 8 bases upstream of the start codon AUG. This sequence helps recruit the ribosome to the mRNA to initiate protein synthesis by aligning it with the start codon.
Two quite distinct RNA thermometers controlling the bacterial heat-shock response have been characterised. The cellular level of the heat sigma factor σ32 in Escherichia coli is adjusted in part by an RNA structure consisting of two segments (regions A and B) within the coding sequence of the rpoH mRNA. Extensive mutational analyses and structure probing experiments demonstrated that secondary structure formation between the two regions is disrupted by temperature increases. A second well-studied RNA thermometer is the ROSE (repression of heat shock gene expression) element, which was discovered in the nitrogen-fixing soybean symbiont Bradyrhizobium japonicum. Meanwhile, 17 ROSE-like sequences have been described in different Rhizobium species and in Agrobacterium tumefaciens. They are all located in the 5'-UTR of small heat-shock genes and are 70–120 nt long. Computer-assisted secondary structure predictions suggest a conserved RNA structure composed of three or four hairpins. In each case, the SD sequence and the AUG start codon are masked by imperfect base pairing in the highly conserved 3’-proximal stem-loop. A complex architecture comprising paired regions, internal loops and a bulged G residue is necessary for proper thermosensing. Spectro-scopic studies and RNaseH experiments using a synthetic ROSE sequence revealed temperature-mediated structural alterations in the RNA thermometer. There is no evidence for the contribution of protein factors in temperature sensing, either in vivo or in vitro but many signals to the RNA . In the present study we present evidence that ROSE-like thermometers are not Rhizobium-specific regulatory elements, but regulate translation of small heat-shock genes in numerous other bacteria.
Approximately 120 completed genome sequences from archaea and bacteria were searched for genes encoding small heat-shock proteins. Different serotypes of the same species were not considered. As observed previously, some bacteria do not contain any small heat-shock genes. Most known ROSE elements are between 60 and 100 nt long and bear significant sequence similarity in the second half. A total of 27 previously unknown structures originating from 18 different bacterial genomes showed the typical characteristics of ROSE-like RNA thermometers (a-proteobacteria: Bartonella henselae ibpA2, Bartonella quintana ibpA2, Brucella suis ibpA and hspA, Caulobacter crescentus CC2258 and CC3592, Rhodopseudomonas palustris RPA0054 and hspD, Sinorhizobium meliloti ibpA and b21295; g-proteobacteria: Erwinia carotovora ibpA and ibpB, Escherichia coli ibpA and ibpB, Pseudomonas aeruginosa ibpA, Pseudomonas putida ibpA, Pseudomonas syringae PSPT02170, Salmonella typhimurium ibpA and ibpB, Shewanella oneidensis ibpA, Shigella flexneri ibpA and ibpB, Vibrio cholerae hspA, Vibrio parahaemolyticus hspA, Vibrio vulnificus hspA, Yersinia pestis ibpA and ibpB). Although the overall sequence similarity to known ROSE elements is low, the core region forming this functionally critical part of the structure is highly conserved in sequence. Nucleotides involved in base-pairing are conserved, while loop-forming sequences are diverse. Most interestingly, ROSE-like structures are not limited to Rhizobiaceae. Similar structures were identified in other α-proteobacteria such as Brucella suis and Caulobacter crescentus. Potential ROSE elements were also found upstream of small heat-shock genes in enteric bacteria and other important representatives of the γ-proteobacteria such as Vibrio and Pseudomonas. Regulation by ROSE-like thermometers might be restricted to these two bacterial lineages, as related structures were not found in the genome sequences of any other group of bacteria.
Another feature that is shared by all ROSE-type RNA thermometers is the final hairpin, which contains the SD sequence and a bulged G residue.
Mutations in the Shine-Dalgarno sequence can reduce or increase translation. This change is due to a reduced or increased mRNA-ribosome pairing efficiency, as evidenced by the fact that complementary mutations in the anti-Shine-Dalgarno sequence can restore translation.
The critical importance of this substructure is supported by the finding that the four bases involved in masking of the SD sequences are conserved in all currently known ROSE elements. Apparently, the ROSE element is the most widespread RNA thermometer known to date. Other RNA thermometers are either restricted to closely related species, such as the intragenic RNA thermometer in rpoH genes of g-proteobacteria, or are unique, such as the 5’- UTRs of the lcrF and prfA genes in Yersinia pestis and Listeria monocytogenes, respectively. The diversity among the currently known examples suggests that there is potential for many new RNA thermometers to be discovered.
RNA thermometers have an interesting novel feature in common. In addition to the thermometer-mediated translational control, the corresponding genes are also regulated at the transcriptional level by a σ32 promoter. The dependence of the E. coli ipbAB operon on σ32 has been described previously. The similarity of the S. typhimurium ibpA and C. crescentus CC2258 promoters with the σ32 con- sensus sequence is strongly indicative of an equivalent regulation. In contrast, the hspArpoH1 operon of B. japonicum is transcribed from a temperature-independent s70-type promoter. Transcription of ROSE-controlled genes from housekeeping promoters has also been suggested for other rhizobia and A. tumefaciens. Regardless of the promoter used, the transcript levels of ROSE-controlled genes correlates with temperature, i.e., small heat-shock gene transcripts are abundant only at high temperatures. While temperature-dependent transcript levels are easily conceivable in the case of s32-controlled promoters, it is more difficult to explain the s70-dependent promoters. It has been suggested that transcripts carrying the highly structured ROSE sequence are prone to ribonucleolytic attack at low temperatures. The question remains as to why bacteria such as E. coli, S. typhimurium and C. crescentus use a dual system to control expression of their small heat-shock genes. Combining transcriptional and translational control might offer at least three advantages:
(i) Allowing transcription of a gene only when the corresponding product is needed certainly saves resources. In this respect, solely posttranscriptional control mechanisms are expected to pose a metabolic burden on the cell, as any RNA that is not translated would raise the metabolic costs.
(ii) Recent mathematical models of heat-shock response networks suggest that simple systems based on an RNA thermometer are sufficient to respond adequately to temperature changes. However, to achieve robustness in a fluctuating environment, more complex systems, including feedback control mechanisms, are required. Combining an RNA thermometer with a temperature-regulated σ32 promoter results in a transient heat-shock response, since expression of the σ32 regulon is feedback-controlled by a circuit involving cellular chaperones and proteases.
(iii) Implementing a σ32 promoter upstream of the ROSE element might offer yet another benefit. While an RNA thermometer monitors temperature deviations directly, the σ32 regulon integrates signals associated with a heat shock, such as denatured proteins. Such signals can also be induced independently of temperature changes, for example by protein overexpression or treatment with antibiotics. Thus, the σ32 promoter upstream of an RNA thermometer provides an additional input module allowing the cell to integrate multiple, and not necessarily related signals in a coordinated fashion.
"
