Tackling Non-Canonical Splicing in Arrhythmogenic Cardiomyopathy to Reduce the Uncertain Significance Variants Burden

Tackling Non-Canonical Splicing in Arrhythmogenic Cardiomyopathy to Reduce the Uncertain Significance Variants Burden | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Tackling Non-Canonical Splicing in Arrhythmogenic Cardiomyopathy to Reduce the Uncertain Significance Variants Burden Rudy Celeghin, Giulia Tosato, Serena Pinci, Francesca Dalla Zanna, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9022128/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Background . Splice-altering variants (SAVs), particularly those outside canonical splice sites, are an underappreciated contributor to inherited cardiovascular diseases. In arrhythmogenic cardiomyopathy (ACM), these variants frequently remain classified as of uncertain significance (VUS) due to limited predictive power and lack of transcript-level evidence, constraining genetic yield and clinical management. Our study aimed to determine the functional impact of SAVs in ACM genes and refine their classification using ACMG/AMP and ClinGen SVI criteria. Methods . SAVs identified in 200 ACM probands underwent SpliceAI prediction, GTEx cardiac exon-usage annotation, and functional assessment using pSPL3-based minigene assays. Aberrant transcripts were quantified using Percent Splicing Alteration (PSA). Segregation data and ACMG/AMP criteria refined by ClinGen SVI were applied to integrate functional and clinical evidence for classification. Results. Aberrant splicing was confirmed in 9/20 variants (45%), including synonymous, missense, and non-canonical intronic changes. SpliceAI scores correlated strongly with PSA values (R²=0.86). Case–control burden testing revealed significant enrichment of splice-altering variants in DSP , DSG2 , DSC2 and FLNC . Integrating predictive algorithms with experimental validation and segregation analysis markedly enhances reclassification of 16/20 variants (80%). Conclusion. Splicing defects beyond canonical sites significantly shape ACM genetic landscape. Integrating predictive models with experimental validation clarifies uncertain variants bridging the gap between genomic uncertainty and clinical decision-making. Non-Canonical Splicing Variants Arrhythmogenic Cardiomyopathy Genetics in vitro assay Figures Figure 1 Figure 2 Figure 3 Key points ♣ Non-canonical splice-altering variants contribute significantly to the genetic architecture of arrhythmogenic cardiomyopathy (ACM) and are frequently missed or misclassified by standard DNA-based diagnostic approaches. ♣ Functional splicing assays resolve variant uncertainty, enabling reclassification of 80% of previously uncertain variants and substantially improving diagnostic yield across major ACM-associated genes. ♣ Integration of SpliceAI prediction, minigene assays, and tissue-specific exon usage (GTEx) provides a robust, evidence-based framework for interpreting splicing variants in ACM according to ACMG/ClinGen SVI guidelines. ♣ Aberrant splicing events were experimentally confirmed in 45% of tested variants, including synonymous and non-canonical changes, reinforcing the need for transcript-level evaluation in inherited cardiomyopathies. ♣ Tissue-specific exon expression analysis refines pathogenicity assessment, distinguishing variants that disrupt clinically relevant cardiac isoforms from those unlikely to affect myocardial biology. Background Variants that alter RNA splicing represent a substantial and frequently underestimated cause of monogenic disease, including inherited cardiomyopathies. While clinical genetic testing has traditionally focused on protein-altering variants, splicing defects account for an estimated 10% of pathogenic findings and often escape detection or correct interpretation.[ 1 – 4 ] Canonical splice-site changes are typically well captured by current American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) guidelines,[ 5 ] yet variants affecting non-canonical positions, such as deep intronic regions, last-nucleotide exonic substitutions, or regulatory cis-elements, remain challenging to classify.[ 1 , 3 , 5 – 8 ] These elusive alterations may generate exon skipping, intron retention, or pseudoexon activation, ultimately producing aberrant transcripts with potentially deleterious functional consequences.[ 9 – 11 ] Arrhythmogenic cardiomyopathy (ACM) is a heritable myocardial disorder characterized by fibrofatty replacement, ventricular arrhythmias, and an elevated risk of sudden cardiac death.[ 12 ] Advances in next-generation sequencing have broadened the genetic landscape of ACM, extending beyond the desmosomal complex to include an expanding set of non-desmosomal genes.[ 13 ] , [ 14 ] Despite this progress, the overall diagnostic yield remains constrained by a persistent and substantial burden of variants of uncertain significance (VUS), which continues to limit definitive molecular diagnosis and clinical translation. Among these, non-canonical splice-altering variants are likely under-recognized contributors to disease, owing both to insufficient predictive power of computational tools and to the limited availability of transcript-based assays in routine diagnostics. Recent functional studies across inherited cardiac conditions have revealed that such non-canonical splicing defects represent a non-negligible fraction of pathogenic variants,[ 15 – 19 ] reaffirming the need for systematic transcript-level evaluation. Tools such as SpliceAI have improved the prediction of splicing disruption, but in vitro assays, including reverse-transcription PCR (RT-PCR)[ 20 ] and minigene systems,[ 21 ] are recommended by ClinGen Sequence Variant Interpretation (SVI) to confirm the biological impact of candidate variants.[ 22 ] Using minigene assays, complemented by cardiac tissue RNA analysis, we aimed to quantify non-canonical splicing alterations in ACM-associated genes, evaluating the concordance between predicted and observed effects, and refine variant classification under ACMG/AMP and ClinGen SVI frameworks. By uncovering splicing defects that would otherwise remain hidden behind ambiguous sequencing results, our work seeks to enhance diagnostic precision and reduce the persistent VUS burden in ACM genomic medicine. Methods Study population A retrospective analysis was conducted on a cohort of patients diagnosed with ACM, aiming to identify splicing site variants in genes primarily associated with the disease. Patients were enrolled through the Cardiovascular Pathology Unit (UOC Patologia Cardiovascolare) of the University Hospital of Padua and the Regional Registry for Cardio-Cerebro-Vascular Diseases. Clinical diagnosis of ACM was established based on current guidelines.[ 23 ] Ethical approval and informed consent were obtained in accordance with institutional guidelines. Genetic screening Genetic screening of the enrolled patients was previously performed by the Cardiovascular Genetics Laboratory at the Cardiovascular Pathology Unit, University Hospital of Padua. Whole-exome sequencing (WES) was initially carried out on all patients, and identified variants were subsequently confirmed by Sanger sequencing. Population frequency and expression datasets Population frequencies were extracted from the Genome Aggregation Database (gnomAD v2.1, controls only) in order to perform the Burden test. Gene- and exon-specific cardiac expression profiles were retrieved from the Genotype-Tissue Expression (GTEx) database to assess tissue-relevant isoform usage. Splicing effect prediction The predicted impact of each variant on RNA splicing was evaluated using SpliceAI (GRCh38). For each variant, delta scores for acceptor gain/loss and donor gain/loss were retrieved. Following ClinGen SVI recommendations, variants with a SpliceAI score > 0.25 were considered at risk for splicing disruption and prioritized for functional validation.[ 4 ] Minigene constructs Candidate variants were tested using a modified version of pSPL3-based minigene constructs [ 24 ] containing the exon of interest and flanking intronic regions. Wild-type (WT) and mutant (MUT) inserts were cloned into the vector and transfected into Human embryonic kidney 293 cells (HEK-293T). After RNA extraction and reverse transcription, splicing products were analyzed by Sanger sequencing ( Tab.S1 , Fig. S1 , Fig. S2 ). Cardiac tissue RNA analysis Myocardial RNA obtained during heart transplantation from an affected proband was available for direct transcript analysis by RT-PCR. Quantification of splicing alteration Splicing efficiency was quantified using the Percent Splicing Alteration (PSA) metric, defined as the proportion of transcripts exhibiting a specific alternative splicing event relative to the total transcript pool. In minigene assays, PSA was calculated as follows: $$\text{P}\text{S}\text{A}=\frac{\text{P}\text{S}\text{A}}{\text{P}\text{S}\text{A}+\text{P}\text{S}\text{I}}$$ where PSA and the Percentage Spliced-In (PSI) represent the relative intensities of bands corresponding to alternatively and canonically spliced transcripts, respectively, calculated from TapeStation peak intensities. Variant re-classification A comprehensive evaluation of each variant’s pathogenic potential was performed using three main bioinformatics platforms: Franklin ( https://franklin.genoox.com ), VarSome ( https://varsome.com ), and Alamut (Sophia Genetics, Switzerland), which integrate a wide range of public and proprietary databases, as well as multiple in silico predictors. Variant interpretation followed ACMG/AMP guidelines as refined by the ClinGen SVI Splicing Subgroup:[ 25 ] PS3/BS3 and PP3/BP4 were not applied because functional assays and in silico splicing predictions were already incorporated through the PVS1/BP7 framework. PVS1 strength was assigned based on predicted effects of the splicing alteration on the protein and evidence of loss of function (LoF). BP7_Strong was applied to synonymous or intronic variants without experimentally detectable splicing impact. Furthermore, PP1 was included when segregation supported variant-phenotype cosegregation[ 5 ] and its strengh was determined based on the probability of segregation as previously proposed.[ 26 ] Statistical analysis Categorical variables were reported as counts and percentages, while continuous variables were expressed as appropriate. Case-control burden testing was performed by calculating odds ratios (OR) with 95% confidence intervals (CI) and corresponding p-values. Correlation between SpliceAI prediction scores and experimentally derived PSA values was assessed using simple linear regression, and the strength of association was expressed as the coefficient of determination (R²). A p-value < 0.05 was considered statistically significant. Detailed methods are provided in Supplementary Materials . Results Variant identification and localization A comprehensive variant screening was performed in a cohort of 200 probands with a diagnosis of ACM,[ 23 ] revealing 20 splicing variants across the following genes: desmoplakin ( DSP NM_004415.4; n = 8), desmoglein-2 ( DSG2 NM_001943.5; n = 4), filamin-C ( FLNC NM_001458.5; n = 4), desmocollin-2 ( DSC2 NM_024422.6; n = 3), and plakofilin-2 ( PKP2 NM_001005242.3; n = 1) (Fig. 1 ). Among these variants, two were missense variants, five affected canonical splice sites (± 1 or 2 base pairs from exon-intron boundaries), while the remaining were located in non-canonical splicing regions (see Fig.S3 ). Burden of Splice-Altering VUS Among ACM Cases vs Controls To evaluate whether putative splice-altering variants were enriched in ACM, we performed a burden test comparing their allele frequency in our 200 ACM probands (400 alleles) with that observed in gnomAD v2.1 controls. Across desmosomal genes and FLNC , rare splice-site VUS were significantly over-represented in cases. For DSC2 , we observed 2 ACM alleles versus 68 in gnomAD, corresponding to an OR of 7.744 (95% CI 1.891–31.70; p = 0.0008). A similar enrichment was found for DSG2 (3 ACM vs 60 control alleles; OR of 12.8, 95% CI 4.00-40.99; p < 0.0001) and was even more pronounced for DSP , where 6 ACM alleles and 121 gnomAD alleles yielded an OR of 13.59 (95% CI 5.950–31.02; p < 0.0001). FLNC splice VUS were less frequent (3 ACM vs 236 control alleles) but still significantly enriched (OR of 3.35, 95% CI 1.07–10.52; p = 0.0276). In contrast, the frequency of PKP2 splice VUS (1 ACM vs 140 control alleles) was not different from controls (OR 1.88, 95% CI 0.26–13.44; p = 0.5244), and the wide CI likely reflects the limited sample size and insufficient power to detect a significant burden. Overall, these data indicate a non-random enrichment of rare splice-disrupting variants in DSC2, DSG2, DSP and FLNC in ACM, mirroring the increased burden of non-canonical splice variants reported for TTN in dilated cardiomyopathy ( Tab.S2 ).[ 16 ] Genotype-Tissue Expression (GTEx) An additional analysis was carried out using the GTEx Project, which provides detailed information on transcript isoform usage across human tissues, including the heart. Inspection of the splicing profiles highlights that the myocardium expresses a specific subset of isoforms, characterized by the consistent inclusion of certain exons and the systematic skipping of others ( Fig. S4 ). This allows us to determine which exons are actually used in the adult cardiac transcriptome, and therefore which splice-affecting variants are most likely to have functional relevance. Notably, the GTEx data indicate that all exons of the analyzed genes are expressed in heart tissue except for PKP2 exon 6, which is not represented in the predominant cardiac isoforms. The distribution of isoforms thus enables the identification of “critical” exons whose disrupted splicing may exert a pathogenic effect. These data effectively complement genetic analyses and functional assays by providing essential biological context for interpreting variants that alter normal splicing. SpliceAI-based splicing impact prediction In silico splicing impact prediction was performed on all variants using SpliceAI (Table 1). According to ClinGen SVI recommendations,[ 25 ] a score threshold of > 0.25 was set to indicate potential splicing disruption. A total of 9/20 variants (including 2 missense, 5 located in canonical sites and 3 in non-canonical sites) exceeded this threshold, suggesting a probable effect on RNA splicing. Notably, variants with SpliceAI scores below 0.25 (n = 11) were also experimentally evaluated to validate and refine in silico predictions. Minigene assay The minigene assay revealed aberrant splicing events in 9 out of 20 variants tested ( Fig.S5 ). Observed splicing alterations included exon skipping (n = 4), intron retention (n = 3), and the generation of multiple aberrant transcript isoforms (n = 2), frequently involving the activation of cryptic splice sites within exonic regions (Fig. 1 ). Detailed description of variants effect is provided in Supplementary Materials . RNA analysis from cardiac tissue A cardiac tissue sample obtained during heart transplantation from a proband carrying the DSP c.597G > A variant was available in our institutional archive and used for RNA extraction to validate the minigene assay findings. The extracted RNA was reverse-transcribed, and PCR amplification was performed using primers specifically targeting the exon of interest. Consistent with the minigene splicing assay, the DSP c.597G > A variant generated two distinct transcripts, one of which lacked exon 4. These results reinforce the reliability of the minigene system and underscores its value as a functional assay when patient-derived cardiac samples are not accessible ( Fig.S6 ). Correlation between Percentage Splicing Alteration vs SpliceAI Correlation analysis between SpliceAI-prediced scores and experimentally derived PSA values demonstrated a strong concordance (R 2 = 0.86), supporting the predictive accuracy of SpliceAI in estimating splicing-disruptive potential. Most variants displayed consistent trends between predicted and observed effects, while three variants ( DSP c.273 + 5G > A, DSP c. 597G > A, FLNC c.3790G > A) showed discrepancies (Fig. 2 ). Segregation analysis The cosegregation analysis was feasible in two families with more than two relatives carrying the variants and displaying the clinical phenotype (Fig. 3 ). In Family A, where proband (III-7) carries the FLNC c.5398 + 1G > T resulting in complete exon 32 skipping, the variants co-segregates with the disease phenotype in eight additional relatives, supporting its role in disease pathogenesis. In Family B, where proband (III-5) carries the DSP c.939 + 1G > A leading to intron 7 retention, the variants co-segregates with the disease phenotype in four affected relatives. Accordingly, in our cohort, only these two families provide genetic evidence consistent with ACMG/AMP guidelines, justifying the application of the PP1 criterion, at either supporting or moderate strength, depending on the number of informative meioses. Reclassification Five of the twenty (variants 1 to 5; Table 2 ) analyzed variants were used for validation purposes: two previously classified as Pathogenic (P), two as Likely Pathogenic (LP), and one as Benign (B). Variants initially categorized as P and B served as reference controls in the in vitro assays, providing a direct comparison to in vivo observations. The in vitro splicing results were fully concordant with the prior classifications, thereby confirming the accuracy of the functional assessment. Notably, the pathogenic variants demonstrated clear splicing disruptions consistent with prior evidence obtained by an independent laboratory using complementary approaches, including RNA analysis from patient-derived tissues. For one of the six control variants ( FLNC c.5398 + 1G > T), integration of in vitro minigene assay data enabled refinement of the classification from LP to P. This upgrade was justified by combining the in vitro evidence supporting a null effect on splicing (PVS1 criterion already applied) with segregation data fulfilling the PP1_Strong criterion. Moreover, the minigene assay provided decisive functional evidence for the reclassification of 15 out of the 20 variants initially reported as VUS. Application of the ACMG/AMP criteria led to the reclassification of 5 variants as LP through PVS1, and 10 variants as LB based on the absence of detectable splicing alterations and the application of BP7. Overall, this study enabled the reclassification of 16 variants out of 20 variants (80%) located in genes associated with ACM, highlighting the pivotal contribution of functional in vitro assays in resolving variant interpretation and improving clinical genetic diagnosis. Discussion Recent studies emphasized the emerging role of RNA analysis in inherited cardiomyopathies,[ 15 – 19 ] yet an ACM-focused multi-genic splicing analysis is still missing. In silico tools have traditionally shown modest accuracy in predicting these events.[ 27 – 29 ] Although integrating machine-learning approaches with large-scale RNA-seq datasets has boosted predictive accuracy, the interpretation of non-canonical splice-site variants remains indeterminate.[ 30 , 31 ] Our study provides a systematic evaluation of splice-altering variants across key-ACM genes, identified in a cohort of 200 Italian probands, integrating computational predictions and functional assays. By combining SpliceAI-based prediction with minigene experimental validation, we demonstrated that splicing-disruptive variants, including those located outside canonical GT/AG dinucleotides, contribute substantially to ACM pathogenesis, a mechanism that is often under recognized in clinical diagnostics. This reinforces the concept that splicing disruption represents a major contributor to the allelic spectrum of ACM, alongside more traditionally recognized variant types such as missense or truncating. Predictive tools and functional validation SpliceAI predictions correlated strongly with experimental results, confirming its utility as a first-line prioritization tool for potential splice-altering variants. The quantitative agreement between SpliceAI scores and PSA metrics (R 2 = 0.86) validates the predictive framework of deep-learning models trained on large transcriptomic datasets. However, three outlier variants (15%) revealed limited accuracy in estimating the relative abundance of aberrant transcripts versus the WT isoform. These discrepancies suggest that, although SpliceAI reliably estimates the likelihood of splice site perturbation, it does not fully capture the complex quantitative dynamics underlying alternative splicing outcomes. Consequently, these findings emphasize the necessity of integrating complementary experimental approaches, such as minigene assays or RNA sequencing, to achieve a comprehensive and quantitative characterization of variant-induced splicing alterations.[ 20 , 21 ] Tissue-specific transcript context An additional consideration emerging from our analysis is the importance of interpreting splicing defects within the tissue-specific transcriptome. The biological relevance of an aberrant splicing event depends not only on the presence of an altered transcript but also on whether the affected exon is actually incorporated into the major isoforms expressed in the myocardium. GTEx shows that essentially all exons of the analyzed genes are represented in the predominant myocardial transcripts, with one notable exception: PKP2 exon 6, which is absent in major heart isoforms. Thus, combining GTEx exon usage with our functional data helps distinguish critical, constitutively expressed exons, where splicing disruption is most likely pathogenic, from exons with limited cardiac relevance, strengthening evidence-based variant interpretation. Genetic burden The burden analysis revealed a significant enrichment of rare splice-altering VUS in DSC2, DSG2, DSP , and FLNC among ACM cases compared with gnomAD controls, mirroring previous observations in TTN -associated cardiomyopathy.[ 16 ] This supports the notion that splice-disrupting events are not incidental but represent a mechanistically relevant class of ACM variants. In contrast, PKP2 splice VUS were not enriched in cases, with one ACM allele compared to 140 control alleles. Although the calculated OR (1.88) suggests no meaningful difference, the wide CI (95% CI 0.26–13.44) reflects the very limited number of PKP2 splice variants in our cohort. The small sample size severely reduces statistical power and prevents meaningful burden inference for PKP2 . Thus, the absence of enrichment in this gene should be interpreted cautiously, as a true signal, if present, may be obscured by insufficient allele counts. Clinical and translational implications By applying the ACMG/AMP framework refined by ClinGen SVI,[ 25 ] we leveraged minigene results to reclassify several variants previously labelled as VUS. Functional confirmation of splicing defects enabled reclassification of 80% of previously uncertain variants, shifting them toward LP or LB categories: variants demonstrating strong functional evidence of splicing disruption warrant pathogenic classification (PVS1), particularly when supported by LoF mechanisms and familial segregation (PP1). Conversely, absence of any detectable splicing alteration supports a benign interpretation (BP7). The ability to reclassify a substantial proportion of variants from VUS to (likely) pathogenic or (likely) benign has profound clinical consequences. Such reclassification does not merely refine variant labels, it directly shapes genetic counseling, guides cascade testing, and informs risk stratification for entire families. By providing concrete, experimentally derived evidence, this study illustrates how functional assays transform ambiguous findings into decisive, actionable information for clinicians and patients alike. This impact is particularly striking for non-canonical splice variants, a category notoriously difficult to interpret due to the absence of supporting evidence from segregation, population frequency, or protein-level analyses. Without functional data, these variants are often condemned to remain VUS indefinitely, limiting their clinical utility. Our results clearly show that a systematic evaluation of these variants, using robust splicing assays, can substantially boost the diagnostic yield of ACM genetic testing, unlocking clinically relevant answers that would otherwise remain inaccessible. In this context, our study exemplifies how functional assays can be powerfully integrated into the ACMG/AMP and ClinGen SVI frameworks. Rather than relying solely on indirect inference, variant interpretation becomes grounded in biological reality. This approach effectively bridges the gap between genomic uncertainty and clinical decision-making, ensuring that patients and their families benefit from the most accurate and informative genetic assessments possible. Limitations Minigene assays remain artificial systems that cannot fully replicate tissue-specific regulatory contexts or three-dimensional chromatin architecture. Although validation in cardiac tissue for one variant confirmed assay fidelity, access to myocardial samples is limited. The burden analysis was performed on a relatively limited cohort, and comparison relied on gnomAD controls rather than ancestry-matched screened cardiomyopathy-free individuals. This may introduce population stratification bias and limits the precision of OR estimates. Finally, our experimental design focused on variants located within canonical splice junctions and near-exon intronic regions included in minigene constructs, while deep intronic variants, capable of pseudoexon activation, were not systematically explored. Conclusions Our study highlights the hidden but significant contribution of splicing disruption to the genetic architecture of ACM. The integration of SpliceAI predictions with functional validation not only refines variant interpretation but also enhances diagnostic precision in inherited cardiomyopathies. By systematically uncovering splicing defects in both canonical and non-canonical regions, we provide a framework for improving variant classification and patient care within the era of precision genomic medicine. Abbreviations • ACM Arrhythmogenic Cardiomyopathy • ACMG American College of Medical Genetics and Genomics • AG Acceptor Gain • AL Acceptor Loss • AMP Association for Molecular Pathology • B Benign • CI Confidence Interval • ClinGen Clinical Genome Resource • DG Donor Gain • DL Donor Loss • DSC2 Desmocollin-2 • DSG2 Desmoglein-2 • DSP Desmoplakin • FLNC Filamin-C • gnomAD Genome Aggregation Database • GTEx Genotype-Tissue Expression • HEK-293T Human Embryonic Kidney 293T cells • LB Likely Benign • LoF Loss of Function • LP Likely Pathogenic • MAF Minor Allele Frequency • MUT Mutant • OR Odds Ratio • P Pathogenic • PKP2 Plakophilin-2 • PMID PubMed Identifier • PSA Percent Splicing Alteration • PSI Percent Spliced-In • PVS1 Pathogenic Very Strong • RT-PCR Reverse Transcription Polymerase Chain Reaction • SAVs Splice-Altering Variants • SD Splice Donor • SA Splice Acceptor • SVI Sequence Variant Interpretation • VUS Variant(s) of Uncertain Significance • WES Whole-Exome Sequencing • WT Wild Type Declarations Funding This work was supported by the Italian Ministry of University and Research (MUR) (PRIN grant 20229FE439 - CUP C53D23004670006 " The mechanistic link between genetic substrate and immune reactions in inflammatory cardiomyopathies ". Rome; the Registry for Cardio-cerebro-vascular Pathology, Veneto region, Venice. RC and MBM are Postdoc Fellows supported by the Italian Ministry of Health, PNRR Next-Generation EU grant PNRR-MR1-2022-12376614. GT is a PhD student supported by Italian Ministry of Health DM 630 (PNRR). Ethical Approval All patients provided written informed consent, in accordance with the protocol approved by the ethics committee, including for publication of pedigree charts, of the Azienda Ospedale Università Padova. All investigations were carried out according to the Declaration of Helsinki. Approval Committee Comitato Etico per la Sperimentazione Clinica. Azienda Ospedale Università Padova. Data and materials availability All raw data are available upon request to the corresponding author. Competing Interests All other authors declare no competing interest. Author contributions Conceptualization: R.C, G.T.; Formal analysis, R.C, G.T; Investigation: R.C, G.T, S.P.; Methodology: R.C, G.T, S.P.; Supervision: K.P.; Writing original draft: R.C, G.T.; Writing review & editing, R.C, G.T, S.P., F.DZ., M.C., M.BM., C.B. and K.P. Funding acquisition: K.P., C.B. All authors have read and agreed to the published version of the manuscript. References Anna A, Monika G. Splicing mutations in human genetic disorders: examples, detection, and confirmation. J Appl Genet. 2018;59:253–68. Scotti MM, Swanson MS. RNA mis-splicing in disease. Nat Rev Genet. 2016;17:19–32. Lord J, Gallone G, Short PJ, McRae JF, Ironfield H, Wynn EH, et al. Pathogenicity and selective constraint on variation near splice sites. Genome Res. 2019;29:159–70. Jaganathan K, Kyriazopoulou Panagiotopoulou S, McRae JF, Darbandi SF, Knowles D, Li YI, et al. Predicting Splicing from Primary Sequence with Deep Learning. Cell. 2019;176:535–e548524. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405–24. Gloss BS, Dinger ME. Realizing the significance of noncoding functionality in clinical genomics. Exp Mol Med. 2018;50:1–8. French JD, Edwards SL. The Role of Noncoding Variants in Heritable Disease. Trends Genet. 2020;36:880–91. Christiaans I, Mook ORF, Alders M, Bikker H, Lekanne Dit Deprez RH. Large next-generation sequencing gene panels in genetic heart disease: challenges in clinical practice. Neth Heart J. 2019;27:299–303. van den Hoogenhof MMG, Beqqali A, Amin AS, van der Made I, Aufiero S, Khan MAF, et al. RBM20 Mutations Induce an Arrhythmogenic Dilated Cardiomyopathy Related to Disturbed Calcium Handling. Circulation. 2018;138:1330–42. Vanichkina DP, Schmitz U, Wong JJ, Rasko JEJ. Challenges in defining the role of intron retention in normal biology and disease. Semin Cell Dev Biol. 2018;75:40–9. Wang Y, Wang Z. Systematical identification of splicing regulatory cis-elements and cognate trans-factors. Methods. 2014;65:350–8. Pilichou K, Thiene G, Bauce B, Rigato I, Lazzarini E, Migliore F, et al. Arrhythmogenic cardiomyopathy. Orphanet J Rare Dis. 2016;11:33. James CA, Jongbloed JDH, Hershberger RE, Morales A, Judge DP, Syrris P, et al. International Evidence Based Reappraisal of Genes Associated With Arrhythmogenic Right Ventricular Cardiomyopathy Using the Clinical Genome Resource Framework. Circ Genom Precis Med. 2021;14:e003273. Bueno Marinas M, Cason M, Bariani R, Celeghin R, De Gaspari M, Pinci S et al. A Comprehensive Analysis of Non-Desmosomal Rare Genetic Variants in Arrhythmogenic Cardiomyopathy: Integrating in Padua Cohort Literature-Derived Data. Int J Mol Sci 2024, 25. O'Neill MJ, Wada Y, Hall LD, Mitchell DW, Glazer AM, Roden DM. Functional Assays Reclassify Suspected Splice-Altering Variants of Uncertain Significance in Mendelian Channelopathies. Circ Genom Precis Med. 2022;15:e003782. Patel PN, Ito K, Willcox JAL, Haghighi A, Jang MY, Gorham JM, et al. Contribution of Noncanonical Splice Variants to TTN Truncating Variant Cardiomyopathy. Circ Genom Precis Med. 2021;14:e003389. Singer ES, Ingles J, Semsarian C, Bagnall RD. Key Value of RNA Analysis of MYBPC3 Splice-Site Variants in Hypertrophic Cardiomyopathy. Circ Genom Precis Med. 2019;12:e002368. Ito K, Patel PN, Gorham JM, McDonough B, DePalma SR, Adler EE, et al. Identification of pathogenic gene mutations in LMNA and MYBPC3 that alter RNA splicing. Proc Natl Acad Sci U S A. 2017;114:7689–94. Singer ES, Crowe J, Holliday M, Isbister JC, Lal S, Nowak N, et al. The burden of splice-disrupting variants in inherited heart disease and unexplained sudden cardiac death. NPJ Genom Med. 2023;8:29. Wai HA, Lord J, Lyon M, Gunning A, Kelly H, Cibin P, et al. Blood RNA analysis can increase clinical diagnostic rate and resolve variants of uncertain significance. Genet Med. 2020;22:1005–14. Raponi M, Smith LD, Silipo M, Stuani C, Buratti E, Baralle D. BRCA1 exon 11 a model of long exon splicing regulation. RNA Biol. 2014;11:351–9. Brnich SE, Abou Tayoun AN, Couch FJ, Cutting GR, Greenblatt MS, Heinen CD, et al. Recommendations for application of the functional evidence PS3/BS3 criterion using the ACMG/AMP sequence variant interpretation framework. Genome Med. 2019;12:3. Corrado D, Anastasakis A, Basso C, Bauce B, Blomstrom-Lundqvist C, Bucciarelli-Ducci C, et al. Proposed diagnostic criteria for arrhythmogenic cardiomyopathy: European Task Force consensus report. Int J Cardiol. 2024;395:131447. Reuter P, Walter M, Kohl S, Weisschuh N. Systematic analysis of CNGA3 splice variants identifies different mechanisms of aberrant splicing. Sci Rep. 2023;13:2896. Walker LC, Hoya M, Wiggins GAR, Lindy A, Vincent LM, Parsons MT, et al. Using the ACMG/AMP framework to capture evidence related to predicted and observed impact on splicing: Recommendations from the ClinGen SVI Splicing Subgroup. Am J Hum Genet. 2023;110:1046–67. Jarvik GP, Browning BL. Consideration of Cosegregation in the Pathogenicity Classification of Genomic Variants. Am J Hum Genet. 2016;98:1077–81. Jian X, Boerwinkle E, Liu X. In silico tools for splicing defect prediction: a survey from the viewpoint of end users. Genet Med. 2014;16:497–503. Desmet FO, Hamroun D, Lalande M, Collod-Beroud G, Claustres M, Beroud C. Human Splicing Finder: an online bioinformatics tool to predict splicing signals. Nucleic Acids Res. 2009;37:e67. Houdayer C, Caux-Moncoutier V, Krieger S, Barrois M, Bonnet F, Bourdon V, et al. Guidelines for splicing analysis in molecular diagnosis derived from a set of 327 combined in silico/in vitro studies on BRCA1 and BRCA2 variants. Hum Mutat. 2012;33:1228–38. Riolo G, Cantara S, Ricci C. What's Wrong in a Jump? Prediction and Validation of Splice Site Variants. Methods Protoc 2021, 4. Riepe TV, Khan M, Roosing S, Cremers FPM, t Hoen PAC. Benchmarking deep learning splice prediction tools using functional splice assays. Hum Mutat. 2021;42:799–810. Tables Tables are available in the Supplementary Files section. Supplementary Files SupplementaryMaterials.docx Tab.1.xlsx Tab.1 | Splice-site variants identified in a cohort of 200 ACM patients. Twenty splice-site variants were identified across five genes most commonly associated with ACM: DSC2, DSG2, DSP, FLNC, and PKP2. The figure reports for each variant the corresponding gene, its position within the cDNA and annotated with the corresponding intron/exon junction, the Minor Allele Frequency (MAF), the conservation score of the affected nucleotides (phyloP100) and any prior literature references (PMID: PubMed ID). Variants identified in ACM patients were analyzed using SpliceAI to assess their potential impact on splicing. Effects are categorized as acceptor gain (AG), acceptor loss (AL), donor gain (DG), and donor loss (DL). Predicted probabilities of splice site gain or loss, and the overall SpliceAI score, defined as the maximum of the four delta scores. Tab.2.xlsx Tab.2 | Reclassification of identified variants according to ClinGen SVI recommendations. Functional studies revealed aberrant splicing in 9 variants (1 canonical, 2 missense and 2 non-canonical), leading to application of the PVS1 in accordance with ClinGen SVI guidelines. Co-segregation data enabled application of PP1 for two canonical variants. PM2, PP5 and BP6 were applied not over supporting. Colors correspond to ACMG/AMP categories: red, pathogenic (P); orange, likely pathogenic (LP); yellow, variant of uncertain significance (VUS); light green, likely benign (LB); green, benign (B). Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 28 Apr, 2026 Reviewers invited by journal 29 Mar, 2026 Editor assigned by journal 06 Mar, 2026 First submitted to journal 04 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9022128","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":614091624,"identity":"c86bcb69-15ee-4eeb-913a-893b1f3fb75b","order_by":0,"name":"Rudy Celeghin","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Rudy","middleName":"","lastName":"Celeghin","suffix":""},{"id":614091625,"identity":"60bdcb3c-244e-47dd-bb32-3558d42c01a8","order_by":1,"name":"Giulia Tosato","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Giulia","middleName":"","lastName":"Tosato","suffix":""},{"id":614091626,"identity":"d6cb63d0-9c64-4898-8694-931509355074","order_by":2,"name":"Serena Pinci","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Serena","middleName":"","lastName":"Pinci","suffix":""},{"id":614091627,"identity":"2bc6dfd2-8a27-4221-8349-04dd2c31451e","order_by":3,"name":"Francesca Dalla Zanna","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Francesca","middleName":"Dalla","lastName":"Zanna","suffix":""},{"id":614091628,"identity":"1862b81c-371b-4269-979b-a030de51c084","order_by":4,"name":"Maria Bueno Marinas","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"Bueno","lastName":"Marinas","suffix":""},{"id":614091629,"identity":"ecabe3af-337b-4c11-bddd-f366d59cd7a3","order_by":5,"name":"Marco Cason","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Marco","middleName":"","lastName":"Cason","suffix":""},{"id":614091630,"identity":"c679dbfd-a7f8-4b95-af3b-3bb2363ceebf","order_by":6,"name":"Cristina Basso","email":"","orcid":"","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":false,"prefix":"","firstName":"Cristina","middleName":"","lastName":"Basso","suffix":""},{"id":614091631,"identity":"af4d4598-96f3-439e-95bd-800b2ff57f43","order_by":7,"name":"Kalliopi Pilichou","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABA0lEQVRIie2RsWrDMBCGf2NQF5GsMg7OKxgCyZAhryJTaJfMJUMGhUD9CnmMTCGjgqBdDF0b4iFaMvUB0q1SXEOaWp4z6EPDf4KPu+MAj+cOYQgXkgMUCGyQwIMAOEIQpxKIWhGVQuWv4nAuSl3Iy2Pc5hCuNlG+E/I4K3uIlybMyyTdfw2PejtGpy8alZhmZp7iRNEzLn87DdLDdJRmxbNzsARGyV4VBbNBqGx9mA6Z+XErXX2r7It2JWb/unzSdiVa6WoXUu2iBlHx9MLMLpQQ3qiwj8edPs/KSTfOtT7PVdJ5V5voeztO+kvZ3KZCgrDrOhD2uK1Ie9S/isfj8XhqfgAA92paExF9JQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-3646-0669","institution":"University of Padua: Universita degli Studi di Padova","correspondingAuthor":true,"prefix":"","firstName":"Kalliopi","middleName":"","lastName":"Pilichou","suffix":""}],"badges":[],"createdAt":"2026-03-03 16:07:49","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9022128/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9022128/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105982151,"identity":"283936d3-1448-4226-b5e8-0fe0b453feac","added_by":"auto","created_at":"2026-04-02 06:57:44","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":553448,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional characterization of variant-induced splicing alterations and resulting mRNA isoforms.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSchematic representation of the splicing assays performed using minigene constructs. On the left of each panel, the minigene are shown, containing the pSPL3 exon (splice donor (SD) and splice acceptor (SA)), flanking the gene-specific exon of interest and its adjacent intronic regions carrying the tested variant. These constructs reproduce the endogenous pre-mRNA splicing context. On the right, Sanger sequencing of RT-PCR products from transfected HEK-293T cells reveals alternative transcripts. These results provide functional evidence that the tested variants disrupt normal splice site recognition, resulting in exon skipping (A), intron retention (B) or the production of multiple aberrant transcript (C). These findings support the pathogenic relevance of the investigated splicing defects.\u003c/p\u003e","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/901e1844a94f99b48bf83282.jpg"},{"id":105982153,"identity":"8f0afe0b-f488-4470-a0ed-b84b83c36451","added_by":"auto","created_at":"2026-04-02 06:57:45","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":461410,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional effects and protein outcomes of splice site variants in ACM genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe RNA splicing outcome are detailed alongside their predicted impact on protein structure using SpliceAI while the PSA values quantify the relative abundance of aberrant splice isoforms. Variants were assessed for their potential to disrupt normal splicing using SpliceAI predictions and experimental validated through minigene assays. R\u003csup\u003e2\u003c/sup\u003e= coefficient of determination for linear regression.\u003c/p\u003e","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/7879e4a00b07b54a5b45960d.jpg"},{"id":105982141,"identity":"f8e0cb59-569c-48e6-aeb2-2690dce75656","added_by":"auto","created_at":"2026-04-02 06:57:40","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":246762,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSegregation analysis of splice site variants in two families affected by ACM.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePedigree diagrams of two unrelated families, each with a proband (indicated by an arrow) affected by ACM and carrying a splice site variant. In both pedigrees, squares and cycle represent respectively and females. Filled symbols denote clinically affected individuals, while open symbols indicate unaffected individuals. A diagonal line indicates deceased individuals. In Family A, the FLNC c.5398+1G\u0026gt;T variant (highlighted by green dots) segregates in eight relatives and in Family B, the DSP c.939+1G\u0026gt;A variant (highlighted by orange dots) segregates in four relatives.\u003c/p\u003e","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/34e6e73d98c506279dac9e09.jpg"},{"id":106093476,"identity":"794fff5b-310f-41e0-99ac-a85fbb3cdb52","added_by":"auto","created_at":"2026-04-03 11:37:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2271455,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/0a43713b-f5f9-43a8-b9c6-9360bbeddd4e.pdf"},{"id":105982144,"identity":"4fe94cd0-8d6f-4651-8932-a53977dfed93","added_by":"auto","created_at":"2026-04-02 06:57:41","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1454031,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/185a8ae412fcb93ba20bf417.docx"},{"id":105982154,"identity":"77095d2c-6d0d-475e-aa34-a53f5d7c9f9f","added_by":"auto","created_at":"2026-04-02 06:57:45","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":12593,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTab.1 | Splice-site variants identified in a cohort of 200 ACM patients.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTwenty splice-site variants were identified across five genes most commonly associated with ACM: DSC2, DSG2, DSP, FLNC, and PKP2. The figure reports for each variant the corresponding gene, its position within the cDNA and annotated with the corresponding intron/exon junction, the Minor Allele Frequency (MAF), the conservation score of the affected nucleotides (phyloP100) and any prior literature references (PMID: PubMed ID). Variants identified in ACM patients were analyzed using SpliceAI to assess their potential impact on splicing. Effects are categorized as acceptor gain (AG), acceptor loss (AL), donor gain (DG), and donor loss (DL). Predicted probabilities of splice site gain or loss, and the overall SpliceAI score, defined as the maximum of the four delta scores.\u003c/p\u003e","description":"","filename":"Tab.1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/fb457e5edafe136a3504f00c.xlsx"},{"id":105982149,"identity":"03230af7-432e-4ca9-9c0b-ed78ec4b9537","added_by":"auto","created_at":"2026-04-02 06:57:44","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":11021,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTab.2 | Reclassification of identified variants according to ClinGen SVI recommendations.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFunctional studies revealed aberrant splicing in 9 variants (1 canonical, 2 missense and 2 non-canonical), leading to application of the PVS1 in accordance with ClinGen SVI guidelines. Co-segregation data enabled application of PP1 for two canonical variants. PM2, PP5 and BP6 were applied not over supporting. Colors correspond to ACMG/AMP categories: red, pathogenic (P); orange, likely pathogenic (LP); yellow, variant of uncertain significance (VUS); light green, likely benign (LB); green, benign (B).\u003c/p\u003e","description":"","filename":"Tab.2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9022128/v1/70ac93f1c787447ac441b9ab.xlsx"}],"financialInterests":"","formattedTitle":"Tackling Non-Canonical Splicing in Arrhythmogenic Cardiomyopathy to Reduce the Uncertain Significance Variants Burden","fulltext":[{"header":"Key points","content":"\u003cp\u003e\u0026clubs; Non-canonical splice-altering variants contribute significantly to the genetic architecture of arrhythmogenic cardiomyopathy (ACM) and are frequently missed or misclassified by standard DNA-based diagnostic approaches.\u003c/p\u003e\u003cp\u003e\u0026clubs; Functional splicing assays resolve variant uncertainty, enabling reclassification of 80% of previously uncertain variants and substantially improving diagnostic yield across major ACM-associated genes.\u003c/p\u003e\u003cp\u003e\u0026clubs; Integration of SpliceAI prediction, minigene assays, and tissue-specific exon usage (GTEx) provides a robust, evidence-based framework for interpreting splicing variants in ACM according to ACMG/ClinGen SVI guidelines.\u003c/p\u003e\u003cp\u003e\u0026clubs; Aberrant splicing events were experimentally confirmed in 45% of tested variants, including synonymous and non-canonical changes, reinforcing the need for transcript-level evaluation in inherited cardiomyopathies.\u003c/p\u003e\u003cp\u003e\u0026clubs; Tissue-specific exon expression analysis refines pathogenicity assessment, distinguishing variants that disrupt clinically relevant cardiac isoforms from those unlikely to affect myocardial biology.\u003c/p\u003e"},{"header":"Background","content":"\u003cp\u003eVariants that alter RNA splicing represent a substantial and frequently underestimated cause of monogenic disease, including inherited cardiomyopathies. While clinical genetic testing has traditionally focused on protein-altering variants, splicing defects account for an estimated 10% of pathogenic findings and often escape detection or correct interpretation.[\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e] Canonical splice-site changes are typically well captured by current American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) guidelines,[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] yet variants affecting non-canonical positions, such as deep intronic regions, last-nucleotide exonic substitutions, or regulatory cis-elements, remain challenging to classify.[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] These elusive alterations may generate exon skipping, intron retention, or pseudoexon activation, ultimately producing aberrant transcripts with potentially deleterious functional consequences.[\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eArrhythmogenic cardiomyopathy (ACM) is a heritable myocardial disorder characterized by fibrofatty replacement, ventricular arrhythmias, and an elevated risk of sudden cardiac death.[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] Advances in next-generation sequencing have broadened the genetic landscape of ACM, extending beyond the desmosomal complex to include an expanding set of non-desmosomal genes.[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003csup\u003e,\u003c/sup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] Despite this progress, the overall diagnostic yield remains constrained by a persistent and substantial burden of variants of uncertain significance (VUS), which continues to limit definitive molecular diagnosis and clinical translation. Among these, non-canonical splice-altering variants are likely under-recognized contributors to disease, owing both to insufficient predictive power of computational tools and to the limited availability of transcript-based assays in routine diagnostics.\u003c/p\u003e \u003cp\u003eRecent functional studies across inherited cardiac conditions have revealed that such non-canonical splicing defects represent a non-negligible fraction of pathogenic variants,[\u003cspan additionalcitationids=\"CR16 CR17 CR18\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] reaffirming the need for systematic transcript-level evaluation. Tools such as SpliceAI have improved the prediction of splicing disruption, but \u003cem\u003ein vitro\u003c/em\u003e assays, including reverse-transcription PCR (RT-PCR)[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] and minigene systems,[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] are recommended by ClinGen Sequence Variant Interpretation (SVI) to confirm the biological impact of candidate variants.[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eUsing minigene assays, complemented by cardiac tissue RNA analysis, we aimed to quantify non-canonical splicing alterations in ACM-associated genes, evaluating the concordance between predicted and observed effects, and refine variant classification under ACMG/AMP and ClinGen SVI frameworks. By uncovering splicing defects that would otherwise remain hidden behind ambiguous sequencing results, our work seeks to enhance diagnostic precision and reduce the persistent VUS burden in ACM genomic medicine.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy population\u003c/h2\u003e \u003cp\u003eA retrospective analysis was conducted on a cohort of patients diagnosed with ACM, aiming to identify splicing site variants in genes primarily associated with the disease. Patients were enrolled through the Cardiovascular Pathology Unit (UOC Patologia Cardiovascolare) of the University Hospital of Padua and the Regional Registry for Cardio-Cerebro-Vascular Diseases. Clinical diagnosis of ACM was established based on current guidelines.[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] Ethical approval and informed consent were obtained in accordance with institutional guidelines.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGenetic screening\u003c/h3\u003e\n\u003cp\u003eGenetic screening of the enrolled patients was previously performed by the Cardiovascular Genetics Laboratory at the Cardiovascular Pathology Unit, University Hospital of Padua. Whole-exome sequencing (WES) was initially carried out on all patients, and identified variants were subsequently confirmed by Sanger sequencing.\u003c/p\u003e\n\u003ch3\u003ePopulation frequency and expression datasets\u003c/h3\u003e\n\u003cp\u003ePopulation frequencies were extracted from the Genome Aggregation Database (gnomAD v2.1, controls only) in order to perform the Burden test. Gene- and exon-specific cardiac expression profiles were retrieved from the Genotype-Tissue Expression (GTEx) database to assess tissue-relevant isoform usage.\u003c/p\u003e\n\u003ch3\u003eSplicing effect prediction\u003c/h3\u003e\n\u003cp\u003eThe predicted impact of each variant on RNA splicing was evaluated using SpliceAI (GRCh38). For each variant, delta scores for acceptor gain/loss and donor gain/loss were retrieved. Following ClinGen SVI recommendations, variants with a SpliceAI score\u0026thinsp;\u0026gt;\u0026thinsp;0.25 were considered at risk for splicing disruption and prioritized for functional validation.[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/p\u003e\n\u003ch3\u003eMinigene constructs\u003c/h3\u003e\n\u003cp\u003eCandidate variants were tested using a modified version of pSPL3-based minigene constructs [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] containing the exon of interest and flanking intronic regions. Wild-type (WT) and mutant (MUT) inserts were cloned into the vector and transfected into Human embryonic kidney 293 cells (HEK-293T). After RNA extraction and reverse transcription, splicing products were analyzed by Sanger sequencing (\u003cb\u003eTab.S1\u003c/b\u003e, \u003cb\u003eFig.\u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e, \u003cb\u003eFig.\u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e\u003c/b\u003e).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCardiac tissue RNA analysis\u003c/h2\u003e \u003cp\u003eMyocardial RNA obtained during heart transplantation from an affected proband was available for direct transcript analysis by RT-PCR.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eQuantification of splicing alteration\u003c/h3\u003e\n\u003cp\u003eSplicing efficiency was quantified using the Percent Splicing Alteration (PSA) metric, defined as the proportion of transcripts exhibiting a specific alternative splicing event relative to the total transcript pool. In minigene assays, PSA was calculated as follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\text{P}\\text{S}\\text{A}=\\frac{\\text{P}\\text{S}\\text{A}}{\\text{P}\\text{S}\\text{A}+\\text{P}\\text{S}\\text{I}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere PSA and the Percentage Spliced-In (PSI) represent the relative intensities of bands corresponding to alternatively and canonically spliced transcripts, respectively, calculated from TapeStation peak intensities.\u003c/p\u003e\n\u003ch3\u003eVariant re-classification\u003c/h3\u003e\n\u003cp\u003eA comprehensive evaluation of each variant\u0026rsquo;s pathogenic potential was performed using three main bioinformatics platforms: Franklin (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://franklin.genoox.com\u003c/span\u003e\u003cspan address=\"https://franklin.genoox.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), VarSome (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://varsome.com\u003c/span\u003e\u003cspan address=\"https://varsome.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and Alamut (Sophia Genetics, Switzerland), which integrate a wide range of public and proprietary databases, as well as multiple \u003cem\u003ein silico\u003c/em\u003e predictors.\u003c/p\u003e \u003cp\u003eVariant interpretation followed ACMG/AMP guidelines as refined by the ClinGen SVI Splicing Subgroup:[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] PS3/BS3 and PP3/BP4 were not applied because functional assays and \u003cem\u003ein silico\u003c/em\u003e splicing predictions were already incorporated through the PVS1/BP7 framework.\u003c/p\u003e \u003cp\u003ePVS1 strength was assigned based on predicted effects of the splicing alteration on the protein and evidence of loss of function (LoF). BP7_Strong was applied to synonymous or intronic variants without experimentally detectable splicing impact.\u003c/p\u003e \u003cp\u003eFurthermore, PP1 was included when segregation supported variant-phenotype cosegregation[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] and its strengh was determined based on the probability of segregation as previously proposed.[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eCategorical variables were reported as counts and percentages, while continuous variables were expressed as appropriate. Case-control burden testing was performed by calculating odds ratios (OR) with 95% confidence intervals (CI) and corresponding p-values. Correlation between SpliceAI prediction scores and experimentally derived PSA values was assessed using simple linear regression, and the strength of association was expressed as the coefficient of determination (R\u0026sup2;). A p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e \u003cp\u003eDetailed methods are provided in \u003cb\u003eSupplementary Materials\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eVariant identification and localization\u003c/h2\u003e \u003cp\u003eA comprehensive variant screening was performed in a cohort of 200 probands with a diagnosis of ACM,[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] revealing 20 splicing variants across the following genes: desmoplakin (\u003cem\u003eDSP\u003c/em\u003e NM_004415.4; n\u0026thinsp;=\u0026thinsp;8), desmoglein-2 (\u003cem\u003eDSG2\u003c/em\u003e NM_001943.5; n\u0026thinsp;=\u0026thinsp;4), filamin-C (\u003cem\u003eFLNC\u003c/em\u003e NM_001458.5; n\u0026thinsp;=\u0026thinsp;4), desmocollin-2 (\u003cem\u003eDSC2\u003c/em\u003e NM_024422.6; n\u0026thinsp;=\u0026thinsp;3), and plakofilin-2 (\u003cem\u003ePKP2\u003c/em\u003e NM_001005242.3; n\u0026thinsp;=\u0026thinsp;1) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAmong these variants, two were missense variants, five affected canonical splice sites (\u0026plusmn;\u0026thinsp;1 or 2 base pairs from exon-intron boundaries), while the remaining were located in non-canonical splicing regions (see \u003cb\u003eFig.S3\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eBurden of Splice-Altering VUS Among ACM Cases vs Controls\u003c/h2\u003e \u003cp\u003eTo evaluate whether putative splice-altering variants were enriched in ACM, we performed a burden test comparing their allele frequency in our 200 ACM probands (400 alleles) with that observed in gnomAD v2.1 controls. Across desmosomal genes and \u003cem\u003eFLNC\u003c/em\u003e, rare splice-site VUS were significantly over-represented in cases. For \u003cem\u003eDSC2\u003c/em\u003e, we observed 2 ACM alleles versus 68 in gnomAD, corresponding to an OR of 7.744 (95% CI 1.891\u0026ndash;31.70; p\u0026thinsp;=\u0026thinsp;0.0008). A similar enrichment was found for \u003cem\u003eDSG2\u003c/em\u003e (3 ACM vs 60 control alleles; OR of 12.8, 95% CI 4.00-40.99; p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) and was even more pronounced for \u003cem\u003eDSP\u003c/em\u003e, where 6 ACM alleles and 121 gnomAD alleles yielded an OR of 13.59 (95% CI 5.950\u0026ndash;31.02; p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). \u003cem\u003eFLNC\u003c/em\u003e splice VUS were less frequent (3 ACM vs 236 control alleles) but still significantly enriched (OR of 3.35, 95% CI 1.07\u0026ndash;10.52; p\u0026thinsp;=\u0026thinsp;0.0276). In contrast, the frequency of \u003cem\u003ePKP2\u003c/em\u003e splice VUS (1 ACM vs 140 control alleles) was not different from controls (OR 1.88, 95% CI 0.26\u0026ndash;13.44; p\u0026thinsp;=\u0026thinsp;0.5244), and the wide CI likely reflects the limited sample size and insufficient power to detect a significant burden. Overall, these data indicate a non-random enrichment of rare splice-disrupting variants in \u003cem\u003eDSC2, DSG2, DSP\u003c/em\u003e and \u003cem\u003eFLNC\u003c/em\u003e in ACM, mirroring the increased burden of non-canonical splice variants reported for \u003cem\u003eTTN\u003c/em\u003e in dilated cardiomyopathy (\u003cb\u003eTab.S2\u003c/b\u003e).[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eGenotype-Tissue Expression (GTEx)\u003c/h2\u003e \u003cp\u003eAn additional analysis was carried out using the GTEx Project, which provides detailed information on transcript isoform usage across human tissues, including the heart. Inspection of the splicing profiles highlights that the myocardium expresses a specific subset of isoforms, characterized by the consistent inclusion of certain exons and the systematic skipping of others (\u003cb\u003eFig. S4\u003c/b\u003e). This allows us to determine which exons are actually used in the adult cardiac transcriptome, and therefore which splice-affecting variants are most likely to have functional relevance. Notably, the GTEx data indicate that all exons of the analyzed genes are expressed in heart tissue except for \u003cem\u003ePKP2\u003c/em\u003e exon 6, which is not represented in the predominant cardiac isoforms. The distribution of isoforms thus enables the identification of \u0026ldquo;critical\u0026rdquo; exons whose disrupted splicing may exert a pathogenic effect. These data effectively complement genetic analyses and functional assays by providing essential biological context for interpreting variants that alter normal splicing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eSpliceAI-based splicing impact prediction\u003c/h2\u003e \u003cp\u003e \u003cem\u003eIn silico\u003c/em\u003e splicing impact prediction was performed on all variants using SpliceAI (Table\u0026nbsp;1). According to ClinGen SVI recommendations,[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] a score threshold of \u0026gt;\u0026thinsp;0.25 was set to indicate potential splicing disruption. A total of 9/20 variants (including 2 missense, 5 located in canonical sites and 3 in non-canonical sites) exceeded this threshold, suggesting a probable effect on RNA splicing. Notably, variants with SpliceAI scores below 0.25 (n\u0026thinsp;=\u0026thinsp;11) were also experimentally evaluated to validate and refine \u003cem\u003ein silico\u003c/em\u003e predictions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eMinigene assay\u003c/h2\u003e \u003cp\u003eThe minigene assay revealed aberrant splicing events in 9 out of 20 variants tested (\u003cb\u003eFig.S5\u003c/b\u003e). Observed splicing alterations included exon skipping (n\u0026thinsp;=\u0026thinsp;4), intron retention (n\u0026thinsp;=\u0026thinsp;3), and the generation of multiple aberrant transcript isoforms (n\u0026thinsp;=\u0026thinsp;2), frequently involving the activation of cryptic splice sites within exonic regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Detailed description of variants effect is provided in \u003cb\u003eSupplementary Materials\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eRNA analysis from cardiac tissue\u003c/h2\u003e \u003cp\u003eA cardiac tissue sample obtained during heart transplantation from a proband carrying the \u003cem\u003eDSP\u003c/em\u003e c.597G\u0026thinsp;\u0026gt;\u0026thinsp;A variant was available in our institutional archive and used for RNA extraction to validate the minigene assay findings. The extracted RNA was reverse-transcribed, and PCR amplification was performed using primers specifically targeting the exon of interest. Consistent with the minigene splicing assay, the \u003cem\u003eDSP\u003c/em\u003e c.597G\u0026thinsp;\u0026gt;\u0026thinsp;A variant generated two distinct transcripts, one of which lacked exon 4. These results reinforce the reliability of the minigene system and underscores its value as a functional assay when patient-derived cardiac samples are not accessible (\u003cb\u003eFig.S6\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eCorrelation between Percentage Splicing Alteration vs SpliceAI\u003c/h2\u003e \u003cp\u003eCorrelation analysis between SpliceAI-prediced scores and experimentally derived PSA values demonstrated a strong concordance (R\u003csup\u003e2\u003c/sup\u003e= 0.86), supporting the predictive accuracy of SpliceAI in estimating splicing-disruptive potential. Most variants displayed consistent trends between predicted and observed effects, while three variants (\u003cem\u003eDSP\u003c/em\u003e c.273\u0026thinsp;+\u0026thinsp;5G\u0026thinsp;\u0026gt;\u0026thinsp;A, \u003cem\u003eDSP\u003c/em\u003e c. 597G\u0026thinsp;\u0026gt;\u0026thinsp;A, \u003cem\u003eFLNC\u003c/em\u003e c.3790G\u0026thinsp;\u0026gt;\u0026thinsp;A) showed discrepancies (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eSegregation analysis\u003c/h2\u003e \u003cp\u003eThe cosegregation analysis was feasible in two families with more than two relatives carrying the variants and displaying the clinical phenotype (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In Family A, where proband (III-7) carries the \u003cem\u003eFLNC\u003c/em\u003e c.5398\u0026thinsp;+\u0026thinsp;1G\u0026thinsp;\u0026gt;\u0026thinsp;T resulting in complete exon 32 skipping, the variants co-segregates with the disease phenotype in eight additional relatives, supporting its role in disease pathogenesis. In Family B, where proband (III-5) carries the \u003cem\u003eDSP\u003c/em\u003e c.939\u0026thinsp;+\u0026thinsp;1G\u0026thinsp;\u0026gt;\u0026thinsp;A leading to intron 7 retention, the variants co-segregates with the disease phenotype in four affected relatives.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAccordingly, in our cohort, only these two families provide genetic evidence consistent with ACMG/AMP guidelines, justifying the application of the PP1 criterion, at either supporting or moderate strength, depending on the number of informative meioses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eReclassification\u003c/h2\u003e \u003cp\u003eFive of the twenty (variants 1 to 5; \u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e) analyzed variants were used for validation purposes: two previously classified as Pathogenic (P), two as Likely Pathogenic (LP), and one as Benign (B). Variants initially categorized as P and B served as reference controls in the \u003cem\u003ein vitro\u003c/em\u003e assays, providing a direct comparison to \u003cem\u003ein vivo\u003c/em\u003e observations. The \u003cem\u003ein vitro\u003c/em\u003e splicing results were fully concordant with the prior classifications, thereby confirming the accuracy of the functional assessment. Notably, the pathogenic variants demonstrated clear splicing disruptions consistent with prior evidence obtained by an independent laboratory using complementary approaches, including RNA analysis from patient-derived tissues.\u003c/p\u003e \u003cp\u003eFor one of the six control variants (\u003cem\u003eFLNC\u003c/em\u003e c.5398\u0026thinsp;+\u0026thinsp;1G\u0026thinsp;\u0026gt;\u0026thinsp;T), integration of \u003cem\u003ein vitro\u003c/em\u003e minigene assay data enabled refinement of the classification from LP to P. This upgrade was justified by combining the \u003cem\u003ein vitro\u003c/em\u003e evidence supporting a null effect on splicing (PVS1 criterion already applied) with segregation data fulfilling the PP1_Strong criterion.\u003c/p\u003e \u003cp\u003eMoreover, the minigene assay provided decisive functional evidence for the reclassification of 15 out of the 20 variants initially reported as VUS. Application of the ACMG/AMP criteria led to the reclassification of 5 variants as LP through PVS1, and 10 variants as LB based on the absence of detectable splicing alterations and the application of BP7.\u003c/p\u003e \u003cp\u003eOverall, this study enabled the reclassification of 16 variants out of 20 variants (80%) located in genes associated with ACM, highlighting the pivotal contribution of functional \u003cem\u003ein vitro\u003c/em\u003e assays in resolving variant interpretation and improving clinical genetic diagnosis.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eRecent studies emphasized the emerging role of RNA analysis in inherited cardiomyopathies,[\u003cspan additionalcitationids=\"CR16 CR17 CR18\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] yet an ACM-focused multi-genic splicing analysis is still missing. \u003cem\u003eIn silico\u003c/em\u003e tools have traditionally shown modest accuracy in predicting these events.[\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] Although integrating machine-learning approaches with large-scale RNA-seq datasets has boosted predictive accuracy, the interpretation of non-canonical splice-site variants remains indeterminate.[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eOur study provides a systematic evaluation of splice-altering variants across key-ACM genes, identified in a cohort of 200 Italian probands, integrating computational predictions and functional assays. By combining SpliceAI-based prediction with minigene experimental validation, we demonstrated that splicing-disruptive variants, including those located outside canonical GT/AG dinucleotides, contribute substantially to ACM pathogenesis, a mechanism that is often under recognized in clinical diagnostics. This reinforces the concept that splicing disruption represents a major contributor to the allelic spectrum of ACM, alongside more traditionally recognized variant types such as missense or truncating.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003ePredictive tools and functional validation\u003c/h2\u003e \u003cp\u003eSpliceAI predictions correlated strongly with experimental results, confirming its utility as a first-line prioritization tool for potential splice-altering variants. The quantitative agreement between SpliceAI scores and PSA metrics (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.86) validates the predictive framework of deep-learning models trained on large transcriptomic datasets. However, three outlier variants (15%) revealed limited accuracy in estimating the relative abundance of aberrant transcripts versus the WT isoform. These discrepancies suggest that, although SpliceAI reliably estimates the likelihood of splice site perturbation, it does not fully capture the complex quantitative dynamics underlying alternative splicing outcomes. Consequently, these findings emphasize the necessity of integrating complementary experimental approaches, such as minigene assays or RNA sequencing, to achieve a comprehensive and quantitative characterization of variant-induced splicing alterations.[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eTissue-specific transcript context\u003c/h2\u003e \u003cp\u003eAn additional consideration emerging from our analysis is the importance of interpreting splicing defects within the tissue-specific transcriptome. The biological relevance of an aberrant splicing event depends not only on the presence of an altered transcript but also on whether the affected exon is actually incorporated into the major isoforms expressed in the myocardium. GTEx shows that essentially all exons of the analyzed genes are represented in the predominant myocardial transcripts, with one notable exception: \u003cem\u003ePKP2\u003c/em\u003e exon 6, which is absent in major heart isoforms. Thus, combining GTEx exon usage with our functional data helps distinguish critical, constitutively expressed exons, where splicing disruption is most likely pathogenic, from exons with limited cardiac relevance, strengthening evidence-based variant interpretation.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eGenetic burden\u003c/h2\u003e \u003cp\u003eThe burden analysis revealed a significant enrichment of rare splice-altering VUS in \u003cem\u003eDSC2, DSG2, DSP\u003c/em\u003e, and \u003cem\u003eFLNC\u003c/em\u003e among ACM cases compared with gnomAD controls, mirroring previous observations in \u003cem\u003eTTN\u003c/em\u003e-associated cardiomyopathy.[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] This supports the notion that splice-disrupting events are not incidental but represent a mechanistically relevant class of ACM variants.\u003c/p\u003e \u003cp\u003eIn contrast, \u003cem\u003ePKP2\u003c/em\u003e splice VUS were not enriched in cases, with one ACM allele compared to 140 control alleles. Although the calculated OR (1.88) suggests no meaningful difference, the wide CI (95% CI 0.26\u0026ndash;13.44) reflects the very limited number of \u003cem\u003ePKP2\u003c/em\u003e splice variants in our cohort. The small sample size severely reduces statistical power and prevents meaningful burden inference for \u003cem\u003ePKP2\u003c/em\u003e. Thus, the absence of enrichment in this gene should be interpreted cautiously, as a true signal, if present, may be obscured by insufficient allele counts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eClinical and translational implications\u003c/h2\u003e \u003cp\u003eBy applying the ACMG/AMP framework refined by ClinGen SVI,[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] we leveraged minigene results to reclassify several variants previously labelled as VUS. Functional confirmation of splicing defects enabled reclassification of 80% of previously uncertain variants, shifting them toward LP or LB categories: variants demonstrating strong functional evidence of splicing disruption warrant pathogenic classification (PVS1), particularly when supported by LoF mechanisms and familial segregation (PP1). Conversely, absence of any detectable splicing alteration supports a benign interpretation (BP7).\u003c/p\u003e \u003cp\u003eThe ability to reclassify a substantial proportion of variants from VUS to (likely) pathogenic or (likely) benign has profound clinical consequences. Such reclassification does not merely refine variant labels, it directly shapes genetic counseling, guides cascade testing, and informs risk stratification for entire families. By providing concrete, experimentally derived evidence, this study illustrates how functional assays transform ambiguous findings into decisive, actionable information for clinicians and patients alike.\u003c/p\u003e \u003cp\u003eThis impact is particularly striking for non-canonical splice variants, a category notoriously difficult to interpret due to the absence of supporting evidence from segregation, population frequency, or protein-level analyses. Without functional data, these variants are often condemned to remain VUS indefinitely, limiting their clinical utility. Our results clearly show that a systematic evaluation of these variants, using robust splicing assays, can substantially boost the diagnostic yield of ACM genetic testing, unlocking clinically relevant answers that would otherwise remain inaccessible.\u003c/p\u003e \u003cp\u003eIn this context, our study exemplifies how functional assays can be powerfully integrated into the ACMG/AMP and ClinGen SVI frameworks. Rather than relying solely on indirect inference, variant interpretation becomes grounded in biological reality. This approach effectively bridges the gap between genomic uncertainty and clinical decision-making, ensuring that patients and their families benefit from the most accurate and informative genetic assessments possible.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section3\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eMinigene assays remain artificial systems that cannot fully replicate tissue-specific regulatory contexts or three-dimensional chromatin architecture. Although validation in cardiac tissue for one variant confirmed assay fidelity, access to myocardial samples is limited.\u003c/p\u003e \u003cp\u003eThe burden analysis was performed on a relatively limited cohort, and comparison relied on gnomAD controls rather than ancestry-matched screened cardiomyopathy-free individuals. This may introduce population stratification bias and limits the precision of OR estimates.\u003c/p\u003e \u003cp\u003eFinally, our experimental design focused on variants located within canonical splice junctions and near-exon intronic regions included in minigene constructs, while deep intronic variants, capable of pseudoexon activation, were not systematically explored.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eOur study highlights the hidden but significant contribution of splicing disruption to the genetic architecture of ACM. The integration of SpliceAI predictions with functional validation not only refines variant interpretation but also enhances diagnostic precision in inherited cardiomyopathies. By systematically uncovering splicing defects in both canonical and non-canonical regions, we provide a framework for improving variant classification and patient care within the era of precision genomic medicine.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; ACM\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eArrhythmogenic Cardiomyopathy\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; ACMG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAmerican College of Medical Genetics and Genomics\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; AG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAcceptor Gain\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; AL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAcceptor Loss\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; AMP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAssociation for Molecular Pathology\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; B\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eBenign\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; CI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eConfidence Interval\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; ClinGen\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eClinical Genome Resource\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; DG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDonor Gain\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; DL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDonor Loss\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; DSC2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDesmocollin-2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; DSG2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDesmoglein-2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; DSP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDesmoplakin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; FLNC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eFilamin-C\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; gnomAD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGenome Aggregation Database\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; GTEx\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGenotype-Tissue Expression\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; HEK-293T\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHuman Embryonic Kidney 293T cells\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; LB\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLikely Benign\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; LoF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLoss of Function\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; LP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLikely Pathogenic\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; MAF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMinor Allele Frequency\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; MUT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMutant\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; OR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eOdds Ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; P\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePathogenic\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; PKP2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePlakophilin-2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; PMID\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePubMed Identifier\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; PSA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePercent Splicing Alteration\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; PSI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePercent Spliced-In\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; PVS1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePathogenic Very Strong\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; RT-PCR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eReverse Transcription Polymerase Chain Reaction\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; SAVs\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSplice-Altering Variants\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; SD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSplice Donor\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; SA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSplice Acceptor\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; SVI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSequence Variant Interpretation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; VUS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eVariant(s) of Uncertain Significance\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; WES\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eWhole-Exome Sequencing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003e\u0026bull; WT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eWild Type\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":" \u003cp\u003e \u003cb\u003eFunding\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThis work was supported by the Italian Ministry of University and Research (MUR) (PRIN grant 20229FE439 - CUP C53D23004670006 \"\u003cem\u003eThe mechanistic link between genetic substrate and immune reactions in inflammatory cardiomyopathies\u003c/em\u003e\". Rome; the Registry for Cardio-cerebro-vascular Pathology, Veneto region, Venice. RC and MBM are Postdoc Fellows supported by the Italian Ministry of Health, PNRR Next-Generation EU grant PNRR-MR1-2022-12376614. GT is a PhD student supported by Italian Ministry of Health DM 630 (PNRR).\u003c/p\u003e\u003cp\u003e \u003ch2\u003eEthical Approval\u003c/h2\u003e \u003cp\u003eAll patients provided written informed consent, in accordance with the protocol approved by the ethics committee, including for publication of pedigree charts, of the Azienda Ospedale Universit\u0026agrave; Padova. All investigations were carried out according to the Declaration of Helsinki.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eApproval Committee\u003c/strong\u003e \u003cp\u003eComitato Etico per la Sperimentazione Clinica. Azienda Ospedale Universit\u0026agrave; Padova.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eData and materials availability\u003c/h2\u003e \u003cp\u003eAll raw data are available upon request to the corresponding author.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eAll other authors declare no competing interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eConceptualization: R.C, G.T.; Formal analysis, R.C, G.T; Investigation: R.C, G.T, S.P.; Methodology: R.C, G.T, S.P.; Supervision: K.P.; Writing original draft: R.C, G.T.; Writing review \u0026amp; editing, R.C, G.T, S.P., F.DZ., M.C., M.BM., C.B. and K.P. Funding acquisition: K.P., C.B. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAnna A, Monika G. Splicing mutations in human genetic disorders: examples, detection, and confirmation. J Appl Genet. 2018;59:253\u0026ndash;68.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScotti MM, Swanson MS. RNA mis-splicing in disease. Nat Rev Genet. 2016;17:19\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLord J, Gallone G, Short PJ, McRae JF, Ironfield H, Wynn EH, et al. Pathogenicity and selective constraint on variation near splice sites. Genome Res. 2019;29:159\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaganathan K, Kyriazopoulou Panagiotopoulou S, McRae JF, Darbandi SF, Knowles D, Li YI, et al. Predicting Splicing from Primary Sequence with Deep Learning. Cell. 2019;176:535\u0026ndash;e548524.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGloss BS, Dinger ME. Realizing the significance of noncoding functionality in clinical genomics. Exp Mol Med. 2018;50:1\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrench JD, Edwards SL. The Role of Noncoding Variants in Heritable Disease. Trends Genet. 2020;36:880\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChristiaans I, Mook ORF, Alders M, Bikker H, Lekanne Dit Deprez RH. Large next-generation sequencing gene panels in genetic heart disease: challenges in clinical practice. Neth Heart J. 2019;27:299\u0026ndash;303.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan den Hoogenhof MMG, Beqqali A, Amin AS, van der Made I, Aufiero S, Khan MAF, et al. RBM20 Mutations Induce an Arrhythmogenic Dilated Cardiomyopathy Related to Disturbed Calcium Handling. Circulation. 2018;138:1330\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVanichkina DP, Schmitz U, Wong JJ, Rasko JEJ. Challenges in defining the role of intron retention in normal biology and disease. Semin Cell Dev Biol. 2018;75:40\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Wang Z. Systematical identification of splicing regulatory cis-elements and cognate trans-factors. Methods. 2014;65:350\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePilichou K, Thiene G, Bauce B, Rigato I, Lazzarini E, Migliore F, et al. Arrhythmogenic cardiomyopathy. Orphanet J Rare Dis. 2016;11:33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJames CA, Jongbloed JDH, Hershberger RE, Morales A, Judge DP, Syrris P, et al. International Evidence Based Reappraisal of Genes Associated With Arrhythmogenic Right Ventricular Cardiomyopathy Using the Clinical Genome Resource Framework. Circ Genom Precis Med. 2021;14:e003273.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBueno Marinas M, Cason M, Bariani R, Celeghin R, De Gaspari M, Pinci S et al. A Comprehensive Analysis of Non-Desmosomal Rare Genetic Variants in Arrhythmogenic Cardiomyopathy: Integrating in Padua Cohort Literature-Derived Data. Int J Mol Sci 2024, 25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO'Neill MJ, Wada Y, Hall LD, Mitchell DW, Glazer AM, Roden DM. Functional Assays Reclassify Suspected Splice-Altering Variants of Uncertain Significance in Mendelian Channelopathies. Circ Genom Precis Med. 2022;15:e003782.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatel PN, Ito K, Willcox JAL, Haghighi A, Jang MY, Gorham JM, et al. Contribution of Noncanonical Splice Variants to TTN Truncating Variant Cardiomyopathy. Circ Genom Precis Med. 2021;14:e003389.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSinger ES, Ingles J, Semsarian C, Bagnall RD. Key Value of RNA Analysis of MYBPC3 Splice-Site Variants in Hypertrophic Cardiomyopathy. Circ Genom Precis Med. 2019;12:e002368.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIto K, Patel PN, Gorham JM, McDonough B, DePalma SR, Adler EE, et al. Identification of pathogenic gene mutations in LMNA and MYBPC3 that alter RNA splicing. Proc Natl Acad Sci U S A. 2017;114:7689\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSinger ES, Crowe J, Holliday M, Isbister JC, Lal S, Nowak N, et al. The burden of splice-disrupting variants in inherited heart disease and unexplained sudden cardiac death. NPJ Genom Med. 2023;8:29.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWai HA, Lord J, Lyon M, Gunning A, Kelly H, Cibin P, et al. Blood RNA analysis can increase clinical diagnostic rate and resolve variants of uncertain significance. Genet Med. 2020;22:1005\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaponi M, Smith LD, Silipo M, Stuani C, Buratti E, Baralle D. BRCA1 exon 11 a model of long exon splicing regulation. RNA Biol. 2014;11:351\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrnich SE, Abou Tayoun AN, Couch FJ, Cutting GR, Greenblatt MS, Heinen CD, et al. Recommendations for application of the functional evidence PS3/BS3 criterion using the ACMG/AMP sequence variant interpretation framework. Genome Med. 2019;12:3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorrado D, Anastasakis A, Basso C, Bauce B, Blomstrom-Lundqvist C, Bucciarelli-Ducci C, et al. Proposed diagnostic criteria for arrhythmogenic cardiomyopathy: European Task Force consensus report. Int J Cardiol. 2024;395:131447.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReuter P, Walter M, Kohl S, Weisschuh N. Systematic analysis of CNGA3 splice variants identifies different mechanisms of aberrant splicing. Sci Rep. 2023;13:2896.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalker LC, Hoya M, Wiggins GAR, Lindy A, Vincent LM, Parsons MT, et al. Using the ACMG/AMP framework to capture evidence related to predicted and observed impact on splicing: Recommendations from the ClinGen SVI Splicing Subgroup. Am J Hum Genet. 2023;110:1046\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJarvik GP, Browning BL. Consideration of Cosegregation in the Pathogenicity Classification of Genomic Variants. Am J Hum Genet. 2016;98:1077\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJian X, Boerwinkle E, Liu X. In silico tools for splicing defect prediction: a survey from the viewpoint of end users. Genet Med. 2014;16:497\u0026ndash;503.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesmet FO, Hamroun D, Lalande M, Collod-Beroud G, Claustres M, Beroud C. Human Splicing Finder: an online bioinformatics tool to predict splicing signals. Nucleic Acids Res. 2009;37:e67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoudayer C, Caux-Moncoutier V, Krieger S, Barrois M, Bonnet F, Bourdon V, et al. Guidelines for splicing analysis in molecular diagnosis derived from a set of 327 combined in silico/in vitro studies on BRCA1 and BRCA2 variants. Hum Mutat. 2012;33:1228\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiolo G, Cantara S, Ricci C. What's Wrong in a Jump? Prediction and Validation of Splice Site Variants. Methods Protoc 2021, 4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiepe TV, Khan M, Roosing S, Cremers FPM, t Hoen PAC. Benchmarking deep learning splice prediction tools using functional splice assays. Hum Mutat. 2021;42:799\u0026ndash;810.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-translational-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jtrm","sideBox":"Learn more about [Journal of Translational Medicine](http://translational-medicine.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/jtrm/default.aspx","title":"Journal of Translational Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Non-Canonical Splicing Variants, Arrhythmogenic Cardiomyopathy, Genetics, in vitro assay","lastPublishedDoi":"10.21203/rs.3.rs-9022128/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9022128/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e. Splice-altering variants (SAVs), particularly those outside canonical splice sites, are an underappreciated contributor to inherited cardiovascular diseases. In arrhythmogenic cardiomyopathy (ACM), these variants frequently remain classified as of uncertain significance (VUS) due to limited predictive power and lack of transcript-level evidence, constraining genetic yield and clinical management. Our study aimed to determine the functional impact of SAVs in ACM genes and refine their classification using ACMG/AMP and ClinGen SVI criteria.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e. SAVs identified in 200 ACM probands underwent SpliceAI prediction, GTEx cardiac exon-usage annotation, and functional assessment using pSPL3-based minigene assays. Aberrant transcripts were quantified using Percent Splicing Alteration (PSA). Segregation data and ACMG/AMP criteria refined by ClinGen SVI were applied to integrate functional and clinical evidence for classification.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults.\u003c/strong\u003e Aberrant splicing was confirmed in 9/20 variants (45%), including synonymous, missense, and non-canonical intronic changes. SpliceAI scores correlated strongly with PSA values (R²=0.86). Case–control burden testing revealed significant enrichment of splice-altering variants in \u003cem\u003eDSP\u003c/em\u003e, \u003cem\u003eDSG2\u003c/em\u003e, \u003cem\u003eDSC2\u003c/em\u003e and \u003cem\u003eFLNC\u003c/em\u003e. Integrating predictive algorithms with experimental validation and segregation analysis markedly enhances reclassification of 16/20 variants (80%).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion.\u003c/strong\u003eSplicing defects beyond canonical sites significantly shape ACM genetic landscape. Integrating predictive models with experimental validation clarifies uncertain variants bridging the gap between genomic uncertainty and clinical decision-making.\u003c/p\u003e","manuscriptTitle":"Tackling Non-Canonical Splicing in Arrhythmogenic Cardiomyopathy to Reduce the Uncertain Significance Variants Burden","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-02 06:57:30","doi":"10.21203/rs.3.rs-9022128/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-04-28T20:07:06+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-29T16:45:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-06T11:47:00+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Translational Medicine","date":"2026-03-05T04:25:54+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-translational-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jtrm","sideBox":"Learn more about [Journal of Translational Medicine](http://translational-medicine.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/jtrm/default.aspx","title":"Journal of Translational Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b11470c8-d0f1-4cc3-a18a-025936250892","owner":[],"postedDate":"April 2nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-02T06:57:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-02 06:57:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9022128","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9022128","identity":"rs-9022128","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}