An in vivo parallelized reporter assay to uncover tissue-specific splicing regulatory sequences in a multicellular animal

Sanjana Bhatnagar; Jade Ho; Nour H Sadek; Isha Singh; Michael Zoberman; Bina Koterniak; Yufang Liu; Daniel D Fusca; Asher D Cutter; Alan M Moses; John A Calarco

doi:10.1093/nar/gkag451

. 2026 May 25;54(10):gkag451. doi: 10.1093/nar/gkag451

An in vivo parallelized reporter assay to uncover tissue-specific splicing regulatory sequences in a multicellular animal

Sanjana Bhatnagar ¹, Jade Ho ², Nour H Sadek ³, Isha Singh ⁴, Michael Zoberman ⁵, Bina Koterniak ⁶, Yufang Liu ⁷, Daniel D Fusca ⁸, Asher D Cutter ⁹, Alan M Moses ¹⁰, John A Calarco ^11,^✉

¹ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

² Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

³ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

⁴ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

⁵ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

⁶ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

⁷ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

⁸ Department of Ecology and Evolutionary Biology, 25 Willcocks Street, Toronto, Ontario, M5S 3B2, Canada

⁹ Department of Ecology and Evolutionary Biology, 25 Willcocks Street, Toronto, Ontario, M5S 3B2, Canada

¹⁰ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

¹¹ Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada

^✉

To whom correspondence should be addressed. Email: john.calarco@utoronto.ca

Roles

Sanjana Bhatnagar: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Writing - original draft, Writing - review & editing

Jade Ho: Investigation

Nour H Sadek: Formal analysis, Investigation, Validation

Isha Singh: Investigation, Validation

Michael Zoberman: Data curation, Formal analysis, Investigation

Bina Koterniak: Data curation, Formal analysis, Investigation

Yufang Liu: Investigation

Daniel D Fusca: Data curation, Formal analysis, Investigation

Asher D Cutter: Supervision, Writing - review & editing

Alan M Moses: Conceptualization, Funding acquisition, Methodology, Resources, Software, Writing - review & editing

John A Calarco: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Visualization, Writing - original draft, Writing - review & editing

PMCID: PMC13199693 PMID: 42179040

Abstract

Introns play a critical role in regulating alternative splicing. However, identifying functional intronic motifs is challenging due to their short and degenerate sequence composition. Massively parallel reporter assays have provided insights into cis-regulatory logic governing alternative splicing, but these approaches are generally performed in cell culture, limiting their ability to capture tissue-specific contexts. Here, we implemented in vivo parallelized reporter assays (PRA) in Caenorhabditis elegans neurons and muscle cells to screen for intronic enhancer and silencer motifs among thousands of randomized sequences. We identified nearly 200 sequences regulating splicing in these tissues. We uncovered core sub-sequences with tissue-biased enhancing and silencing activity, including motifs recognized by well-characterized RNA-binding proteins, and unmapped motifs with no obvious cognate binding protein. Mapping our PRA-derived motifs to native introns flanking tissue-biased alternative exons revealed their conservation across nematodes, supporting their functional relevance. Additionally, individual intronic regions frequently contained diverse combinations of these motifs, indicative of complex engagement of these sequences by RNA-binding proteins. Finally, we performed targeted mutagenesis of PRA-derived intronic enhancers flanking a neuronal microexon, identifying key cis-regulatory determinants of microexon splicing. Together, our study provides a framework to explore the role of intronic regions in tissue-specific splicing regulation within a multicellular organism.

Graphical Abstract

For image description, please refer to the figure legend and surrounding text.

Introduction

Pre-messenger RNA (mRNA) alternative splicing is a fundamental co-/post-transcriptional process, resulting in the removal of introns and selective joining of exons in various combinations. This mechanism vastly increases transcriptome and proteome diversity, allowing a limited number of genes to generate numerous RNA transcripts. Splice variants can confer specialized functions to widely expressed genes by being differentially expressed in a cell-, tissue-, organ-, developmental-, and physiological state-specific manner [1, 2]. For example, the molecular diversity of neuronal cell types in the nervous system has been attributed, in part, to a multitude of alternative splicing events [3, 4].

Intronic regions, such as the 5′ splice site (5′ss), the 3′ splice site (3′ss), the polypyrimidine tract, and the branch point site, form the basic framework of splicing regulation [2, 5, 6]. In addition to these canonical elements, regulatory sequences further modulate splicing by promoting or inhibiting spliceosomal assembly at exon–intron junctions by making use of exonic and intronic splicing enhancers (ISEs) and intronic splicing silencers (ISSs) [7–11]. These motifs influence cell-specific splicing outcomes by recruiting RNA-binding proteins (RBPs) that are differentially expressed across cell types (for examples see: 12–14). Therefore, characterizing these motifs in their native cellular context is crucial to understanding tissue-differential splicing regulation. Dissecting the specific effect of splicing regulatory motifs, however, remains challenging due to the inherent complexity of splicing regulation. Splicing regulation depends on many factors, including multivalent cis elements that occur in tandem and near alternative exons [12], RNA secondary structures in nascent transcripts [13], and inputs from upstream gene regulatory layers such as transcription dynamics and chromatin states [14]. These layers of complexity make it difficult to isolate the contribution of individual cis regulatory intronic motifs to splicing in native gene contexts.

Minigene splicing reporters, containing only alternative exons and their flanking intronic/exonic regions, have provided simpler sequence contexts to study the effect of cis regulatory elements on splicing patterns (for examples see: [2, 14–26]). referen For instance, early screens of randomized nucleotide sequences embedded within fluorescent reporters have successfully identified exonic splicing silencers [16]. More recently, deep sequencing has provided a direct readout of splicing patterns from reporters carrying libraries of randomized, rationally designed, or naturally occurring variant sequences [14, 17–21]. These massively parallel reporter assays (MPRAs) have garnered immense interest for their potential to reduce regulatory complexity by testing small regions of nucleotide sequence in parallel for cis regulatory activity while keeping other variables unchanged.

While powerful, MPRAs have been largely limited to experiments in mammalian cell cultures, given their amenability for effective transfection and scalability. This limits the generalization of the identified cis regulatory elements to in vivo contexts and often fails to address the context-dependent regulation of biological processes, which is crucial for understanding the development of distinct tissues from the same genome. In a few recent studies, MPRAs have been implemented in organisms to study transcription factor binding sites in promoter regions and mRNA decay dynamics systematically [22, 27–29]. However, adaptation of MPRAs to a multicellular organism to screen for sequences with tissue-biased splicing regulatory activity in vivo is lacking. Caenorhabditis elegans, with an intron-rich genome and RBPs conserved across metazoa, is an attractive model system to study tissue-differential alternative splicing [30, 31]. From a technical perspective, C. elegans also offers the unique capacity to accommodate large libraries of injected reporters in the form of extrachromosomal arrays [32], making it an attractive model for in vivo MPRA screens.

In this study, we developed a parallelized reporter assay (PRA) to identify sequences with splicing regulatory activity in tissues of the nematode C. elegans. We used our PRA design to survey the activity of several thousand splicing reporter plasmids containing random intronic decamer sequences expressed in C. elegans neuronal and muscle cells. We recovered 192 ISEs and ISSs with activity in one or both tissues. We further employed univariate linear regression (ULR) to identify the core 4–7mer (k-mer) subsequences responsible for the splicing regulatory properties of these synthetic intronic elements. Our k-mer analysis identified sequences similar to known C. elegans RBP consensus motifs but also identified uncharacterized sequences with broad or tissue-biased splicing regulatory activity. Finally, we mapped our PRA-identified intronic silencer and enhancer sequences onto endogenous introns flanking tissue-biased alternative exons. We found that these cis elements are conserved and confirmed the neuronal splicing-enhancing activity of two such motifs in a model microexon splicing decision. Taken together, our results highlight the utility of our PRA approach in understanding the regulatory logic of tissue-biased alternative splicing regulation in vivo.

Materials and methods

Caenorhabditis elegans strains used

We used N2 animals grown on OP50 Escherichia coli for this study using standard procedures [33]. All animals were grown in a 21°C incubator.

Construction and sequencing of synthetic reporter libraries for PRA

We utilized a previously developed unc-16 exon-16 splicing reporter system, containing mutagenized ISEs [34]. These enhancers are originally recognized by the trans factors UNC-75 (UGUUGUG) and EXC-7 (UAAGUU) [34]. This mutagenized reporter served as a reference reporter for our experiments. Synthetic PRA element reporters were generated using forward primers containing randomized 10-nucleotide oligonucleotides, which anneal upstream of the mutated intronic cis elements and a reverse primer containing a 10-nucleotide barcode (Integrated DNA Techologies, IDT), annealing within the 3′UTR. The resulting reporter fragments were then ligated using Gibson Assembly into digested vector backbones containing the pan-neuronal rgef-1 neuron-specific promoter and the myo-3 muscle-specific promoter and the unc-54 3′UTR.

The tissue-specific reference reporter was incorporated into pools as an internal control for comparison of splicing patterns from reporters containing synthetic 10-mer PRA elements. The plasmid libraries were pooled into five groups to potentially achieve a total of 800 vectors per pool. The libraries were prepared for sequencing, and an 8-nucleotide degenerate unique molecular identifier (UMI) was introduced downstream of the barcode within the 3′UTR using a reverse primer (5′-GAACAGTACGGTACTAAGCAAC NNNNNNNN GAAGAGTAATTGGACTCAGAAG-3′). These samples were further purified using the MagBind Total Pure NGS (Omega Bio-Tek) protocol to enrich for fragments of the correct length, containing the synthetic PRA element, barcode, and UMI. The enriched fragments were further amplified and prepared for multiplexed plasmid sequencing using the NEBNext Ultra II DNA Library Prep Kit for Illumina. The prepared samples were assessed using Bioanalyzer and quantified using quantitative polymerase chain reaction (qPCR). These were then subjected to paired-end MiSeq sequencing (2 × 150 bp). The sequencing data was further analyzed using custom Python scripts to extract true PRA element-barcode pairs using UMI counts (https://doi.org/10.5281/zenodo.17593183; see README.md). The intronic PRA elements and corresponding barcodes were extracted using the adjacent regions in the reporter. All primers and oligonucleotides used in this study are listed in Supplemental Table S1, and plasmid maps are available upon request.

Microinjection, RNA extraction, amplification, and sequencing of tissue-specific PRA reporter libraries in C. elegans

The tissue-specific reporter libraries were introduced by microinjection into Inline graphic adult animals as 50 pools, each containing 80 reporters, such that targeted cells carried reporters as extrachromosomal arrays. We harvested total RNA from a mixed population of the F₁ progeny of injected mothers after 3 days using TRI reagent (Sigma), followed by DNA digestion and cleanup using Zymo RNA Clean and Concentrator (+DNaseI) kit, according to manufacturer’s conditions. RNA was extracted from the injected population without prior screening for array-expressing animals, as co-injection markers were excluded from these mixes to maximize the copy number of reporters in arrays.

The RNA libraries were pooled into 5 groups, such that each pool consisted of RNA from up to 800 injected vectors. The purified RNA was reverse transcribed using Invitrogen™ SuperScript™ IV and oSB29 reverse primer (Supplemental Table S1), which introduced a UMI in each complementary DNA (cDNA) molecule. The oSB29 primer anneals to the barcode flanking region of the minigene reporter transcript, enabling selective reverse transcription of the reporter transcripts. The resulting reporter cDNA samples were cleaned up using 1.8× MagBind Total Pure NGS (Omega Bio-Tek) magnetic beads to cDNA ratio to remove the reverse transcription (RT) primer. We selectively enriched for reporter transcripts using Q5^® High-Fidelity 2× Master Mix using primers oSB30 and oSB31 primers (Supplemental Table S1). This was followed by another round of cleanup using MagBind Total Pure NGS (Omega Bio-Tek) in a 1.4× beads to cDNA ratio. These samples were prepared for multiplexed amplicon sequencing using the NEBNext^® Ultra™ II for DNA Library Prep kit as recommended by the manufacturer. We further assessed the samples using a Bioanalyzer (Agilent) and quantified them using qPCR for accurate multiplexing before sequencing. The mixed samples were then subjected to paired-end MiSeq sequencing (2 × 150 bp) with 20% PhiX spike-in as an internal control to improve the accuracy of base calling by increasing diversity for sequencing.

High-throughput classification and quantification of splicing isoforms in tissue-specific PRA libraries

Each tissue-specific PRA library yielded ∼12 million reads, with a mean sequence quality (Phred score) of 34. The reads were filtered based on the quality scores and then subjected to downstream analysis. A custom Python-based splicing isoform classifier algorithm was developed to accurately categorize reads by utilizing sequence motifs such as barcodes and UMIs, and exon–exon and exon–intron junctional features (https://doi.org/10.5281/zenodo.17593183; see README.md). Initially, we extracted barcodes and UMIs from the reads by considering the flanking regions. Only UMIs that met specific length criteria (8 nucleotides) were included, while excluding a small proportion of improperly extracted UMIs and longer sequences that could result in inaccurate counts. The classifier identifies five isoform categories: included, skipped, cryptic, unspliced, and unknown. Reads are classified as spliced-in isoforms if both the Exon 1–Alternate Exon junction (TTAAAGCTAA) and the Alternate Exon–Exon3 junction (TGAAAAAAGAAGATCCA) are detected. If the Exon1–Exon3 junction (TAAAGAAGAT) is present, and the AlternateExon–Exon3 junction (TGAAAAAAGAAGATCCA) is absent, the read is classified as a spliced-out isoform. Furthermore, unspliced isoforms are detected based on the portions of intron 1 (AATTTTTTAG) or intron 2 (CAAATTTTTTCAG) present in the reads. The classifier further calculates Hamming distance for regions within each read against the abovementioned reference motifs with a mismatch threshold of <2 nucleotides. This accounts for sequencing errors while maintaining specificity and further ensures robust identification of splicing events. Reads that were not classified as spliced isoforms or unspliced isoforms due to an exceeding number of mismatches with the expected junction sequences and were thus lacking a predefined junction were categorized as unknown isoforms. This serves as a fallback category for ambiguous isoforms. Interestingly, a subset of these unknown isoforms appeared to result from competitive splice sites (CSSs). We then optimized the splicing isoform classifier algorithm to classify such events as cryptic splicing isoforms. Cryptic splicing isoforms were identified by the presence of the Exon 1–Alternate Exon junction (TTAAAGCTAA) and Exon 3 (AAGATCCAT), along with a sequence upstream of the inserted PRA element (GTTTCAAAT or the shorter sequence AAATTGG).

After classification of reads, UMIs were collapsed into read counts by ensuring that only unique UMI-barcode pairs were counted; duplicate pairs were removed to avoid inflation of read numbers within a category. Additionally, the reporters with >20 read counts were selected for downstream analysis. The percent spliced-in (PSI) values were calculated using read counts of different isoforms corresponding to each PRA element and barcode. Specifically, the PSI value represents the percentage of spliced-in isoforms in the population of canonically spliced isoforms [35].

After calculation of PSI values, Fisher’s exact test was applied using spliced-in and spliced-out read counts for each PRA element reporter compared to the reference reporter. To control for false discovery rates (FDR), the Benjamini–Hochberg procedure was applied, with a threshold set at 0.01.

Similarly, cryptic isoform % represents the percentage of cryptic-splicing isoforms in the total population of spliced isoforms. To group different PRA CSS inducers, these were ranked from highest to lowest based on cryptic isoform %. We used quantile-based binning to divide the data into five bins containing approximately equal numbers of observations, after sorting the elements in ascending order based on their cryptic isoform %.

Individual reporter plasmid construction and semi-quantitative reverse transcriptase-polymerase chain reaction validations

We prepared plasmids for a subset of selected reporters for validations and follow up experiments. In some cases, we also constructed plasmids with mutated 10-mers to confirm that ISE activity was due to the intronic sequence. Reporter fragments were amplified in two parts, the upstream portion of the reporter extending to the intronic 10-mer was amplified using TN nested F primer (for neuronal-targeted reporters) or Muscle F (for muscle-targeted reporters) along with common reverse primer unc-16 USR (Supplemental Table S1). The downstream portion of the reporter containing the intronic 10-mer, as well as the 10 nucleotide barcode, was amplified using a forward primer consisting the intronic PRA element and a reverse primer containing the barcode. These fragments were synthesized through PCR (Phusion polymerase, NEB) by using a neuronal rgef-1 promoter:unc-16 minigene reference reporter plasmid or a muscle myo-3 promoter:unc-16 minigene reference reporter plasmid as a template (Supplemental Table S2).

These fragments were resolved on gel and further purified for Gibson Assembly into either the neuronal unc-16 minigene vector backbone, digested with SnaBI and MluI or the muscle unc-16 minigene vector backbone, digested with SnaBI and XhoI restriction enzymes. Plasmids were purified from DH5-alpha bacteria cultures using the Presto Mini Plasmid kit from GeneAid. The isolated reporter plasmids were verified by Sanger sequencing (oYL 13 reverse primer), and the sequences were assessed for proper insertion of the intronic PRA elements and the barcodes (Supplemental Table S1).

Plasmids were delivered into Inline graphic adult animals by microinjection. After 3 days, F₁ progeny were harvested, and we used the previously stated protocol for RNA extraction and purification. This was followed by a semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) assay using the QIAGEN OneStep RT-PCR Kit, as recommended by the manufacturer. The forward primer annealed at the start of exon 1 (unc-16 RT F) and the reverse primer (unc-16 RT R) annealed at the end of exon 3 (Supplemental Table S1). In general, 25–50 ng of total RNA was used as input and 30 cycles of amplification were performed. This was followed by performing 2% agarose gel electrophoresis on the obtained cDNA samples. We further performed densitometry measurements on the gel bands using Fiji ImageJ software ( Inline graphic [36]. PSI values were calculated after correcting for heteroduplex signal by adding half of the heteroduplex intensity to the background-corrected spliced-in band (numerator) and adding the total heteroduplex intensity to the combined background-corrected spliced-in and spliced-out intensities (denominator).

All primer sequences used are available in Supplemental Table S1, and all plasmids created are available upon request.

Mapping PRA elements functioning as competitive 5′ss to native C. elegans intronic regions

We used the C. elegans transcript annotation file (wormbase release WBPS15, GFF, and GTF files) [37] and inferred the intron coordinates [38]. This modified GFF with intronic coordinates was then processed to obtain the intron bed file. Then we used the genome sequence FASTA file from the same wormbase release WBPS15, and using the bedtools getfasta command [39], we obtained a customized FASTA file containing intronic sequences.

Next, we used the Find Individual Motifs Occurrence (FIMO) from the MEME suite [40], to scan the intronic FASTA file with 54 PRA-derived competitive 5′ss. Each motif was represented as a 9-mer spanning positions −3 to +6, capturing the competing splice site. FIMO was run with a P-value threshold of 1 × 10^-4. Further, default settings were used for background model estimation and q-value computation using Benjamini Hochberg procedure. Additionally, scanning was performed on the coding strand.

The strength of competitive 5′ss was evaluated using MaxEntScan scores to quantify their complementarity to U1 small nuclear RNA (snRNA) [41]. The maxentscan package was installed via Bioconda and applied to the 9-mer sequences corresponding to the CSSs (http://bioconda.github.io/recipes/maxentscan/README.html).

k-mer count ULR to characterize core regulatory subsequences

To generate different k-mer count matrices, we calculated the counts of each type of k-mer Inline graphic in a 16 bp region, including the intronic 10-mer and the flanking 3 nucleotides on each side. We limited the flanking length to 3 bp on both sides to control for the false overrepresentation of k-mers that overlap more with the flanking region relative to the element. Each k-mer is treated as an independent predictor of the PSI value of the synthetic PRA element it belongs to. We performed ULR using the counts of only those k-mers that exist in at least five distinct synthetic PRA elements. This filter reduces noise from sparsely occurring k-mers that might not yield reliable estimates. We then performed ULR using the sklearn.linear_model package and further used scipy.stats for statistical functions in Python. We performed linear regression individually for each predictor to establish the association between the k-mer and the observed PSI value (https://doi.org/10.5281/zenodo.17593183; see README.md).

To account for measurement noise, we computed t-statistics, which inherently quantify variability by normalizing regression coefficients by their standard errors. This approach allows for a standardized assessment of the significance of each regression coefficient, facilitating comparisons across different models and datasets. The calculation of t-stat values was followed by selection of the k-mers with significantly strong effect on splicing regulation (Benjamini–Hochberg FDR, q-value = 0.05).

Clustering tissue-biased PRA k-mers based on sequence similarity

The highly similar k-mers with a significantly strong effect on splicing in both tissues were clustered using the method described in [42]. The k-mers are further fragmented into subsequences, and the counts of the resulting subsequences are taken across the k-mer catalogs to generate a count matrix. This represents the sequence properties of each PRA intronic k-mer, and similarity is evaluated by calculating the Jaccard distance. Next, affinity propagation was performed using the same parameters as in [42], on the obtained Jaccard distance matrix to identify the appropriate number of clusters present in the data using a custom python script. This was used to guide the number of clusters defined for k-means clustering, which was subsequently performed to extract clusters of highly similar sequences (https://doi.org/10.5281/zenodo.17593183).

Once the k-mer clusters were obtained, they were processed to generate position weight matrices (PWMs) using R as previously described [43]. Pairwise alignments were performed within each cluster, followed by arranging them by scores (in descending order). Then, multiple sequence alignment using ClustalW [44] was performed. Later, PWMs [43] were inferred from the multiple sequence alignment for each k-mer cluster, as well as consensus sequences and corresponding logos were obtained using the R package seqlogo in both tissue libraries (https://doi.org/10.5281/zenodo.17593183).

Motif comparison of significant splicing enhancer and silencer k-mers with RBP motifs across tissues

We performed motif comparison analysis to compare splicing enhancing and splicing silencing k-mers against the latest C. elegans RBP motif catalog from the RNAcompete dataset [45] using TOMTOM, a MEME suite motif tool [46]. We used the Sandelin-Wasserman similarity approach to calculate motif similarity between splicing enhancing and silencing kmer cluster PWMs from both muscle and neuronal tissues by disabling the reverse complement strand search parameter. We further chose a statistical cutoff of Inline graphic to report any significant matches.

Phylogenetically averaged motif score analysis on C. elegans neuronal/muscle switch-like splicing events

In order to identify phylogenetically conserved cis elements by phylogenetically averaged motif (PAM) score analysis, we developed a suite of scripts to extract and catalog orthologous intronic sequences based on transcript annotation files. We began by annotating all alternative and constitutive splice junctions in C. elegans using the coordinate and transcript information present in a recent transcript annotation file (wormbase release WBPS15, GFF, and GTF files) [37]. These were classified into various splicing categories: constitutively spliced exons (always included in mRNA), alternatively spliced exons such as skipped exons (exons that are either included or excluded from the mRNA), alternative 5′ss exons (exons with two donor sites, where one is selected), alternative 3′ss exons (exons with two acceptor sites, where one is selected), retained introns (introns incorporated into mRNA), and composite exons (exons exhibiting >1 type of alternative splicing, e.g. alternative 5′ss exons that are also skipped). Additionally, since the GFF file does not include information on introns, we inferred the intron coordinates using AGAT software [38]. The output file contained intron coordinates inferred from exon coordinates in the GFF file. Hence, we developed a custom pipeline to filter out introns and further add intron metadata by leveraging the annotation of the exons they flanked (https://doi.org/10.5281/zenodo.17593183). Notably, this pipeline infers splicing event metadata directly from genomic coordinates in the GFF file, without requiring expression datasets for classification. This makes it a versatile and broadly applicable tool for splicing analysis across species—applicable wherever a GFF annotation is available and not limited to the context of this study.

For this analysis, we focused on identifying orthologous flanking intron sequences from 47 switch-like splicing events differentially spliced in neuron and muscle cells (Supplemental Table S3; B.K and J.A.C., unpublished results) using sequences from up to 50 nematode species per gene. A minimum of five orthologous sequences per intronic fragment was required for PAMS computation to be feasible, based on orthologous groups generated in [47] using OrthoFinder [48], and the ortholog calls came from genomes obtained from WormBase ParaSite [37] and the Caenorhabditis Genomes Project (https://zenodo.org/records/12633738). These comprised of orthologs for the longest isoform of a given gene in C. elegans from the following species—Caenorhabditis afra, C. angaria, Caenorhabditis astrocarya, Caenorhabditis becei, Caenorhabditis bovis, Caenorhabditis brenneri, Caenorhabditis briggsae, Caenorhabditis castelli, Caenorhabditis dolens, Caenorhabditis doughertyi, Caenorhabditis drosophilae, Caenorhabditis imperialis, Caenorhabditis inopinata, Caenorhabditis japonica, Caenorhabditis kamaaina, Caenorhabditis latens, Caenorhabditis macrosperma, Caenorhabditis monodelphis, Caenorhabditis nigoni, Caenorhabditis nouraguensis, Caenorhabditis oiwi, Caenorhabditis panamensis, Caenorhabditis parvicauda, Caenorhabditis plicata, Caenorhabditis portoensis, Caenorhabditis quiockensis, Caenorhabditis remanei, Caenorhabditis sinica, Caenorhabditis sp. 2, Caenorhabditis sp. 24, Caenorhabditis sp. 25, Caenorhabditis sp. 27, Caenorhabditis sp. 30, Caenorhabditis sp. 33, Caenorhabditis sp. 48, Caenorhabditis sp. 49, Caenorhabditis sp. 51, Caenorhabditis sp. 54, Caenorhabditis sp. 56, Caenorhabditis sp. 8, Caenorhabditis sulstoni, Caenorhabditis tribulationis, Caenorhabditis tropicalis, Caenorhabditis uteleia, Caenorhabditis virilis, Caenorhabditis waitukubuli, Caenorhabditis wallacei, Caenorhabditis yunquensis, Caenorhabditis zanzibari.

We identified most similar homologs for each C. elegans gene by first aligning the protein sequences for all sequences per orthology group (MAFFT) [49], followed by computing percent identity with the C. elegans protein and selecting the gene with the encoded protein sequence with the highest score per species. After identifying these groups, we selected the corresponding whole gene sequences from the species and performed whole gene sequence alignment (MAFFT). For each MAFFT run, we selected parameters –adjustdirectionaccurately to account for strandedness. This meant that all sequences will be on the same strand within an orthology group but might vary from one group to another, keeping the C. elegans sequence strand as reference. MAFFT adjusts the strandedness within ortholog groups as per C. elegans reference sequence.

Due to the unavailability of whole-genome alignments for all of these Caenorhabditis species, we could not leverage conventional genome liftover methods. Hence, we developed a custom alignment-based pipeline to project reference exon–intron coordinates onto orthologous sequences. This coordinate projection method employs exons as the reference in the multiple sequence alignments. The alignment approach assumes that an exon in C. elegans aligns to the corresponding exon in other species, and that the flanking introns align exclusively to intronic regions. Given that the alignment includes both exonic and intronic regions, the flanking intronic sequences surrounding an exon can be directly extracted and analyzed.

Consequently, the exon coordinates in the gapped sequences in alignment are deduced and further used to extract upstream and downstream intron fragments for C. elegans as well as the aligned orthologous sequences from other species. The length of C. elegans intron fragments are determined dynamically (for introns >70 bp in size, a 70 bp fragment is extracted, whereas for introns shorter than 70 bp, the entire intron is considered). The intron fragment length is standardized across orthologous sequences to match the C. elegans intron length. This ensures that the extracted fragments from other species align with the reference species both in size and positional context. We further filtered out the orthologous sequences that are too similar (percent identity = 0.95) to the reference and ensure scores are being calculated using sequences with higher sequence divergence. This was done to avoid overestimating conservation PAM scores due to duplicated sequences between species within these groups, which may not be under evolutionary constraint but rather have not had sufficient time to diverge. Additionally, only those C. elegans intron fragments that had a minimum of five orthologous sequences were used for the analysis. In this case, all the intron fragments adjacent to the selected 47 switch exons had >5 orthologous sequences within the ortholog group after filtering, thus none were dropped out of the analysis. We further adopted the pipeline previously developed [50] to compute PAM scores for our identified tissue-biased ISE and ISS k-mer motif PWMs in intron fragment ortholog groups (https://doi.org/10.5281/zenodo.17593183). This resulted in a matrix with PAM scores for all PRA splicing regulatory motifs across all intron fragments. We further performed hierarchical clustering on motifs as well as intronic fragments using cluster3 [51] to group intronic fragments with similar motif signatures and co-occurring PRA splicing regulatory motifs in a group of sequences. The resulting clustered file was visualized using Java TreeView [52].

Targeted mutagenesis of neuronal ISE k-mers in dpy-23 minigene reporter

We targeted neuronal ISE motifs 2 and 3 that were found to be conserved in the downstream intron of dpy-23 microexon 4 through PAM analysis. The dpy-23 minigene reporters with the minimal gene fragment—including the tissue-differentially spliced microexon and adjacent constitutive exons and introns—were selectively expressed in C. elegans neuronal tissues using the pan-neuronal promoter rgef-1.

The mutated dpy-23 minigene plasmids were prepared in two parts: upstream portion was synthesized through PCR using forward primer oMZ127 and reverse primer oSB266 (Supplemental Table S1) and neuronal rgef1:dpy-23 minigene reporter as a template. The downstream portions for three types of mutated reporters were ordered in the form of gBlocks (Supplemental Table S1). The dpy-23 minigene reporter plasmid was digested using Not1 and Bsu36I restriction enzymes and Fast Digest Green buffer.

Next, we used Gibson Assembly to assemble plasmids using these fragments and the digested neuronal dpy-23 reporter vector backbone. The assembled plasmids were purified as described above. Subsequently, we prepared microinjection mixes containing mutated neuronal dpy-23 minigene reporters, wild-type neuronal and muscle-expressed dpy-23 minigene reporters separately. We added pCFJ90 (myo-2 promoter::mCherry) at 2.5 ng/ul, and pCFJ104 (myo-3 promoter::mCherry) at 5 ng/μl to the microinjection mixes as co-injection markers. Microinjection mixes for each reporter were injected into 6 N2 adults each. The lines were established by picking and maintaining mCherry positive animals with stabilized extrachromosomal arrays. We then prepared three replicates with four to five plates of transgenic animals each. Each plate was seeded with three adults, and plates were harvested after 3 days, and RNA was extracted using TRI reagent (Sigma), followed by DNA digestion and cleanup using Zymo RNA Clean and Concentrator (+DNaseI) kit, according to manufacturer conditions. We further performed targeted RT-PCR, with forward primer oMZ 266 and reverse primer oMZ 175. The amplified cDNA samples were further run on 3% agarose gel for 2.5–3 h for high resolution of microexon spliced-in and spliced-out bands. Later, the gels from these samples were imaged and PSI densitometry values were calculated using Fiji ImageJ software [36].

Results

Implementation of a parallelized splicing reporter assay in C. elegans neuronal and muscle tissue

To screen for intronic elements capable of modulating splicing in different tissues, we developed a PRA in C. elegans. Two synthetic splicing reporter libraries were constructed under the control of promoters that drive expression in either neurons or body wall muscle cells (Fig. 1A). Each reporter library was prepared to include up to 4000 random 10 nucleotide (10-mer) sequences strategically located in the downstream intron flanking an alternative exon from a previously characterized unc-16 minigene [14] (Fig. 1A). Specifically, the random 10-mer sequence was positioned adjacent to a pair of mutated ISEs recognized by the RBPs UNC-75/CELF and EXC-7/Hu [14]. We reasoned that by weakening exon inclusion in this manner, we could screen the inserted random sequences for activity from a location known to harbor bona fide enhancers. Each reporter was indexed with a unique 10 nucleotide barcode in the 3′UTR, and UMI were introduced in reporter transcript cDNA for de-multiplexing and precise quantification of reporter transcripts, respectively [53] (Fig. 1B). This approach allowed us to measure splicing outcomes for different reporters using PSI, which measures the percentage of exon-included isoforms within the overall population of spliced transcripts produced by each reporter.

Furthermore, to identify cis-element and barcode pairs within each reporter, we performed multiplexed plasmid sequencing of the reporter libraries (see the ‘Materials and methods’ section). Similar to the cDNA libraries, we indexed each plasmid with a UMI to quantify its actual representation within the plasmid pools (see the ‘Materials and methods’ section). In addition, in both library expression experiments, we also spiked in reference reporters, which lacked an inserted 10-mer sequence, to serve as a baseline for reporter splicing patterns in each tissue. We recovered 2222 reporters and 3886 reporters from the neuron-expressed and muscle-expressed PRA libraries, respectively, containing unique cis-elements and barcode pairs (Supplemental Fig. S1). To investigate this discrepancy in library sizes, we quantified the relative abundance of our reference reporter (which also served as the parental vector used to construct our libraries), compared to the randomized PRA reporters. The neuronal library consistently showed a higher ratio of reference reporters to randomized PRA reporters, as measured using plasmid UMI counts. In contrast, the muscle library exhibited substantially lower reference reporter ratios, (Supplemental Fig. S1A). As a result, a larger fraction of the neuronal library contained background reference reporter, rather than the randomized reporters, thus reducing the number of recoverable unique cis-element and barcode pairs.

After injecting each library into animals, purifying RNA, and sequencing cDNA libraries, we successfully recovered isoform reads for 1,299 and 3,526 reporters from the neuron- and muscle-expressed libraries, respectively (Supplemental Fig. S1C and Supplemental Tables S4 and S5; and see the ‘Materials and methods’ section). We assessed RNA recovery efficiency for each reporter plasmid pool, sequenced as five pools of ∼800 reporter plasmids each, in both the neuronal and muscle libraries. Reporters with at least 20 reads were considered successfully recovered. Overall, neuronal reporter pools showed reduced RNA recovery (RNA recovered for 53% reporters on average) compared to muscle reporter pools (RNA recovered for 93% reporters on average). (Supplemental Fig. S1B).

We further confirmed that each nucleotide position of the inserted 10-mers was represented by all four bases without any systematic bias (Supplemental Fig. S2A). To demonstrate reproducibility of our PRA results, two biological replicates of a smaller-scale PRA were also generated using a subset of 71 randomly selected reporters from the larger neuronal tissue reporter library (Supplemental Table S6). A strong correlation was observed between replicate PSI values from the 71 reporters when they were expressed in neurons or in muscle cells ( Inline graphic (Supplemental Fig. S2B and C). PSI value measurements were also positively correlated between the 71-reporter library and the original 4000-neuronal reporter library ( (Supplemental Fig. S2D). Taken together, our results indicate that C. elegans can be used to survey thousands of reporters in parallel to monitor the impact of introduced random sequences on alternative splicing patterns in vivo.

PRA sequences resembling a strong 5′ss consensus lead to competing splice site usage

Alternative splicing occurs through the relative ability of the spliceosome to recognize competing splice sites [54, 55]. One determinant of splice site usage is the relative strength of the 5′ss, generally dictated by their complementarity to the first nine bases of U1 snRNA [56, 57]. Prior PRA experiments have noted that random sequences inserted into reporters can occasionally trigger competing splice site usage [14, 58]. Given the location of our inserted 10-mer sequences downstream of the native 5′ss flanking the reporter alternative exon, we reasoned that some 10-mer sequences could similarly generate competing splice sites in our reporter pre-mRNA.

To characterize this possibility, we searched our reporter library sequencing data for cDNAs with unexpected exon junction connectivity (Fig. 2). The neuronal- and muscle-expressed reporter transcript populations were primarily composed of canonical spliced transcripts (90% and 93.22%, respectively; Fig. 2A and B). However, we also identified noncanonical splice-isoforms, representing cryptic splicing events at novel intronic 5′ss introduced by a subset of synthetic 10-mer elements (Fig. 2A and B). The overall prevalence of such transcripts was limited to 1.95% of all neuronal transcripts and 1.21% of all muscle transcripts (Fig. 2A and B).

The neuronal and muscle PRA elements exhibiting cryptic splicing activity were not directly comparable due to the absence of overlap between the two sets. Additionally, cryptic splicing was rare in the muscle dataset (6 of 3526 elements > 5%) compared to the neuronal dataset; hence, cryptic splice site usage was not further analyzed in this comparative context. At a single reporter level, we identified a total of 54 10-mer sequences generating unanticipated cryptic isoforms at a rate of ≥5% (Fig. 2C and Supplemental Table S7) from both the neuron- and muscle-expressed libraries. We independently verified cryptic splice site usage resulting from reporters containing the top four CSS sequences by semi-quantitative RT-PCR, consistent with our analysis of the sequencing data (Fig. 2D). To assess sequence characteristics of competitive 5′ss of varying strengths, we grouped the CSS inducing reporters into five bins based on cryptic isoform usage, ranging from 5% to 80% (Fig. 2E). The sequences in each bin were aligned at the 5′ss to generate PWMs and consensus sequences. As expected, the bin containing reporters with the highest cryptic isoform usage rates (59.7%–75.6%) most closely matched the 5′ss consensus (5′-C/A)AG|GUAAGU-3′) [59]. Moreover, bins with lower cryptic isoform usage rates displayed progressively weaker matches to the consensus 5′ss sequence (Fig. 2E). Interestingly, in Bin 3, sequences enabling moderate levels of CSS usage included noncanonical nucleotides such as a C at intronic position +2 (Supplemental Table S7). These PRA elements containing competitive 5′ss were further searched against native C. elegans intronic sequences. Notably, all 54 competitive 5′ss mapped to C. elegans intronic regions ( Inline graphic ; see the ‘Materials and methods’ section), with 51 of them located within 50 bp of the canonical 5′ss within these introns (Supplemental Table S8). Additionally, we assessed the affinity of these competitive 5′ss for U1 snRNA consensus motif. While sequences with high scores generally aligned well with the consensus, their functional impact on splicing (cryptic%) did not always correlate with the score, indicating that sequence context and other regulatory elements influence splice-site utilization (Supplemental Table S7).

Collectively, these results demonstrate that sequences resembling consensus 5′ss are sufficient to compete with weak alternative splice sites for recruitment of the splicing machinery. Our results further indicate that our PRA can recover sequence features dictating CSS usage.

Synthetic PRA sequences can act as ISEs or ISSs

We next determined whether our PRA approach could identify intronic 10-mers in our libraries that significantly altered PSI values when compared to a reference reporter control ( Inline graphic , Fisher’s exact test followed by Benjamini–Hochberg FDR . We also required that any candidate regulatory 10-mers led to minimal (<5%) cryptic splicing in reporter transcripts. Importantly, using these criteria, we recovered 56 10-mers with putative ISE activity, and 29 10-mers acting as putative ISSs from our neuronal-expressed reporter library (Supplemental Table S1). From our muscle-expressed reporter library, we recovered seventy-three 10-mers with putative ISE activity, and two 10-mers acting as potential ISSs (Fig. 3A and Supplemental Table S2). To assess the robustness of the PRA in recovering ISEs and ISSs, we independently measured the splicing patterns of reporters from the neuron-expressed library by semi-quantitative RT-PCR and densitometry (Fig. 3B). We chose 10 reporters containing 10-mers with measured ISE activity ranging from strong to more modest. All the reporters tested displayed increased alternative exon inclusion compared to the reference reporter control and generally correlated with PSI values measured by our PRA (Fig. 3B and C).

Inline graphic — Identification of tissue-specific ISEs and ISSs. (A) Sequencing data analysis resulted in the identification of tissue-specific PRA cis-elements with PSI values significantly higher than the reference reporter (PRA ISEs) and with PSI values significantly lower than the reference reporter (PRA ISSs). (; P <.05, Fisher’s exact test followed by Benjamini–Hochberg FDR critical value < 0.01). (B) Ten neuronal PRA ISEs were selected for semi-quantitative RT-PCR validation. The bands shown on the gel correspond to exon-included isoforms, exon-skipped isoforms, and a heteroduplex—a hybrid of exon-included and excluded species (marked with an asterisk). Gel densitometry-based PSI values are displayed along with RNA-Seq derived PSI values. The sequences of these PRA ISEs are presented in the accompanying table. The gel densitometry–based PSI values were calculated from three replicates and corrected for heteroduplex band intensity; these values are shown in the plot below. The table shows the PSI densitometry values (averaged standard deviation across replicates) and the PSI RNA seq values for the reporters. (C) The linear regression plot shows correlation between PSI densitometry (averaged across replicates) and PSI RNA seq values.

To test whether the observed splicing regulatory effects above can be confidently attributed to the introduced intronic 10-mers versus the ten nucleotide barcodes located in the reporter 3′UTRs, we arbitrarily and minimally mutated the intronic sequences of 5 reporters with neuronal PRA ISEs (Supplemental Fig. S3). In all cases tested, we observed a reduction in exon inclusion in reporters with mutated 10-mers compared with reporters containing the putative enhancer sequences (Supplemental Fig. S3). Taken together, our results indicate that our PRAs can recover intronic sequences with splicing regulatory activity in vivo.

Regulatory activity of recovered PRA cis elements can be attributed to core subsequences

Motifs associated with alternative splicing regulation in vivo are generally 4–7 nucleotides long [60–62]. Therefore, we tested shorter subsequences within our library of 10-mer intronic elements for their regulatory potential (Fig. 4) and assessed if these short sequences were associated with tissue-biased splicing activity (Fig. 5 below). We used k-mer count ULR to comprehensively capture the contribution of 4–7 nucleotide subsequences to observed splicing outcomes. K-mer count matrices were generated for the 16 bp region containing the randomized 10 bp synthetic PRA sequence and 3 bp of flanking intronic region on both sides to account for regulatory motifs created due to adjacent sequences in the reporter (Fig. 4A). Based on the diversity of our PRA libraries, 100% of all possible 4-mers and, 87.8% and 99.7% of all possible 5-mers in neuronal and muscle libraries, respectively, were tested (Fig. 4B). We used the t-stat value as a measure to quantify the effect of each k-mer on the splicing outcome as it inherently accounts for the uncertainty in the regression coefficient estimates by incorporating standard error (Fig. 4C and D; see the ‘Materials and methods’ section for details).

Using this approach, a total of 72 and 185 intronic PRA element k-mers with significant splicing regulatory effects were identified from our neuron- and muscle-expressed reporter libraries, respectively (two-tailed t-test, P-value <.05; Benjamini–Hochberg FDR < 0.05) (Fig. 4C and D and Supplemental Tables S9 and S10). Here, we focus on reporting features of 6-mers, as most identified 4-mers and 5-mers could be mapped within a 6-mer context (Supplemental Tables S9 and S10, and Fig. 5 below). We assessed whether these 6-mers could be mapped to experimentally characterized C. elegans RBP motifs using TOMTOM (statistical cutoff of q-value < 0.05) [46] with C. elegans RBP PWMs derived from RNAcompete datasets (Fig. 4E) [63]. Interestingly, we recovered several 6-mers bearing similarity to known RBP consensus motifs. From our library expressed in neurons, the 6-mer with the strongest ISE activity was GCAUGC (t-stat = 5.82), followed by UUGUGC (t-stat = 5.55) (Fig. 4C). These sequences bear a strong resemblance to the consensus sequences of FOX-1/ASD-1 [64, 65] and UNC-75 [34, 66–68], known RBP splicing regulators expressed in neurons (Fig. 4E). From our muscle-expressed library, the 6-mer with the strongest ISE activity was GUGUGU (t-stat = 6.9), which closely resembles the consensus sequence recognized by SUP-12, a known muscle-expressed RBP splicing regulator [69, 70] (Fig. 4D and E).

In addition to recovering sequences resembling characterized RBP consensus motifs, our k-mer analysis also identified sequences with putative ISE and ISS activity not presently mapped to a cognate C. elegans RBP. For example, the most potent splicing-silencing 6-mers, UGGCGC (neuron-expressed library; t-stat = −3.81) and UCCUGC (muscle-expressed library; t-stat = −3.98) shared no resemblance with any known RBP motifs (Fig. 4C and D). GA-rich 6-mers with ISE activity were also identified from both tissue-expressed libraries. Additionally, AAGGAU (neuron-expressed library t-stat = 4.37) and GAAGGA (muscle-expressed library t-stat = 5.04) were identified; however, they did not match any of the available C. elegans RBP motifs (Fig. 4C). Collectively, our results demonstrate that our PRA can facilitate the identification of both known and novel cis elements with splicing regulatory activity.

Direct comparison of intronic enhancer and silencer activity in neuronal and muscle tissue reveals tissue-biased roles for sequences

Tissue-biased splicing patterns are generated through the cell-specific activity of RBPs binding to cognate cis elements [12, 71, 72]. As such, we conducted a direct comparative analysis of the identified k-mers to investigate the differential effects of sequences on tissue-specific splicing regulatory activity, using the tissue-specific t-stat as a metric (Fig. 5; see the ‘Materials and methods’ section). In total, 21 k-mers exhibited similar splicing effects in both tissues and were classified as shared splicing regulators (Fig. 5A and Supplemental Table S11). In contrast, 119 k-mers (4–7 nucleotides) displayed tissue-biased splicing effects, with 82 k-mers (68 enhancing; 15 silencing) exhibiting muscle-biased and 37 k-mers (33 enhancing; 4 silencing) exhibiting neuron-biased splicing effects (Fig. 5A and Supplemental Table S11).

As mentioned above, a subset of smaller k-mers with tissue-biased activity were subsequences of larger 6-mers and 7-mers. To further group the sequences, k-mers in each tissue-biased regulatory category were clustered based on sequence similarity using a previously described k-mer clustering and alignment method (see the ‘Materials and methods’ section) [42]. This approach yielded 20 intronic splicing regulatory motif clusters, including 5 neuron-biased ISE clusters, 10 muscle-biased ISE clusters, 1 neuron-biased ISS cluster, and 4 muscle-biased ISS clusters (Fig. 5B and Supplemental Table S11). We assessed whether these tissue-biased PRA k-mer motif PWMs could be mapped to experimentally characterized C. elegans RBP motifs (Supplemental Fig. S4). Consistent with our earlier analysis, and with recovering bona fide tissue-biased splicing regulatory elements, five PRA k-mer motifs were matched with presently known C. elegans RBP splicing regulator motifs. Neuron-biased ISE motif 2 (DUUGUG-) resembled the UNC-75 consensus sequence, while motif 4 (–GCAUG-) closely matched the FOX1/ASD-1 consensus motif (Supplemental Fig. S4). Similarly, muscle-biased ISE motif 7 (-YGUGUG-) mapped to the consensus sequence recognized by SUP-12 and motif 10 (-UAAYW-) mapped to B0280.17 or K07H8.9 motifs. Both B0280.17 and K07H8.9 are putative uncharacterized homologs of the mammalian QKI splicing factor and display somewhat distinct sequence preferences to another worm QKI paralog, ASD-2 [73] (Supplemental Fig. S3).

Interestingly, the other 16 PRA k-mer motifs with significant tissue-biased splicing regulatory effects did not map to any motifs within the C. elegans motif catalog (Supplemental Table S11). This latter group of cis elements represent uncharacterized motifs that warrant further study for their role in splicing regulation. Collectively, these results indicate that our PRA data can be used as a discovery tool to extract meaningful cis regulatory signals with tissue-biased activity within the context of our minigene reporter

PRA-derived ISE and ISS motifs can be mapped to endogenous conserved intronic regions flanking tissue-biased switch-like exons

Given that our PRA-derived ISE and ISS k-mer motifs were recovered from engineered splicing reporters in a single minigene context, it is difficult to determine whether these cis-elements have a more general role in promoting tissue-biased splicing patterns. Thus, we wanted to determine whether these motifs are also functional in native sequence contexts. Specifically, we wanted to assess if the identified tissue-specific ISE and ISS sequences are conserved in intronic regions flanking tissue-biased alternative exons of Caenorhabditis genes. However, noncoding regions such as introns are highly diverged and can be difficult to align with traditional alignment software. For instance, several studies have demonstrated that Caenorhabditis nematodes undergo extensive turnover of introns [74–76]. However, prior work has demonstrated that intronic regions flanking alternative exons tend to display elevated levels of sequence conservation compared to regions flanking constitutive exons, likely reflecting selective pressures acting to retain splicing regulatory sequences [31, 77–79].

To measure conservation of specific ISE and ISS motifs, we used phylogenetic average motif (PAM) score representations for intronic fragments [50]. This approach averages the best match scores for a provided motif (represented as a PWM) across sets of unaligned orthologous noncoding regions. The scores are further normalized by scrambling the motifs 100 times and calculating a z-score between the real match scores and the distribution of scrambled match scores serving as a background [50] (see the ‘Materials and methods’ section). These z-scores are hereafter referred as PAM scores. An advantage of this approach is that it leverages information about conservation of motifs while also accommodating some flexibility in their positioning within a noncoding region.

Mining existing transcriptome data from C. elegans tissues, we have identified a set of 47 alternative exons in genes that display switch-like splicing patterns, with a strong bias toward inclusion in either neurons or muscle cells (Supplemental Table S3; B.K. and J.A.C., unpublished). The PAM scores for our 20 PRA-identified ISE and ISS k-mer motifs were calculated for 5′ and 3′ orthologous intronic regions flanking these switch-like exons from up to 50 Caenorhabditis species [47]. We then performed hierarchical clustering on PRA ISE/ISS motifs and intronic fragments based on PAM scores results to identify potential co-regulated exons (Fig. 6A).

Among the 20 PRA-identified tissue-biased ISE and ISS k-mer motifs, 19 were found to be conserved in intronic regions adjacent to switch-like exons and the orthologous sequences of these intronic regions in other Caenorhabditis species with a PAM score > 2 (Fig. 6A). As expected, the neuron-biased ISE motif 4 (–GCAUG–), resembling the FOX-1/ASD-1 motif (Fig. 6A), was found highly conserved (PAM score Inline graphic 2) in 20 intronic regions from 18 splicing events. Similarly, neuron-biased ISE motif 2 (-UUGUG) that resembles the core UNC-75 motif (Supplemental Fig. S3), and muscle-biased ISE motif 7 (YGUGUG) that closely matches the SUP-12 motif, were found conserved in 9 and 8 intronic regions, respectively (Fig. 6A).

Interestingly, in addition to introns containing single highly conserved motifs, a subset (20%) involved two or more conserved k-mer motifs, suggesting combinatorial control (Fig. 6A). Focusing on one example involving robust neuronal tissue-specific exon inclusion, we present the PAM map for the intronic regions flanking alternative exon 4 in the dpy-23 gene. The dpy-23 gene is known to enable clathrin adaptor activity, playing a crucial role in trafficking and endocytosis [80, 81]. Interestingly, the fourth exon in the dpy-23 gene is a microexon (18 nucleotides) that undergoes pronounced neuron-biased inclusion [31] (neuronal PSI = 82% and muscle PSI 10%; Supplemental Table S3 and Fig. 6B and C). Notably, the PAM signatures for the intronic regions flanking the microexon revealed that two motifs with neuron-biased ISE activity (motif 2 and motif 3) were found to be conserved in the downstream intron towards the 5′ss end (Fig. 6B and C). The presence of these motifs with neuron-biased ISE activity at relevant intronic locations suggests that they may serve a key role in promoting neuron-biased inclusion of alternative exon 4 in this gene.

Taken together, our results demonstrate that ISE and ISS motifs identified by our PRA analysis are conserved across nematodes in intronic regions flanking tissue-biased alternative exons, consistent with a role in alternative splicing regulation of endogenous transcripts. Moreover, our data suggest that the splicing of these switch-like exons can be regulated through diverse combinations of cis elements. Finally, by combining PAM score representations, PRA activity measurements, and empirically determined splicing outcomes, testable predictions can be made about how the presence of specific cis elements can lead to tissue-biased splicing patterns.

Mutation of PRA-identified ISEs disrupts the splicing of an endogenous neuronal microexon

To further demonstrate the activity of the PRA-identified ISE and ISS motifs, we performed targeted mutagenesis experiments in the context of a native tissue-biased alternative exon. Since the PAM score profile of introns flanking the dpy-23 microexon revealed two motifs with neuron-biased ISE activity within the downstream intron (Fig. 6C), it led us to hypothesize that these cis elements may play a key role in promoting inclusion of this microexon in neurons.

To test this hypothesis, we introduced site-specific mutations within the mapped motifs, either alone or in combination, in a dpy-23 minigene splicing reporter that recapitulates tissue-biased microexon inclusion (Fig. 7A and B). Notably, all neuron-expressed dpy-23 minigene reporter plasmids with mutated PRA splicing motifs tested exhibited reduced microexon inclusion (Fig. 7B). The reporter plasmid with mutated motif 3 exhibited a moderate decrease in microexon inclusion with the average PSI value dropping from 74.9 ± 2.2% for the wild-type neuronal dpy-23 minigene reporter to 65.1 ± 2.7%. In contrast, mutation of motif 2 resulted in a substantially stronger effect (PSI = 15.9 ± 3.7%, mean dPSI compared with wild-type = −59%; Fig. 7B), highlighting a critical role of this motif in dpy-23 microexon splicing regulation. The reporter with both motif 2 and motif 3 mutated exhibited a further reduction in microexon inclusion (PSI = 10.1 ± 6%), resembling the muscle-tissue specific splicing pattern of the microexon (PSI = 10.1%; Fig. 7C). However, effects were comparable to those observed with mutation of motif 2, suggesting a dominant role of this motif in dpy-23 microexon splicing regulation (Fig. 7B).

Since motif 2 strongly resembles the UNC-75 motif (Supplemental Fig. S3), we further aimed to validate the role of UNC-75 in dpy-23 microexon splicing regulation in neuronal tissue. We introduced the wild-type, neuron-expressed, dpy-23 minigene reporter into unc-75 knockout animals (see the ‘Materials and methods’ section). Intriguingly, we observed a striking drop in PSI values similar to that obtained with the dpy-23 minigene reporter carrying mutated motif 2 compared to the reporter carrying mutated motif 3 (wild-type N2 versus unc-75 knockout dPSI = 71%; Fig. 7C). Collectively, these results indicate that UNC-75 is likely the factor engaging with this PRA-identified cis element. More broadly, these data also highlight the ability to generate testable hypotheses from our PAM scoring analysis of PRA-derived motifs.

Discussion

A screening tool to identify splicing regulatory sequences with tissue-biased activity in vivo

In this study we have successfully implemented an in vivo PRA to investigate the splicing regulatory effects of thousands of synthetic intronic sequences in a quantifiable manner in vivo. We further demonstrate that by adapting this methodology for a multicellular animal, cis elements with tissue-biased activity can be extracted from the data. Our approach thus enables the interrogation of splicing regulatory signals in their physiologically relevant cellular environment, rather than in cell lines, complementing ongoing PRA-based efforts to study cis regulatory logic of gene expression in live organisms [82–84].

We anticipate that our in vivo PRA approach can be expanded as a screening tool to identify splicing regulatory sequences with diverse spatio-temporal activity in the context of animal development. Moreover, given the wealth of environmental stimuli and signals known to impact gene expression in C. elegans [85–87], our PRA will also prove useful to identify intronic sequences that show signal-dependent splicing activity in the context of these paradigms.

We do note some limitations to this study. The splicing regulatory effects of the PRA elements were tested only within the context of a single minigene reporter (derived from the unc-16 locus) and at a fixed position within the downstream intron, which may not fully capture their behavior in other genomic contexts or intron locations. To partially mitigate this limitation, we mapped the identified tissue-biased PRA elements to their native genes, which display strong tissue-differential splicing, using comparative genomics approach. However, further work is needed to evaluate these elements in varied splicing events, gene contexts, and intronic locations to fully understand their regulatory impact. In addition, the number of randomized minigene reporters in each neuronal library pool was lower compared to the muscle library pools. This was partly due to a higher prevalence of reference reporters (Supplementary Fig. S1). The neuronal PRA also showed reduced reporter RNA recovery compared to the muscle PRA which may stem from the higher ratio of reference reporters to randomized reporters in the neuronal library. It has also been reported that the robust epigenetic silencing mechanisms within neuronal cells can suppress expression from foreign and repetitive sequences [1]. This could make the extrachromosomal arrays carrying the minigene reporters particularly susceptible to this repression. In the future, it will be interesting to sequence the reporters found in arrays recovered from the transgenic animals to get a real estimate of the number of minigene reporters successfully introduced and transcribed.

The ISE and silencer elements we have identified provide a subset of the parts list contributing to how context-dependent splicing is achieved. Our work thus complements ongoing efforts to develop interpretable models for the rules governing splicing outcomes across organisms, also known as the splicing code [88–90]. Among the sequences we found with regulatory activity are those recognized by RNA binding proteins with known tissue-biased impacts on alternative splicing [34, 64, 91]. However, we also discovered sequences with strong effects on splicing without clear cognate RBP partners (‘unmapped elements’). One possibility is that these unmapped elements might be regulating splicing by influencing RNA secondary structures that further affect spliceosome assembly on adjacent splice junctions [13, 92]. In addition, it is possible that the splicing effects of these elements, observed in vivo are further modulated by RNA modification mechanisms. For example, A-to-I editing, which is abundant in the C. elegans nervous system [93] could create, enhance, or disrupt binding sites of the cognate RBPs, thereby modifying splicing outcomes in different tissues. Hence, future work will be directed to better characterize these unmapped cis elements through biochemical (e.g mRNP mapping or RNA structure mapping) or reverse genetic screening approaches, identifying novel regulators of splicing in native tissue contexts. Moreover, future libraries could be made that introduce longer randomized sequences, which would not only increase our ability to detect linear seqeunce elements, but could favour the presence of additional secondary structures to be screened for splicing regulatory activity.

In addition to identifying intronic enhancers and silencers of alternative splicing, our screen identified candidate 5′ss that effectively competed with the naturally occurring 5′ss flanking the reporter alternative exon (Fig. 2). Consistent with early studies [94] we identify that a GUAAG pentamer, which is perfectly complementary to nucleotides in the U1 snRNA, is often sufficient to induce high levels of splice site utilization at the expense of the nearby native alternative splice site. However, a somewhat more surprising result was that even noncanonical splice sites, including those with GC at positions +1 and +2 (GC-splice donors), led to intermediate levels of competitive splicing when positioned downstream of the native alternative splice donor. Consistent with this result, a small proportion of introns in C. elegans and other metazoa contain GC-splice donors associated with both constitutive and alternative spliced exons [95]. Additionally, our results support growing evidence that effective splice donor sequences can exhibit more flexibility in sequence composition than previously appreciated, creating additional opportunities for context-dependent splicing regulation when combined with other enhancer and silencer motifs [96–99]. Finally, our results also predict that identified competitive splice-site donor sequences should be selected against and depleted from introns, particularly those flanking alternative exons with weakened 5′ss.

With the rise of single cell RNA-seq (scRNA-seq) approaches in C. elegans [100–104], future iterations of our in vivo PRA can also be modified to survey the effects of randomized sequences on splicing outcomes in single cells. Indeed, such methods coupling in vivo PRAs with scRNA-seq are being developed for other layers of gene expression [105, 106]. We note that at present, especially for the nervous system, which contains diverse neuronal subtypes, our broad tissue in vivo PRA is likely enriching for signals that act broadly in neurons. However, analogous to vertebrates, recent studies have shown alternative splicing differences between individual neuronal subtypes [107–109]. We thus anticipate that sampling the activity of randomized sequences at single cell resolution will be of relevance to the nervous system.

Extensive combinatorial control of alternative splicing supported by evolutionary signatures

Mapping our in vivo PRA-derived cis elements onto native intronic sequences that flank switch-like exons revealed evolutionary-conserved signals for these motifs (Fig. 6). Our work builds upon previous studies demonstrating that comparative genomics can be useful to identify splicing regulatory sequences in intronic regions [110–113]. Perhaps most striking, although clustering identified several shared intronic motif signatures, we often observed that diverse and unique combinations of cis elements are enriched within individual introns.

These results have several broad implications. First, the data indicate that there are many paths to achieve tissue-biased splicing outcomes, as indicated by the growing repertoire of nuclear RBPs in multicellular animals, including C. elegans, that can serve this role [31, 91, 114, 115]. Such flexibility would accommodate the gain and loss of individual sequences through evolution, supporting compensatory mechanisms to preserve regulatory function (similar to ‘developmental systems drift’) [116], or contributing to novel splicing patterns [117, 118]. Second, the utilization of several distinct cis elements would also enable multiple RBPs to work additively, antagonistically, or synergistically, to create robust splicing outcomes during development and between tissues, which we and others have observed in vivo [12, 34, 91, 119–121]. Finally, these results are reminiscent of ‘phenotypic convergence’, where within a given promoter or intronic region, distinct cis elements are recognized by their corresponding trans-acting factors in different cell types, to achieve the same regulatory outcome [122, 123].

In addition, we identified tissue-biased PRA motifs that bind RBPs homologous to mammalian RBPs, but whose motif sequences show subtle differences from their mammalian counterparts or instance, B0280.7 and K07H8.9 represent two out of six C. elegans RBPs which are homologous to the mammalian splicing factor QKI, yet the motifs these two RBPs recognize show subtle variations [73]. These divergent motifs are consistent with a model of duplication followed by neofunctionalization, in which one paralog retains the ancestral QKI motif, while others are released from evolutionary constraint and acquire modified binding preferences [73]. Notably, the muscle-biased PRA motif 10 showed a strong match to the binding motifs of B0280.7 and K07H8.9. The role of QKI in muscle-specific alternative splicing regulation has been well established, with QKI acting as a key regulator of splicing programs that generate contractile and structural protein isoforms essential for cardiomyocyte function [124]. It will therefore be interesting to test these C. elegans QKI homologs for a role in promoting muscle-biased splicing. Our study thus provides a framework to integrate PRA screens with comparative genomics to make informative and testable mechanistic predictions.

Two intronic enhancers regulate the inclusion of a neuronal microexon in C. elegans

Guided by our PRA and PAM score analysis, we validated the impact of two neuron-biased ISEs on the splicing of exon 4, an 18 nucleotide microexon, from the dpy-23 gene (Fig. 7). Microexons (<27 nucleotides) have emerged as an interesting class of exons, with an increased frequency in vertebrate genes compared to invertebrates. Aberrant regulation of these small exons has been associated with autism spectrum disorders and pancreatic cancer [125, 126]. The increased occurrence of microexons in vertebrates is thought to be associated with the emergence of the SR-related splicing activator proteins SRRM3 and SRRM4 [126, 127]. However, perturbation of SRRM4 does not eliminate inclusion of all microexons, and other factors have also been found to work together with this protein [128]. Caenorhabditis elegans lacks an SRRM4 ortholog, but its genome does contain microexons which can be spliced in a tissue-biased manner [129, 130]. Interestingly, the ISE that we validated to have a strong impact on dpy-23 microexon splicing matches the core UUGUG consensus sequence recognized by the CELF ortholog UNC-75, and splicing of the microexon is severely [34, 68]. This result is consistent with previous studies suggesting an evolutionary conserved role for CELF proteins in regulating microexon splicing in planaria and in vertebrates [131]. In C. elegans, we add UNC-75/CELF to an emerging list of conserved intron-binding RBPs that likely work combinatorially with PRP-40 (a U1 snRNP complex-interacting protein) to regulate microexon inclusion [129].

The second neuron-biased ISE we identified had a subtle but significant effect on microexon splicing (Fig. 6). At present, it is unclear what factor may be recognizing the core AAGUU sequence in this enhancer, although a likely candidate could be the Hu ortholog EXC-7, which recognizes a similar motif [14]. Consistent with this possibility, EXC-7 has been found to regulate a subset of microexons in C. elegans [129] and is also known to co-regulate alternative splicing with UNC-75 [14]. The proximity of this enhancer to the 5′ss flanking the microexon may create an environment that favours tissue-biased inclusion in neurons, possibly through stimulating or stabilizing U1 snRNP recruitment, analogous to the TIA1/TIAL proteins [132]. Further exploration of this enhancer will shed additional mechanistic insight into its role in tissue-biased splicing.

Importantly, we attempted to validate additional, conserved muscle-biased enhancer motifs at another locus (kvs-5; Supplemental Fig. S5). However, when designing minigene splicing reporters to test these cis-elements for splicing activity, our reporters were unable to recapitulate splicing patterns of the endogenous locus (Supplemental Fig. S5). These results highlight a limitation of our current experimental system, and more broadly reflect that additional, more distantly-located sequence features, can contribute to tissue-biased splicing outcomes. In future studies, candidate splicing regulatory elements of interest can be further tested within endogenous contexts by using techniques such as CRISPR/Cas9-mediated genome editing [133, 134]. Studying cis-element mutations in a native context could also uncover the longer-distance impacts of these mutations, including their effect on more distally positioned exons, connection with co-transcriptional splicing of the corresponding gene locus, or a role in modulating RNA secondary structures forming to promote or repress exon inclusion [92, 135, 136]. Collectively, these results highlight the broader utility of intersecting our PRA-derived regulatory sequences with endogenous alternatively spliced loci to inform novel regulatory landscapes. We anticipate that continued use of these screening approaches will prove useful in our efforts to understand the code governing context-specific splicing in multicellular animals.

Supplementary Material

gkag451_Supplemental_Files

gkag451_supplemental_files.zip^{(4.4MB, zip)}

Acknowledgements

We thank members of the Calarco and Moses labs for critical discussions throughout the course of this study. We also acknowledge Santiago Vargas and Alissa Hofmann for early work on this project.

Author contributions: Sanjana Bhatnagar (Conceptualization [equal], Data curation [lead], Formal analysis [lead], Investigation [lead], Methodology [lead], Software [lead], Validation [lead], Writing—original draft [lead], Writing—review & editing [equal]), Jade Ho (Investigation [supporting]), Nour H Sadek (Formal analysis [supporting], Investigation [supporting], Validation [supporting]), Isha Singh (Investigation [supporting], Validation [supporting]), Michael Zoberman (Data curation [supporting], Formal analysis [supporting], Investigation [supporting]), Bina Koterniak (Data curation [supporting], Formal analysis [supporting], Investigation [supporting]), Yufang Liu (Investigation [supporting]), Daniel D Fusca (Data curation [supporting], Formal analysis [supporting], Investigation [supporting]), Asher D Cutter (Supervision [supporting], Writing—review & editing [supporting]), Alan Moses (Conceptualization [supporting], Funding acquisition [supporting], Methodology [supporting], Resources [supporting], Software [supporting], Writing—review & editing [supporting]), John Calarco (Conceptualization [lead], Funding acquisition [lead], Methodology [supporting], Project administration [equal], Resources [supporting], Supervision [lead], Visualization [supporting], Writing—original draft [supporting], Writing—review & editing [supporting]).

Notes

Present address: Mayo Clinic Alix School of Medicine, Mayo Clinic College of Medicine and Science 200 1st SW, Rochester MN, 55905, USA

Present address: Genetics and Genome Biology Program, The Hospital for Sick Children, 170 Elizabeth St, Toronto, Ontario, M5G 1E8, Canada

Contributor Information

Sanjana Bhatnagar, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Jade Ho, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Nour H Sadek, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Isha Singh, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Michael Zoberman, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Bina Koterniak, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Yufang Liu, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Daniel D Fusca, Department of Ecology and Evolutionary Biology, 25 Willcocks Street, Toronto, Ontario, M5S 3B2, Canada.

Asher D Cutter, Department of Ecology and Evolutionary Biology, 25 Willcocks Street, Toronto, Ontario, M5S 3B2, Canada.

Alan M Moses, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

John A Calarco, Department of Cell and Systems Biology, 25 Harbord Street, Toronto, Ontario, M5S 3G5, Canada.

Supplementary data

Supplementary data is available at NAR online.

Conflict of interest

None declared.

Funding

This work was supported by grants from the Canadian Institutes of Health Research [grant numbers 156300 and 180365] and the Natural Sciences and Engineering Council of Canada [grant RGPIN-2017-06573] to J.A.C. Canadian Institutes of Health Research [180365].

Data availability

The raw sequence data (fastq files) are available at the NCBI SRA:

https://www.ncbi.nlm.nih.gov/sra/PRJNA1359015

The code and additional data are available through Zenodo:

https://doi.org/10.5281/zenodo.17593183

Inferring Exon and Intron Metadata from .gff file (v1.0.1).

https://doi.org/10.5281/zenodo.15757714

References

1. Pan Q, Shai O, Lee LJ et al. Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008;40:1413–5. 10.1038/ng.259 [DOI] [PubMed] [Google Scholar]
2. Wang Z, Burge CB. Splicing regulation: from a parts list of regulatory elements to an integrated splicing code. RNA. 2008;14:802–13. 10.1261/rna.876308 [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Zheng S, Black DL. Alternative pre-mRNA splicing in neurons: growing up and extending its reach. Trends Genet. 2013;29:442–8. 10.1016/j.tig.2013.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Raj B, Blencowe BJ. Alternative splicing in the mammalian nervous system: recent insights into mechanisms and functional roles. Neuron. 2015;87:14–27. 10.1016/j.neuron.2015.05.004 [DOI] [PubMed] [Google Scholar]
5. Warf MB, Berglund JA. Role of RNA structure in regulating pre-mRNA splicing. Trends Biochem Sci. 2010;35:169–78. 10.1016/j.tibs.2009.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Baralle M, Baralle FE. The splicing code. Biosystems. 2018;164:39–48. 10.1016/j.biosystems.2017.11.002 [DOI] [PubMed] [Google Scholar]
7. Blencowe BJ. Exonic splicing enhancers: mechanism of action, diversity and role in human genetic diseases. Trends Biochem Sci. 2000;25:106–10. 10.1016/S0968-0004(00)01549-8 [DOI] [PubMed] [Google Scholar]
8. Black DL. Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003;72:291–336. 10.1146/annurev.biochem.72.121801.161720 [DOI] [PubMed] [Google Scholar]
9. Wang Y, Ma M, Xiao X et al. Intronic splicing enhancers, cognate splicing factors and context dependent regulation rules. Nat Struct Mol Biol. 2012;19:1044–52. 10.1038/nsmb.2377 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Fairbrother WG, Yeh R-F, Sharp PA et al. Predictive identification of exonic splicing enhancers in human genes. Science. 2002;297:1007–13. 10.1126/science.1073774 [DOI] [PubMed] [Google Scholar]
11. Gao Y, Lin K-T, Jiang T et al. Systematic characterization of short intronic splicing-regulatory elements in SMN2 pre-mRNA. Nucleic Acids Res. 2022;50:731–49. 10.1093/nar/gkab1280 [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Ule J, Blencowe BJ. Alternative splicing regulatory networks: functions, mechanisms, and evolution. Mol Cell. 2019;76:329–45. 10.1016/j.molcel.2019.09.017 [DOI] [PubMed] [Google Scholar]
13. Bartys N, Kierzek R, Lisowiec-Wachnicka J. The regulation properties of RNA secondary structure in alternative splicing. Biochim Biophys Acta Gene Regul Mech. 2019;1862:194401. 10.1016/j.bbagrm.2019.07.002 [DOI] [PubMed] [Google Scholar]
14. Mikl M, Hamburg A, Pilpel Y et al. Dissecting splicing decisions and cell-to-cell variability with designed sequence libraries. Nat Commun. 2019;10:4572. 10.1038/s41467-019-12642-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Rodrigues F, Thuma L, Klämbt C. The regulation of glial-specific splicing of Neurexin IV requires HOW and Cdk12 activity. Dev Camb Engl. 2012;139:1765–76. [DOI] [PubMed] [Google Scholar]
16. Wang Z, Rolish ME, Yeo G et al. Systematic identification and analysis of exonic splicing silencers. Cell. 2004;119:831–45. 10.1016/j.cell.2004.11.010 [DOI] [PubMed] [Google Scholar]
17. Ke S, Shang S, Kalachikov SM et al. Quantitative evaluation of all hexamers as exonic splicing elements. Genome Res. 2011;21:1360–74. 10.1101/gr.119628.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Rosenberg AB, Patwardhan RP, Shendure J et al. Learning the sequence determinants of alternative splicing from millions of random sequences. Cell. 2015;163:698–711. 10.1016/j.cell.2015.09.054 [DOI] [PubMed] [Google Scholar]
19. Baeza-Centurion P, Miñana B, Schmiedel JM et al. Combinatorial genetics reveals a scaling law for the effects of mutations on splicing. Cell. 2019;176:549–63.e23. 10.1016/j.cell.2018.12.010 [DOI] [PubMed] [Google Scholar]
20. Rhine CL, Neil C, Wang J et al. Massively parallel reporter assays discover de novo exonic splicing mutants in paralogs of Autism genes. PLoS Genet. 2022;18:e1009884. 10.1371/journal.pgen.1009884 [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Rong S, Neil CR, Welch A et al. Large-scale functional screen identifies genetic variants with splicing effects in modern and archaic humans. Proc Natl Acad Sci USA. 2023;120:e2218308120. 10.1073/pnas.2218308120 [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Rabani M, Pieper L, Chew G-L et al. A massively parallel reporter assay of 3’ UTR sequences identifies In vivo rules for mRNA degradation. Mol Cell. 2017;68:1083–94. 10.1016/j.molcel.2017.11.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Cooper TA. Use of minigene systems to dissect alternative splicing elements. Methods San Diego Calif. 2005;37:331–40. 10.1016/j.ymeth.2005.07.015 [DOI] [PubMed] [Google Scholar]
24. Wang Z, Burge CB. Splicing regulation: from a parts list of regulatory elements to an integrated splicing code. RNA. 2008;14:802–13. 10.1261/rna.876308 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Gaildrat P, Killian A, Martins A et al. Use of splicing reporter minigene assay to evaluate the effect on splicing of unclassified genetic variants. Methods Mol Biol Clifton NJ. 2010;653:249–57. [DOI] [PubMed] [Google Scholar]
26. Smith SA, Lynch KW. Cell-based splicing of minigenes. Methods Mol Biol. 2014;1126:243–55. [DOI] [PubMed] [Google Scholar]
27. Kreimer A, Ashuach T, Inoue F et al. Massively parallel reporter perturbation assays uncover temporal regulatory architecture during neural differentiation. Nat Commun. 2022;13:1504. 10.1038/s41467-022-28659-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
28. McAfee JC, Bell JL, Krupa O et al. Focus on your locus with a massively parallel reporter assay. J Neurodev Disord. 2022;14:50. 10.1186/s11689-022-09461-x [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Agarwal V, Inoue F, Schubach M et al. Massively parallel characterization of transcriptional regulatory elements. Nature. 2025;639:411–20. 10.1038/s41586-024-08430-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Gracida X, Norris AD, Calarco JA. Regulation of tissue-specific alternative splicing: C. elegans as a model system. Adv Exp Med Biol. 2016;907:229–61. [DOI] [PubMed] [Google Scholar]
31. Koterniak B, Pilaka PP, Gracida X et al. Global regulatory features of alternative splicing across tissues and within the nervous system of C. elegans. Genome Res. 2020;30:1766–80. 10.1101/gr.267328.120 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Mello CC, Kramer JM, Stinchcomb D et al. Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 1991;10:3959–70. 10.1002/j.1460-2075.1991.tb04966.x [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94. 10.1093/genetics/77.1.71 [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Norris AD, Gao S, Norris ML et al. A pair of RNA-binding proteins controls networks of splicing events contributing to specialization of neural cell types. Mol Cell. 2014;54:946–59. 10.1016/j.molcel.2014.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Schafer S, Miao K, Benson CC et al. Alternative splicing signatures in RNA-seq data: percent spliced in (PSI). Curr Protoc Hum Genet. 2015;87:11.16.1–11.16.14. [DOI] [PubMed] [Google Scholar]
36. Schindelin J, Arganda-Carreras I, Frise E et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9:676–82. 10.1038/nmeth.2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Howe KL, Bolt BJ, Cain S et al. WormBase 2016: expanding to enable helminth genomic research. Nucleic Acids Res. 2016;44:D774–80. 10.1093/nar/gkv1217 [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Dainat J, Hereñú D, Pucholt P. AGAT: another GFF analysis toolkit to handle annotations in any GTF/GFF format. (Version v0.7.0). Zenodo. 10.5281/zenodo.3552717 [DOI]
39. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–2. 10.1093/bioinformatics/btq033 [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Grant CE, Bailey TL, Noble WS. FIMO: scanning for occurrences of a given motif. Bioinformatics. 2011;27:1017–8. 10.1093/bioinformatics/btr064 [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Yeo G, Burge CB. Maximum entropy modeling of short sequence motifs with applications to RNA splicing signals. J Comput Biol J Comput Mol Cell Biol. 2004;11:377–94. 10.1089/1066527041410418 [DOI] [PubMed] [Google Scholar]
42. Kuret K, Amalietti AG, Jones DM et al. Positional motif analysis reveals the extent of specificity of protein–RNA interactions observed by CLIP. Genome Biol. 2022;23:191. 10.1186/s13059-022-02755-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Tremblay BJ-M. universalmotif: an R package for biological motif analysis. J Open Source Softw. 2024;9:7012. 10.21105/joss.07012 [DOI] [Google Scholar]
44. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22:4673–80. 10.1093/nar/22.22.4673 [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Ray D, Kazan H, Chan ET et al. Rapid and systematic analysis of the RNA recognition specificities of RNA-binding proteins. Nat Biotechnol. 2009;27:667–70. 10.1038/nbt.1550 [DOI] [PubMed] [Google Scholar]
46. Gupta S, Stamatoyannopoulos JA, Bailey TL et al. Quantifying similarity between motifs. Genome Biol. 2007;8:R24. 10.1186/gb-2007-8-2-r24 [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Fusca DD, Kasimatis KR, Zhu HV et al. Dynamic birth and death of argonaute gene Family functional repertoire across Caenorhabditis nematodes. Genome Biol Evol. 2025;17:evaf106. 10.1093/gbe/evaf106 [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Emms DM, Kelly S. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 2019;20:238. 10.1186/s13059-019-1832-y [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Katoh K, Rozewicki J, Yamada KD. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief Bioinform. 2019;20:1160–6. 10.1093/bib/bbx108 [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Alam A, Duncan AG, Mitchell JA et al. Functional similarity of non-coding regions is revealed in phylogenetic average motif score representations. bioRxiv, 10.1101/2023.04.09.536185, 11 April 2023, preprint: not peer reviewed. [DOI]
51. de Hoon MJL, Imoto S, Nolan J et al. Open source clustering software. Bioinformatics. 2004;20:1453–4. [DOI] [PubMed] [Google Scholar]
52. Saldanha AJ. Java treeview–extensible visualization of microarray data. Bioinforma Oxf Engl. 2004;20:3246–8. 10.1093/bioinformatics/bth349 [DOI] [PubMed] [Google Scholar]
53. Kivioja T, Vähärautio A, Karlsson K et al. Counting absolute numbers of molecules using unique molecular identifiers. Nat Methods. 2012;9:72–4. 10.1038/nmeth.1778 [DOI] [PubMed] [Google Scholar]
54. Sibley CR, Blazquez L, Ule J. Lessons from non-canonical splicing. Nat Rev Genet. 2016;17:407–21. 10.1038/nrg.2016.46 [DOI] [PMC free article] [PubMed] [Google Scholar]
55. Aldalaqan S, Dalgliesh C, Luzzi S et al. Cryptic splicing: common pathological mechanisms involved in male infertility and neuronal diseases. Cell Cycle Georget Tex. 2022;21:219–27. 10.1080/15384101.2021.2015672 [DOI] [PMC free article] [PubMed] [Google Scholar]
56. Rinke J, Appel B, Blöcker H et al. The 5’-terminal sequence of U1 RNA complementary to the consensus 5’ splice site of hnRNA is single-stranded in intact U1 snRNP particles. Nucleic Acids Res. 1984;12:4111–26. 10.1093/nar/12.10.4111 [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Roca X, Krainer AR, Eperon IC. Pick one, but be quick: 5′ splice sites and the problems of too many choices. Genes Dev. 2013;27:129–44. 10.1101/gad.209759.112 [DOI] [PMC free article] [PubMed] [Google Scholar]
58. Perchlik M, Sasse A, Mostafavi S et al. Impact on splicing in Saccharomyces cerevisiae of random 50-base sequences inserted into an intron. RNA N Y N. 2023;30:52–67. 10.1261/rna.079752.123 [DOI] [PMC free article] [PubMed] [Google Scholar]
59. Buratti E, Baralle M, Baralle FE. Defective splicing, disease and therapy: searching for master checkpoints in exon definition. Nucleic Acids Res. 2006;34:3494–510. 10.1093/nar/gkl498 [DOI] [PMC free article] [PubMed] [Google Scholar]
60. Liu H-X, Zhang M, Krainer AR. Identification of functional exonic splicing enhancer motifs recognized by individual SR proteins. Genes Dev. 1998;12:1998–2012. 10.1101/gad.12.13.1998 [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Chen M, Manley JL. Mechanisms of alternative splicing regulation: insights from molecular and genomics approaches. Nat Rev Mol Cell Biol. 2009;10:741–54. 10.1038/nrm2777 [DOI] [PMC free article] [PubMed] [Google Scholar]
62. Lee Y, Rio DC. Mechanisms and regulation of alternative pre-mRNA splicing. Annu Rev Biochem. 2015;84:291–323. 10.1146/annurev-biochem-060614-034316 [DOI] [PMC free article] [PubMed] [Google Scholar]
63. Ray D, Kazan H, Cook KB et al. A compendium of RNA-binding motifs for decoding gene regulation. Nature. 2013;499:172–7. 10.1038/nature12311 [DOI] [PMC free article] [PubMed] [Google Scholar]
64. Jin Y, Suzuki H, Maegawa S et al. A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG. EMBO J. 2003;22:905–12. 10.1093/emboj/cdg089 [DOI] [PMC free article] [PubMed] [Google Scholar]
65. Sun S, Zhang Z, Fregoso O et al. Mechanisms of activation and repression by the alternative splicing factors RBFOX1/2. RNA. 2012;18:274–83. 10.1261/rna.030486.111 [DOI] [PMC free article] [PubMed] [Google Scholar]
66. Kuroyanagi H, Watanabe Y, Suzuki Y et al. Position-dependent and neuron-specific splicing regulation by the CELF family RNA-binding protein UNC-75 in Caenorhabditis elegans. Nucleic Acids Res. 2013;41:4015–25. 10.1093/nar/gkt097 [DOI] [PMC free article] [PubMed] [Google Scholar]
67. Chen L, Liu Z, Zhou B et al. CELF RNA binding proteins promote axon regeneration in C. elegans and mammals through alternative splicing of Syntaxins. eLife. 2016;5:e16072. 10.7554/eLife.16072 [DOI] [PMC free article] [PubMed] [Google Scholar]
68. Pilaka-Akella P, Sadek NH, Fusca D et al. Neuron-specific repression of alternative splicing by the conserved CELF protein UNC-75 in Caenorhabditis elegans. Genetics. 2025;229:iyaf025. 10.1093/genetics/iyaf025 [DOI] [PMC free article] [PubMed] [Google Scholar]
69. Anyanful A, Ono K, Johnsen RC et al. The RNA-binding protein SUP-12 controls muscle-specific splicing of the ADF/cofilin pre-mRNA in C. elegans. J Cell Biol. 2004;167:639–47. 10.1083/jcb.200407085 [DOI] [PMC free article] [PubMed] [Google Scholar]
70. Kuwasako K, Takahashi M, Unzai S et al. RBFOX and SUP-12 sandwich a G base to cooperatively regulate tissue-specific splicing. Nat Struct Mol Biol. 2014;21:778–86. 10.1038/nsmb.2870 [DOI] [PubMed] [Google Scholar]
71. Fu X-D, Ares M. Context-dependent control of alternative splicing by RNA-binding proteins. Nat Rev Genet. 2014;15:689–701. 10.1038/nrg3778 [DOI] [PMC free article] [PubMed] [Google Scholar]
72. Yang Y, Lee GC, Nakagaki-Silva E et al. Cell-type specific regulator RBPMS switches alternative splicing via higher-order oligomerization and heterotypic interactions with other splicing regulators. Nucleic Acids Res. 2023;51:9961–82. 10.1093/nar/gkad652 [DOI] [PMC free article] [PubMed] [Google Scholar]
73. Sasse A, Ray D, Laverty KU et al. A resource of RNA-binding protein motifs across eukaryotes reveals evolutionary dynamics and gene-regulatory function. Nat Biotechnol. 2025. 10.1038/s41587-025-02733-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
74. Roy SW, Penny D. Smoke without fire: most reported cases of intron gain in nematodes instead reflect intron losses. Mol Biol Evol. 2006;23:2259–62. 10.1093/molbev/msl098 [DOI] [PubMed] [Google Scholar]
75. Ma M-Y, Xia J, Shu K-X et al. Intron losses and gains in the nematodes. Biol Direct. 2022;17:13. 10.1186/s13062-022-00328-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
76. Teterina AA, Willis JH, Baer CF et al. Pervasive conservation of intron number and other genetic elements revealed by a chromosome-level genome assembly of the hyper-polymorphic nematode Caenorhabditis brenneri. Genome Biol Evol. 2025;17:evaf037. 10.1093/gbe/evaf037 [DOI] [PMC free article] [PubMed] [Google Scholar]
77. Sugnet CW, Kent WJ, Ares M et al. Transcriptome and genome conservation of alternative splicing events in humans and mice. Pac Symp Biocomput. 2004;66–77. 10.1142/9789812704856_0007 [DOI] [PubMed] [Google Scholar]
78. Fagnani M, Barash Y, Ip JY et al. Functional coordination of alternative splicing in the mammalian central nervous system. Genome Biol. 2007;8:R108. 10.1186/gb-2007-8-6-r108 [DOI] [PMC free article] [PubMed] [Google Scholar]
79. Baek D, Green P. Sequence conservation, relative isoform frequencies, and nonsense-mediated decay in evolutionarily conserved alternative splicing. Proc Natl Acad Sci USA. 2005;102:12813–8. 10.1073/pnas.0506139102 [DOI] [PMC free article] [PubMed] [Google Scholar]
80. Pan C-L, Baum PD, Gu M et al. C. elegans AP-2 and Retromer control wnt secretion by regulating MIG-14/Wntless. Dev Cell. 2008;14:132–9. 10.1016/j.devcel.2007.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
81. Kaplan OI, Doroquez DB, Cevik S et al. Endocytosis genes facilitate protein and membrane transport in C. elegans sensory cilia. Curr Biol. 2012;22:451–60. 10.1016/j.cub.2012.01.060 [DOI] [PMC free article] [PubMed] [Google Scholar]
82. Brown AR, Fox GA, Kaplow IM et al. An in vivo systemic massively parallel platform for deciphering animal tissue-specific regulatory function. Front Genet. 2025;16:1–19. 10.3389/fgene.2025.1533900 [DOI] [PMC free article] [PubMed] [Google Scholar]
83. Lambert JT, Su-Feher L, Cichewicz K et al. Parallel functional testing identifies enhancers active in early postnatal mouse brain. eLife. 2021;10:e69479. 10.7554/eLife.69479 [DOI] [PMC free article] [PubMed] [Google Scholar]
84. Zheng Y, VanDusen NJ. Massively parallel reporter assays for high-throughput in vivo analysis of cis-regulatory elements. J Cardiovasc Dev Dis. 2023;10:144. [DOI] [PMC free article] [PubMed] [Google Scholar]
85. Martinez BA, Reis Rodrigues P, Nuñez Medina RM et al. An alternatively spliced, non-signaling insulin receptor modulates insulin sensitivity via insulin peptide sequestration in C. elegans. eLife. 2020;9:e49917. 10.7554/eLife.49917 [DOI] [PMC free article] [PubMed] [Google Scholar]
86. Lee RY, Hench J, Ruvkun G. Regulation of C. elegans DAF-16 and its human ortholog FKHRL1 by the daf-2 insulin-like signaling pathway. Curr Biol. 2001;11:1950–7. 10.1016/S0960-9822(01)00595-4 [DOI] [PubMed] [Google Scholar]
87. Tamburino AM, Ryder SP, Walhout AJM. A compendium of Caenorhabditis elegans RNA binding proteins predicts extensive regulation at multiple levels. G3 (Bethesda). 2013;3:297–304. 10.1534/g3.112.004390 [DOI] [PMC free article] [PubMed] [Google Scholar]
88. Liao SE, Sudarshan M, Regev O. Deciphering RNA splicing logic with interpretable machine learning. Proc Natl Acad Sci USA. 2023;120:e2221165120. 10.1073/pnas.2221165120 [DOI] [PMC free article] [PubMed] [Google Scholar]
89. La Fleur A, Shi Y, Seelig G. Decoding biology with massively parallel reporter assays and machine learning. Genes Dev. 2024;38:843–65. 10.1101/gad.351800.124 [DOI] [PMC free article] [PubMed] [Google Scholar]
90. Gupta K, Yang C, McCue K et al. Improved modeling of RNA-binding protein motifs in an interpretable neural model of RNA splicing. Genome Biol. 2024;25:23. 10.1186/s13059-023-03162-x [DOI] [PMC free article] [PubMed] [Google Scholar]
91. Tan JH, Fraser AG. The combinatorial control of alternative splicing in C. elegans. PLoS Genet. 2017;13:e1007033. 10.1371/journal.pgen.1007033 [DOI] [PMC free article] [PubMed] [Google Scholar]
92. Li H, Ding Z, Fang Z-Y et al. Conserved intronic secondary structures with concealed branch sites regulate alternative splicing of poison exons. Nucleic Acids Res. 2024;52:6002–16. 10.1093/nar/gkae185 [DOI] [PMC free article] [PubMed] [Google Scholar]
93. Rajendren S, Dhakal A, Vadlamani P et al. Profiling neural editomes reveals a molecular mechanism to regulate RNA editing during development. Genome Res. 2021;31:27–39. 10.1101/gr.267575.120 [DOI] [PMC free article] [PubMed] [Google Scholar]
94. Eperon LP, Estibeiro JP, Eperon IC. The role of nucleotide sequences in splice site selection in eukaryotic pre-messenger RNA. Nature. 1986;324:280–2. 10.1038/324280a0 [DOI] [PubMed] [Google Scholar]
95. Farrer T, Roller AB, Kent WJ et al. Analysis of the role of Caenorhabditis elegans GC-AG introns in regulated splicing. Nucleic Acids Res. 2002;30:3360–7. 10.1093/nar/gkf465 [DOI] [PMC free article] [PubMed] [Google Scholar]
96. Wong MS, Kinney JB, Krainer AR. Quantitative activity profile and context dependence of all human 5’ splice sites. Mol Cell. 2018;71:1012–26. 10.1016/j.molcel.2018.07.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
97. Roca X, Akerman M, Gaus H et al. Widespread recognition of 5’ splice sites by noncanonical base-pairing to U1 snRNA involving bulged nucleotides. Genes Dev. 2012;26:1098–109. 10.1101/gad.190173.112 [DOI] [PMC free article] [PubMed] [Google Scholar]
98. Roca X, Sachidanandam R, Krainer AR. Intrinsic differences between authentic and cryptic 5’ splice sites. Nucleic Acids Res. 2003;31:6321–33. 10.1093/nar/gkg830 [DOI] [PMC free article] [PubMed] [Google Scholar]
99. Tan J, Ho JXJ, Zhong Z et al. Noncanonical registers and base pairs in human 5’ splice-site selection. Nucleic Acids Res. 2016;44:3908–21. 10.1093/nar/gkw163 [DOI] [PMC free article] [PubMed] [Google Scholar]
100. Taylor SR, Santpere G, Weinreb A et al. Molecular topography of an entire nervous system. Cell. 2021;184:4329–47. 10.1016/j.cell.2021.06.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
101. Barrett A, Varol E, Weinreb A et al. Integrating bulk and single cell RNA-seq refines transcriptomic profiles of individual C. elegans neurons. eLife. 2025;14:RP106183. 10.7554/eLife.106183.2 [DOI] [Google Scholar]
102. Cao J, Packer JS, Ramani V et al. Comprehensive single-cell transcriptional profiling of a multicellular organism. Science. 2017;357:661–7. 10.1126/science.aam8940 [DOI] [PMC free article] [PubMed] [Google Scholar]
103. Packer JS, Zhu Q, Huynh C et al. A lineage-resolved molecular atlas of C. elegans embryogenesis at single cell resolution. Science. 2019;365:eaax1971. 10.1126/science.aax1971 [DOI] [PMC free article] [PubMed] [Google Scholar]
104. Roux AE, Yuan H, Podshivalova K et al. Individual cell types in C. elegans age differently and activate distinct cell-protective responses. Cell Rep. 2023;42:112902. 10.1016/j.celrep.2023.112902 [DOI] [PubMed] [Google Scholar]
105. Zhao S, Hong CKY, Myers CA et al. A single-cell massively parallel reporter assay detects cell-type-specific gene regulation. Nat Genet. 2023;55:346–54. 10.1038/s41588-022-01278-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
106. Zhang H, Li X, Song D et al. Worm perturb-seq: massively parallel whole-animal RNAi and RNA-seq. Nat Commun. 2025;16:4785. 10.1038/s41467-025-60154-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
107. Feng H, Moakley DF, Chen S et al. Complexity and graded regulation of neuronal cell-type-specific alternative splicing revealed by single-cell RNA sequencing. Proc Natl Acad Sci USA. 2021;118:e2013056118. 10.1073/pnas.2013056118 [DOI] [PMC free article] [PubMed] [Google Scholar]
108. Weinreb A, Varol E, Barrett A et al. Alternative splicing across the C. elegans nervous system. Nat Commun. 2025;16:4508. 10.1038/s41467-025-58293-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
109. Thompson M, Bixby R, Dalton R et al. Splicing in a single neuron is coordinately controlled by RNA binding proteins and transcription factors. eLife. 2019;8:e46726. 10.7554/eLife.46726 [DOI] [PMC free article] [PubMed] [Google Scholar]
110. Bortfeldt R, Schindler S, Szafranski K et al. Comparative analysis of sequence features involved in the recognition of tandem splice sites. BMC Genomics. 2008;9:202. 10.1186/1471-2164-9-202 [DOI] [PMC free article] [PubMed] [Google Scholar]
111. Kabat JL, Barberan-Soler S, McKenna P et al. Intronic alternative splicing regulators identified by comparative genomics in nematodes. PLoS Comput Biol. 2006;2:e86. 10.1371/journal.pcbi.0020086 [DOI] [PMC free article] [PubMed] [Google Scholar]
112. Kol G, Lev-Maor G, Ast G. Human–mouse comparative analysis reveals that branch-site plasticity contributes to splicing regulation. Hum Mol Genet. 2005;14:1559–68. 10.1093/hmg/ddi164 [DOI] [PubMed] [Google Scholar]
113. Yeo GW, Nostrand ELV, Liang TY. Discovery and analysis of evolutionarily conserved intronic splicing regulatory elements. PLoS Genet. 2007;3:e85. 10.1371/journal.pgen.0030085 [DOI] [PMC free article] [PubMed] [Google Scholar]
114. Barberan-Soler S, Zahler AM. Alternative splicing regulation during C. elegans development: splicing factors as regulated targets. PLoS Genet. 2008;4:e1000001. 10.1371/journal.pgen.1000001 [DOI] [PMC free article] [PubMed] [Google Scholar]
115. Laver JD, Duncan M, Calarco JA. A global view of the RNA-binding and regulatory protein landscape in Caenorhabditis elegans. bioRxiv, 10.1101/2024.11.07.622524, 11 November 2024, preprint: not peer reviewed. [DOI]
116. True JR, Haag ES. Developmental system drift and flexibility in evolutionary trajectories. Evol Dev. 2001;3:109–19. 10.1046/j.1525-142x.2001.003002109.x [DOI] [PubMed] [Google Scholar]
117. Barbosa-Morais NL, Irimia M, Pan Q et al. The evolutionary landscape of alternative splicing in vertebrate species. Science. 2012;338:1587–93. 10.1126/science.1230612 [DOI] [PubMed] [Google Scholar]
118. Jelen N, Ule J, Zivin M et al. Evolution of Nova-dependent splicing regulation in the brain. PLoS Genet. 2007;3:1838–47. 10.1371/journal.pgen.0030173 [DOI] [PMC free article] [PubMed] [Google Scholar]
119. Dvinge H. Regulation of alternative mRNA splicing: old players and new perspectives. FEBS Lett. 2018;592:2987–3006. 10.1002/1873-3468.13119 [DOI] [PubMed] [Google Scholar]
120. Iadevaia V, Gerber AP. Combinatorial control of mRNA fates by RNA-binding proteins and non-coding RNAs. Biomolecules. 2015;5:2207–22. 10.3390/biom5042207 [DOI] [PMC free article] [PubMed] [Google Scholar]
121. Lin J-C, Tarn W-Y. Exon selection in alpha-tropomyosin mRNA is regulated by the antagonistic action of RBM4 and PTB. Mol Cell Biol. 2005;25:10111–21. 10.1128/MCB.25.22.10111-10121.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
122. Carrasco J, Rauer M, Hummel B et al. ELAV and FNE determine neuronal transcript signatures through EXon-activated rescue. Mol Cell. 2020;80:156–63. 10.1016/j.molcel.2020.09.011 [DOI] [PubMed] [Google Scholar]
123. Kuroyanagi H, Kobayashi T, Mitani S et al. Transgenic alternative-splicing reporters reveal tissue-specific expression profiles and regulation mechanisms in vivo. Nat Methods. 2006;3:909–15. 10.1038/nmeth944 [DOI] [PubMed] [Google Scholar]
124. Montañés-Agudo P, Aufiero S, Schepers EN et al. The RNA-binding protein QKI governs a muscle-specific alternative splicing program that shapes the contractile function of cardiomyocytes. Cardiovasc Res. 2023;119:1161–74. 10.1093/cvr/cvad007 [DOI] [PMC free article] [PubMed] [Google Scholar]
125. Gonatopoulos-Pournatzis T, Blencowe BJ. Microexons: at the nexus of nervous system development, behaviour and autism spectrum disorder. Curr Opin Genet Dev. 2020;65:22–33. 10.1016/j.gde.2020.03.007 [DOI] [PubMed] [Google Scholar]
126. Irimia M, Weatheritt RJ, Ellis JD et al. A highly conserved program of neuronal microexons is misregulated in autistic brains. Cell. 2014;159:1511–23. 10.1016/j.cell.2014.11.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
127. Head SA, Hernandez-Alias X, Yang J-S et al. Silencing of SRRM4 suppresses microexon inclusion and promotes tumor growth across cancers. PLoS Biol. 2021;19:e3001138. 10.1371/journal.pbio.3001138 [DOI] [PMC free article] [PubMed] [Google Scholar]
128. Gupta T, Margolin G, Burgess HA. Mutations in the microexon splicing regulator srrm4 have minor phenotypic effects on zebrafish neural development. G3 (Bethesda). 2025;15:jkaf052. 10.1093/g3journal/jkaf052 [DOI] [PMC free article] [PubMed] [Google Scholar]
129. Choudhary B, Marx O, Norris AD. Spliceosomal component PRP-40 is a central regulator of microexon splicing. Cell Rep. 2021;36:109464. 10.1016/j.celrep.2021.109464 [DOI] [PMC free article] [PubMed] [Google Scholar]
130. Torres-Méndez A, Bonnal S, Marquez Y et al. A novel protein domain in an ancestral splicing factor drove the evolution of neural microexons. Nat Ecol Evol. 2019;3:691–701. 10.1038/s41559-019-0813-6 [DOI] [PubMed] [Google Scholar]
131. Solana J, Irimia M, Ayoub S et al. Conserved functional antagonism of CELF and MBNL proteins controls stem cell-specific alternative splicing in planarians. eLife. 2016;5:e16797. 10.7554/eLife.16797 [DOI] [PMC free article] [PubMed] [Google Scholar]
132. Förch P, Puig O, Martínez C et al. The splicing regulator TIA-1 interacts with U1-C to promote U1 snRNP recruitment to 5′ splice sites. EMBO J. 2002;21:6882–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
133. Dickinson DJ, Ward JD, Reiner DJ et al. Engineering the Caenorhabditis elegans genome using Cas9-triggered homologous recombination. Nat Methods. 2013;10:1028–34. 10.1038/nmeth.2641 [DOI] [PMC free article] [PubMed] [Google Scholar]
134. Martin CJ, Calarco JA. Approaches for CRISPR/Cas9 genome editing in C. elegans. Methods Mol Biol. 2022;2468:215–37. [DOI] [PubMed] [Google Scholar]
135. Raker VA, Mironov AA, Gelfand MS et al. Modulation of alternative splicing by long-range RNA structures in Drosophila. Nucleic Acids Res. 2009;37:4533–44. 10.1093/nar/gkp407 [DOI] [PMC free article] [PubMed] [Google Scholar]
136. Bartys N, Kierzek R, Lisowiec-Wachnicka J. The regulation properties of RNA secondary structure in alternative splicing. Biochim Biophys Acta Gene Regul Mech. 2019;1862:194401. 10.1016/j.bbagrm.2019.07.002 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

Dainat J, Hereñú D, Pucholt P. AGAT: another GFF analysis toolkit to handle annotations in any GTF/GFF format. (Version v0.7.0). Zenodo. 10.5281/zenodo.3552717 [DOI]

Supplementary Materials

gkag451_Supplemental_Files

gkag451_supplemental_files.zip^{(4.4MB, zip)}

Data Availability Statement

The raw sequence data (fastq files) are available at the NCBI SRA:

https://www.ncbi.nlm.nih.gov/sra/PRJNA1359015

The code and additional data are available through Zenodo:

https://doi.org/10.5281/zenodo.17593183

Inferring Exon and Intron Metadata from .gff file (v1.0.1).

https://doi.org/10.5281/zenodo.15757714

[B1] 1. Pan Q, Shai O, Lee LJ et al. Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008;40:1413–5. 10.1038/ng.259 [DOI] [PubMed] [Google Scholar]

[B2] 2. Wang Z, Burge CB. Splicing regulation: from a parts list of regulatory elements to an integrated splicing code. RNA. 2008;14:802–13. 10.1261/rna.876308 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B3] 3. Zheng S, Black DL. Alternative pre-mRNA splicing in neurons: growing up and extending its reach. Trends Genet. 2013;29:442–8. 10.1016/j.tig.2013.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B4] 4. Raj B, Blencowe BJ. Alternative splicing in the mammalian nervous system: recent insights into mechanisms and functional roles. Neuron. 2015;87:14–27. 10.1016/j.neuron.2015.05.004 [DOI] [PubMed] [Google Scholar]

[B5] 5. Warf MB, Berglund JA. Role of RNA structure in regulating pre-mRNA splicing. Trends Biochem Sci. 2010;35:169–78. 10.1016/j.tibs.2009.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B6] 6. Baralle M, Baralle FE. The splicing code. Biosystems. 2018;164:39–48. 10.1016/j.biosystems.2017.11.002 [DOI] [PubMed] [Google Scholar]

[B7] 7. Blencowe BJ. Exonic splicing enhancers: mechanism of action, diversity and role in human genetic diseases. Trends Biochem Sci. 2000;25:106–10. 10.1016/S0968-0004(00)01549-8 [DOI] [PubMed] [Google Scholar]

[B8] 8. Black DL. Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003;72:291–336. 10.1146/annurev.biochem.72.121801.161720 [DOI] [PubMed] [Google Scholar]

[B9] 9. Wang Y, Ma M, Xiao X et al. Intronic splicing enhancers, cognate splicing factors and context dependent regulation rules. Nat Struct Mol Biol. 2012;19:1044–52. 10.1038/nsmb.2377 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B10] 10. Fairbrother WG, Yeh R-F, Sharp PA et al. Predictive identification of exonic splicing enhancers in human genes. Science. 2002;297:1007–13. 10.1126/science.1073774 [DOI] [PubMed] [Google Scholar]

[B11] 11. Gao Y, Lin K-T, Jiang T et al. Systematic characterization of short intronic splicing-regulatory elements in SMN2 pre-mRNA. Nucleic Acids Res. 2022;50:731–49. 10.1093/nar/gkab1280 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B12] 12. Ule J, Blencowe BJ. Alternative splicing regulatory networks: functions, mechanisms, and evolution. Mol Cell. 2019;76:329–45. 10.1016/j.molcel.2019.09.017 [DOI] [PubMed] [Google Scholar]

[B13] 13. Bartys N, Kierzek R, Lisowiec-Wachnicka J. The regulation properties of RNA secondary structure in alternative splicing. Biochim Biophys Acta Gene Regul Mech. 2019;1862:194401. 10.1016/j.bbagrm.2019.07.002 [DOI] [PubMed] [Google Scholar]

[B14] 14. Mikl M, Hamburg A, Pilpel Y et al. Dissecting splicing decisions and cell-to-cell variability with designed sequence libraries. Nat Commun. 2019;10:4572. 10.1038/s41467-019-12642-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B15] 15. Rodrigues F, Thuma L, Klämbt C. The regulation of glial-specific splicing of Neurexin IV requires HOW and Cdk12 activity. Dev Camb Engl. 2012;139:1765–76. [DOI] [PubMed] [Google Scholar]

[B16] 16. Wang Z, Rolish ME, Yeo G et al. Systematic identification and analysis of exonic splicing silencers. Cell. 2004;119:831–45. 10.1016/j.cell.2004.11.010 [DOI] [PubMed] [Google Scholar]

[B17] 17. Ke S, Shang S, Kalachikov SM et al. Quantitative evaluation of all hexamers as exonic splicing elements. Genome Res. 2011;21:1360–74. 10.1101/gr.119628.110 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B18] 18. Rosenberg AB, Patwardhan RP, Shendure J et al. Learning the sequence determinants of alternative splicing from millions of random sequences. Cell. 2015;163:698–711. 10.1016/j.cell.2015.09.054 [DOI] [PubMed] [Google Scholar]

[B19] 19. Baeza-Centurion P, Miñana B, Schmiedel JM et al. Combinatorial genetics reveals a scaling law for the effects of mutations on splicing. Cell. 2019;176:549–63.e23. 10.1016/j.cell.2018.12.010 [DOI] [PubMed] [Google Scholar]

[B20] 20. Rhine CL, Neil C, Wang J et al. Massively parallel reporter assays discover de novo exonic splicing mutants in paralogs of Autism genes. PLoS Genet. 2022;18:e1009884. 10.1371/journal.pgen.1009884 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B21] 21. Rong S, Neil CR, Welch A et al. Large-scale functional screen identifies genetic variants with splicing effects in modern and archaic humans. Proc Natl Acad Sci USA. 2023;120:e2218308120. 10.1073/pnas.2218308120 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B22] 22. Rabani M, Pieper L, Chew G-L et al. A massively parallel reporter assay of 3’ UTR sequences identifies In vivo rules for mRNA degradation. Mol Cell. 2017;68:1083–94. 10.1016/j.molcel.2017.11.014 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B23] 23. Cooper TA. Use of minigene systems to dissect alternative splicing elements. Methods San Diego Calif. 2005;37:331–40. 10.1016/j.ymeth.2005.07.015 [DOI] [PubMed] [Google Scholar]

[B24] 24. Wang Z, Burge CB. Splicing regulation: from a parts list of regulatory elements to an integrated splicing code. RNA. 2008;14:802–13. 10.1261/rna.876308 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B25] 25. Gaildrat P, Killian A, Martins A et al. Use of splicing reporter minigene assay to evaluate the effect on splicing of unclassified genetic variants. Methods Mol Biol Clifton NJ. 2010;653:249–57. [DOI] [PubMed] [Google Scholar]

[B26] 26. Smith SA, Lynch KW. Cell-based splicing of minigenes. Methods Mol Biol. 2014;1126:243–55. [DOI] [PubMed] [Google Scholar]

[B27] 27. Kreimer A, Ashuach T, Inoue F et al. Massively parallel reporter perturbation assays uncover temporal regulatory architecture during neural differentiation. Nat Commun. 2022;13:1504. 10.1038/s41467-022-28659-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B28] 28. McAfee JC, Bell JL, Krupa O et al. Focus on your locus with a massively parallel reporter assay. J Neurodev Disord. 2022;14:50. 10.1186/s11689-022-09461-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[B29] 29. Agarwal V, Inoue F, Schubach M et al. Massively parallel characterization of transcriptional regulatory elements. Nature. 2025;639:411–20. 10.1038/s41586-024-08430-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B30] 30. Gracida X, Norris AD, Calarco JA. Regulation of tissue-specific alternative splicing: C. elegans as a model system. Adv Exp Med Biol. 2016;907:229–61. [DOI] [PubMed] [Google Scholar]

[B31] 31. Koterniak B, Pilaka PP, Gracida X et al. Global regulatory features of alternative splicing across tissues and within the nervous system of C. elegans. Genome Res. 2020;30:1766–80. 10.1101/gr.267328.120 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B32] 32. Mello CC, Kramer JM, Stinchcomb D et al. Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 1991;10:3959–70. 10.1002/j.1460-2075.1991.tb04966.x [DOI] [PMC free article] [PubMed] [Google Scholar]

[B33] 33. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94. 10.1093/genetics/77.1.71 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B34] 34. Norris AD, Gao S, Norris ML et al. A pair of RNA-binding proteins controls networks of splicing events contributing to specialization of neural cell types. Mol Cell. 2014;54:946–59. 10.1016/j.molcel.2014.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B35] 35. Schafer S, Miao K, Benson CC et al. Alternative splicing signatures in RNA-seq data: percent spliced in (PSI). Curr Protoc Hum Genet. 2015;87:11.16.1–11.16.14. [DOI] [PubMed] [Google Scholar]

[B36] 36. Schindelin J, Arganda-Carreras I, Frise E et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9:676–82. 10.1038/nmeth.2019 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B37] 37. Howe KL, Bolt BJ, Cain S et al. WormBase 2016: expanding to enable helminth genomic research. Nucleic Acids Res. 2016;44:D774–80. 10.1093/nar/gkv1217 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B38] 38. Dainat J, Hereñú D, Pucholt P. AGAT: another GFF analysis toolkit to handle annotations in any GTF/GFF format. (Version v0.7.0). Zenodo. 10.5281/zenodo.3552717 [DOI]

[B39] 39. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–2. 10.1093/bioinformatics/btq033 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B40] 40. Grant CE, Bailey TL, Noble WS. FIMO: scanning for occurrences of a given motif. Bioinformatics. 2011;27:1017–8. 10.1093/bioinformatics/btr064 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B41] 41. Yeo G, Burge CB. Maximum entropy modeling of short sequence motifs with applications to RNA splicing signals. J Comput Biol J Comput Mol Cell Biol. 2004;11:377–94. 10.1089/1066527041410418 [DOI] [PubMed] [Google Scholar]

[B42] 42. Kuret K, Amalietti AG, Jones DM et al. Positional motif analysis reveals the extent of specificity of protein–RNA interactions observed by CLIP. Genome Biol. 2022;23:191. 10.1186/s13059-022-02755-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B43] 43. Tremblay BJ-M. universalmotif: an R package for biological motif analysis. J Open Source Softw. 2024;9:7012. 10.21105/joss.07012 [DOI] [Google Scholar]

[B44] 44. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22:4673–80. 10.1093/nar/22.22.4673 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B45] 45. Ray D, Kazan H, Chan ET et al. Rapid and systematic analysis of the RNA recognition specificities of RNA-binding proteins. Nat Biotechnol. 2009;27:667–70. 10.1038/nbt.1550 [DOI] [PubMed] [Google Scholar]

[B46] 46. Gupta S, Stamatoyannopoulos JA, Bailey TL et al. Quantifying similarity between motifs. Genome Biol. 2007;8:R24. 10.1186/gb-2007-8-2-r24 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B47] 47. Fusca DD, Kasimatis KR, Zhu HV et al. Dynamic birth and death of argonaute gene Family functional repertoire across Caenorhabditis nematodes. Genome Biol Evol. 2025;17:evaf106. 10.1093/gbe/evaf106 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B48] 48. Emms DM, Kelly S. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 2019;20:238. 10.1186/s13059-019-1832-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[B49] 49. Katoh K, Rozewicki J, Yamada KD. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief Bioinform. 2019;20:1160–6. 10.1093/bib/bbx108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B50] 50. Alam A, Duncan AG, Mitchell JA et al. Functional similarity of non-coding regions is revealed in phylogenetic average motif score representations. bioRxiv, 10.1101/2023.04.09.536185, 11 April 2023, preprint: not peer reviewed. [DOI]

[B51] 51. de Hoon MJL, Imoto S, Nolan J et al. Open source clustering software. Bioinformatics. 2004;20:1453–4. [DOI] [PubMed] [Google Scholar]

[B52] 52. Saldanha AJ. Java treeview–extensible visualization of microarray data. Bioinforma Oxf Engl. 2004;20:3246–8. 10.1093/bioinformatics/bth349 [DOI] [PubMed] [Google Scholar]

[B53] 53. Kivioja T, Vähärautio A, Karlsson K et al. Counting absolute numbers of molecules using unique molecular identifiers. Nat Methods. 2012;9:72–4. 10.1038/nmeth.1778 [DOI] [PubMed] [Google Scholar]

[B54] 54. Sibley CR, Blazquez L, Ule J. Lessons from non-canonical splicing. Nat Rev Genet. 2016;17:407–21. 10.1038/nrg.2016.46 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B55] 55. Aldalaqan S, Dalgliesh C, Luzzi S et al. Cryptic splicing: common pathological mechanisms involved in male infertility and neuronal diseases. Cell Cycle Georget Tex. 2022;21:219–27. 10.1080/15384101.2021.2015672 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B56] 56. Rinke J, Appel B, Blöcker H et al. The 5’-terminal sequence of U1 RNA complementary to the consensus 5’ splice site of hnRNA is single-stranded in intact U1 snRNP particles. Nucleic Acids Res. 1984;12:4111–26. 10.1093/nar/12.10.4111 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B57] 57. Roca X, Krainer AR, Eperon IC. Pick one, but be quick: 5′ splice sites and the problems of too many choices. Genes Dev. 2013;27:129–44. 10.1101/gad.209759.112 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B58] 58. Perchlik M, Sasse A, Mostafavi S et al. Impact on splicing in Saccharomyces cerevisiae of random 50-base sequences inserted into an intron. RNA N Y N. 2023;30:52–67. 10.1261/rna.079752.123 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B59] 59. Buratti E, Baralle M, Baralle FE. Defective splicing, disease and therapy: searching for master checkpoints in exon definition. Nucleic Acids Res. 2006;34:3494–510. 10.1093/nar/gkl498 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B60] 60. Liu H-X, Zhang M, Krainer AR. Identification of functional exonic splicing enhancer motifs recognized by individual SR proteins. Genes Dev. 1998;12:1998–2012. 10.1101/gad.12.13.1998 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B61] 61. Chen M, Manley JL. Mechanisms of alternative splicing regulation: insights from molecular and genomics approaches. Nat Rev Mol Cell Biol. 2009;10:741–54. 10.1038/nrm2777 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B62] 62. Lee Y, Rio DC. Mechanisms and regulation of alternative pre-mRNA splicing. Annu Rev Biochem. 2015;84:291–323. 10.1146/annurev-biochem-060614-034316 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B63] 63. Ray D, Kazan H, Cook KB et al. A compendium of RNA-binding motifs for decoding gene regulation. Nature. 2013;499:172–7. 10.1038/nature12311 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B64] 64. Jin Y, Suzuki H, Maegawa S et al. A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG. EMBO J. 2003;22:905–12. 10.1093/emboj/cdg089 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B65] 65. Sun S, Zhang Z, Fregoso O et al. Mechanisms of activation and repression by the alternative splicing factors RBFOX1/2. RNA. 2012;18:274–83. 10.1261/rna.030486.111 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B66] 66. Kuroyanagi H, Watanabe Y, Suzuki Y et al. Position-dependent and neuron-specific splicing regulation by the CELF family RNA-binding protein UNC-75 in Caenorhabditis elegans. Nucleic Acids Res. 2013;41:4015–25. 10.1093/nar/gkt097 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B67] 67. Chen L, Liu Z, Zhou B et al. CELF RNA binding proteins promote axon regeneration in C. elegans and mammals through alternative splicing of Syntaxins. eLife. 2016;5:e16072. 10.7554/eLife.16072 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B68] 68. Pilaka-Akella P, Sadek NH, Fusca D et al. Neuron-specific repression of alternative splicing by the conserved CELF protein UNC-75 in Caenorhabditis elegans. Genetics. 2025;229:iyaf025. 10.1093/genetics/iyaf025 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B69] 69. Anyanful A, Ono K, Johnsen RC et al. The RNA-binding protein SUP-12 controls muscle-specific splicing of the ADF/cofilin pre-mRNA in C. elegans. J Cell Biol. 2004;167:639–47. 10.1083/jcb.200407085 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B70] 70. Kuwasako K, Takahashi M, Unzai S et al. RBFOX and SUP-12 sandwich a G base to cooperatively regulate tissue-specific splicing. Nat Struct Mol Biol. 2014;21:778–86. 10.1038/nsmb.2870 [DOI] [PubMed] [Google Scholar]

[B71] 71. Fu X-D, Ares M. Context-dependent control of alternative splicing by RNA-binding proteins. Nat Rev Genet. 2014;15:689–701. 10.1038/nrg3778 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B72] 72. Yang Y, Lee GC, Nakagaki-Silva E et al. Cell-type specific regulator RBPMS switches alternative splicing via higher-order oligomerization and heterotypic interactions with other splicing regulators. Nucleic Acids Res. 2023;51:9961–82. 10.1093/nar/gkad652 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B73] 73. Sasse A, Ray D, Laverty KU et al. A resource of RNA-binding protein motifs across eukaryotes reveals evolutionary dynamics and gene-regulatory function. Nat Biotechnol. 2025. 10.1038/s41587-025-02733-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B74] 74. Roy SW, Penny D. Smoke without fire: most reported cases of intron gain in nematodes instead reflect intron losses. Mol Biol Evol. 2006;23:2259–62. 10.1093/molbev/msl098 [DOI] [PubMed] [Google Scholar]

[B75] 75. Ma M-Y, Xia J, Shu K-X et al. Intron losses and gains in the nematodes. Biol Direct. 2022;17:13. 10.1186/s13062-022-00328-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B76] 76. Teterina AA, Willis JH, Baer CF et al. Pervasive conservation of intron number and other genetic elements revealed by a chromosome-level genome assembly of the hyper-polymorphic nematode Caenorhabditis brenneri. Genome Biol Evol. 2025;17:evaf037. 10.1093/gbe/evaf037 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B77] 77. Sugnet CW, Kent WJ, Ares M et al. Transcriptome and genome conservation of alternative splicing events in humans and mice. Pac Symp Biocomput. 2004;66–77. 10.1142/9789812704856_0007 [DOI] [PubMed] [Google Scholar]

[B78] 78. Fagnani M, Barash Y, Ip JY et al. Functional coordination of alternative splicing in the mammalian central nervous system. Genome Biol. 2007;8:R108. 10.1186/gb-2007-8-6-r108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B79] 79. Baek D, Green P. Sequence conservation, relative isoform frequencies, and nonsense-mediated decay in evolutionarily conserved alternative splicing. Proc Natl Acad Sci USA. 2005;102:12813–8. 10.1073/pnas.0506139102 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B80] 80. Pan C-L, Baum PD, Gu M et al. C. elegans AP-2 and Retromer control wnt secretion by regulating MIG-14/Wntless. Dev Cell. 2008;14:132–9. 10.1016/j.devcel.2007.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B81] 81. Kaplan OI, Doroquez DB, Cevik S et al. Endocytosis genes facilitate protein and membrane transport in C. elegans sensory cilia. Curr Biol. 2012;22:451–60. 10.1016/j.cub.2012.01.060 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B82] 82. Brown AR, Fox GA, Kaplow IM et al. An in vivo systemic massively parallel platform for deciphering animal tissue-specific regulatory function. Front Genet. 2025;16:1–19. 10.3389/fgene.2025.1533900 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B83] 83. Lambert JT, Su-Feher L, Cichewicz K et al. Parallel functional testing identifies enhancers active in early postnatal mouse brain. eLife. 2021;10:e69479. 10.7554/eLife.69479 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B84] 84. Zheng Y, VanDusen NJ. Massively parallel reporter assays for high-throughput in vivo analysis of cis-regulatory elements. J Cardiovasc Dev Dis. 2023;10:144. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B85] 85. Martinez BA, Reis Rodrigues P, Nuñez Medina RM et al. An alternatively spliced, non-signaling insulin receptor modulates insulin sensitivity via insulin peptide sequestration in C. elegans. eLife. 2020;9:e49917. 10.7554/eLife.49917 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B86] 86. Lee RY, Hench J, Ruvkun G. Regulation of C. elegans DAF-16 and its human ortholog FKHRL1 by the daf-2 insulin-like signaling pathway. Curr Biol. 2001;11:1950–7. 10.1016/S0960-9822(01)00595-4 [DOI] [PubMed] [Google Scholar]

[B87] 87. Tamburino AM, Ryder SP, Walhout AJM. A compendium of Caenorhabditis elegans RNA binding proteins predicts extensive regulation at multiple levels. G3 (Bethesda). 2013;3:297–304. 10.1534/g3.112.004390 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B88] 88. Liao SE, Sudarshan M, Regev O. Deciphering RNA splicing logic with interpretable machine learning. Proc Natl Acad Sci USA. 2023;120:e2221165120. 10.1073/pnas.2221165120 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B89] 89. La Fleur A, Shi Y, Seelig G. Decoding biology with massively parallel reporter assays and machine learning. Genes Dev. 2024;38:843–65. 10.1101/gad.351800.124 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B90] 90. Gupta K, Yang C, McCue K et al. Improved modeling of RNA-binding protein motifs in an interpretable neural model of RNA splicing. Genome Biol. 2024;25:23. 10.1186/s13059-023-03162-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[B91] 91. Tan JH, Fraser AG. The combinatorial control of alternative splicing in C. elegans. PLoS Genet. 2017;13:e1007033. 10.1371/journal.pgen.1007033 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B92] 92. Li H, Ding Z, Fang Z-Y et al. Conserved intronic secondary structures with concealed branch sites regulate alternative splicing of poison exons. Nucleic Acids Res. 2024;52:6002–16. 10.1093/nar/gkae185 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B93] 93. Rajendren S, Dhakal A, Vadlamani P et al. Profiling neural editomes reveals a molecular mechanism to regulate RNA editing during development. Genome Res. 2021;31:27–39. 10.1101/gr.267575.120 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B94] 94. Eperon LP, Estibeiro JP, Eperon IC. The role of nucleotide sequences in splice site selection in eukaryotic pre-messenger RNA. Nature. 1986;324:280–2. 10.1038/324280a0 [DOI] [PubMed] [Google Scholar]

[B95] 95. Farrer T, Roller AB, Kent WJ et al. Analysis of the role of Caenorhabditis elegans GC-AG introns in regulated splicing. Nucleic Acids Res. 2002;30:3360–7. 10.1093/nar/gkf465 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B96] 96. Wong MS, Kinney JB, Krainer AR. Quantitative activity profile and context dependence of all human 5’ splice sites. Mol Cell. 2018;71:1012–26. 10.1016/j.molcel.2018.07.033 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B97] 97. Roca X, Akerman M, Gaus H et al. Widespread recognition of 5’ splice sites by noncanonical base-pairing to U1 snRNA involving bulged nucleotides. Genes Dev. 2012;26:1098–109. 10.1101/gad.190173.112 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B98] 98. Roca X, Sachidanandam R, Krainer AR. Intrinsic differences between authentic and cryptic 5’ splice sites. Nucleic Acids Res. 2003;31:6321–33. 10.1093/nar/gkg830 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B99] 99. Tan J, Ho JXJ, Zhong Z et al. Noncanonical registers and base pairs in human 5’ splice-site selection. Nucleic Acids Res. 2016;44:3908–21. 10.1093/nar/gkw163 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B100] 100. Taylor SR, Santpere G, Weinreb A et al. Molecular topography of an entire nervous system. Cell. 2021;184:4329–47. 10.1016/j.cell.2021.06.023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B101] 101. Barrett A, Varol E, Weinreb A et al. Integrating bulk and single cell RNA-seq refines transcriptomic profiles of individual C. elegans neurons. eLife. 2025;14:RP106183. 10.7554/eLife.106183.2 [DOI] [Google Scholar]

[B102] 102. Cao J, Packer JS, Ramani V et al. Comprehensive single-cell transcriptional profiling of a multicellular organism. Science. 2017;357:661–7. 10.1126/science.aam8940 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B103] 103. Packer JS, Zhu Q, Huynh C et al. A lineage-resolved molecular atlas of C. elegans embryogenesis at single cell resolution. Science. 2019;365:eaax1971. 10.1126/science.aax1971 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B104] 104. Roux AE, Yuan H, Podshivalova K et al. Individual cell types in C. elegans age differently and activate distinct cell-protective responses. Cell Rep. 2023;42:112902. 10.1016/j.celrep.2023.112902 [DOI] [PubMed] [Google Scholar]

[B105] 105. Zhao S, Hong CKY, Myers CA et al. A single-cell massively parallel reporter assay detects cell-type-specific gene regulation. Nat Genet. 2023;55:346–54. 10.1038/s41588-022-01278-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B106] 106. Zhang H, Li X, Song D et al. Worm perturb-seq: massively parallel whole-animal RNAi and RNA-seq. Nat Commun. 2025;16:4785. 10.1038/s41467-025-60154-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B107] 107. Feng H, Moakley DF, Chen S et al. Complexity and graded regulation of neuronal cell-type-specific alternative splicing revealed by single-cell RNA sequencing. Proc Natl Acad Sci USA. 2021;118:e2013056118. 10.1073/pnas.2013056118 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B108] 108. Weinreb A, Varol E, Barrett A et al. Alternative splicing across the C. elegans nervous system. Nat Commun. 2025;16:4508. 10.1038/s41467-025-58293-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B109] 109. Thompson M, Bixby R, Dalton R et al. Splicing in a single neuron is coordinately controlled by RNA binding proteins and transcription factors. eLife. 2019;8:e46726. 10.7554/eLife.46726 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B110] 110. Bortfeldt R, Schindler S, Szafranski K et al. Comparative analysis of sequence features involved in the recognition of tandem splice sites. BMC Genomics. 2008;9:202. 10.1186/1471-2164-9-202 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B111] 111. Kabat JL, Barberan-Soler S, McKenna P et al. Intronic alternative splicing regulators identified by comparative genomics in nematodes. PLoS Comput Biol. 2006;2:e86. 10.1371/journal.pcbi.0020086 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B112] 112. Kol G, Lev-Maor G, Ast G. Human–mouse comparative analysis reveals that branch-site plasticity contributes to splicing regulation. Hum Mol Genet. 2005;14:1559–68. 10.1093/hmg/ddi164 [DOI] [PubMed] [Google Scholar]

[B113] 113. Yeo GW, Nostrand ELV, Liang TY. Discovery and analysis of evolutionarily conserved intronic splicing regulatory elements. PLoS Genet. 2007;3:e85. 10.1371/journal.pgen.0030085 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B114] 114. Barberan-Soler S, Zahler AM. Alternative splicing regulation during C. elegans development: splicing factors as regulated targets. PLoS Genet. 2008;4:e1000001. 10.1371/journal.pgen.1000001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B115] 115. Laver JD, Duncan M, Calarco JA. A global view of the RNA-binding and regulatory protein landscape in Caenorhabditis elegans. bioRxiv, 10.1101/2024.11.07.622524, 11 November 2024, preprint: not peer reviewed. [DOI]

[B116] 116. True JR, Haag ES. Developmental system drift and flexibility in evolutionary trajectories. Evol Dev. 2001;3:109–19. 10.1046/j.1525-142x.2001.003002109.x [DOI] [PubMed] [Google Scholar]

[B117] 117. Barbosa-Morais NL, Irimia M, Pan Q et al. The evolutionary landscape of alternative splicing in vertebrate species. Science. 2012;338:1587–93. 10.1126/science.1230612 [DOI] [PubMed] [Google Scholar]

[B118] 118. Jelen N, Ule J, Zivin M et al. Evolution of Nova-dependent splicing regulation in the brain. PLoS Genet. 2007;3:1838–47. 10.1371/journal.pgen.0030173 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B119] 119. Dvinge H. Regulation of alternative mRNA splicing: old players and new perspectives. FEBS Lett. 2018;592:2987–3006. 10.1002/1873-3468.13119 [DOI] [PubMed] [Google Scholar]

[B120] 120. Iadevaia V, Gerber AP. Combinatorial control of mRNA fates by RNA-binding proteins and non-coding RNAs. Biomolecules. 2015;5:2207–22. 10.3390/biom5042207 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B121] 121. Lin J-C, Tarn W-Y. Exon selection in alpha-tropomyosin mRNA is regulated by the antagonistic action of RBM4 and PTB. Mol Cell Biol. 2005;25:10111–21. 10.1128/MCB.25.22.10111-10121.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B122] 122. Carrasco J, Rauer M, Hummel B et al. ELAV and FNE determine neuronal transcript signatures through EXon-activated rescue. Mol Cell. 2020;80:156–63. 10.1016/j.molcel.2020.09.011 [DOI] [PubMed] [Google Scholar]

[B123] 123. Kuroyanagi H, Kobayashi T, Mitani S et al. Transgenic alternative-splicing reporters reveal tissue-specific expression profiles and regulation mechanisms in vivo. Nat Methods. 2006;3:909–15. 10.1038/nmeth944 [DOI] [PubMed] [Google Scholar]

[B124] 124. Montañés-Agudo P, Aufiero S, Schepers EN et al. The RNA-binding protein QKI governs a muscle-specific alternative splicing program that shapes the contractile function of cardiomyocytes. Cardiovasc Res. 2023;119:1161–74. 10.1093/cvr/cvad007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B125] 125. Gonatopoulos-Pournatzis T, Blencowe BJ. Microexons: at the nexus of nervous system development, behaviour and autism spectrum disorder. Curr Opin Genet Dev. 2020;65:22–33. 10.1016/j.gde.2020.03.007 [DOI] [PubMed] [Google Scholar]

[B126] 126. Irimia M, Weatheritt RJ, Ellis JD et al. A highly conserved program of neuronal microexons is misregulated in autistic brains. Cell. 2014;159:1511–23. 10.1016/j.cell.2014.11.035 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B127] 127. Head SA, Hernandez-Alias X, Yang J-S et al. Silencing of SRRM4 suppresses microexon inclusion and promotes tumor growth across cancers. PLoS Biol. 2021;19:e3001138. 10.1371/journal.pbio.3001138 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B128] 128. Gupta T, Margolin G, Burgess HA. Mutations in the microexon splicing regulator srrm4 have minor phenotypic effects on zebrafish neural development. G3 (Bethesda). 2025;15:jkaf052. 10.1093/g3journal/jkaf052 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B129] 129. Choudhary B, Marx O, Norris AD. Spliceosomal component PRP-40 is a central regulator of microexon splicing. Cell Rep. 2021;36:109464. 10.1016/j.celrep.2021.109464 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B130] 130. Torres-Méndez A, Bonnal S, Marquez Y et al. A novel protein domain in an ancestral splicing factor drove the evolution of neural microexons. Nat Ecol Evol. 2019;3:691–701. 10.1038/s41559-019-0813-6 [DOI] [PubMed] [Google Scholar]

[B131] 131. Solana J, Irimia M, Ayoub S et al. Conserved functional antagonism of CELF and MBNL proteins controls stem cell-specific alternative splicing in planarians. eLife. 2016;5:e16797. 10.7554/eLife.16797 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B132] 132. Förch P, Puig O, Martínez C et al. The splicing regulator TIA-1 interacts with U1-C to promote U1 snRNP recruitment to 5′ splice sites. EMBO J. 2002;21:6882–92. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B133] 133. Dickinson DJ, Ward JD, Reiner DJ et al. Engineering the Caenorhabditis elegans genome using Cas9-triggered homologous recombination. Nat Methods. 2013;10:1028–34. 10.1038/nmeth.2641 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B134] 134. Martin CJ, Calarco JA. Approaches for CRISPR/Cas9 genome editing in C. elegans. Methods Mol Biol. 2022;2468:215–37. [DOI] [PubMed] [Google Scholar]

[B135] 135. Raker VA, Mironov AA, Gelfand MS et al. Modulation of alternative splicing by long-range RNA structures in Drosophila. Nucleic Acids Res. 2009;37:4533–44. 10.1093/nar/gkp407 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B136] 136. Bartys N, Kierzek R, Lisowiec-Wachnicka J. The regulation properties of RNA secondary structure in alternative splicing. Biochim Biophys Acta Gene Regul Mech. 2019;1862:194401. 10.1016/j.bbagrm.2019.07.002 [DOI] [PubMed] [Google Scholar]

PERMALINK

An in vivo parallelized reporter assay to uncover tissue-specific splicing regulatory sequences in a multicellular animal

Sanjana Bhatnagar

Jade Ho

Nour H Sadek

Isha Singh

Michael Zoberman

Bina Koterniak

Yufang Liu

Daniel D Fusca

Asher D Cutter

Alan M Moses

John A Calarco

Roles

Abstract

Graphical Abstract

Graphical Abstract.

Introduction

Materials and methods

Caenorhabditis elegans strains used

Construction and sequencing of synthetic reporter libraries for PRA

Microinjection, RNA extraction, amplification, and sequencing of tissue-specific PRA reporter libraries in C. elegans

High-throughput classification and quantification of splicing isoforms in tissue-specific PRA libraries

Individual reporter plasmid construction and semi-quantitative reverse transcriptase-polymerase chain reaction validations

Mapping PRA elements functioning as competitive 5′ss to native C. elegans intronic regions

k-mer count ULR to characterize core regulatory subsequences

Clustering tissue-biased PRA k-mers based on sequence similarity

Motif comparison of significant splicing enhancer and silencer k-mers with RBP motifs across tissues

Phylogenetically averaged motif score analysis on C. elegans neuronal/muscle switch-like splicing events

Targeted mutagenesis of neuronal ISE k-mers in dpy-23 minigene reporter

Results

Implementation of a parallelized splicing reporter assay in C. elegans neuronal and muscle tissue

Figure 1.

PRA sequences resembling a strong 5′ss consensus lead to competing splice site usage

Figure 2.

Synthetic PRA sequences can act as ISEs or ISSs

Figure 3.

Regulatory activity of recovered PRA cis elements can be attributed to core subsequences

Figure 4.

Figure 5.

Direct comparison of intronic enhancer and silencer activity in neuronal and muscle tissue reveals tissue-biased roles for sequences

PRA-derived ISE and ISS motifs can be mapped to endogenous conserved intronic regions flanking tissue-biased switch-like exons

Figure 6.

Mutation of PRA-identified ISEs disrupts the splicing of an endogenous neuronal microexon

Figure 7.

Discussion

A screening tool to identify splicing regulatory sequences with tissue-biased activity in vivo

Extensive combinatorial control of alternative splicing supported by evolutionary signatures

Two intronic enhancers regulate the inclusion of a neuronal microexon in C. elegans

Supplementary Material

Acknowledgements

Notes

Contributor Information

Supplementary data

Conflict of interest

Funding

Data availability

References

Associated Data

Data Citations

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases