Cooperative engagement and subsequent selective displacement of SR proteins define the pre-mRNA 3D structural scaffold for early spliceosome assembly

We recently reported that serine–arginine-rich (SR) protein-mediated pre-mRNA structural remodeling generates a pre-mRNA 3D structural scaffold that is stably recognized by the early spliceosomal components. However, the intermediate steps between the free pre-mRNA and the assembled early spliceosome are not yet characterized. By probing the early spliceosomal complexes in vitro and RNA-protein interactions in vivo, we show that the SR proteins bind the pre-mRNAs cooperatively generating a substrate that recruits U1 snRNP and U2AF65 in a splice signal-independent manner. Excess U1 snRNP selectively displaces some of the SR protein molecules from the pre-mRNA generating the substrate for splice signal-specific, sequential recognition by U1 snRNP, U2AF65 and U2AF35. Our work thus identifies a novel function of U1 snRNP in mammalian splicing substrate definition, explains the need for excess U1 snRNP compared to other U snRNPs in vivo, demonstrates how excess SR proteins could inhibit splicing, and provides a conceptual basis to examine if this mechanism of splicing substrate definition is employed by other splicing regulatory proteins.     

Temperature-dependent splicing of beta-globin pre-mRNA

A T→G mutation at nucleotide 705 of human β-globin intron 2 creates an aberrant 5′ splice site and activates a cryptic 3′ splice site upstream. In consequence, the pre-mRNA is spliced via aberrant splice sites, despite the presence of the still functional correct sites. Surprisingly, when IVS2-705 HeLa or K562 cells were cultured at temperatures below 30°C, aberrant splicing was inhibited and correct splicing was restored. Similar temperature effects were seen for another β-globin pre-mRNA, IVS2-745, and in a construct in which a β-globin intron was inserted into a coding sequence of EGFP. Temperature-induced alternative splicing was affected by the nature of the internal aberrant splice sites flanking the correct sites and by exonic sequences. The results indicate that in the context of thalassemic splicing mutations and possibly in other alternatively spliced pre-mRNAs, temperature is one of the parameters that affect splice site selection.     

An exonic splicing silencer in the testes-specific DNA ligase III beta exon

Alternative pre-mRNA splicing of two terminal exons (α and β) regulates the expression of the human DNA ligase III gene. In most tissues, the α exon is expressed. In testes and during spermatogenesis, the β exon is used instead. The α exon encodes the interaction domain with a scaffold DNA repair protein, XRCC1, while the β exon-encoded C-terminal does not. Sequence elements regulating the alternative splicing pattern were mapped by in vitro splicing assays in HeLa nuclear extracts. Deletion of a region beginning in the β exon and extending into the downstream intron derepressed splicing to the β exon. Two silencing elements were found within this 101 nt region: a 16 nt exonic splicing silencer immediately upstream of the β exon polyadenylation signal and a 45 nt intronic splicing silencer. The exonic splicing silencer inhibited splicing, even when the polyadenylation signal was deleted or replaced by a 5′ splice site. This element also enhanced polyadenylation under conditions unfavourable to splicing. The splicing silencer partially inhibited assembly of spliceosomal complexes and functioned in an adenoviral pre-mRNA context. Silencing of splicing by the element was associated with cross-linking of a 37 kDa protein to the RNA substrate. The element exerts opposite functions in splicing and polyadenylation.     

Human ribosomal protein S13 regulates expression of its own gene at the splicing step by a feedback mechanism

The expression of ribosomal protein (rp) genes is regulated at multiple levels. In yeast, two genes are autoregulated by feedback effects of the protein on pre-mRNA splicing. Here, we have investigated whether similar mechanisms occur in eukaryotes with more complicated and highly regulated splicing patterns. Comparisons of the sequences of ribosomal protein S13 gene (RPS13) among mammals and birds revealed that intron 1 is more conserved than the other introns. Transfection of HEK 293 cells with a minigene-expressing ribosomal protein S13 showed that the presence of intron 1 reduced expression by a factor of four. Ribosomal protein S13 was found to inhibit excision of intron 1 from rpS13 pre-mRNA fragment in vitro. This protein was shown to be able to specifically bind the fragment and to confer protection against ribonuclease cleavage at sequences near the 5′ and 3′ splice sites. The results suggest that overproduction of rpS13 in mammalian cells interferes with splicing of its own pre-mRNA by a feedback mechanism.     

Intron Definition and a Branch Site Adenosine at nt 385 Control RNA Splicing of HPV16 E6*I and E7 Expression

HPV16 E6 and E7, two viral oncogenes, are expressed from a single bicistronic pre-mRNA. In this report, we provide the evidence that the bicistronic pre-mRNA intron 1 contains three 5′ splice sites (5′ ss) and three 3′ splice sites (3′ ss) normally used in HPV16⁺ cervical cancer and its derived cell lines. The choice of two novel alternative 5′ ss (nt 221 5′ ss and nt 191 5′ ss) produces two novel isoforms of E6E7 mRNAs (E6*V and E6*VI). The nt 226 5′ ss and nt 409 3′ ss is preferentially selected over the other splice sites crossing over the intron to excise a minimal length of the intron in RNA splicing. We identified AACAAAC as the preferred branch point sequence (BPS) and an adenosine at nt 385 (underlined) in the BPS as a branch site to dictate the selection of the nt 409 3′ ss for E6*I splicing and E7 expression. Introduction of point mutations into the mapped BPS led to reduced U2 binding to the BPS and thereby inhibition of the second step of E6E7 splicing at the nt 409 3′ ss. Importantly, the E6E7 bicistronic RNA with a mutant BPS and inefficient splicing makes little or no E7 and the resulted E6 with mutations of ⁹¹QYNK⁹⁴ to ⁹¹PSFW⁹⁴ displays attenuate activity on p53 degradation. Together, our data provide structural basis of the E6E7 intron 1 for better understanding of how viral E6 and E7 expression is regulated by alternative RNA splicing. This study elucidates for the first time a mapped branch point in HPV16 genome involved in viral oncogene expression.     

Proteomic analysis of in vivo-assembled pre-mRNA splicing complexes expands the catalog of participating factors

Previous compositional studies of pre-mRNA processing complexes have been performed in vitro on synthetic pre-mRNAs containing a single intron. To provide a more comprehensive list of polypeptides associated with the pre-mRNA splicing apparatus, we have determined the composition of the bulk pre-mRNA processing machinery in living cells. We purified endogenous nuclear pre-mRNA processing complexes from human and chicken cells comprising the massive (>200S) supraspliceosomes (a.k.a. polyspliceosomes). As expected, RNA components include a heterogeneous mixture of pre-mRNAs and the five spliceosomal snRNAs. In addition to known pre-mRNA splicing factors, 5′ end binding factors, 3′ end processing factors, mRNA export factors, hnRNPs and other RNA binding proteins, the protein components identified by mass spectrometry include RNA adenosine deaminases and several novel factors. Intriguingly, our purified supraspliceosomes also contain a number of structural proteins, nucleoporins, chromatin remodeling factors and several novel proteins that were absent from splicing complexes assembled in vitro. These in vivo analyses bring the total number of factors associated with pre-mRNA to well over 300, and represent the most comprehensive analysis of the pre-mRNA processing machinery to date.     

The Peculiarities of Large Intron Splicing in Animals

In mammals a considerable 92% of genes contain introns, with hundreds and hundreds of these introns reaching the incredible size of over 50,000 nucleotides. These “large introns” must be spliced out of the pre-mRNA in a timely fashion, which involves bringing together distant 5′ and 3′ acceptor and donor splice sites. In invertebrates, especially Drosophila, it has been shown that larger introns can be spliced efficiently through a process known as recursive splicing—a consecutive splicing from the 5′-end at a series of combined donor-acceptor splice sites called RP-sites. Using a computational analysis of the genomic sequences, we show that vertebrates lack the proper enrichment of RP-sites in their large introns, and, therefore, require some other method to aid splicing. We analyzed over 15,000 non-redundant, large introns from six mammals, 1,600 from chicken and zebrafish, and 560 non-redundant large introns from five invertebrates. Our bioinformatic investigation demonstrates that, unlike the studied invertebrates, the studied vertebrate genomes contain consistently abundant amounts of direct and complementary strand interspersed repetitive elements (mainly SINEs and LINEs) that may form stems with each other in large introns. This examination showed that predicted stems are indeed abundant and stable in the large introns of mammals. We hypothesize that such stems with long loops within large introns allow intron splice sites to find each other more quickly by folding the intronic RNA upon itself at smaller intervals and, thus, reducing the distance between donor and acceptor sites.     

Functionally antagonistic sequences are required for normal autoregulation of Drosophila tra-2 pre-mRNA splicing

Expression of functional TRA-2 protein in the male germline of Drosophila is regulated through a negative feedback mechanism in which a specific TRA-2 isoform represses splicing of the M1 intron in the TRA-2 pre-mRNA. We have previously shown that the mechanism of M1 splicing repression is conserved between distantly related Drosophila species. Using transgenic fly strains, we have examined the effects on regulation of mutations in two conserved features of the M1 intron. Our results show that TRA-2-dependent repression of M1 splicing depends on the presence of a suboptimal non-consensus 3′ splice site. Substitution of this 3′ splice site with a strong splice site resulted in TRA-2 independent splicing, while substitution with an unrelated weak 3′ splice site was compatible with repression, implying that reduced basal splicing efficiency is important for regulation. A second conserved element internal to the intron was found to be essential for efficient M1 splicing in the soma where the intron is not normally retained. We show that the role of this element is to enhance splicing and overcome the reduction in efficiency caused by the intron’s suboptimal 3′ splice site. Our results indicate that antagonistic elements in the M1 intron act together to establish a context that is permissive for repression of splicing by TRA-2 while allowing efficient splicing in the absence of a repressor.     

Function of a Bovine Papillomavirus Type 1 Exonic Splicing Suppressor Requires a Suboptimal Upstream 3′ Splice Site

Alternative splicing is an important mechanism for the regulation of bovine papillomavirus type 1 (BPV-1) gene expression during the virus life cycle. Previous studies in our laboratory have identified two purine-rich exonic splicing enhancers (ESEs), SE1 and SE2, located between two alternative 3′ splice sites at nucleotide (nt) 3225 and nt 3605. Further analysis of BPV-1 late-pre-mRNA splicing in vitro revealed a 48-nt pyrimidine-rich region immediately downstream of SE1 that inhibits utilization of the nt 3225 3′ splice site. This inhibitory element, which we named an exonic splicing suppressor (ESS), has a U-rich 5′ end, a C-rich central part, and an AG-rich 3′ end (Z. M. Zheng, P. He, and C. C. Baker, J. Virol. 70:4691–4699, 1996). The present study utilized in vitro splicing of both homologous and heterologous pre-mRNAs to further characterize the ESS. The BPV-1 ESS was inserted downstream of the 3′ splice site in the BPV-1 late pre-mRNA, Rous sarcoma virus src pre-mRNA, human immunodeficiency virus tat-rev pre-mRNA, and Drosophila dsx pre-mRNA, all containing a suboptimal 3′ splice site, and in the human β-globin pre-mRNA, which contains a constitutive 3′ splice site. These studies demonstrated that suppression of splicing by the BPV-1 ESS requires an upstream suboptimal 3′ splice site but not an upstream ESE. Furthermore, the ESS functions when located either upstream or downstream of BPV-1 SE1. Mutational analyses demonstrated that the function of the ESS is sequence dependent and that only the C-rich region of the ESS is essential for suppression of splicing in all the pre-mRNAs tested.     

Alternative splicing and bioinformatic analysis of human U12-type introns

U12-type introns exist, albeit rarely, in a variety of multicellular organisms. Splicing of U12 intron-containing precursor mRNAs takes place in the U12-type spliceosome that is distinct from the major U2-type spliceosome. Due to incompatibility of these two spliceosomes, alternative splicing involving a U12-type intron may give rise to a relatively complicated impact on gene expression. We studied alternative U12-type intron splicing in an attempt to gain more mechanistic insights. First, we characterized mutually exclusive exon selection of the human JNK2 gene, which involves an unusual intron possessing the U12-type 5′ splice site and the U2-type 3′ splice site. We demonstrated that the long and evolutionary conserved polypyrimidine tract of this hybrid intron provides important signals for inclusion of its downstream alternative exon. In addition, we examined the effects of single nucleotide polymorphisms in the human WDFY1 U12-type intron on pre-mRNA splicing. These results provide mechanistic implications on splice-site selection of U12-type intron splicing. We finally discuss the potential effects of splicing of a U12-type intron with genetic defects or within a set of genes encoding RNA processing factors on global gene expression.     

Functional studies on the ATM intronic splicing processing element

In disease-associated genes, the understanding of the functional significance of deep intronic nucleotide variants may represent a difficult challenge. We have previously reported a new disease-causing mechanism that involves an intronic splicing processing element (ISPE) in ATM, composed of adjacent consensus 5′ and 3′ splice sites. A GTAA deletion within ISPE maintains potential adjacent splice sites, disrupts a non-canonical U1 snRNP interaction and activates an aberrant exon. In this paper, we demonstrate that binding of U1 snRNA through complementarity within a ~40 nt window downstream of the ISPE prevents aberrant splicing. By selective mutagenesis at the adjacent consensus ISPE splice sites, we show that this effect is not due to a resplicing process occurring at the ISPE. Functional comparison of the ATM mouse counterpart and evaluation of the pre-mRNA splicing intermediates derived from affected cell lines and hybrid minigene assays indicate that U1 snRNP binding at the ISPE interferes with the cryptic acceptor site. Activation of this site results in a stringent 5′–3′ order of intron sequence removal around the cryptic exon. Artificial U1 snRNA loading by complementarity to heterologous exonic sequences represents a potential therapeutic method to prevent the usage of an aberrant CFTR cryptic exon. Our results suggest that ISPE-like intronic elements binding U1 snRNPs may regulate correct intron processing.     

The Pre-mRNA 5′ Cap Determines Whether U6 Small Nuclear RNA Succeeds U1 Small Nuclear Ribonucleoprotein Particle at 5′ Splice Sites

Efficient splicing of the 5′-most intron of pre-mRNA requires a 5′ m⁷G(5′)ppp(5′)N cap, which has been implicated in U1 snRNP binding to 5′ splice sites. We demonstrate that the cap alters the kinetic profile of U1 snRNP binding, but its major effect is on U6 snRNA binding. With two alternative wild-type splice sites in an adenovirus pre-mRNA, the cap selectively alters U1 snRNA binding at the site to which cap-independent U1 snRNP binding is stronger and that is used predominantly in splicing; with two consensus sites, the cap acts on both, even though one is substantially preferred for splicing. However, the most striking quantitative effect of the 5′ cap is neither on U1 snRNP binding nor on the assembly of large complexes but on the replacement of U1 snRNP by U6 snRNA at the 5′ splice site. Inhibition of splicing by a cap analogue is correlated with the loss of U6 interactions at the 5′ splice site and not with any loss of U1 snRNP binding.     

Cooperative interaction of branch signals in the actin intron of Saccharomyces cerevisiae.

In pre-mRNA splicing, specific spliceosomal components recognize key intron sequences, but the mechanisms by which splice sites are selected arenot completely understood. In the Saccharomyces cerevisiae actin intron a silent branch point-like sequence (UACUAAG) is located 7 nt upstream of the canonical sequence. Mutation of the canonicalUACUAAC sequence to UAAUAAC reduces utilization of this signal and activates the cryptic UACUAAG. Splicing-dependent beta-galactosidase assays have shown that these two splice signals cooperate to enhance splicing. Analyses of several variants of this double branch point intron demonstrate that the upstream UACUAAG sequence significantly increases usage of the UAAUAAC as a site of lariat formation. This activation is sequence-specific and unidirectional. However the ability of the UACUAAG signal to activate the downstream branch point is dependent on the presence of a short non-conserved sequence located a few nucleotides upstream of the UACUAAG. Mutation of this sequence leads to the disappearance of the cooperative interactions between the two branch signals. Our results show that this non-conserved sequence and the UACUAAG signal must both be present to achieve activation of the downstream branch point and suggest that a specific structure may be necessary to allow efficient recognition of the UAAUAAC.     

Rearrangements within human spliceosomes captured after exon ligation

In spliceosomes, dynamic RNA/RNA and RNA/protein interactions position the pre-mRNA substrate for the two chemical steps of splicing. Not all of these interactions have been characterized, in part because it has not been possible to arrest the complex at clearly defined states relative to chemistry. Previously, it was shown in yeast that the DEAD/H-box protein Prp22 requires an extended 3′ exon to promote mRNA release from the spliceosome following second-step chemistry. In line with that observation, we find that shortening the 3′ exon blocks cleaved lariat intron and mRNA release in human splicing extracts, which allowed us to stall human spliceosomes in a new post-catalytic complex (P complex). In comparison to C complex, which is blocked at a point following first-step chemistry, we detect specific differences in RNA substrate interactions near the splice sites. These differences include extended protection across the exon junction and changes in protein crosslinks to specific sites in the 5′ and 3′ exons. Using selective reaction monitoring (SRM) mass spectrometry, we quantitatively compared P and C complex proteins and observed enrichment of SF3b components and loss of the putative RNA-dependent ATPase DHX35. Electron microscopy revealed similar structural features for both complexes. Notably, additional density is present when complexes are chemically fixed, which reconciles our results with previously reported C complex structures. Our ability to compare human spliceosomes before and after second-step chemistry has opened a new window to rearrangements near the active site of spliceosomes, which may play roles in exon ligation and mRNA release.     

Mutation of putative branchpoint consensus sequences in plant introns reduces splicing efficiency

Intron lariat formation between the 5' end of an intron and a branchpoint adenosine is a fundamental aspect of the first step in animal and yeast nuclear pre-mRNA splicing. Despite similarities in intron sequence requirements and the components of splicing, differences exist between the splicing of plant and vertebrate introns. The identification of AU-rich sequences as major functional elements in plant introns and the demonstration that a branchpoint consensus sequence was not required for splicing have led to the suggestion that the transition from AU-rich intron to GC-rich exon is a major potential signal by which plant pre-mRNA splice sites are recognized. The role of putative branchpoint sequences as an internal signal in plant intron recognition/definition has been re-examined. Single nucleotide mutations in putative branchpoint adenosines contained within CUNAN sequences in four different plant introns all significantly reduced splicing efficiency. These results provide the most direct evidence to date for preferred branchpoint sequences being required for the efficient splicing of at least some plant introns in addition to the important role played by AU sequences in dicot intron recognition. The observed patterns of 3' splice site selection in the introns studied are consistent with the scanning model described for animal intron 3' splice site selection. It is suggested that, despite the clear importance of AU sequences for plant intron splicing, the fundamental processes of splice site selection and splicing in plants are similar to those in animals.     

AUG sequences are required to sustain nonsense-codon-mediated suppression of splicing

More than 90% of human genes are rich in intronic latent 5′ splice sites whose utilization in pre-mRNA splicing would introduce in-frame stop codons into the resultant mRNAs. We have therefore hypothesized that suppression of splicing (SOS) at latent 5′ splice sites regulates alternative 5′ splice site selection in a way that prevents the production of toxic nonsense mRNAs and verified this idea by showing that the removal of such in-frame stop codons is sufficient to activate latent splicing. Splicing control by SOS requires recognition of the mRNA reading frame, presumably recognizing the start codon sequence. Here we show that AUG sequences are indeed essential for SOS. Although protein translation does not seem to be required for SOS, the first AUG is shown here to be necessary but not sufficient. We further show that latent splicing can be elicited upon treatment with pactamycin—a drug known to block translation by its ability to recognize an RNA fold—but not by treatment with other drugs that inhibit translation through other mechanisms. The effect of pactamycin on SOS is dependent neither on steady-state translation nor on the pioneer round of translation. This effect is found for both transfected and endogenous genes, indicating that SOS is a natural mechanism.     

Interaction of the yeast DExH-box RNA helicase prp22p with the 3' splice site during the second step of nuclear pre-mRNA splicing

Using site-specific incorporation of the photochemical cross-linking reagent 4-thiouridine, we demonstrate the previously unknown association of two proteins with yeast 3′ splice sites. One of these is an unidentified ~122 kDa protein that cross-links to 3′ splice sites during formation of the pre-spliceosome. The other factor is the DExH-box RNA helicase, Prp22p. With substrates functional in the second step of splicing, only very weak cross-linking of Prp22p to intron sequences at the 3′ splice site is observed. In contrast, substrates blocked at the second step exhibit strong cross-linking of Prp22 to intron sequences at the 3′ splice site, but not to adjacent exon sequences. In vitro reconstitution experiments also show that the association of Prp22p with intron sequences at the 3′ splice site is dependent on Prp16p and does not persist when release of mature mRNA from the spliceosome is blocked. Taken together, these results suggest that the 3′ splice site of yeast introns is contacted much earlier than previously envisioned by a protein of ~120 kDa, and that a transient association of Prp22p with the 3′ splice site occurs between the first and second catalytic steps.