2Riboregulators are short RNA sequences that, upon binding to a ligand, change 3 their secondary structure and influence the expression rate of a downstream gene. 4 They constitute an attractive alternative to transcription factors for building synthetic 5 gene regulatory networks because they can be engineered de novo and they have a fast 6 turnover and a low metabolic burden. However, riboregulators are generally designed in 7 silico and tested in vivo, which only provides a yes/no evaluation of their performances, 8 thus hindering the improvement of design algorithms. Here we show that a cell-free 9 transcription-translation (TX-TL) system provides valuable quantitative information 10 about the performances of in silico designed riboregulators. In particular, we use the 11 ribosome as an exquisite molecular machine that detects functional riboregulators, pre-12 cisely measures their concentration and linearly amplifies the signal by generating a 13 fluorescent protein. We apply this method to characterize two types of translational 14 riboregulators composed of a cis-repressed (cr) and a trans-activating (ta) strand. At 15 the DNA level we demonstrate that high concentrations of taDNA poisoned the acti-16 vator until total shut off. At the RNA level, we show that this approach provides a 17 fast and simple way to measure dissociation constants of functional riboregulators, in 18 contrast to standard mobility-shift assays. Our method opens the route for using cell-19 free TX-TL systems for the quantitative characterization of functional riboregulators 20 in order to improve their design in silico.
21Keywords 22 in vitro synthetic biology, RNA translational riboregulator, cell-free protein synthesis 23 2 During the early wave of synthetic biology (1 , 2 ), known transcription factors were wired 25 to their corresponding promoter sequences to control the expression of other transcription 26 factors or effector proteins. While this approach has been very successful in engineering 27 gene regulatory networks (GRNs) (3 ) with few nodes, the number of different elements in 28 synthetic GRNs has stagnated at 5-6 (4 ). Two arguments may explain this limit. First, 29 protein-DNA interactions are very difficult to design, although very promising computation 30 methods are arising (5 ); the engineer must thus choose well-known transcription factor-31 promoter pairs. Second, the expression of these transcription factors imposes a metabolic 32 burden to the cells (6 ).
33Implementing regulatory circuits at the RNA level may help solving these issues because 34 RNA-RNA interactions can be predicted from the sequence (7 -9 ) and protein expression is 35 not needed for regulation, which lowers the metabolic burden. The principal component of 36 an RNA-regulated GRN is the riboregulator: an RNA sequence in the 5' untranslated region 37 (UTR) of a gene of interest that has an effect on its expression rate. Since they were first used 38 in synthetic biology more than a decade ago (10 ), several riboregulators have been desi...