Exonic Splicing Mutations Are More Prevalent than Currently Estimated and Can Be Predicted by Using In Silico Tools

The identification of a causal mutation is essential for molecular diagnosis and clinical management of many genetic disorders. However, even if next-generation exome sequencing has greatly improved the detection of nucleotide changes, the biological interpretation of most exonic variants remains challenging. Moreover, particular attention is typically given to protein-coding changes often neglecting the potential impact of exonic variants on RNA splicing. Here, we used the exon 10 of MLH1, a gene implicated in hereditary cancer, as a model system to assess the prevalence of RNA splicing mutations among all single-nucleotide variants identified in a given exon. We performed comprehensive minigene assays and analyzed patient’s RNA when available. Our study revealed a staggering number of splicing mutations in MLH1 exon 10 (77% of the 22 analyzed variants), including mutations directly affecting splice sites and, particularly, mutations altering potential splicing regulatory elements (ESRs). We then used this thoroughly characterized dataset, together with experimental data derived from previous studies on BRCA1, BRCA2, CFTR and NF1, to evaluate the predictive power of 3 in silico approaches recently described as promising tools for pinpointing ESR-mutations. Our results indicate that ΔtESRseq and ΔHZ_EI-based approaches not only discriminate which variants affect splicing, but also predict the direction and severity of the induced splicing defects. In contrast, the ΔΨ-based approach did not show a compelling predictive power. Our data indicates that exonic splicing mutations are more prevalent than currently appreciated and that they can now be predicted by using bioinformatics methods. These findings have implications for all genetically-caused diseases.     

Human Splicing Finder: an online bioinformatics tool to predict splicing signals

Thousands of mutations are identified yearly. Although many directly affect protein expression, an increasing proportion of mutations is now believed to influence mRNA splicing. They mostly affect existing splice sites, but synonymous, non-synonymous or nonsense mutations can also create or disrupt splice sites or auxiliary cis-splicing sequences. To facilitate the analysis of the different mutations, we designed Human Splicing Finder (HSF), a tool to predict the effects of mutations on splicing signals or to identify splicing motifs in any human sequence. It contains all available matrices for auxiliary sequence prediction as well as new ones for binding sites of the 9G8 and Tra2-β Serine-Arginine proteins and the hnRNP A1 ribonucleoprotein. We also developed new Position Weight Matrices to assess the strength of 5′ and 3′ splice sites and branch points. We evaluated HSF efficiency using a set of 83 intronic and 35 exonic mutations known to result in splicing defects. We showed that the mutation effect was correctly predicted in almost all cases. HSF could thus represent a valuable resource for research, diagnostic and therapeutic (e.g. therapeutic exon skipping) purposes as well as for global studies, such as the GEN2PHEN European Project or the Human Variome Project.     

An Alu-derived intronic splicing enhancer facilitates intronic processing and modulates aberrant splicing in ATM

We have previously reported a natural GTAA deletion within an intronic splicing processing element (ISPE) of the ataxia telangiectasia mutated (ATM) gene that disrupts a non-canonical U1 snRNP interaction and activates the excision of the upstream portion of the intron. The resulting pre-mRNA splicing intermediate is then processed to a cryptic exon, whose aberrant inclusion in the final mRNA is responsible for ataxia telangiectasia. We show here that the last 40 bases of a downstream intronic antisense Alu repeat are required for the activation of the cryptic exon by the ISPE deletion. Evaluation of the pre-mRNA splicing intermediate by a hybrid minigene assay indicates that the identified intronic splicing enhancer represents a novel class of enhancers that facilitates processing of splicing intermediates possibly by recruiting U1 snRNP to defective donor sites. In the absence of this element, the splicing intermediate accumulates and is not further processed to generate the cryptic exon. Our results indicate that Alu-derived sequences can provide intronic splicing regulatory elements that facilitate pre-mRNA processing and potentially affect the severity of disease-causing splicing mutations.     

In silico prediction of splice-altering single nucleotide variants in the human genome

In silico tools have been developed to predict variants that may have an impact on pre-mRNA splicing. The major limitation of the application of these tools to basic research and clinical practice is the difficulty in interpreting the output. Most tools only predict potential splice sites given a DNA sequence without measuring splicing signal changes caused by a variant. Another limitation is the lack of large-scale evaluation studies of these tools. We compared eight in silico tools on 2959 single nucleotide variants within splicing consensus regions (scSNVs) using receiver operating characteristic analysis. The Position Weight Matrix model and MaxEntScan outperformed other methods. Two ensemble learning methods, adaptive boosting and random forests, were used to construct models that take advantage of individual methods. Both models further improved prediction, with outputs of directly interpretable prediction scores. We applied our ensemble scores to scSNVs from the Catalogue of Somatic Mutations in Cancer database. Analysis showed that predicted splice-altering scSNVs are enriched in recurrent scSNVs and known cancer genes. We pre-computed our ensemble scores for all potential scSNVs across the human genome, providing a whole genome level resource for identifying splice-altering scSNVs discovered from large-scale sequencing studies.     

Defective pre-mRNA splicing in PKD1 due to presumed missense and synonymous mutations causing autosomal dominant polycystic disease

Autosomal dominant polycystic kidney disease is the most common human monogenic disorder and is caused by mutations in the PKD1 or PKD2 genes. Most patients with the disease present mutations in PKD1, and a considerable number of these alterations are single base substitutions within the coding sequence that are usually predicted to lead to missense or synonymous mutations. There is growing evidence that some of these mutations can be detrimental by affecting the pre-mRNA splicing process. The aim of our study was to test PKD1 mutations, described as missense or synonymous in the literature or databases, for their effects on exon inclusion. Bioinformatics tools were used to select mutations with a potential effect on pre-mRNA splicing. Mutations were experimentally tested using minigene assays. Exons and adjacent intronic sequences were PCR-amplified and cloned in the splicing reporter minigene, and selected mutations were introduced by site-directed mutagenesis. Minigenes were transfected into kidney derived cell lines. RNA from cultured cells was analyzed by RT-PCR and DNA sequencing. Analysis of thirty-three PKD1 exonic mutations revealed three mutations that induce splicing defects. The substitution c.11156G > A, previously predicted as missense mutation p.R3719Q, abolished the donor splice site of intron 38 and resulted in the incorporation of exon 38 with 117 bp of intron 38 and skipping of exon 39. Two synonymous variants, c.327A > T (p.G109G) and c.11257C > A (p.R3753R), generated strong donor splice sites within exons 3 and 39 respectively, resulting in incorporation of incomplete exons. These three nucleotide substitutions represent the first PKD1 exonic mutations that induce aberrant mRNAs. Our results strengthen the importance to evaluate the consequences of presumed missense and synonymous mutations at the mRNA level.     

Regulation of glutamate receptor B pre-mRNA splicing by RNA editing

RNA-editing enzymes of the ADAR family convert adenosines to inosines in double-stranded RNA substrates. Frequently, editing sites are defined by base-pairing of the editing site with a complementary intronic region. The glutamate receptor subunit B (GluR-B) pre-mRNA harbors two such exonic editing sites termed Q/R and R/G. Data from ADAR knockout mice and in vitro editing assays suggest an intimate connection between editing and splicing of GluR-B pre-mRNA.

By comparing the events at the Q/R and R/G sites, we can show that editing can both stimulate and repress splicing efficiency. The edited nucleotide, but not ADAR binding itself, is sufficient to exert this effect. The presence of an edited nucleotide at the R/G site reduces splicing efficiency of the adjacent intron facilitating alternative splicing events occurring downstream of the R/G site.

Lack of editing inhibits splicing at the Q/R site. Editing of both the Q/R nucleotide and an intronic editing hotspot are required to allow efficient splicing. Inefficient intron removal may ensure that only properly edited mRNAs become spliced and exported to the cytoplasm.

     

Functional Analysis of Mutations in Exon 9 of NF1 Reveals the Presence of Several Elements Regulating Splicing

Neurofibromatosis type 1 (NF1) is one of the most common human hereditary disorders, predisposing individuals to the development of benign and malignant tumors in the nervous system, as well as other clinical manifestations. NF1 is caused by heterozygous mutations in the NF1 gene and around 25% of the pathogenic changes affect pre-mRNA splicing. Since the molecular mechanisms affected by these mutations are poorly understood, we have analyzed the splicing mutations identified in exon 9 of NF1, which is particularly prone to such changes, to better define the possible splicing regulatory elements. Using a minigene approach, we studied the effect of five splicing mutations in this exon described in patients. These highlighted three regulatory motifs within the exon. An in vivo splicing analysis of an extensive collection of changes generated in the minigene demonstrated that the CG motif at c.910-911 is critical for the recognition of exon 9. We also found that the GC motif at c.945-946 is involved in exon recognition through SRSF2 and that this motif is part of a Composite Exon Splicing Regulatory Element made up of physically overlapping enhancer and silencer elements. Finally, through an in vivo splicing analysis and in vitro binding assays, we demonstrated that the c.1007G>A mutation creates an Exonic Splicing Silencer element that binds the hnRNPA1 protein. The complexity of the splicing regulatory elements present in exon 9 is most likely responsible for the fact that mutations in this region represent 25% of all exonic changes that affect splicing in the NF1 gene.     

In vitro and in silico analysis reveals an efficient algorithm to predict the splicing consequences of mutations at the 5' splice sites

We have found that two previously reported exonic mutations in the PINK1 and PARK7 genes affect pre-mRNA splicing. To develop an algorithm to predict underestimated splicing consequences of exonic mutations at the 5′ splice site, we constructed and analyzed 31 minigenes carrying exonic splicing mutations and their derivatives. We also examined 189 249 U2-dependent 5′ splice sites of the entire human genome and found that a new variable, the SD-Score, which represents a common logarithm of the frequency of a specific 5′ splice site, efficiently predicts the splicing consequences of these minigenes. We also employed the information contents (R_i) to improve the prediction accuracy. We validated our algorithm by analyzing 32 additional minigenes as well as 179 previously reported splicing mutations. The SD-Score algorithm predicted aberrant splicings in 198 of 204 sites (sensitivity = 97.1%) and normal splicings in 36 of 38 sites (specificity = 94.7%). Simulation of all possible exonic mutations at positions −3, −2 and −1 of the 189 249 sites predicts that 37.8, 88.8 and 96.8% of these mutations would affect pre-mRNA splicing, respectively. We propose that the SD-Score algorithm is a practical tool to predict splicing consequences of mutations affecting the 5′ splice site.     

MutPred Splice: machine learning-based prediction of exonic variants that disrupt splicing

We have developed a novel machine-learning approach, MutPred Splice, for the identification of coding region substitutions that disrupt pre-mRNA splicing. Applying MutPred Splice to human disease-causing exonic mutations suggests that 16% of mutations causing inherited disease and 10 to 14% of somatic mutations in cancer may disrupt pre-mRNA splicing. For inherited disease, the main mechanism responsible for the splicing defect is splice site loss, whereas for cancer the predominant mechanism of splicing disruption is predicted to be exon skipping via loss of exonic splicing enhancers or gain of exonic splicing silencer elements. MutPred Splice is available at http://mutdb.org/mutpredsplice.     

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.     

Mutations in U5 snRNA loop 1 influence the splicing of different genes in vivo

The U5 snRNA loop 1 is characterized by the conserved sequence G₁C₂C₃U₄U₅U₆Y₇A₈Y₉ and is essential for the alignment of exons during the second step of pre-mRNA splicing in Saccharo myces cerevisiae. Despite this sequence conservation the size, rather than sequence, of loop 1 is critical for exon alignment in vitro. To determine the in vivo requirements for U5 loop 1 a library of loop 1 sequences was transformed into a yeast strain where the endogenous U5 gene was deleted. Comparison of viable mutations in loop 1 revealed that position 6 was invariant and positions 5 and 7 displayed some sequence conservation. These data indicate positions 5, 6 and 7 in loop 1 are important for U5 function in vivo. A screen for mutations that suppress the temperature-sensitive phenotype of three loop 1 mutants produced eight intragenic suppressors all containing alterations in loop 1. Further analysis of these temperature-sensitive mutants revealed that each displayed distinct cell cycle arrest phenotypes and pre-mRNA splicing inhibition patterns. The cell cycle arrest is likely attributed to inefficient splicing of α-tubulin pre-mRNA in one mutant and actin pre-mRNA in another. These results suggest that various mutations in loop 1 may affect the splicing of different pre-mRNAs in vivo.     

Aberrant splicing of tau pre-mRNA caused by intronic mutations associated with the inherited dementia frontotemporal dementia with parkinsonism linked to chromosome 17

Frontotemporal dementia accounts for a significant fraction of dementia cases. Frontotemporal dementia with parkinsonism linked to chromosome 17 is associated with either exonic or intronic mutations in the tau gene. This highlights the involvement of aberrant pre-mRNA splicing in the pathogenesis of neurodegenerative disorders. Little is known about the molecular mechanisms of the splicing defects underlying these diseases. To establish a model system for studying the role of pre-mRNA splicing in neurodegenerative diseases, we have constructed a tau minigene that reproduces tau alternative splicing in both cultured cells and in vitro biochemical assays. We demonstrate that mutations in a nonconserved intronic region of the human tau gene lead to increased splicing between exon 10 and exon 11. Systematic biochemical analyses indicate the importance of U1 snRNP and, to a lesser extent, U6 snRNP in differentially recognizing wild-type versus intron mutant tau pre-mRNAs. Gel mobility shift assays with purified U1 snRNP and oligonucleotide-directed RNase H cleavage experiments support the idea that the intronic mutations destabilize a stem-loop structure that sequesters the 5' splice site downstream of exon 10 in tau pre-mRNA, leading to increases in U1 snRNP binding and in splicing between exon 10 and exon 11. Thus, mutations in nonconserved intronic regions that increase rather than decrease alternative splicing can be an important pathogenic mechanism for the development of human diseases.     

Global control of aberrant splice-site activation by auxiliary splicing sequences: evidence for a gradient in exon and intron definition

Auxiliary splicing signals play a major role in the regulation of constitutive and alternative pre-mRNA splicing, but their relative importance in selection of mutation-induced cryptic or de novo splice sites is poorly understood. Here, we show that exonic sequences between authentic and aberrant splice sites that were activated by splice-site mutations in human disease genes have lower frequencies of splicing enhancers and higher frequencies of splicing silencers than average exons. Conversely, sequences between authentic and intronic aberrant splice sites have more enhancers and less silencers than average introns. Exons that were skipped as a result of splice-site mutations were smaller, had lower SF2/ASF motif scores, a decreased availability of decoy splice sites and a higher density of silencers than exons in which splice-site mutation activated cryptic splice sites. These four variables were the strongest predictors of the two aberrant splicing events in a logistic regression model. Elimination or weakening of predicted silencers in two reporters consistently promoted use of intron-proximal splice sites if these elements were maintained at their original positions, with their modular combinations producing expected modification of splicing. Together, these results show the existence of a gradient in exon and intron definition at the level of pre-mRNA splicing and provide a basis for the development of computational tools that predict aberrant splicing outcomes.