RNA thermometers (RNATs) are RNA sequences that range from 40 to more than a 100 nucleotides commonly found in the 5' untranslated region of some genes and that regulate in cis their translation without the need of other factors [Kortmann and Narberhaus, (2012); Narberhaus, (2009)]. These RNAT sequences show certain three dimensional structures, some of which interact with the ribosome binding site (RBS) of their regulated genes and hinders the proccessivity of the ribosome complex at certain temperatures. The dynamics of the formation of these structures is temperature dependent and is the basis of the regulation of the translation rate of a given transcript [Chowdhury, S., et al.,(2006); Narberhaus, F., et al.,(2006)].

Functional RNAT have been found in different organisms, mainly pathogenic bacteria, and many others have been predicted in almost everyfrom a number of bioinformatic studies. They have been found to regulate the expression of virulence factors, heat and cold shock response factors and even proteins involved the development of some bacteriophages.

Their apparent widespread presence in living organisms has made RNATs attractive for some applications, specially the ones related to the replacement of chemical inducers and for the development of new drugs.

However, from the experience of those who have been working extensively with RNAT in the later years, the accurate bioinformatic prediction of functional RNAT has proven to be an exceptionally difficult task; the reasons for this are pointed to be the poor sequence conservation observed among RNATs and the gaps in our current understanding of the RNAT function, their structural diversity and the effect of other regulatory sequences far from the RBS region [Kortmann and Narberhaus, (2012); Waldminghaus, et al., (2007)].

The discovery of new RNATs has relied on a mixed approach that involves bioinformatics and experimental validation, as well as approaches that involve mutational libraries, synthetic constructions and directed evolution.

Even when the naturally found RNATs usually regulate the expression of transcription factors, the synthetic constructions made so far have focused mainly to characterize the effect of a given RNAT using a reporter protein (LacZ or a fluorescent protein) directly downstream of a RNAT. In our work, we intend to prove that RNATs can also be employed to effectively regulate the expression of transcription factors in synthetic circuits and point at possible applications for the circuit topologies that would be made feasible with this new kind of synthetic regulatory device.

Although RNATs show almost no sequence similarity among them, a number of structural features can be used to classify them. Here we enlist the most described RNATs structural groups described to date:

1. ROSE.- ROSE stands for "Regulation Of heat Shock Expression". ROSE elements are 60 to >100 nucleotide sequences found upstream of heat shock proteins. They have been found to be conserved in alpha and gamma-proteobacteria. Among the structural features of the ROSE element family are: a) their folding in 2 to 4 stemloop structures; b) a short conserved sequence (UU/CGCU) near the Shine-Dalgarno sequence; and c) the presence of a number of non-cannonical base interactions (the G83-G94 pair; a triple bair among U96-C80-C81; the U79-U97; and the interaction of the AUG codon and C71, G72 and U73.
2. FourU elements
3. Synechocystis element