Team:Peking R/Project/RNAToolkit3

Basal-Level-Free Genetic Rheostats

Introduction
The relatively simple and repetitive composition of RNA molecules makes them easy to synthesize and manipulate. Evidence shows that RNA also has the structural sophistication needed for dynamic interplay between functionally distinct catalytic and receptor domains[1]. In our project, we tried to utilize these small genetic rheostats to tune translation rate of target gene quantitatively.

Numerous riboswitches and hammerhead ribozymes have been depicted previously [2,3]. However, they all function via hiding or exposing the ribosomal binding site (RBS, which is a major determinant of translation rate) through different mechanisms (Figure 1a and 1b) [3,4]. This will probably lead to a relatively high basal level because of the conservation of intact RBS. We speculated that, with the self-splicing Group I intron ribozymes that will be further described later, we could interrupt the RBS sequence via inserting a ribozyme to remove the basal level of translation rate and manipulate translation rate via modulating fraction of ligand-induced self-splicing (Figure 1c).

	Figure 1. Predicted mechanisms for different allosteric RNA-mediated gene expression control. (a) Allosteric riboswitch. In the absence of effector molecules----theophylline in this example (left), the 5' region can adopt a highly folded structure that extensively pairs the region that includes ribosomal binding site to part of the aptamer sequence (shown in green). In the presence of theophylline (right), the secondary structure shifts, such that the ribosomal binding site is exposed [high]. (Sean A. Lynch, Shawn K. Desai, Hari Krishna Sajja, and Justin P. Gallivan. (2006). A high-throughput screen for synthetic riboswitches reveals mechanistic insights into their function. Chemistry & Biology 14, 173-184) (b) Hammerhead aptazyme. Without the inducer (left)----thiamine pyrophosphate (TPP), hammerhead ribozyme is stabilized in a misfolded state with the ribosomal binding site sequestered by the ribozyme stem I. The addition of TPP (right) unmasks the ribosomal binding site through self-cleaving (Markus Wieland, Armin Benz, Benedikt Klauser, and JÖrg S. Hartig. (2009). Artificial ribozyme switches containing natural riboswitch aptamer domains. RNA technology 121, 2753-2756). (c) Group I intron aptazyme. Precursor mRNA (right) with the ribosomal binding site separated by the group I intron. RNA processed in the presence of c-di-GMP (left) creates a perfect ribosomal binding site (Elaine R.Lee, et al. (2010). An allosteric self-splicing ribozyme triggered by a bacterial second messenger. Science 329, 845-848). Green: ribosomal binding site, black: open reading frame (ORF), red arrow: hammerhead ribozyme cleavage site, 5'-ss and 3'-ss: group I self-splicing ribozyme corresponding splicing site. Start codons and effector molecule theophylline are indicated in the figure. Watson-Crick base pairs are denoted by dashes, all other base pairs are indicated as dots. All of these three elements are used in our project deriving from corresponding reference listed after the legends. The structures shown upper might be a little different from the original design in the listed paper, such us the start codon----AUG instead of the atypical ones----UUG in (c).

In fact, this mechanism has been depicted in the pathogenic bacterium Clostridium difficile in nature with an allosteric group I ribozyme inside the RBS, wherein self-splicing is regulated by a distinct aptamer class that senses the bacterial second messenger c-di-GMP [5]. Here, we regard this ribozyme as a prototype to evolve ligand-regulated self-splicing ribozymes. Before those, we will first make a brief introduction about group I intron ribozyme.

Group I intron ribozymes
Group I intron ribozymes consist of two parts----an aptamer which directly senses the small effector molecule and an intron which self-splices in response to the conformational change transmitted from the aptamer.

An aptamer is a piece of nucleic acid or peptide that binds to specific ligand. Here, we refer to RNA aptamers. Various kinds of aptamers have been discovered, among which is the one responses to 3'-5'-cyclic diguanylic acid (c-di-GMP). C-di-GMP is a second messenger signaling molecule that regulates many vital processes within the bacterial kingdom including motility, pathogenesis and biofilm formation [6,7]. Its corresponding aptamers can be categorized into two classes and the binding mechanism of class I is illustrated in Figure 2 and Figure 3 [7].

	Figure 2. Comparison of the consensus sequences and structural models for class I (left) and class II (right) c-di-GMP aptamers. (Elaine R.Lee, et al. (2010). An allosteric self-splicing ribozyme triggered by a bacterial second messenger. Science 329, 845-848)

	Figure 3.  Structure of the class I c-di-GMP aptamer. (a) Secondary structure representation of the crystallized c-di-GMP aptamer. Helices P1, P2, and P3 are colored purple, blue, and green respectively. C-di-GMP is shown in red. Nucleotides that directly contact c-di-GMP are shown in orange. The U1A cocrystallization protein is shown in gray. Watson-Crick base pairs are denoted by dashes, all other base pairs are indicated using the nomenclature of Westhof and Leontis. A* indicates a Watson-Crick base pair in the tertiary structure. The two guanines of c-di-GMP are denoted Gα and Gβ for clarity. (b) Representation of crystal structure of the Vc2 c-di-GMP riboswitch aptamer from V. cholera. Coloring is the same as used in (a). (Kathryn D Smith, Sarah V Lipchock, Tyler D Ames, Jimin Wang, Ronald R Breaker, and Scott A Strobel. (2009). Structural basis of ligand binding by a c-di-GMP riboswitch. Nat Struct Mol Biol. 16, 1218-1223)

An intron is a non-coding sequence found within the genes of all three domains of life. Introns are classified into four groups based on splicing mechanisms. Group I and group II introns are able to self-splice using different mechanisms [8]. However, assistance from proteins is required for in vivo splicing for Group II intron. All Group I introns share a conserved secondary structure consists of nine paired regions (P1-P9 shown in Figure 4) [9]. Splicing of Group I introns is processed by two sequential ester-transfer reactions. The exogenous guanosine or guanosine nucleotide first docks onto the active G-binding site located in P7, and its 3'-OH is aligned to attack the phosphodiester bond at the 5' splice site located in P1, resulting in a free 3'-OH group at the upstream exon and the exoG being attached to the 5' end of the intron. Then the terminal G (ΩG) of the intron swaps the exoG and occupies the G-binding site to organize the second ester-transfer reaction, the 3'-OH group of the upstream exon in P1 is aligned to attacks the 3' splice site in P10, leading to the ligation of the adjacent upstream and downstream exons and free of the catalytic intron [10] (Figure 5).

	Figure 4. Overall secondary and tertiary structure of the Azoarcus pre-2S intron splicing complex. (a) Intron secondary structure. The intron sequences, exon sequences and structural elements (P and J elements) are depicted in the colours. The ligation reaction catalyzed by this complex is shown by black arrows. The RNA transcript (UP62) comprising the majority of the intron is shown with capital letters, while residues derived from the two chimaeric oligonucleotides dCIRC (intron/3'-exon segment) and CAT (5'-exon segment) are shown in lower-case letters. The break between dCIRC and UP62 is indicated with a jagged line. (b) Overall structure of the intron. The backbone is depicted with a ribbon and individual bases depicted as cylinders. The two tetraloop-tetraloop receptor interactions on the periphery of the intron are indicated. (Peter L. Adams, Mary R. Stahley, Anne B. Kosek, Jimin Wang, and Scott A. Strobel. (2004). Crystal structure of a self-splicing group I intron with both exons. Nature 430, 45-50)

	Figure 5. Mechanism of group I intron splicing. 5' and 3' exons are in red and blue, respectively. ΩG, terminal intron guanine. G*, exogenous guanosine. (Step I) Nucleophilic attack on the 5' splice site by the 3'-OH of G* in guanosine-binding site (GBS). (Step 2) Nucleophilic attack on the 3' splice site by the free 3'-OH of the 5' exon. (Step 3) Free intron and spliced exons. (Rahul Raghavan and Michael F. Minnick (2009). Group I Introns and Inteins: Disparate Origins but Convergent Parasitic Strategies. American Society for Microbiology 191, 6193-6202)

Engineering of Ligand-Responsive Self-splicing
We firstly examined the utility of naturally emerging c-di-GMP ribozyme in E.coli with a circuit design shown in Figure 6. It was anticipated that the self-splicing could be chemically induced by c-di-GMP in vivo and the expression level of GFP would increase with increasing ligand concentration.

	Figure 6. The plasmid used in this project (left) and key features of the c-di-GMP aptamer-ribozyme construct (right), including validated splice and GTP attack sites. GTP1 and GTP2 indicate two different GTP attacking sites. Alternative base-pair interactions guiding allosteric function are shaded blue and green. (Elaine R.Lee, et al. (2010). An allosteric self-splicing ribozyme triggered by a bacterial second messenger. Science 329, 845-848)

However, no change in GFP expression was observed during in vivo chemical induction (Figure 7). This is probably due to impenetrability of the second messenger c-di-GMP when it diffuses through cell membrane. Therefore, it's not proper to use c-di-GMP as a ligand to carry experiments in vivo.

	Diagram 1. C-di-GMP inducing experiment. Activity of c-di-GMP-dependent allosteric ribozyme reported by green fluorescent protein (GFP). All are incubated for 120 minutes in the presence of various concentrations of the second messenger c-di-GMP and 1mM arabinose(Only the highest concentration used in this work is shown). C-di-GMP ribozyme used here has been described in Figure 6a previously. The numbers over the column is corresponding c-di-GMP concentration used. All are excitated at 470nm with investigating their absorption at 509nm.

We then swapped c-di-GMP aptamer with a theophylline hammerhead ribozyme which was found by Ronald R. Breaker in 2000 [3]. We designed three alternative base-pairing structures, the ribozyme P1 stem (Figure 7 dark blue boxing), hammerhead stem I (orange shading) and the extending P10 stem (blue shading), which may explain theophylline inducing control. Misfolded aptamer formation disrupts base pairing of the ribozyme P1 stem and further affects the structure of the group I intron. Because theophylline binding stabilizes aptamer substructures, presence of theophylline should favor the formation of hammerhead ribozyme stem I and induced the cleavage of the hammerhead. This process induced the correct formation of the group I intron thus endow it with the self-splicing activity which finally brings on the formation of RBS----AGGAGG.

	Figure 7. Rational design of engineered theophylline ribozyme. Proposed mechanism for engineered theophylline ribozyme-mediated gene control (a and b). (a-b) In the absence of theophylline, the correct structure of group I intron is disrupted. As theophylline binding stabilizes aptamer substructures, presence of theophylline should favor the formation of hammerhead ribozyme stem I. This leads to cleavage of the hammerhead ribozyme and further splicing of the group I intron which finally brings on the formation of RBS----AGGAGG (shown in red letters separated). (c) Proposed secondary structure of the positive control in theophylline inducing experiment. (d) Map of the plasmid using in this part of the project. Orange: hammerhead stem I, blue: ribozyme P1 stem, black: open reading frame (ORF), dark blue box: anti RBS stem, red letters separated: ribosomal binding site (RBS), 5'ss and 3'ss: group I self-splicing ribozyme corresponding splicing site, red arrow: hammerhead ribozyme cleavage site, theophylline aptamer, hammerhead ribozyme and the elements in plasmid map are illustrated in this figure. Watson-Crick base pairs are denoted by dashes, all other base pairs are indicated as dots.

As the Diagram 2 shows below, addition of theophylline indeed facilitated the expression of GFP, which implies that the domain swapping succeeded.

	Diagram 2. Theophylline inducing experiment. Activity of theophylline-dependent engineered group I intron reported by green fluorescent protein (GFP). All are incubated for 120 minutes in the presence or absence of 10mM theophylline and 1mM arabinose. The plasmid used in this work is theophylline-dependent engineered group I intron described in Figure 7a,7b,7d. * is the positive control ---- hammerhead-group I ribozyme without specific response described in Figure 7c. All are excitated at 470nm with investigation their absorption at 509nm.

Discussion
Genetic rheostats are promising gene regulatory elements for efficient operation of cellular processes due to their modularity and pre-translational fast-response. More and more RNA regulatory elements have been studied and an easier approach to combine them and cluster their properties is need urgently. Breaker and his co-workers have previously demonstrated that the cleavage activity of hammerhead ribozyme can be coupled with the binding of molecular effectors to corresponding aptamer by simple expedient of appending aptamers to non-essential stem regions [3,11,12]. However, in addition to hammerhead ribozyme, there are still lots of available RNA regulatory elements with complicated structures that are difficult to engineer [13].

Here, we proposed a design in which the activity of other complicated ribozyme, including group I intron, can be easily modulated, combining with engineered hammerhead aptazymes. Hammerhead ribozyme here was used as an “Adapter” to link the binding of effector molecules to specific aptamer to the self-splicing activity of group I intron as Figure 8 and Figure 9 shows previously in this project. Under this design, we can easily swap the hammerhead aptazyme, the adapter, to construct engineered group I introns that self-splice in response to different effector molecules, without further in vitro selection which is time consuming and labor resume. In summary, this design is fast, affordable and more predictable with general interest.

Reference:

1. Ronald R Breaker. (2002). Engineered allosteric ribozymes as biosensor components. Current Opinion in Biotechnology 13, 31-39
2. Markus Wieland, Armin Benz, Benedikt Klauser, and JÖrg S. Hartig. (2009). Artificial ribozyme switches containing natural riboswitch aptamer domains. RNA technology 121, 2753-2756
3. Garrett A. Soukup, Gail A. M. Emilsson and Ronald R. Breaker. (2000). Altering molecular recognition of RNA aptamers by allosteric selection. J. Mol. Biol. 298, 623-632
4. Sean A. Lynch, Shawn K. Desai, Hari Krishna Sajja, and Justin P. Gallivan. (2006). A high-throughput screen for synthetic riboswitches reveals mechanistic insights into their function. Chemistry & Biology 14, 173-184
5. Elaine R.Lee, et al. (2010). An allosteric self-splicing ribozyme triggered by a bacterial second messenger. Science 329, 845-848
6. Regine Hengge. (2009). Principles of c-di-GMP signaling in bacteria. Nature 7, 263-273
7. Kathryn D Smith, Sarah V Lipchock, Tyler D Ames, Jimin Wang, Ronald R Breaker, and Scott A Strobel. (2009). Structural basis of ligand binding by a c-di-GMP riboswitch. Nat Struct Mol Biol. 16, 1218-1223
8. Rahul Raghavan and Michael F. Minnick (2009). Group I Introns and Inteins: Disparate Origins but Convergent Parasitic Strategies. American Society for Microbiology 191, 6193-6202
9. Sarah A Woodson. (2005). Structure and assembly of group I introns. Sciencedirect 15, 324-330
10. Peter L. Adams, Mary R. Stahley, Anne B. Kosek, Jimin Wang, and Scott A. Strobel. (2004). Crystal structure of a self-splicing group I intron with both exons. Nature 430, 45-50
11. Michael P. Robertson and Andrew D. Ellington. (1999). Design and optimization of effector-activated ribozyme ligases. Nucleic Acids Research 28, 1751-1759
12. Michael P. Robertson and Andrew D. Ellington. (1999). In vitro selection of an allosteric ribozyme that transduces analytes to amplicons. Nature Biotech 17, 62-66
13. Kristin M Thompson, Heather A Syrett, Scott M Knudsen and Andrew D Ellington. (2002). Group I aptazymes as genetic regulatory switches. BMC biotechnology 2, 21