Figure 1. Structural change of RNAT’s according to the environmental temperature. The SD stands for Shine-Dalgarno sequence, which is recognized and bind by ribosome to initiate translation. The AUG stands for start codon, from where the translation begins.
One example for this mechanism is the regulation of E.Coli’s rpoH gene (Figure 2). Responding to environmental temperature change, rpoH gene regulates the expression of the heat shock protein. Low temperature (30℃) induces a bend in the ribosome-binding site (RBS)-associated downstream box (DB) region, thereby interfering with ribosome binding. High temperature (42℃) disrupts the bend and initiates the process of translation (3).
Figure 2. a. Formation of stem III in the rpoH transcript at low temperatures (30 °C) induces a bend in the ribosome-binding site (RBS)-associated downstream box (DB) region, thereby interfering with ribosome binding. b. A rise in temperature to 42 °C opens stem III and stem I of the rpoH mRNA, liberates the AUG start codon and DB region, facilitates ribosome binding.
Inspired by such a mechanism, our group designed a series of RNATs whose SD sequence will have trap-release structural change according to the environmental temperature. The following is the schematic diagram of the RNATs we designed (Figure 3):
Figure 3. Schematic diagram of the RNATs we designed. The red box indicates the SD sequence.
The design of RNATs based on physical principles
To give RNAT sequences that meet the given parameters, the central problem is to predict RNATs’ secondary structure at a given temperature. Two methods are adapted according to the computer algorithm’s requirement.
One principle adapted in predicting RNA secondary structure is free energy minimization (4). A secondary structure with the least free energy is regarded to be the optimal solution (5).
Another principle adapted here is the partition function method (6). Rather than give one definite structure as the free energy minimization method, the partition function gives the probability of each secondary structure’s appearance. In the following equation, Q stands for the partition function and P (structure) stands for the probability of one specific structure’s appearance.
The design of RNATs adapting computer algorithms
The software RNAThermo for designing an RNA thermometer is presented. RNAThermo is based on the Vienna RNA Package, which is a program for RNA Secondary Structure Prediction and Comparison (7).
Before introducing RNAThermo, this is the basic RNAT design problem: given regulation temperature, the structure of the RNAT (both unmelted structure S1 and melted structure S2) and the SD sequence position of the RNAT, find an RNAT whose SD sequence will be released from the hairpin structure above the desired temperature.
A three-stage algorithm is designed to solve the problem (workflow is present in Figure 4): A set of sequences that fold into the S1 structure at low temperature is found. Sequences that cannot fold into the S2 structure at the high temperature are screened out. After these two stages, sequences meet the structural requirement are obtained. However, regulation temperatures of the obtained sequences remain unmeasured. The third stage is designed to screen RNATs to meet the temperature requirement.
Figure 4. Workflow of designing an RNA thermometer
Verification of the designed RNATs’ secondary structure
The first step in verification the in silico design is testing the designed structure in vitro. In-line probing method is adapted to measure the RNATs’ structure (10). The results are as shown in Figure 5.
Figure 5 Result of the in-line probing. The sequence of the RNAT is 5’-GAAUACAUGUUAAUUAUGCCAUCCAGGCAUACAGAAGAAGUUAAU-3’ and the regulation temperature of the RNAT is 39.5℃. RNAT loaded in lane 1, 2, 3 was incubated at 46℃ for 20h, 26h and 32h. RNAT loaded in lane 4, 5, 6 was incubated at 42℃ for 20h, 26h and 32h. RNAT loaded in lane 7, 8, 9 was incubated at 37℃ for 20h, 26h and 32h. The red boxes mark sections that melt when temperature rises.
When temperature rises, sections marked by the red boxes melt thus bands appear. The results show strong evidence that the designed RNATs can fold into desired secondary structure.
Verification of the designed RNATs’ temperature-sensing regulatory function
Then, rectification of the temperature-response regulatory function in vivo should be taken in verification of the in silico design. GFP is adapted as reporter gene in measuring the RNATs’ temperature-response regulatory function. The results are shown in Figure 6.
Figure 6 Schematic diagram of ‘RNAT + GFP’ gene.
E.Coli were cultured in 30℃ until they reached stationary phrase. Then the E.Coli were divided into two flasks. For the experimental group, a 45℃ heat shock was exerted to the E.Coli. For the control group, the temperature remained 30℃. Photos were taken after a two-hour adjustment. Two RNAT sequences were tested and the results are shown in Figure 8 and Figure 9.
Figure 8 Result of the verification of the RNAT’s regulatory function of the RNAT. The sequence is 5’-ACACGGAUCUACUAGCGUGAAUUUAUCACGGGAAGAAGUCGCCGUAA-3’. a. RNAT + GFP at 30℃. b. RNAT + GFP at 45℃. c. RNAT Only at 30℃. d. RNATOnly at 30℃. e. Histogram shows average intensity of the GFP’s luminance.
Figure 9. Result of the verification of the RNAT’s regulatory function of the RNAT. The sequence is 5’-GAAUACAUGUUAAUUAUGCCAUCCAGGCAUACAGAAGAAGUUAAT-3’. a. RNAT + GFP at 30℃. b. RNAT + GFP at 45℃. c. RNAT Only at 30℃. d. RNAT Only at 30℃. e. Histogram shows average intensity of the GFP’s luminance.
The results show strong evidence that the designed RNATs can function as desired.
Potential Application in Fermentation Industry
Computer aided RNAT design provides a new method for achieving controlled expression of products in fermentation industry. Engineered microorganisms sense a temperature signal and initiate the regulation.(Figure 10).
Figure 10. Schematic diagram of ‘RNAT + Signal Peptide + Lysozyme’ gene.
Reference
(1). Jens Kortmann and Franz Narberhaus. Bacterial RNA thermometers: molecular zippers and switches. NATURE REVIEWS MICROBIOLOGY, VOLUME 10, 265, APRIL 2012
(2). Birgit Klinkert and Franz Narberhaus. Microbial thermosensors. Cell. Mol. Life Sci. (2009) 66:2661–2676
(3). Miyo Terao Morita, Yoshiyuki Tanaka, Takashi S. Kodama, Yoshimasa Kyogoku,
Hideki Yanagi and Takashi Yura. Translational induction of heat shock transcription factor sigma32: evidence for a built-in RNA thermosensor.. Genes Dev. 1999 13: 655-665
(4). David H. Mathews. Revolutions in RNA Secondary Structure Prediction. J. Mol. Biol. (2006) 359, 526–532
(5). David H Mathews and Douglas H Turner. Prediction of RNA secondary structure by free energy minimization. Current Opinion in Structural Biology 2006, 16:270–278
(6). J. S. McCASKlLL. The Equilibrium Partition Function and Base Pair Binding Probabilities for RNA Secondary Structure. Biopolymers, Vol. 29,1105-1119 (1990)
(7). http://www.tbi.univie.ac.at/~ivo/RNA/
(8). http://www.tbi.univie.ac.at/~ivo/RNA/man/RNAfold.html
(9). L. Hofacker, W. Fontan. Fast folding and comparison of RNA secondary structures. Monatshefte fur Chemie , 125, 167-188.
(10). In-Line Probing Analysis of Riboswitches.Elizabeth E. Regulski and Ronald R. Breaker. NATURE PROTOCOL EXCHANGE http://www.nature.com/protocolexchange/protocols/1889
Acknowledgement
Thank Prof. CHEN Guoqiang, Prof. SUN Zhirong and Prof. DAI Junbiao for their guidance in the project. Thank Prof. Tom Kellie for his revision of the PPT and the report. Thanks Dr. YIN Ping and Dr. QU Peng for their help in the RNA experiments. Thanks FU Xiaozhi and LI Teng for their help in the molecular biology experiments.
"
