Dynamic imaging of nascent RNA reveals general principles of transcription dynamics and stochastic splice site selection.

The activities of RNA polymerase and the spliceosome are responsible for the heterogeneity in the abundance and isoform composition of mRNA in human cells. However, the dynamics of these megadalton enzymatic complexes working in concert on endogenous genes have not been described. Here, we establish a quasi-genome-scale platform for observing synthesis and processing kinetics of single nascent RNA molecules in real time. We find that all observed genes show transcriptional bursting. We also observe large kinetic variation in intron removal for single introns in single cells, which is inconsistent with deterministic splice site selection. Transcriptome-wide footprinting of the U2AF complex, nascent RNA profiling, long-read sequencing, and lariat sequencing further reveal widespread stochastic recursive splicing within introns. We propose and validate a unified theoretical model to explain the general features of transcription and pervasive stochastic splice site selection.

Splicing Enhancers at Intron-Exon Borders Participate in Acceptor Splice Sites Recognition.

Acceptor splice site recognition (3' splice site: 3'ss) is a fundamental step in precursor messenger RNA (pre-mRNA) splicing. Generally, the U2 small nuclear ribonucleoprotein (snRNP) auxiliary factor (U2AF) heterodimer recognizes the 3'ss, of which U2AF35 has a dual function: (i) It binds to the intron-exon border of some 3'ss and (ii) mediates enhancer-binding splicing activators' interactions with the spliceosome. Alternative mechanisms for 3'ss recognition have been suggested, yet they are still not thoroughly understood. Here, we analyzed 3'ss recognition where the intron-exon border is bound by a ubiquitous splicing regulator SRSF1. Using the minigene analysis of two model exons and their mutants, BRCA2 exon 12 and VARS2 exon 17, we showed that the exon inclusion correlated much better with the predicted SRSF1 affinity than 3'ss quality, which were assessed using the Catalog of Inferred Sequence Binding Preferences of RNA binding proteins (CISBP-RNA) database and maximum entropy algorithm (MaxEnt) predictor and the U2AF35 consensus matrix, respectively. RNA affinity purification proved SRSF1 binding to the model 3'ss. On the other hand, knockdown experiments revealed that U2AF35 also plays a role in these exons' inclusion. Most probably, both factors stochastically bind the 3'ss, supporting exon recognition, more apparently in VARS2 exon 17. Identifying splicing activators as 3'ss recognition factors is crucial for both a basic understanding of splicing regulation and human genetic diagnostics when assessing variants' effects on splicing.

Impact of acceptor splice site NAGTAG motif on exon recognition.

Pre-mRNA splicing is an essential step in gene expression, when introns are removed and exons joined by the complex of proteins called spliceosome. Correct splicing requires a precise exon/intron junction definition, which is determined by a consensual donor and acceptor splice site at the 5' and 3' end, respectively. An acceptor splice site (3'ss) consists of highly conserved AG nucleotides in positions E-2 and E-1. These nucleotides can appear in tandem, located 3 bp from each other. Then they are referred to as NAGNAG or tandem 3'ss, which can be alternatively spliced. NAG/TAG 3'ss motif abundance is extremely low and cannot be easily explained by just a nucleotide preference in this position. We tested artificial NAG/TAG motif's potential negative effect on exon recognition using a minigene assay. Introducing the NAG/TAG motif into seven different exons revealed no general negative effect on exon recognition. The only observed effect was the partial use of the newly formed distal 3'ss. We can conclude that this motif's extremely low preference in a natural 3'ss is not a consequence of the NAG/TAG motif's negative effect on exon recognition, but more likely the result of other RNA processing aspects, such as an alternative 3'ss choice, decreased 3'ss strength, or incorporating an amber stop codon.

Predicting Splicing from Primary Sequence with Deep Learning.

The splicing of pre-mRNAs into mature transcripts is remarkable for its precision, but the mechanisms by which the cellular machinery achieves such specificity are incompletely understood. Here, we describe a deep neural network that accurately predicts splice junctions from an arbitrary pre-mRNA transcript sequence, enabling precise prediction of noncoding genetic variants that cause cryptic splicing. Synonymous and intronic mutations with predicted splice-altering consequence validate at a high rate on RNA-seq and are strongly deleterious in the human population. De novo mutations with predicted splice-altering consequence are significantly enriched in patients with autism and intellectual disability compared to healthy controls and validate against RNA-seq in 21 out of 28 of these patients. We estimate that 9%-11% of pathogenic mutations in patients with rare genetic disorders are caused by this previously underappreciated class of disease variation.

Exon Junction Complexes Suppress Spurious Splice Sites to Safeguard Transcriptome Integrity.

Productive splicing of human precursor messenger RNAs (pre-mRNAs) requires the correct selection of authentic splice sites (SS) from the large pool of potential SS. Although SS consensus sequence and splicing regulatory proteins are known to influence SS usage, the mechanisms ensuring the effective suppression of cryptic SS are insufficiently explored. Here, we find that many aberrant exonic SS are efficiently silenced by the exon junction complex (EJC), a multi-protein complex that is deposited on spliced mRNA near the exon-exon junction. Upon depletion of EJC proteins, cryptic SS are de-repressed, leading to the mis-splicing of a broad set of mRNAs. Mechanistically, the EJC-mediated recruitment of the splicing regulator RNPS1 inhibits cryptic 5'SS usage, while the deposition of the EJC core directly masks reconstituted 3'SS, thereby precluding transcript disintegration. Thus, the EJC protects the transcriptome of mammalian cells from inadvertent loss of exonic sequences and safeguards the expression of intact, full-length mRNAs.

Exon Junction Complex Shapes the Transcriptome by Repressing Recursive Splicing.

Recursive splicing (RS) starts by defining an "RS-exon," which is then spliced to the preceding exon, thus creating a recursive 5' splice site (RS-5ss). Previous studies focused on cryptic RS-exons, and now we find that the exon junction complex (EJC) represses RS of hundreds of annotated, mainly constitutive RS-exons. The core EJC factors, and the peripheral factors PNN and RNPS1, maintain RS-exon inclusion by repressing spliceosomal assembly on RS-5ss. The EJC also blocks 5ss located near exon-exon junctions, thus repressing inclusion of cryptic microexons. The prevalence of annotated RS-exons is high in deuterostomes, while the cryptic RS-exons are more prevalent in Drosophila, where EJC appears less capable of repressing RS. Notably, incomplete repression of RS also contributes to physiological alternative splicing of several human RS-exons. Finally, haploinsufficiency of the EJC factor Magoh in mice is associated with skipping of RS-exons in the brain, with relevance to the microcephaly phenotype and human diseases.

Genome-wide CRISPR-Cas9 Interrogation of Splicing Networks Reveals a Mechanism for Recognition of Autism-Misregulated Neuronal Microexons.

Alternative splicing is crucial for diverse cellular, developmental, and pathological processes. However, the full networks of factors that control individual splicing events are not known. Here, we describe a CRISPR-based strategy for the genome-wide elucidation of pathways that control splicing and apply it to microexons with important functions in nervous system development and that are commonly misregulated in autism. Approximately 200 genes associated with functionally diverse regulatory layers and enriched in genetic links to autism control neuronal microexons. Remarkably, the widely expressed RNA binding proteins Srsf11 and Rnps1 directly, preferentially, and frequently co-activate these microexons. These factors form critical interactions with the neuronal splicing regulator Srrm4 and a bi-partite intronic splicing enhancer element to promote spliceosome formation. Our study thus presents a versatile system for the identification of entire splicing regulatory pathways and further reveals a common mechanism for the definition of neuronal microexons that is disrupted in autism.

Quantitative Activity Profile and Context Dependence of All Human 5' Splice Sites.

Pre-mRNA splicing is an essential step in the expression of most human genes. Mutations at the 5' splice site (5'ss) frequently cause defective splicing and disease due to interference with the initial recognition of the exon-intron boundary by U1 small nuclear ribonucleoprotein (snRNP), a component of the spliceosome. Here, we use a massively parallel splicing assay (MPSA) in human cells to quantify the activity of all 32,768 unique 5'ss sequences (NNN/GYNNNN) in three different gene contexts. Our results reveal that although splicing efficiency is mostly governed by the 5'ss sequence, there are substantial differences in this efficiency across gene contexts. Among other uses, these MPSA measurements facilitate the prediction of 5'ss sequence variants that are likely to cause aberrant splicing. This approach provides a framework to assess potential pathogenic variants in the human genome and streamline the development of splicing-corrective therapies.

A donor splice site mutation in CISD2 generates multiple truncated, non-functional isoforms in Wolfram syndrome type 2 patients.

BACKGROUND: Mutations in the gene that encodes CDGSH iron sulfur domain 2 (CISD2) are causative of Wolfram syndrome type 2 (WFS2), a rare autosomal recessive neurodegenerative disorder mainly characterized by diabetes mellitus, optic atrophy, peptic ulcer bleeding and defective platelet aggregation. Four mutations in the CISD2 gene have been reported. Among these mutations, the homozygous c.103 + 1G > A substitution was identified in the donor splice site of intron 1 in two Italian sisters and was predicted to cause a exon 1 to be skipped. METHODS: Here, we employed molecular assays to characterize the c.103 + 1G > A mutation using the patient's peripheral blood mononuclear cells (PBMCs). 5'-RACE coupled with RT-PCR were used to analyse the effect of the c.103 + 1G > A mutation on mRNA splicing. Western blot analysis was used to analyse the consequences of the CISD2 mutation on the encoded protein. RESULTS: We demonstrated that the c.103 + 1G > A mutation functionally impaired mRNA splicing, producing multiple splice variants characterized by the whole or partial absence of exon 1, which introduced amino acid changes and a premature stop. The affected mRNAs resulted in either predicted targets for nonsense mRNA decay (NMD) or non-functional isoforms. CONCLUSIONS: We concluded that the c.103 + 1G > A mutation resulted in the loss of functional CISD2 protein in the two Italian WFS2 patients.

The identification of switch-like alternative splicing exons among multiple samples with RNA-Seq data.

Alternative splicing is an ubiquitous phenomenon in most human genes and has important functions. The switch-like exon is the type of exon that has a high level of usage in some tissues, but has a low level of usage in the other tissues. They usually undergo strong tissue-specific regulations. There is still a lack a systematic method to identify switch-like exons from multiple RNA-seq samples. We proposed a novel method called iterative Tertile Absolute Deviation around the mode (iTAD) to profile the distribution of exon relative usages among multiple samples and to identify switch-like exons and other types of exons using a robust statistic estimator. We validated the method with simulation data, and applied it on RNA-seq data of 16 human body tissues and detected 3,100 switch-like exons. We found that switch-like exons tend to be more associated with Alu elements in their flanking intron regions than other types of exons.

Estimating the prevalence of functional exonic splice regulatory information.

In addition to coding information, human exons contain sequences necessary for correct splicing. These elements are known to be under purifying selection and their disruption can cause disease. However, the density of functional exonic splicing information remains profoundly uncertain. Several groups have experimentally investigated how mutations at different exonic positions affect splicing. They have found splice information to be distributed widely in exons, with one estimate putting the proportion of splicing-relevant nucleotides at >90%. These results suggest that splicing could place a major pressure on exon evolution. However, analyses of sequence conservation have concluded that the need to preserve splice regulatory signals only slightly constrains exon evolution, with a resulting decrease in the average human rate of synonymous evolution of only 1-4%. Why do these two lines of research come to such different conclusions? Among other reasons, we suggest that the methods are measuring different things: one assays the density of sites that affect splicing, the other the density of sites whose effects on splicing are visible to selection. In addition, the experimental methods typically consider short exons, thereby enriching for nucleotides close to the splice junction, such sites being enriched for splice-control elements. By contrast, in part owing to correction for nucleotide composition biases and to the assumption that constraint only operates on exon ends, the conservation-based methods can be overly conservative.

Splicing factors of SR and hnRNP families as regulators of apoptosis in cancer.

SR and hnRNP proteins were initially discovered as regulators of alternative splicing: the process of controlled removal of introns and selective joining of exons through which multiple transcripts and, subsequently, proteins can be expressed from a single gene. Alternative splicing affects genes involved in all crucial cellular processes, including apoptosis. During cancerogenesis impaired apoptotic control facilitates survival of cells bearing molecular aberrations, contributing to their unrestricted proliferation and chemoresistance. Apparently, SR and hnRNP proteins regulate all levels of expression of apoptotic genes, including transcription initiation and elongation, alternative splicing, mRNA stability, translation, and protein degradation. The frequently disturbed expressions of SR/hnRNP proteins in cancers lead to impaired functioning of target apoptotic genes, including regulators of the extrinsic (Fas, caspase-8, caspase-2, c-FLIP) and the intrinsic pathway (Apaf-1, caspase-9, ICAD), genes encoding Bcl-2 proteins, IAPs, and p53 tumor suppressor. Prototypical members of SR/hnRNP families, SRSF1 and hnRNP A1, promote synthesis of anti-apoptotic splice variants of Bcl-x and Mcl-1, which results in attenuation of programmed cell death in breast cancer and chronic myeloid leukemia. SR/hnRNP proteins significantly affect responses to chemotherapy, acting as mediators or modulators of drug-induced apoptosis. Aberrant expression of SRSF1 and hnRNP K can interfere with tumor responses to chemotherapy in pancreatic and liver cancers. Currently, a number of splicing factor inhibitors is being tested in pre-clinical and clinical trials. In this review we discuss recent findings on the role of SR and hnRNP proteins in apoptotic control in cancer cells as well as their significance in anticancer treatments.

Suppression of 5' splice-sites through multiple exonic motifs by hnRNP L.

Selection of 5' splice-sites (5'SS) in alternative splicing plays an important role in gene regulation. Although regulatory mechanisms of heterogeneous nuclear ribonucleoprotein L (hnRNP L), a well-known splicing regulatory protein, have been studied in a substantial level, its role in 5'SS selection is not thoroughly defined. By using a KLF6 pre-mRNA alternative splicing model, we demonstrate in this report that hnRNP L inhibits proximal 5'SS but promotes two consecutive distal 5'SS splicing, antagonizing SRSF1 roles in KLF6 pre-mRNA splicing. In addition, three consecutive CA-rich sequences in a CA cassette immediately upstream of the proximal 5'SS are all required for hnRNP L functions. Importantly, the CA-cassette locations on the proximal exon do not affect hnRNP L roles. We further show that the proximal 5'SS but not the two distal 5'SSs are essential for hnRNP L activities. Notably, in a Bcl-x pre-mRNA model that contains two alternative 5'SS but includes CA-rich elements at distal exon, we demonstrate that hnRNP L also suppresses nearby 5'SS activation. Taken together, we conclude that hnRNP L suppresses 5'SS selection through multiple exonic motifs.

RBM20 and RBM24 cooperatively promote the expression of short enh splice variants.

PDZ-LIM protein ENH1 is a scaffold protein for protein kinases and transcriptional regulators. While ENH1 promotes the hypertrophic growth of cardiomyocytes, its short splice variant (ENH3) prevents the hypertrophic growth. The mechanism underlying the alternative splicing of enh mRNA between ENH short and long isoforms has remained unknown. Here, we found that two splicing factors, RNA-binding motif 20 (RBM20) and RNA-binding motif 24 (RBM24) together promoted the expression of short enh splice variants and bound the 5' intronic region of exon 11 containing an in-phase stop codon. In addition, expression of both RBMs is repressed by hypertrophic stimulations. Collectively, our results suggest that, in healthy conditions, RBM20 and RBM24 cooperate to promote the expression of short ENH isoforms.

RED: A Java-MySQL Software for Identifying and Visualizing RNA Editing Sites Using Rule-Based and Statistical Filters.

RNA editing is one of the post- or co-transcriptional processes that can lead to amino acid substitutions in protein sequences, alternative pre-mRNA splicing, and changes in gene expression levels. Although several methods have been suggested to identify RNA editing sites, there remains challenges to be addressed in distinguishing true RNA editing sites from its counterparts on genome and technical artifacts. In addition, there lacks a software framework to identify and visualize potential RNA editing sites. Here, we presented a software - 'RED' (RNA Editing sites Detector) - for the identification of RNA editing sites by integrating multiple rule-based and statistical filters. The potential RNA editing sites can be visualized at the genome and the site levels by graphical user interface (GUI). To improve performance, we used MySQL database management system (DBMS) for high-throughput data storage and query. We demonstrated the validity and utility of RED by identifying the presence and absence of C→U RNA-editing sites experimentally validated, in comparison with REDItools, a command line tool to perform high-throughput investigation of RNA editing. In an analysis of a sample data-set with 28 experimentally validated C→U RNA editing sites, RED had sensitivity and specificity of 0.64 and 0.5. In comparison, REDItools had a better sensitivity (0.75) but similar specificity (0.5). RED is an easy-to-use, platform-independent Java-based software, and can be applied to RNA-seq data without or with DNA sequencing data. The package is freely available under the GPLv3 license at http://github.com/REDetector/RED or https://sourceforge.net/projects/redetector.

Spliced leader RNA trans-splicing discovered in copepods.

Copepods are one of the most abundant metazoans in the marine ecosystem, constituting a critical link in aquatic food webs and contributing significantly to the global carbon budget, yet molecular mechanisms of their gene expression are not well understood. Here we report the detection of spliced leader (SL) trans-splicing in calanoid copepods. We have examined nine species of wild-caught copepods from Jiaozhou Bay, China that represent the major families of the calanoids. All these species contained a common 46-nt SL (CopepodSL). We further determined the size of CopepodSL precursor RNA (slRNA; 108-158 nt) through genomic analysis and 3'-RACE technique, which was confirmed by RNA blot analysis. Structure modeling showed that the copepod slRNA folded into typical slRNA secondary structures. Using a CopepodSL-based primer set, we selectively enriched and sequenced copepod full-length cDNAs, which led to the characterization of copepod transcripts and the cataloging of the complete set of 79 eukaryotic cytoplasmic ribosomal proteins (cRPs) for a single copepod species. We uncovered the SL trans-splicing in copepod natural populations, and demonstrated that CopepodSL was a sensitive and specific tool for copepod transcriptomic studies at both the individual and population levels and that it would be useful for metatranscriptomic analysis of copepods.

How slow RNA polymerase II elongation favors alternative exon skipping.

Splicing is functionally coupled to transcription, linking the rate of RNA polymerase II (Pol II) elongation and the ability of splicing factors to recognize splice sites (ss) of various strengths. In most cases, slow Pol II elongation allows weak splice sites to be recognized, leading to higher inclusion of alternative exons. Using CFTR alternative exon 9 (E9) as a model, we show here that slowing down elongation can also cause exon skipping by promoting the recruitment of the negative factor ETR-3 onto the UG-repeat at E9 3' splice site, which displaces the constitutive splicing factor U2AF65 from the overlapping polypyrimidine tract. Weakening of E9 5' ss increases ETR-3 binding at the 3' ss and subsequent E9 skipping, whereas strengthening of the 5' ss usage has the opposite effect. This indicates that a delay in the cotranscriptional emergence of the 5' ss promotes ETR-3 recruitment and subsequent inhibition of E9 inclusion.

PP1γ functionally augments the alternative splicing of CaMKIIδ through interaction with ASF.

Protein phosphatase 1 (PP1) and Ca2+/calmodulin-dependent protein kinase δ (CaMKIIδ) are upregulated in heart disorders. Alternative splicing factor (ASF), a major splice factor for CaMKIIδ splicing, can be regulated by both protein kinase and phosphatase. Here we determine the role of PP1 isoforms in ASF-mediated splicing of CaMKIIδ in cells. We found that 1) PP1γ, but not α or ß isoform, enhanced the splicing of CaMKIIδ in HEK293T cells; 2) PP1γ promoted the function of ASF, evidenced by the existence of ASF-PP1γ association as well as the PP1γ overexpression- or silencing-mediated change in CaMKIIδ splicing in ASF-transfected HEK293T cells; 3) CaMKIIδ splicing was promoted by overexpression of PP1γ and impaired by application of PP1 inhibitor 1 (I1PP1) or pharmacological inhibitor tautomycetin in primary cardiomyocytes; 4) CaMKIIδ splicing and enhancement of ASF-PP1γ association induced by oxygen-glucose deprivation followed by reperfusion (OGD/R) were potentiated by overexpression of PP1γ and suppressed by inhibition of PP1γ with I1PP1 or tautomycetin in primary cardiomyocytes; 5) functionally, overexpression and inhibition of PP1γ, respectively, potentiated or suppressed the apoptosis and Bax/Bcl-2 ratio, which were associated with the enhanced activity of CaMKII in OGD/R-stimulated cardiomyocytes; and 6) CaMKII was required for the OGD/R induced- and PP1γ exacerbated-apoptosis of cardiomyocytes, evidenced by a specific inhibitor of CaMKII KN93, but not its structural analog KN92, attenuating the apoptosis and Bax/Bcl-2 ratio in OGD/R and PP1γ-treated cells. In conclusion, our results show that PP1γ promotes the alternative splicing of CaMKIIδ through its interacting with ASF, exacerbating OGD/R-triggered apoptosis in primary cardiomyocytes.

Comparative studies of differential gene calling using RNA-Seq data.

BACKGROUND: With its massive amount of data, gene-expression profiling by RNA-Seq has many advantanges compared with microarray experiments. RNA-Seq analysis, however, is fundamentally different from microarray data analysis. Techniques developed for analyzing microarray data thus cannot be directly applicable for the digital gene expression data. Several statistical methods have been developed for identifying differentially expressed genes specifically from RNA-Seq data over the past few years. RESULTS: In this study, we examined the performance of differential gene-calling methods using RNA-Seq data in practical situations. We focused on two representative methods: one parametric method, DESeq, and one nonparametric method, NOISeq. We examined their performance using both simulated and real datasets. Our simulation followed the RNA-Seq process and produced more realistic short read data. Both DESeq and NOISeq identified over-expressed genes more correctly than under-expressed genes. While DESeq was more likely to call longer genes as differentially expressed than shorter ones, NOISeq did not have such bias. When the underlying variation increased, both methods showed higher rates of false positives. When replicates were not available in the experiments, both methods showed lower rates of true positives and higher rates of false positives. CONCLUSIONS: The level of variation clearly affected the performance of both methods, showing the importance of understanding the variation in the data as well as having replications in RNA-Seq experiments. We showed that it is possible to obtain improved differential gene-calling results by combining the results obtained by the two methods. We suggested strategies to use these two methods individually or combined according to the characteristics of the data.