The 5' and 3' splice sites come together via a three dimensional diffusion mechanism

The mechanisms of a mammalian splicing enhancer

Exonic splicing enhancer (ESE) sequences are bound by serine & arginine-rich (SR) proteins, which in turn enhance the recruitment of splicing factors. It was inferred from measurements of splicing around twenty years ago that Drosophila doublesex ESEs are bound stably by SR proteins, and that the bound proteins interact directly but with low probability with their targets. However, it has not been possible with conventional methods to demonstrate whether mammalian ESEs behave likewise. Using single molecule multi-colour colocalization methods to study SRSF1-dependent ESEs, we have found that that the proportion of RNA molecules bound by SRSF1 increases with the number of ESE repeats, but only a single molecule of SRSF1 is bound. We conclude that initial interactions between SRSF1 and an ESE are weak and transient, and that these limit the activity of a mammalian ESE. We tested whether the activation step involves the propagation of proteins along the RNA or direct interactions with 3' splice site components by inserting hexaethylene glycol or abasic RNA between the ESE and the target 3' splice site. These insertions did not block activation, and we conclude that the activation step involves direct interactions. These results support a model in which regulatory proteins bind transiently and in dynamic competition, with the result that each ESE in an exon contributes independently to the probability that an activator protein is bound and in close proximity to a splice site.

Early history of circular RNAs, children of splicing

In this commentary we briefly summarize early work on circular RNAs derived from spliceosome mediated circularization. We highlight how this early work inspired work on the basic mechanisms of nuclear RNA splicing, the possible function of circular RNAs and the potential uses of circular RNAs as tools in biomedicine. Recent developments in the study of circular RNAs, summarized in this volume, have brought these questions back to the foreground.

An RNA splicing enhancer that does not act by looping

Out of the loop: Do the proteins bound to an enhancer site on pre-mRNA interact directly with the splice site by diffusion (looping), as is generally accepted, or does the intervening RNA play a role? By inserting a PEG linker between an enhancer sequence and alternative splice sites, the interaction of these two elements can be studied. Intervening RNA was essential for the enhancer activity, which rules out the looping model.

Bridged oligonucleotides as molecular probes for investigation of enzyme-substrate interaction and allele-specific analysis of DNA

The efficiency of enzymatic conversion of DNA complexes containing non-nucleotide inserts has been studied. T4 DNA ligase and Taq DNA polymerase have been included in the study as examples of widely used DNA-dependent enzymes. A series of substrate DNA complexes have been formed using native oligonucleotides and bridged ones bearing non-nucleotide inserts based on phosphodiesters of di-, tetra-, or hexaethylene glycol, 1,5-pentanediol, 1,10-decanediol, and 3-hydroxy-2(hydroxymethyl)-tetrahydrofuran. The perturbation in DNA located far from the site of the enzyme action had almost no influence on the substrate properties of the complex, while insertion near this site significantly deteriorated them. The use of a series of modified duplexes allows one to locate the position of the enzyme-binding site on DNA substrate with the accuracy of 1-2 nucleotides. The presence of a non-nucleotide insert in the complex has been also shown to enhance the efficiency of single mismatch discrimination upon both template-directed ligation and extension of oligonucleotides.

Isotope labeling strategies for NMR studies of RNA

The known biological functions of RNA have expanded in recent years and now include gene regulation, maintenance of sub-cellular structure, and catalysis, in addition to propagation of genetic information. As for proteins, RNA function is tightly correlated with structure. Unlike proteins, structural information for larger, biologically functional RNAs is relatively limited. NMR signal degeneracy, relaxation problems, and a paucity of long-range (1)H-(1)H dipolar contacts have limited the utility of traditional NMR approaches. Selective isotope labeling, including nucleotide-specific and segmental labeling strategies, may provide the best opportunities for obtaining structural information by NMR. Here we review methods that have been developed for preparing and purifying isotopically labeled RNAs, as well as NMR strategies that have been employed for signal assignment and structure determination.

Alternative splicing: therapeutic target and tool

Alternative splicing swells the coding capacity of the human genome, expanding the pharmacoproteome, the proteome that provides targets for therapy. Splicing, both constitutive and regulated forms, can itself be targeted by conventional and molecular therapies. This review focuses on splicing as a therapeutic target with a particular emphasis on molecular approaches. The review looks at the use of antisense oligonucleotides, which can be employed to promote skipping of constitutive exons, inhibit inappropriately activated exons, or stimulate exons weakened by mutations. Additionally this manuscript evaluates methods that reprogram RNAs using reactions that recombine RNA molecules in trans. Preliminary, but exciting, results in these areas of investigation suggest that these methods could eventually lead to treatments in heretofore intractable ailments.

Intergenic transcripts in genes with phase I introns

Nonsense-mediated mRNA decay (NMD) is a eukaryotic quality-control mechanism that detects and degrades aberrant transcripts prematurely terminating translation. NMD may be elicited by intergenic transcripts that contain premature termination codons (PTCs), but chimeric mRNAs of genes that have introns of identical phase would be predicted to lack PTCs and escape NMD. We examined intron phase I-containing HLA class II genes for the presence of intergenic mRNAs and found an extraordinary diversity of correctly spliced and polyadenylated intergenic transcripts. They lacked a significant homology at the chimeric joins and had no PTCs. Their expression levels were very low and positively correlated with the expression of natural transcripts. In contrast, pair-wise mixtures of separately transcribed plasmids carrying full-length HLA-DQB1, -DQA1, -DRB1, and -DRA cDNAs produced only hybrid molecules that lacked canonical exon boundaries, had homologous chimeric joins, and occasionally contained PTCs, implicating in vitro artifacts generated by template switching of Taq polymerase and reverse transcriptase. A differential exon structure of hybrid molecules observed in vitro and in cellular RNA preparations suggests that intergenic mRNAs with canonical exon boundaries arise in vivo during exon joining and/or transcription. Since the observed intergenic mRNAs may encode mixed class II heterodimers that were previously shown to present antigens it will be interesting to determine functional properties of such molecules in future studies.

Messenger RNA reprogramming by spliceosome-mediated RNA trans-splicing

In the human genome, the majority of protein-encoding genes are interrupted by introns, which are removed from primary transcripts by a macromolecular enzyme known as the spliceosome. Spliceosomes can constitutively remove all the introns in a primary transcript to yield a fully spliced mRNA or alternatively splice primary transcripts leading to the production of many different mRNAs from one gene. This review examines how spliceosomes can recombine two primary transcripts in trans to reprogram messenger RNAs.

Co-transcriptional splicing of pre-messenger RNAs: considerations for the mechanism of alternative splicing

Nascent transcripts are the true substrates for many splicing events in mammalian cells. In this review we discuss transcription, splicing, and alternative splicing in the context of co-transcriptional processing of pre-mRNA. The realization that splicing occurs co-transcriptionally requires two important considerations: First, the cis-acting elements in the splicing substrate are synthesized at different times in a 5' to 3' direction. This dynamic view of the substrate implies that in a 100 kb intron the 5' splice site will be synthesized as much as an hour before the 3' splice site. Second, the transcription machinery and the splicing machinery, which are both complex and very large, are working in close proximity to each other. It is therefore likely that these two macromolecular machines interact, and recent data supporting this notion is discussed. We propose a model for co-transcriptional pre-mRNA processing that incorporates the concepts of splice site-tethering and dynamic exon definition. Also, we present a dynamic view of the alternative splicing of FGF-R2 and suggest that this view could be generally applicable to many regulated splicing events.

Promiscuity of pre-mRNA spliceosome-mediated trans splicing: a problem for gene therapy?

Trans splicing of messenger RNA has been used in experimental settings to replace mutant RNA sequences. We investigated the feasibility of utilizing trans splicing to replace a mutant RET protooncogene sequence known to inappropriately activate this tyrosine kinase receptor. We constructed a pre-trans-splicing molecule (PTM) consisting of a binding domain complementary to the target intron, the 3' splicing signal sequence (3'ss), derived from adenovirus major late transcript intron 1 and a molecular tag sequence. Accurately targeted trans splicing between the human RET exons and the PTM was demonstrated in NIH 3T3 cells cotransfected with the human RET minigene and the PTM. The efficiency of specific trans splicing was estimated to be no more than 15% in the cotransfection experiment. However, in addition to the targeted trans splicing, nontargeted trans splicing to RET exons was observed. Furthermore, the rapid amplification of 5' cDNA ends (5' RACE) analysis demonstrated that nontargeted trans splicing occurred with endogenously expressed pre-mRNAs in TT cells and that specific trans splicing to RET was a rare event. Attempts to reduce nonspecificity by the addition of a stem-loop to the trans-splicing construct designed to suppress nonspecific splicing failed to have the desired effect. These observations suggest that overexpression of a trans-splicing construct containing a 3'ss results in promiscuous trans splicing and raise significant questions about the specificity and usefulness of currently used trans-splicing approaches. In addition, these findings raise the possibility that nonspecific spliced products may be produced by a variety of gene therapy constructs.

Coupled in vitro synthesis and splicing of RNA polymerase II transcripts

Compelling in vivo studies suggest a tight functional linkage between RNA polymerase II transcription and premessenger RNA splicing. At present, the specific interactions involved in this coupling are poorly understood and deserve investigation. To this end, we developed an in vitro system that permits study of coupled transcription and splicing. Transcripts generated by RNA polymerase II were accurately and efficiently spliced under reaction conditions that permitted both transcription and splicing to occur simultaneously. The splicing of RNA-polymerase-II-driven transcripts was accelerated relative to that of the same transcripts driven by T7 RNA polymerase. Moreover, the product of exon ligation was found associated with the DNA template in reactions driven by RNA polymerase II. These two findings indicate that transcription and splicing were coupled in the in vitro system driven by RNA polymerase II, and suggest that this system will be useful for the biochemical study of this coupling.

Spliceosome-mediated RNA trans-splicing as a tool for gene therapy

We have developed RNA molecules capable of effecting spliceosome-mediated RNA trans-splicing reactions with a target messenger RNA precursor (pre-mRNA). Targeted trans-splicing was demonstrated in a HeLa nuclear extract, cultured human cells, and H1299 human lung cancer tumors in athymic mice. Trans-splicing between a cancer-associated pre-mRNA encoding the beta-subunit of human chorionic gonadotropin gene 6 and pre-trans-splicing molecule (PTM) RNA was accurate both in vitro and in vivo. Comparison of targeted versus nontargeted trans-splicing revealed a moderate level of specificity, which was improved by the addition of an internal inverted repeat encompassing the PTM splice site. Competition between cis- and trans-splicing demonstrated that cis-splicing can be inhibited by trans-splicing. RNA repair in a splicing model of a nonfunctional lacZ transcript was effected in cells by a PTM, which restored significant beta-galactosidase activity. These observations suggest that spliceosome-mediated RNA trans-splicing may represent a general approach for reprogramming the sequence of targeted transcripts, providing a novel approach to gene therapy.

Positioning of the start site in the initiation of transcription by bacteriophage T7 RNA polymerase

The determination of various polymerase structures has sparked interest in understanding how the polynucleotide template interacts with the active site. In the primer-independent initiation of transcription, an additional question arises as to how the complex directs the first two bases of the template uniquely into the active site. Recent studies in the model RNA polymerase from bacteriophage T7 demonstrate that upstream duplex contacts provide at least some of the binding specificity and suggest that the enzyme interacts with the template strand in a melted context near the start site for transcription. The current work probes the role of the template strand in positioning of the first two templating bases during initiation. The results suggest that such positioning is not rate-limiting in steady-state turnover, and that the insertion of a very large and flexible linker three or four bases upstream of the start site has no significant effect on the fidelity of start site selection. The insertion of linkers immediately adjacent to the start site, however, does significantly decrease the fidelity of start site selection (as evidenced by a large increase in misinitiation at position +2, with little change in the observed rate of correct initiation), suggesting that some of the non-transcribed template DNA does help to position the first two templating bases into the active site of the RNA polymerase. Finally, incorporation of an abasic site at position -1 yields a similar decrease in initiation fidelity, suggesting a role for stacking of the bases at positions -1 and +1.

Analysis of the RPGR gene in 11 pedigrees with the retinitis pigmentosa type 3 genotype: paucity of mutations in the coding region but splice defects in two families

X-linked retinitis pigmentosa (XLRP) is a severe form of inherited progressive retinal degeneration. The RP3 (retinitis pigmentosa type 3) locus at Xp21.1 is believed to account for the disease in the majority of XLRP families. Linkage analysis and identification of patients with chromosomal deletion have refined the location of the RP3 locus and recently have led to the cloning of the RPGR (retinitis pigmentosa GTPase regulator) gene, which has been shown to be mutated in 10%-15% of XLRP patients. In order to systematically characterize the RPGR mutations, we identified 11 retinitis pigmentosa type III (RP3) families by haplotype analysis. Sequence analysis of the PCR-amplified genomic DNA from patients representing these RP3 families did not reveal any causative mutation in RPGR exons 2-19, spanning >98% of the coding region. In patients from two families, we identified transition mutations in the intron region near splice sites (IVS10+3 and IVS13-8). RNA analysis showed that both splice-site mutations resulted in the generation of aberrant RPGR transcripts. Our results support the hypothesis that mutations in the reported RPGR gene are not a common defect in the RP3 subtype of XLRP and that a majority of causative mutations may reside either in as yet unidentified RPGR exons or in another nearby gene at Xp21.1.

Bimolecular exon ligation by the human spliceosome

Intron excision is an essential step in eukaryotic gene expression, but the molecular mechanisms by which the spliceosome accurately identifies splice sites in nuclear precursors to messenger RNAs (pre-mRNAs) are not well understood. A bimolecular assay for the second step of splicing has now revealed that exon ligation by the human spliceosome does not require covalent attachment of a 3' splice site to the branch site. Furthermore, accurate definition of the 3' splice site in this system is independent of either a covalently attached polypyrimidine tract or specific 3' exon sequences. Rather, in this system 3' splice site selection apparently occurs with a 5' --> 3' directionality.

Inverse splicing of a discontinuous pre-mRNA intron generates a circular exon in a HeLa cell nuclear extract

We have recently reported the first example of inverse splicing of a eukaryotic pre-mRNA intron using a whole cell extract from the yeast Saccharomyces cerevisiae. The concomitant circularization of the exon in the course of this splicing reaction gave rise to the hypothesis that the circular RNA species, which had been recently discovered in vivo in mammalian cells, were generated by inverse splicing. Here we report the formation of a circular exon in HeLa cell nuclear extracts by an inverse splicing reaction of the second intron of the human beta-globin gene from a pre-mRNA transcript in which the two intron halves flanked an artificially fused, single exon. Our data demonstrate that the mammalian pre-mRNA splicing system has indeed an intrinsic capability of aligning splice sites in reverse order and that this alignment can be followed by a complete splicing reaction, whereby the discontinuous intron sequences are removed. Thus we propose that circular exons in vivo arise as a result of an inverse splicing reaction following the pairing of a 5' splice site with an upstream 3' splice site and that the frequency of this event is influenced by the presence and strength of other, competing splice sites.