Base pairing between the 3' exon and an internal guide sequence increases 3' splice site specificity in the Tetrahymena self-splicing rRNA intron

Fitness landscape of a dynamic RNA structure

RNA structures are dynamic. As a consequence, mutational effects can be hard to rationalize with reference to a single static native structure. We reasoned that deep mutational scanning experiments, which couple molecular function to fitness, should capture mutational effects across multiple conformational states simultaneously. Here, we provide a proof-of-principle that this is indeed the case, using the self-splicing group I intron from Tetrahymena thermophila as a model system. We comprehensively mutagenized two 4-bp segments of the intron. These segments first come together to form the P1 extension (P1ex) helix at the 5' splice site. Following cleavage at the 5' splice site, the two halves of the helix dissociate to allow formation of an alternative helix (P10) at the 3' splice site. Using an in vivo reporter system that couples splicing activity to fitness in E. coli, we demonstrate that fitness is driven jointly by constraints on P1ex and P10 formation. We further show that patterns of epistasis can be used to infer the presence of intramolecular pleiotropy. Using a machine learning approach that allows quantification of mutational effects in a genotype-specific manner, we demonstrate that the fitness landscape can be deconvoluted to implicate P1ex or P10 as the effective genetic background in which molecular fitness is compromised or enhanced. Our results highlight deep mutational scanning as a tool to study alternative conformational states, with the capacity to provide critical insights into the structure, evolution and evolvability of RNAs as dynamic ensembles. Our findings also suggest that, in the future, deep mutational scanning approaches might help reverse-engineer multiple alternative or successive conformations from a single fitness landscape.

Design and Experimental Evolution of trans-Splicing Group I Intron Ribozymes

Group I intron ribozymes occur naturally as cis-splicing ribozymes, in the form of introns that do not require the spliceosome for their removal. Instead, they catalyze two consecutive trans-phosphorylation reactions to remove themselves from a primary transcript, and join the two flanking exons. Designed, trans-splicing variants of these ribozymes replace the 3′-portion of a substrate with the ribozyme’s 3′-exon, replace the 5′-portion with the ribozyme’s 5′-exon, or insert/remove an internal sequence of the substrate. Two of these designs have been evolved experimentally in cells, leading to variants of group I intron ribozymes that splice more efficiently, recruit a cellular protein to modify the substrate’s gene expression, or elucidate evolutionary pathways of ribozymes in cells. Some of the artificial, trans-splicing ribozymes are promising as tools in therapy, and as model systems for RNA evolution in cells. This review provides an overview of the different types of trans-splicing group I intron ribozymes that have been generated, and the experimental evolution systems that have been used to improve them.

Probing the role of a secondary structure element at the 5'- and 3'-splice sites in group I intron self-splicing: the tetrahymena L-16 ScaI ribozyme reveals a new role of the G.U pair in self-splicing

Several ribozyme constructs have been used to dissect aspects of the group I self-splicing reaction. The Tetrahymena L-21 ScaI ribozyme, the best studied of these intron analogues, catalyzes a reaction analogous to the first step of self-splicing, in which a 5'-splice site analogue (S) and guanosine (G) are converted into a 5'-exon analogue (P) and GA. This ribozyme preserves the active site but lacks a short 5'-terminal segment (called the IGS extension herein) that forms dynamic helices, called the P1 extension and P10 helix. The P1 extension forms at the 5'-splice site in the first step of self-splicing, and P10 forms at the 3'-splice site in the second step of self-splicing. To dissect the contributions from the IGS extension and the helices it forms, we have investigated the effects of each of these elements at each reaction step. These experiments were performed with the L-16 ScaI ribozyme, which retains the IGS extension, and with 5'- and 3'-splice site analogues that differ in their ability to form the helices. The presence of the IGS extension strengthens binding of P by 40-fold, even when no new base pairs are formed. This large effect was especially surprising, as binding of S is essentially unaffected for S analogues that do not form additional base pairs with the IGS extension. Analysis of a U.U pair immediately 3' to the cleavage site suggests that a previously identified deleterious effect from a dangling U residue on the L-21 ScaI ribozyme arises from a fortuitous active site interaction and has implications for RNA tertiary structure specificity. Comparisons of the affinities of 5'-splice site analogues that form only a subset of base pairs reveal that inclusion of the conserved G.U base pair at the cleavage site of group I introns destabilizes the P1 extension >100-fold relative to the stability of a helix with all Watson-Crick base pairs. Previous structural data with model duplexes and the recent intron structures suggest that this effect can be attributed to partial unstacking of the P1 extension at the G.U step. These results suggest a previously unrecognized role of the G.U wobble pair in self-splicing: breaking cooperativity in base pair formation between P1 and the P1 extensions. This effect may facilitate replacement of the P1 extension with P10 after the first chemical step of self-splicing and release of the ligated exons after the second step of self-splicing.

Extraordinarily slow binding of guanosine to the Tetrahymena group I ribozyme: implications for RNA preorganization and function

The Tetrahymena ribozyme derived from the self-splicing group I intron binds a 5'-splice site analog (S) and guanosine (G), catalyzing their conversion to a 5'-exon analog (P) and GA. Herein, we show that binding of guanosine is exceptionally slow, limiting the reaction at near neutral pH. Our results implicate a conformational rearrangement on guanosine binding, likely because the binding site is not prearranged in the absence of ligand. The fast accommodation of guanosine (10(2) to 10(3) x s(-1)) and prior structural data suggest local rather than global rearrangements, raising the possibility that folding of this and perhaps other large RNAs is not fully cooperative. Guanosine binding is accelerated by addition of residues that form helices, referred to as P9.0 and P10, immediately 5' and 3' to the guanosine. These rate enhancements provide evidence for binding intermediates that have the adjacent helices formed before accommodation of guanosine into its binding site. Because the ability to form the P9.0 and P10 helices distinguishes the guanosine at the correct 3'-splice site from other guanosine residues, the faster binding of the correct guanosine can enhance specificity of 3'-splice site selection. Thus, paradoxically, the absence of a preformed binding site and the resulting slow guanosine binding can contribute to splicing specificity by providing an opportunity for the adjacent helices to increase the rate of binding of the guanosine specifying the 3'-splice site.

A Tetrahymena thermophila ribozyme-based indicator gene to detect transposition of marked retroelements in mammalian cells

We devised an indicator gene for retrotransposition based on an autocatalytic ribozyme element--the Tetrahymena thermophila 23S rRNA group I intron--which can self-splice in vitro and does not require--at variance with nuclear mRNA introns--any specific pathway and cellular component for the completion of the splicing process. Several constructs, with the Tetrahymena intron adequately modified so as to be inserted at various positions within a neomycin-containing cassette under conditions that restore the neomycin-coding sequence after splicing out of the intron, were assayed for splicing efficiency in mammalian cells in culture. We show, both by northern blot analysis and by the recovery of neomycin activity upon retroviral transduction of the cassettes, that splicing efficiency depends on both the local base pairing and the global position of the intron within the neomycin transcript, and that some constructs are functional. We further show that they allow the efficient sorting out of retrotransposition events when assayed, as a control, with a human LINE retrotransposon. These indicator genes should be of great help in elucidating the mechanisms of transposition of a series of retroelements associated with transcripts not prone to nuclear mRNA intron splicing and previously not opened to any retrotransposition assay.

Design of highly specific cytotoxins by using trans-splicing ribozymes

We have designed ribozymes based on a self-splicing group I intron that can trans-splice exon sequences into a chosen RNA target to create a functional chimeric mRNA and provide a highly specific trigger for gene expression. We have targeted ribozymes against the coat protein mRNA of a widespread plant pathogen, cucumber mosaic virus. The ribozymes were designed to trans-splice the coding sequence of the diphtheria toxin A chain in frame with the viral initiation codon of the target sequence. Diphtheria toxin A chain catalyzes the ADP ribosylation of elongation factor 2 and can cause the cessation of protein translation. In a Saccharomyces cerevisiae model system, ribozyme expression was shown to specifically inhibit the growth of cells expressing the virus mRNA. A point mutation at the target splice site alleviated this ribozyme-mediated toxicity. Increasing the extent of base pairing between the ribozyme and target dramatically increased specific expression of the cytotoxin and reduced illegitimate toxicity in vivo. Trans-splicing ribozymes may provide a new class of agents for engineering virus resistance and therapeutic cytotoxins.

A protein encoded by a group I intron in Aspergillus nidulans directly assists RNA splicing and is a DNA endonuclease

Some group I introns self-splice in vitro, but almost all are thought to be assisted by proteins in vivo. Mutational analysis has shown that the splicing of certain group I introns depends upon a maturase protein encoded by the intron itself. However the effect of a protein on splicing can be indirect. We now provide evidence that a mitochondrial intron-encoded protein from Aspergillus nidulans directly facilitates splicing in vitro. This demonstrates that a maturase is an RNA splicing protein. The protein-assisted reaction is as fast as that of any other known group I intron. Interestingly the protein is also a DNA endonuclease, an activity required for intron mobilization. Mobile elements frequently encode proteins that promote their propagation. Intron-encoded proteins that also assist RNA splicing would facilitate both the transposition and horizontal transmission of introns.

Expression of a reporter gene interrupted by the Candida albicans group I intron is inhibited by base analogs

We previously reported the identification of an intron (CaLSU) in the 25S ribosomal RNA of some Candida albicans yeast strains. CaLSU was shown to self-splice and has the potential to adopt a secondary structure typical of group I introns. The presence of CaLSU inC. albicans strains correlates with a high degree of susceptibility to base analog antifungal agents, 5-fluorocytosine (5-FC) or 5-fluorouracil (5-FU). Cell death, resulting from addition of base analogs to growing cultures, precluded demonstration of a causal relationship between CaLSU presence and susceptibility to base analogs. In the present study, CaLSU was inserted in a non-essential lacZ reporter gene and expression was examined in Saccharomyces cerevisiae. Different mutations affecting in vitro self-splicing also had similar effects on reporter gene expression in vivo. This indicates that in vivo removal of CaLSU from the reporter gene occurs through the typical self-splicing mechanism of group I introns. Base analogs inhibited expression of the reporter gene product in a concentration-dependent manner upon their addition to the cultures. This supports a model in which disruption of intron secondary structure, consecutive to the incorporation of nucleotide analogs, is a major factor determining the susceptibility of C.albicans cells to base analogs.

An RNA fragment consisting of the P7 and P9.0 stems and the 3'-terminal guanosine of the Tetrahymena group I intron

On the basis of the nucleotide sequence of Tetrahymena group I intron, we constructed a 31 residue RNA that has the P7 stem and the 3'-terminal guanosine residue (3'-G) with a putative stem-loop structure (P9.0) intervening between them. For this model RNA (P7/P9.0/G), four residues around the guanosine binding site (GBS) in the P7 stem were found to exhibit much lower sensitivities to ribonuclease V1 than those of a variant RNA having adenosine in place of the 3'-G, suggesting that the 3'-G contacts around the GBS. NMR analyses of the imino proton resonances of the P7/P9.0/G RNA indicated that the base pairing in the GBS is retained on the interaction with the 3'-G, and that the two base pairs of the putative P9.0 stem-loop are definitely formed. Comparison of the RNA with its variants with either A (3'-A) or a deletion in place of the 3'-G suggested that the stability of the P9.0 stem-loop is affected by the GBS-3'-G interaction. The melting temperatures of the P9.0 stem-loop were determined from the UV absorbances of these RNAs, which quantitatively indicated that the P9.0 stem-loop is significantly stabilized by the interaction of the GBS with the 3'-G, rather than the 3'-A, and also by direct interaction with divalent cations (Mg2+, Ca2+ or Mn2+). Upon replacement of the G-C base pair by C-G in the GBS of the P7/P9.0/G RNA, the specificity was switched from 3'-G to 3'-A, as in the case of the intact intron.

Two group I introns with a C.G basepair at the 5' splice-site instead of the very highly conserved U.G basepair: is selection post-translational?

In virtually all of the 200 group I introns sequenced thus far, the specificity of 5' splice-site cleavage is determined by a basepair between a uracil base at the end of the 5' exon and a guanine in an intron guide sequence which pairs with the nucleotides flanking the splice-site. It has been reported that two introns in the cytochrome oxidase subunit I gene of Aspergillus nidulans and Podospora anserina are exceptions to this rule and have a C.G basepair in this position. We have confirmed the initial reports and shown for one of them that RNA editing does not convert the C to a U. Both introns autocatalytically cleave the 5' splice-site. Mutation of the C to U in one intron reduces the requirement for Mg2+ and leads to an increase in the rate of cleavage. As the C base encodes a highly conserved amino acid, we propose that it is selected post-translationally at the level of protein function, despite its inferior splicing activity.

The in vivo use of alternate 3'-splice sites in group I introns

Alternative splicing of group I introns has been postulated as a possible mechanism that would ensure the translation of proteins encoded into intronic open reading frames, discontinuous with the upstream exon and lacking an initiation signal. Alternate splice sites were previously depicted according to secondary structures of several group I introns. We present here strong evidence that, in the case of Podospora anserina nad 1-i4 and cox1-i7 mitochondrial introns, alternative splicing events do occur in vivo. Indeed, by PCR experiments we have detected molecules whose sequence is precisely that expected if the predicted alternate 3'-splice sites were used.

A phosphorothioate at the 3' splice-site inhibits the second splicing step in a group I intron

RNA polymerases can synthesize RNA containing phosphorothioate linkages in which a sulfur replaces one of the nonbridging oxygens. Only the Rp isomer is generated during transcription. A Rp phosphorothioate at the 5' splice-site of the Tetrahymena group I intron does not inhibit splicing (McSwiggen, J.A. and Cech, T.R. (1989) Science 244, 679). Transcription of mutants in which the first base of the 3' exon, U+1, was mutated to C or G, in the presence, respectively, of either cytosine or guanosine thiotriphosphate, introduced a phosphorothioate at the 3' splice-site. In both cases exon ligation was blocked. In the phosphorothioate substituted U+1G mutant, a new 3' splice-site was selected one base downstream of the correct site; despite the fact that the correct site was selected with very high fidelity in unsubstituted RNA. In contrast, the exon ligation reaction was successfully performed in reverse using unsubstituted intron RNA and ligated exons containing an Rp phosphorothioate at the exon junction site. Chirality was reversed during transesterification as in 5' splice-site cleavage (vide supra). This suggests that one non-bridging oxygen is particularly crucial for both splicing reactions.

Group I permuted intron-exon (PIE) sequences self-splice to produce circular exons

Circularly permuted group I intron precursor RNAs, containing end-to-end fused exons which interrupt half-intron sequences, were generated and tested for self-splicing activity. An autocatalytic RNA can form when the primary order of essential intron sequence elements, splice sites, and exons are permuted in this manner. Covalent attachment of guanosine to the 5' half-intron product, and accurate exon ligation indicated that the mechanism and specificity of splicing were not altered. However, because the exons were fused and the order of the splice sites reversed, splicing released the fused-exon as a circle. With this arrangement of splice sites, circular exon production was a prediction of the group I splicing mechanism. Circular RNAs have properties that would make them attractive for certain studies of RNA structure and function. Reversal of splice site sequences in a context that allows splicing, such as those generated by circularly permuted group I introns, could be used to generate short defined sequences of circular RNA in vitro and perhaps in vivo.

Splice site selection by intron aI3 of the COX1 gene from Saccharomyces cerevisiae

Interactions of the 5' and 3' splice sites with intron internal sequences of the yeast mitochondrial group I intron aI3 were studied using mutation analysis. The results can be fully explained by the splice guide model in which the splice sites are defined by the Internal Guide Sequence. No evidence was found for an alternative interaction between intron nucleotides preceding the 3' splice site and nucleotides in the vicinity of the core region as was found for the Tetrahymena intron. Our results also suggest that binding of the 5' and 3' splice site nucleotides to the IGS can not take place simultaneously. The intron must therefore undergo conformational changes as the reaction proceeds from the first step of self splicing, GTP attack at the 5' splice site, to exon ligation, the second step.

Mutational evidence for competition between the P1 and the P10 helices of a mitochondrial group I intron

A guanosine to cytosine transversion at position 2 of the fifth intron of the mitochondrial gene COB blocks the ligation step of splicing. This mutation prevents the formation of a base pair within the P1 helix of this group I intron--the RNA duplex formed between the 3' end of the upstream exon and the internal guide sequence. The mutation also reduces the rate of the first step of splicing (guanosine addition at the 5' splice junction) while stimulating hydrolysis at the 3' intron-exon boundary. Consequently, the ligation of exons is blocked because the 3' exon is removed prior to cleavage at the 5' splice junction. The lesion can be suppressed by second-site mutations that preserve the potential for base-pairing at this position. Because the P1 duplex and the P10 duplex (between the guide sequence and the 3' exon) overlap at the affected pairings represent alternative structures that do not, indeed cannot, form simultaneously.

The rate and specificity of a group I ribozyme are inversely affected by choice of monovalent salt

The fifth intron of the COB gene of yeast mitochondria splices autocatalytically. The rate of splicing is increased by high concentrations of monovalent salts, but the choice of both cation and anion is significant: The smaller the cation in solution, the faster the reaction (the rate in K+ greater than NH4+ greater than Na+ greater than Li+). Chloride, bromide, iodide and acetate salts enhance autocatalytic processing, but sulfate salts do not and fluoride salts are inhibitory. The choice of monovalent salt affects the KM of the intron for guanosine nucleotide, implying an alteration in the affinity of the RNA for that substrate. Under optimal conditions (1M KCl, 50 mM MgCl2) the catalytic efficiency of this intron exceeds that reported for the ribosomal intron from Tetrahymena, but several side reactions occur, including guanosine-addition within the downstream exon. The site of addition resembles the 5' splice junction, but selection of this site does not involve the internal guide sequence of the intron.

Splicing of COB intron 5 requires pairing between the internal guide sequence and both flanking exons

Group I introns are characterized by a set of conserved sequence elements and secondary structures. Evidence supporting the pairing of certain of these sequences has come from the comparison of intron sequences and from the analysis of mutations that disrupt splicing by interfering with pairing. One of the structures proposed for all group I introns is an internal guide sequence that base pairs with the upstream and the downstream exons, bringing them into alignment for ligation. We made specific mutations in the internal guide sequence and the flanking exons of the fifth intron in the yeast mitochondrial gene for apocytochrome b (COB). Mutations that disrupted the pairing between the internal guide sequence and the upstream exon (the P1 pairing) blocked addition of guanosine to the 5' end of the intron during autocatalytic reactions and prevented formation of the full-length circular intron. In contrast, transcripts containing mutations that disrupted the pairing between the guide sequence and the downstream exon (the P10 helix) initiated splicing but failed to ligate exons. Compensatory mutations that restored helices of normal stability mitigated the effects of the original mutations. These data provide direct evidence for the importance of the base pairing between the internal guide sequence and the downstream exon in the splicing of a wild-type group I intron.