Splicing Analysis of <i>MYO5B</i> Noncanonical Variants in Patients with Low Gamma-Glutamyltransferase Cholestasis

Wang, Li; Qiu, Yi-Ling; Abuduxikuer, Kuerbanjiang; Wang, Neng-Li; Li, Zhong-Die; Cheng, Ye; Lu, Yi; Xie, Xin-Bao; Xing, Qing-He; Wang, Jian-She

doi:https://doi.org/10.1155/2023/8848362

Human Mutation

On this page

Abstract Introduction Methods Results Discussion Abbreviations Data Availability Ethical Approval Conflicts of Interest Authors’ Contributions Acknowledgments Supplementary Materials References Copyright Related Articles

Research Article | Open Access

Volume 2023 | Article ID 8848362 | https://doi.org/10.1155/2023/8848362

Splicing Analysis of MYO5B Noncanonical Variants in Patients with Low Gamma-Glutamyltransferase Cholestasis

Li Wang,¹Yi-Ling Qiu,¹Kuerbanjiang Abuduxikuer,¹Neng-Li Wang,¹Zhong-Die Li,¹Ye Cheng,²Yi Lu,¹Xin-Bao Xie,¹Qing-He Xing,²and Jian-She Wang^1,3

Academic Editor: Nancy Spinner

Received24 Apr 2023

Revised06 Jun 2023

Accepted23 Jun 2023

Published27 Jul 2023

Abstract

Biallelic MYO5B variants have been associated with familial intrahepatic cholestasis (FIC) with low serum gamma-glutamyltransferase (GGT). Intronic or synonymous variants outside of canonical splice sites (hereinafter referred to as noncanonical variants) with uncertain significance were identified in MYO5B posing a challenge in clinical interpretation. This study is aimed at assessing the effects of these variants on premessenger RNA (pre-mRNA) splicing to improve recognition of pathogenic spliceogenic variants in MYO5B and better characterize the MYO5B genetic variation spectrum. Disease-associated MYO5B noncanonical variants were collected from the literature or newly identified low GGT cholestasis patients. In silico splicing predictions were performed to prioritize potential pathogenic variants. Minigene splicing assays were performed to determine their splicing patterns, with confirmation by blood RNA analysis in one case. Eleven (five novel) noncanonical variants with uncertain significance were identified. Minigene splicing assays revealed that three variants (c.2090+3A>T, c.2414+5G>T, and c.613-11G>A) caused complete aberrations, five variants (c.2349A>G/p.(=), c.4221G>A/p.(=), c.1322+5G>A, c.1669-35A>C, and c.3045+3A>T) caused predominant aberrations, and three variants (c.4852+11A>G, c.455+8T>C, and c.2415-6C>G) had no effect on pre-mRNA splicing. Patient-derived RNA analysis showed consistent results. Based on our results, eight variants were reclassified as likely pathogenic and three as likely benign. Combining the clinical features and the above analysis, the diagnosis of MYO5B-associated FIC could be made in three new patients. In conclusion, we characterized the splicing patterns of MYO5B noncanonical variants and suggest that RNA analysis should be routinely included in clinical diagnostics to provide essential evidence for the interpretation of variants.

1. Introduction

Variation in MYO5B, encoding the motor protein myosin Vb (MYO5B), can cause microvillus inclusion disease and/or familial intrahepatic cholestasis (FIC) with normal/low serum gamma-glutamyltransferase (GGT<100 IU/L) activity [1–4]. Microvillus inclusion disease is a rare enteropathy leading to intractable, life-threatening diarrhea in neonates, which is usually associated with null MYO5B variants, such as nonsense, frameshift, and canonical splice site [4–7]. FIC, as a recently identified phenotype, is less lethal and seems to be a continuous phenotypical spectrum, including persistent, recurrent, and transient cholestasis [4]. MYO5B-associated FIC is usually associated with nonnull variants, and intronic or synonymous variants outside of canonical +1/+2 or −1/−2 splice sites (hereinafter referred to as noncanonical variants) with the potential to affect splicing are occasionally identified [7]. To date, MYO5B noncanonical variants account for approximately 10% of the reported disease-associated variants in FIC patients, and approximately 15% of reported MYO5B-associated FIC patients carried at least one such variant [7]. Variant of uncertain significance (VUS) reporting is also becoming commonplace for the routine use of next-generation sequencing technologies in clinical practice. The pathogenicity of canonical splicing variant requires little supporting evidence, while establishing the pathogenicity of other noncanonical variants relies on more supportive evidence. Unfortunately, the supportive evidence is not available specifically for these noncanonical variants in MYO5B, which poses a challenge in clinical diagnostics.

To address this issue, we collected MYO5B noncanonical variants with uncertain significance both from the literature and new patients with suspected MYO5B-associated low GGT cholestasis. In vitro cell-based specific minigene splicing assays, as well as in vivo patient-derived blood RNA splicing, were performed to evaluate their potential damaging effect on splicing. This study is aimed at achieving a more precise classification of noncanonical variants in MYO5B, thereby enhancing our understanding of the genetic variation spectrum in MYO5B and improving genetic diagnosis.

2. Methods

2.1. Subjects

Preexisting whole-exome sequencing (WES) data was reanalyzed in low GGT intrahepatic cholestasis patients without a molecular diagnosis who were referred to Children’s Hospital of Fudan University from January 2015 to March 2022. Inclusion criteria were patients with biallelic MYO5B potential pathogenic variants and at least one of them was noncanonical variant with the potential to affect splicing. MYO5B potential pathogenic variants were defined as follows: (1) allele frequency less than 0.001 in the Genome Aggregation Database (gnomAD, https://gnomad.http://broadinstitute.org, last accessed January 30, 2022) and (2) variants, except for nonsense, frameshift, canonical splice sites, and in-frame variants, should be predicted to be deleterious or splice affecting by at least 1 of 5 in silico tools (detailed in the following methods). Patients were excluded from this study if they met one of the following criteria: (1) the same genotype was also present in the normal family members and (2) presence of pathogenic or likely pathogenic variants classified following the American College of Medical Genetics and Genomics (ACMG) guidelines [8] and a recently proposed ACMG point system [9] in other low GGT intrahepatic cholestasis-related genes (including at least ATP8B1, ABCB11, NR1H4, TJP2, USP53, VPS33B, VIPAR, LSR, CYP27A1, CYP7B1, HSD3B7, and AKR1D1). The flowchart of patient selection is shown in Figure S1. A total of 11 MYO5B noncanonical variants were studied, including 5 newly identified and 6 previously reported (retrieved using electronic searches on PubMed database; search terms: MYO5B, MYO5B variants, and cholestasis; last search was undertaken on 31 March 2022). Minigene splicing assays were performed on all selected variants, and 2 canonical splice site variants were used as positive controls for the assay. Blood RNA splicing analysis was also investigated if samples from patients were available.

2.2. Nomenclature and In Silico Analysis of Variants

All variant nomenclature was checked by the Mutalyzer website (https://www.mutalyzer.nl/) with reference sequence NM_001080467.3 (NP_001073936.1). PROVEAN (http://provean.jcvi.org/index.php), Polyphen2 (http://genetics.bwh.harvard.edu/pph2/), SIFT (https://www.sift.co.uk/), MutationTaster (http://www.mutationtaster.org/), and REVEL (https://sites.google.com/site/revelgenomics/) were used to predict the pathogenicity of missense variants. Human Splicing Finder (http://www.umd.be/HSF), Splice AI (https://spliceailookup.broadinstitute.org/), varSEAK (https://varseak.bio/), MutationTaster, and MaxEntScan (http://genes.mit.edu/burgelab/maxent/Xmaxentscan _scoreseq.html) were applied to predict the spliceogenic effect of noncanonical variants. Variants with Splice AI [10], varSEAK (https://varseak.bio/pdf/SSP-Documentation.pdf), and the change in splice site score of [11] were deemed spliceogenic variants; all interpretations given by HSF and MutationTaster [12, 13], except for “No significant impact on splicing signals” or “No abrogation of potential splice sites,” were deemed to predict altered splicing.

2.3. Minigene Construction

Depending on the location of MYO5B variants, wild-type (WT) MYO5B exons of interest and 500-1000 bp of flanking intronic sequence were amplified by polymerase chain reaction (PCR) from a subject without an identified MYO5B variant using 2xTaq PCR Mix (KT201, TIANGEN BIOTECH). Specific primers, provided in Table S1, were used to introduce additional recombination sequences, and the purified WT amplicons were subcloned into the XhoI site of a modified pcDNA3.1 vector using the ClonExpress® Ultra One Step Cloning Kit (C115, Vazyme). Site-directed mutagenesis introduced c.3538-1G>A [4], c.839-1G>A [7], c.4852+11A>G [4], c.2090+3A>T [7], c.1322+5G>A [7], c.2414+5G>T [4], c.2349A>G [14], c.3045+3A>T [15], c.1669-35A>C, c.455+8T>C, c.2415-6C>G, c.613-11G>A, and c.4221G>A variants individually, using KOD-Plus-Mutagenesis Kit (SMK-101, Toyobo Bio-Technology). All mutagenesis primers are available in Table S1, and all minigenes were verified by the Sanger sequencing (Biosune, China). The workflow of the minigene protocol is outlined in Figure S2.

2.4. Cell Culture and Transfection

Human embryonic kidney 293T and HepG2 cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37°C in 5% CO₂. Cells were seeded in six-well plates and 24 h later were transiently transfected with 1 μg of the WT, mutated, or empty minigenes using jetPRIME® transfection reagent, respectively (Polyplus Transfection, French).

2.5. RT-PCR Analysis

RNA from cells was extracted using the Direct-zol™ RNA Miniprep kits (Cat. R2050, ZYMO RESEARCH), and RNA from fresh whole peripheral blood was extracted using HiPure PX Blood RNA Kit (R4168, Magen). A total of 1 μg RNA was reverse-transcribed into complementary DNA using the PrimeScript™ RT reagent kit with gDNA Eraser (Cat. RR047A, Takara). PCR was then performed with 2xTaq PCR Mix using specific primers (Table S1). PCR products were separated on 2% agarose gels to analyze product size, and the product sequences were then verified by the Sanger sequencing (Biosune, China). For mixed bands, pClone007 Simple Vector Kit (TSV-007S, Tsingke) was used to obtain the single sequence data.

2.6. Quantitative Real-Time PCR (qPCR)

If normal spliced transcript was identified, qPCR was performed with a QuantStudio 3 Real-Time PCR Systems (Applied Biosystems) using TB Green® Premix Ex Taq™ (Cat. RR420A, Takara). The exon-specific primers spanning the exon-exon boundaries of the MYO5B exon and the flanking adenovirus exons are applied and detailed in Table S1.

3. Results

3.1. Clinical Features of New Low GGT Cholestasis Patients with Biallelic MYO5B Variants

All five new patients (Figure 1 and Table 1) were born to healthy nonconsanguineous parents. Four of them were male term babies, and P5 was a boy born prematurely at 29 weeks of gestation. All patients were suspected as having an underlying genetic etiology for low GGT cholestasis. Three patients had extrahepatic abnormalities: P3 had mild congenital heart abnormalities (atrial septal defect and patent foramen ovale) and right hydrocele; P4 had kidney abnormalities including enlargement of the kidney and α1-microglobinuria; P5 had hypothyroidism, severe malnutrition, and a history of self-limited diarrhea (loose stools, 4-5 times/day, and persisted for nearly 1 month at the age of 1 year and 10 months). Table 1 summarizes the clinical findings and biochemical data.

In P1, jaundice first occurred at the age of 4 days and resolved at the age of 20 days but recurred when he was 1 month old. Thereafter, conjugated hyperbilirubinemia and elevated transaminase levels appeared intermittently, but the elevation of TBA persisted during follow-up. Pruritus was observed at the age of 16 months and persisted during follow-up. Hepatomegaly, splenomegaly, and lithiasis were not observed. A needle liver biopsy at the age of 2 years and 1 month showed canalicular cholestasis and mild portal inflammation. Routine treatments, including UDCA and fat-soluble vitamin, were used, and cholestyramine, rifampicin, and phenobarbitone were added as necessary. At the last follow-up (2 years and 4 months), he continues to experience cholestatic symptoms and growth retardation.

P2 presented with jaundice at the age of 2 months. Physical examination and abdominal ultrasonography confirmed hepatomegaly. Pruritus occurred at the age of 7 months. Liver histology at the age of 1 year and 1 month showed features of cholestasis, mild portal inflammation, and slight ductular reaction. Although UDCA and antipruritic medications (cholestyramine, rifampicin, and phenobarbitone) were used, serum bilirubin, transaminase, and bile acid levels remained above normal ranges, and pruritus remained during the follow-up. Additionally, he experienced failure to thrive and growth retardation.

P3 was conceived through in vitro fertilization. He presented at 40 days of age with severe jaundice and a significant prolonged prothrombin time that was unresponsive to parenteral administration of vitamin K. He was admitted to the intensive care unit at his local hospital, where abdominal ultrasonography showed hepatomegaly and splenomegaly. Liver histology was not available. Behavior indicating pruritus was reported by his mother at the age of 2 months; however, we only found mild jaundice without scratches or ceaseless scratching at 4 months of age, when he was first referred to our clinic. He was treated with UDCA and fat-soluble vitamins, and subsequently his jaundice, hepatomegaly, splenomegaly, and all biochemical tests improved. He was free of cholestatic symptoms with all biochemical test results returned to normal at the last follow-up (2 years and 5 months). Height and weight were normal for age.

P4 presented with jaundice and severe pruritus at the age of 1 year and 7 months. He received treatment with prednisone and azathioprine for suspected autoimmune liver disease in a local hospital. During the 1.5 years of immunosuppressive treatment, the biochemical data fluctuated but remained above the normal range most of the time. He was referred to our hospital at the age of 3 years and 3 months for severe jaundice and pruritus attack following a reduced dose of prednisone. Jaundice and scratches were observed, and abdominal ultrasonography confirmed enlarged liver and kidney. Serum laboratory tests showed cholestasis with normal GGT activity and elevated transaminase levels. Kidney function tests were normal, but urine tests showed α1-microglobinuria. Liver histology performed on a prior biopsy (in a local hospital) showed hepatocellular and canalicular cholestasis, ballooning degeneration of hepatocytes, portal fibrosis and mild inflammation, ductular reaction, and reduction of interlobular ductules. The immunosuppressive treatment was stopped, and UDCA therapy was continued. Cholestyramine and rifampicin were also added. He continues to experience cholestatic symptoms (3 years and 4 months).

At the age of 3 months, biochemical tests performed on P5 showed cholestasis with mildly elevated bilirubin, normal GGT activity, elevated transaminase, and total bile acid levels. Pruritus was observed at the age of 1 year. Abdominal ultrasonography showed a normal liver and biliary tree. Liver histology at 1 year and 4 months showed mild portal fibrosis and inflammation and absence of ductular reaction. UDCA, cholestyramine, phenobarbitone, and fat-soluble vitamin were given. At the last follow-up (2 years), he still had persistent cholestasis despite receiving UDCA and cholestyramine therapy and experienced fluctuations in the severity of pruritus and persistent elevation in biochemical data. He suffered from severe malnutrition and poor growth. It should be noted that P5 had hypothyroidism at the first few days of life and has been continuing to receive Euthyrox therapy.

3.2. In Silico Prediction of MYO5B Variants

In the 5 new patients, 9 novel MYO5B potential pathogenic variants were identified, including 4 noncanonical splice site intronic variants, 2 missense variants, 1 synonymous variant, 1 canonical splice site variant, and 1 in-frame deletion. Literature searching yielded 6 MYO5B noncanonical variants [4, 7, 14, 15]. The in silico prediction and allele frequency of the 11 MYO5B noncanonical variants are detailed in Table 2. The in silico prediction and/or allele frequency of other novel variants are described as follows: variant c.335C>G/p.(Pro112Arg) has never been reported in gnomAD, predicted pathogenic by five above mentioned in silico tools; variant c.4637C>T/p.(Thr1546Met) had an allele frequency of 0/135/280948 (the number of homozygotes/heterozygotes/total persons tested) in gnomAD, predicted pathogenic by 2 (Polyphen2 and MutationTaster) out of the 5 in silico tools; c.2996_2998del/p.(Arg999_Lys1000delinsGln) and c.1669-2_1669-1delinsTT have never been reported in gnomAD.

3.3. Splicing Effects of MYO5B Variants in Minigene System

As shown in Figures 2(a) and 2(b) and Figure S3, all MYO5B WT alleles resulted in normal splicing, and two canonical splice site variants (c.3538-1G>A and c.839-1G>A) resulted in pre-mRNA splicing errors as expected, indicating that the minigene RNA splicing assays could accurately evaluate pre-mRNA splicing. Of the 11 analyzed variants, 3 (c.4852+11A>G, c.455+8T>C, and c.2415-6C>G) produced normal slicing, and the other 8 resulted, to some extent, in aberrant splicing.

(a)

(b)

(c)

Figure 2

Splicing effects of MYO5B variants based on minigene splicing assays. (a) Schematic of minigene system. AV-E1 and AV-E2 are adenovirus exon 1 and exon 2, embedded in the pcDNA3.1 vector; amplicons of interest (exons of interest and 500~1000 bp of flanking intronic sequences in this study) were inserted into the pcDNA3.1 vector. SeqF and SeqR are primers for cDNA amplifying and the Sanger sequencing. (b) The splicing products were amplified by PCR and underwent 2% agarose gel electrophoresis. Gray boxes indicate two adenovirus exons; white boxes are MYO5B exons and are marked with an exon number individually; the arrangement of boxes is consistent with the schematic of minigene system. c.3538-1G>A and c.839-1G>A were selected as positive controls for the assay. Δ, deletion; ins, insertion; nt, nucleotides; bp, base pairs. (c) qPCR for relative quantity of the normal splicing in the mutant minigenes with correctly spliced transcript identified. Compared to the wild-type minigenes, c.1322+5G>A, c.2349A>G, c.1669-35A>C, c.3045+3A>T, and c.4221G>A have only 20%, 35%, 17%, 15%, and 30% of normal splicing, respectively. The first variant in each pair represents the common allele, while the second variant represents the rare allele.

The splicing alterations of 8 variants could be subdivided into two groups. The first group consisted of 3 variants, including c.2090+3A>T, c.2414+5G>T, and c.613-11G>A, which led to complete splicing aberrations. Variant c.2090+3A>T caused retention of 185 nucleotides from intron 17 and predicted to encode a truncated protein, MYO5B p.Trp698Leufs48; c.2414+5G>T caused the deletion of the last 71 and 27 nucleotides of exon 19, respectively, by activating a preexisting cryptic 5 splice site within the exon and predicted to encode a truncated protein, MYO5B p.Val782Alafs3 (major product), as well as a protein with 12 amino acids deletion (p.Tyr797_Arg805del); c.613-11G>A caused retention of 9 nucleotides from intron 5 and predicted to encode a protein with 3 amino acids insertion (p.Glu204_Ala205insSerValGln).

In the second group of 5 variants (c.1322+5G>A, c.2349A>G (p.=), c.1669-35A>C, c.3045+3A>T, and c.4221G>A (p.=)), aberrant pre-mRNA splicing and partially normal splicing were simultaneously present. The c.1322+5G>A caused a prominent deletion of the last 134 nucleotides of exon 10 by decreasing the strength of the reference splice site and simultaneously activating a preexisting cryptic 5 splice site within the exon and predicted to encode a truncated protein, MYO5B p.Ile398, which was the major product, and the minor product corresponded to the correct splicing with a mean percentage of 20% WT splicing. Variant c.2349A>G showed two bands: one causing the deletion of the last 71 nucleotides of exon 19 and the second corresponding to the normal splicing with a mean percentage of 35% WT splicing. Variant c.1669-35A>C resulted in predominant retention of 96 nucleotides from intron 13, which was predicted to encode a truncated protein, MYO5B p.Val557Leufs11, and also, a small percentage of the normally spliced product (with a mean percentage of 17% WT splicing) was detected. c.3045+3A>T caused a prominent deletion of the last 125 nucleotides of exon 22 by activating a preexisting cryptic 5 splice site within the exon and predicted to encode a truncated protein, MYO5B p.Gly974Alafs15; a small percentage of products revealed a deletion of the last 4 nucleotides of exon 22 (also expected to cause a premature stop codon) and the normal splicing (with a mean percentage of 10% WT splicing). Variant c.4221G>A showed two bands: one causing the exon 31 skipping and the second corresponding to the normal splicing with a mean percentage of 30% WT splicing. The qPCR results are shown in Figure 2(c), and the possible consequences of aberrant splicing (mRNA changes and predicted protein changes) in the analyzed MYO5B variants were summarized in Figure S4.

3.4. Blood RNA Analysis In Vivo

Fresh peripheral blood samples from P4 and his family members were available for further analysis. RNA was extracted and RT-PCR was performed. The Sanger sequencing showed that P4 and his mother (who carried c.613-11G>A) had two transcripts: one with 9 nucleotides intron 5 retention and another normal transcript, whereas his father (who did not carry c.613-11G>A and served as a normal control) only had the normal transcript (Figure S5). The splicing change observed was consistent with the result obtained from the minigene assay.

3.5. ACMG Reclassification

Supplemented with the functional analysis, the pathogenicity of noncanonical variants was reevaluated according to the ACMG guidelines [8, 9]. Eight previous VUSs (c.2349A>G/p.(=), c.4221G>A/p.(=), c.613-11G>A, c.1322+5G>A, c.1669-35A>C, c.2090+3A>T, c.2414+5G>T, and c.3045+3A>T) were reclassified as likely pathogenic, and three (c.4852+11A>G, c.455+8T>C, and c.2415-6C>G) were reclassified as likely benign (Table 2).

4. Discussion

With the increasing use of next-generation sequencing in clinical practice, including targeted panel sequencing, WES, and even whole-genome sequencing, VUS reporting has become more common. This poses a challenge for clinical management and genetic counseling; for example, a definitive genetic diagnosis cannot be made for some patients with low GGT cholestasis, and biallelic MYO5B VUSs identified due to caution should be exercised in clinical diagnostics regarding VUSs. The reported cases of MYO5B-associated FIC are limited, and a small portion of variants is VUSs. However, variants can be reclassified over time based on functional studies.

Canonical splicing variants are typically located in intronic splicing donor/acceptor sites and often disrupt pre-mRNA splicing, whereas variants outside these sites are frequently classified as benign or VUSs [16]. In addition, synonymous variants are more challenging to determine their pathogenicity than nonsynonymous single nucleotide variants because they do not alter the protein sequence. Several studies have shown that synonymous and intronic variants can also affect gene function by disrupting normal pre-mRNA splicing [17–20]. Therefore, splicing pattern analysis can help reduce uncertainty in clinical pathogenicity interpretation. Minigene splicing assays are widely used in splicing analysis when RNA samples from patients are not available, and their results have been found to closely resemble endogenous splicing patterns identified from patient RNA [21, 22]. Therefore, this study focuses on performing minigene splicing assays and patient-derived RNA analysis on noncanonical variants in MYO5B obtained from a literature review and retrospective clinical collection rather than on more obviously pathogenic variants.

Several noncanonical variants in MYO5B have been associated with low GGT cholestasis in recent years [4, 7, 14, 15]; meanwhile, some new potential pathogenic MYO5B noncanonical variants have also been clinically detected. Although some in silico tools predict them to be pathogenic, the precise effect of these variants on pre-mRNA splicing has not been experimentally determined, except for c.2090+3A>T, which has been analyzed using minigene splicing assays in a recently published study while our experiments were ongoing, and our results are consistent with those reported in the literature [23]. Herein, we investigated eleven MYO5B noncanonical variants, including 6 reported and 5 novel ones. Minigene splicing assays revealed that 8 out of 11 variants, including 5 reported variants (c.2349A>G/p.(=), c.1322+5G>A, c.2090+3A>T, c.2414+5G>T, and c.3045+3A>T) and 3 newly identified variants (c.4221G>A/p.(=), c.613-11G>A, and c.1669-35A>C), disrupted pre-mRNA splicing. RNA samples obtained from P4 with c.613-11G>A were also subjected to RT-PCR and sequencing analysis, and the results were consistent with the minigene assay. The variants c.4852+11A>G, c.455+8T>C, and c.2415-6C>G were not found to significantly affect splicing based on minigene validation. These findings can aid in the reclassification of these variants and contribute to more precise clinical management.

The pathogenicity of a variant depends on both the clinical setting and functional evidence. Based on the functional analysis and following the ACMG guidelines, the biallelic MYO5B variants found in three new patients (P1, P4, and P5) can be classified as likely pathogenic, and therefore, these patients can be genetically diagnosed as MYO5B-associated FIC. However, the diagnosis of a previously reported MYO5B-associated FIC patient [4] may need to be redefined, as the variant c.4852+11A>G has been confirmed not to affect splicing in this study. The results show the ability of RNA splicing analysis to clarify variant interpretation, as abnormal splicing can provide supporting evidence of pathogenicity. It should be noted that the novel variants c.1669-35A>C (carried by P1) and c.613-11G>A (carried by P4), which are located 35 bp and 11 bp from canonical splice site, respectively, would not typically be suspected or classified as pathogenic without further functional studies. Minigene analysis and blood RNA analysis showed these two variants resulted in aberrantly spliced transcripts, and thus, they can be reclassified as likely pathogenic. The novel variants c.335C>G/p.(Pro112Arg) (carried by P1) and c.2996_2998del/p.(Arg999_Lys1000delinsGln) (carried by P4) can also be reclassified as likely pathogenic based on the evidence of a likely pathogenic variant detected in trans. We therefore suggest that functional studies should be introduced to intronic spliceogenic variants if MYO5B-associated FIC is strongly suspected; it has been reported that many deep intronic variants can disrupt splicing and thus can become a cause of genetic disorders [20, 24, 25]. This study also showed that about 73% of the selected MYO5B noncanonical variants can result in aberrantly spliced transcripts.

The definitive diagnosis for the other two patients (P2 and P3) could not be established based on the current evidence. The variants they carry, c.455+8T>C and c.2415-6C>G, respectively, have no effect on splicing and are now being reclassified as likely benign, which means that they cannot be considered as causative variants. Further investigations, such as liver RNA-seq of the patients [26] and new disease-causing gene analysis, may make a difference.

In silico splice prediction tools have been developed to estimate the likelihood of a noncanonical variant being pathogenic, but they are likely making errors, and it has been reported that the choice of the best splice prediction tool may depend on the gene and the type of splice-altering variant [19, 27, 28]. To assess the performance of individual in silico prediction tool, we calculated the accuracy, positive predictive value, and negative predictive value in Table S2. Among the eight variants confirmed to affect splicing, HSF accurately identified all of them, while Splice AI correctly identified seven out of the eight as affecting splicing. Regarding the three variants confirmed to have no effect on splicing, Splice AI correctly identified all of them, with HSF accurately predicted two out of the three variants. There was a tendency for the variants found experimentally to not affect splicing to be more frequently predicted as such. The accuracy of HSF and Splice AI was relatively high at approximately 91%. HSF demonstrated a perfect negative predictive value, while Splice AI achieved a perfect positive predictive value. In contrast, the performance of MutationTaster as a splicing prediction tool is extremely poor, with an accuracy of approximately 45%. Our study’s results partially align with the previous large-scale analyses that showed an accuracy of approximately 91%, a positive predictive value of approximately 84%, and a negative predictive value of around 95% for Splice AI. For HSF, the accuracy was around 58%, with a lower positive predictive value but a higher negative predictive value of approximately 86% [20]. Based on these factors, we recommend utilizing Splice AI and HSF for splice prediction. Our results that suggest in silico tools could assist in the clinical interpretation of variants potentially affecting splicing, but this method cannot substitute for functional analysis since experimental evidence of abnormal splicing is more reliable than splicing predictions according to the guidelines of ACMG. Hence, the incorporation of in silico splicing prediction and experimental splicing analysis may be helpful to ensure accurate clinical interpretation of VUSs. In addition, we observed that two variants (c.455+8T>C and c.2415-6C>G) did not impact pre-mRNA splicing. Although these variants had frequencies below 0.001 in the overall population, they exhibited higher frequencies in the East Asian population, all exceeding 0.001. Assessing the allele frequency of a variant is important when determining whether it is “too common” or “sufficiently rare” for a Mendelian disease. Consequently, we would recommend using as a cutoff the highest frequency seen in any population, rather than the overall frequency in the database.

In conclusion, we have characterized the molecular consequences of 11 MYO5B noncanonical variants, which represent all known MYO5B noncanonical variants with potential pathogenic significance to date. Our functional study found that 8 of these variants resulted in aberrantly spliced transcript, thus contributing to the accurate clinical classification of these variants and improving genetic diagnosis for individuals.

Abbreviations

bp:	Base pair
FIC:	Familial intrahepatic cholestasis
GGT:	Gamma-glutamyltransferase
MYO5B:	Myosin 5B
MVID:	Microvillus inclusion disease
pre-mRNA:	Premessenger RNA
UDCA:	Ursodeoxycholic acid
VUSs:	Variants of uncertain significance
WES:	Whole-exome sequencing.

Data Availability

All data relevant to the study are included in the article or the supplementary materials attached.

Ethical Approval

This study, approved by the ethics committee of Children’s Hospital of Fudan University (No. 2020-402), was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki.

Conflicts of Interest

The authors declare that they have no conflict of interest.

Authors’ Contributions

Jian-She Wang, Qing-He Xing, Li Wang, and Yi-Ling Qiu conceived and designed the work. Jian-She Wang, Qing-He Xing, and Li Wang were involved in analysis and interpretation of data and drafting and revising the manuscript. Jian-She Wang, Yi-Ling Qiu, Kuerbanjiang Abuduxikuer, Neng-Li Wang, Zhong-Die Li, Yi Lu, and Xin-Bao Xie recruited patients and provided clinical information. Li Wang performed the in silico and functional tests. Qing-He Xing and Ye Cheng contributed to interpretation of the genetic data. All authors read the final version and approved the submission.

Acknowledgments

We thank all physicians of the Center for Pediatric Liver Diseases at the Children’s Hospital of Fudan University for their support in collecting patient information. We are also grateful to the studied families for their cooperation. This project was funded by the National Key Research and Development Program of China (Grant No. 2021YFC2700800), the National Natural Science Foundation of China (Grant No. 82171853), and the Shanghai Municipal Commission of Science and Technology Research Project (19JC1411001).

Supplementary Materials

Supplementary 1. Figure S1: the flowchart of patient inclusion.

Supplementary 2. Figure S2: the workflow of the minigene protocol.

Supplementary 3. Figure S3: the Sanger sequencing data of transcripts identified in minigenes.

Supplementary 4. Figure S4: scheme of the possible consequences of aberrant splicing in the analyzed MYO5B variants. Variant names are listed on the left; corresponding mRNA structures are located next to each variant name; the possible consequences of aberrant splicing are listed below each mRNA structure. The purple lines correspond to the locations of nucleotides deletion; the magenta arrows correspond to the locations of nucleotides insertion. ATG, reference start codon; TGA in black, reference stop codon; TGA in gray, disrupted reference stop codon; aa, amino acids; E, exon; nt, nucleotides; del, deletion; ins, insertion.

Supplementary 5. Figure S5: the splicing analysis of c.613-11G>A from the family IV based on blood RNA extraction. Variant c.613-11G>A was inherited from the mother.

Supplementary 6. Table S1: primers used in this research.

Supplementary 7. Table S2: performance assessment of in silico prediction tools on experimentally validated variants.

References

T. Müller, M. W. Hess, N. Schiefermeier et al., “_MYO5B_ mutations cause microvillus inclusion disease and disrupt epithelial cell polarity,” Nature Genetics, vol. 40, no. 10, pp. 1163–1165, 2008.
View at: Publisher Site | Google Scholar
J. T. Roland, D. M. Bryant, A. Datta, A. Itzen, K. E. Mostov, and J. R. Goldenring, “Rab GTPase-Myo5B complexes control membrane recycling and epithelial polarization,” Proceedings of the National Academy of Sciences of the United States of America, vol. 108, no. 7, pp. 2789–2794, 2011.
View at: Publisher Site | Google Scholar
E. Gonzales, S. A. Taylor, A. Davit-Spraul et al., “MYO5B mutations cause cholestasis with normal serum gamma-glutamyl transferase activity in children without microvillous inclusion disease,” Hepatology, vol. 65, no. 1, pp. 164–173, 2017.
View at: Publisher Site | Google Scholar
Y. L. Qiu, J. Y. Gong, J. Y. Feng et al., “Defects in myosin VB are associated with a spectrum of previously undiagnosed low γ-glutamyltransferase cholestasis,” Hepatology, vol. 65, no. 5, pp. 1655–1669, 2017.
View at: Publisher Site | Google Scholar
H. S. Dhekne, O. Pylypenko, A. W. Overeem et al., “MYO5B, STX3, and STXBP2 mutations reveal a common disease mechanism that unifies a subset of congenital diarrheal disorders: a mutation update,” Human Mutation, vol. 39, no. 3, pp. 333–344, 2018.
View at: Publisher Site | Google Scholar
D. Aldrian, G. F. Vogel, T. K. Frey et al., “Congenital diarrhea and cholestatic liver disease: phenotypic spectrum associated with MYO5B mutations,” Journal of Clinical Medicine, vol. 10, no. 3, p. 481, 2021.
View at: Publisher Site | Google Scholar
L. Wang, Y. L. Qiu, H. M. Xu et al., “MYO5B-associated diseases: novel liver-related variants and genotype-phenotype correlation,” Liver International, vol. 42, no. 2, pp. 402–411, 2022.
View at: Publisher Site | Google Scholar
S. Richards, N. Aziz, S. Bale 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,” Genetics in Medicine, vol. 17, no. 5, pp. 405–424, 2015.
View at: Publisher Site | Google Scholar
S. V. Tavtigian, S. M. Harrison, K. M. Boucher, and L. G. Biesecker, “Fitting a naturally scaled point system to the ACMG/AMP variant classification guidelines,” Human Mutation, vol. 41, no. 10, pp. 1734–1737, 2020.
View at: Publisher Site | Google Scholar
K. Jaganathan, S. Kyriazopoulou Panagiotopoulou, J. F. McRae et al., “Predicting splicing from primary sequence with deep learning,” Cell, vol. 176, no. 3, pp. 535–548.e24, 2019.
View at: Publisher Site | Google Scholar
A. Valenzuela-Palomo, E. Bueno-Martínez, L. Sanoguera-Miralles et al., “Splicing predictions, minigene analyses, and ACMG-AMP clinical classification of 42 germline PALB2 splice-site variants,” The Journal of Pathology, vol. 256, no. 3, pp. 321–334, 2022.
View at: Publisher Site | Google Scholar
F. O. Desmet, D. Hamroun, M. Lalande, G. Collod-Béroud, M. Claustres, and C. Béroud, “Human splicing finder: an online bioinformatics tool to predict splicing signals,” Nucleic Acids Research, vol. 37, no. 9, article e67, 2009.
View at: Publisher Site | Google Scholar
J. Zhang, L. L. Liu, J. Y. Gong et al., “TJP2hepatobiliary disorders: novel variants and clinical diversity,” Human Mutation, vol. 41, no. 2, pp. 502–511, 2020.
View at: Publisher Site | Google Scholar
I. Cockar, P. Foskett, S. Strautnieks et al., “Mutations in myosin 5B in children with early-onset cholestasis,” Journal of Pediatric Gastroenterology and Nutrition, vol. 71, no. 2, pp. 184–188, 2020.
View at: Publisher Site | Google Scholar
N. Uyar Aksu, O. Görükmez, Ö. Görükmez, and A. Uncuoğlu, “A novel homozygous mutation in the MYO5B gene associated with normal-gamma-glutamyl transferase progressive familial intrahepatic cholestasis,” Cureus, vol. 13, no. 11, article e19326, 2021.
View at: Publisher Site | Google Scholar
J. Ule and B. J. Blencowe, “Alternative splicing regulatory networks: functions, mechanisms, and evolution,” Molecular Cell, vol. 76, no. 2, pp. 329–345, 2019.
View at: Publisher Site | Google Scholar
N. L. Wang, Y. L. Qiu, W. C. Guan et al., “Splicing analysis of rare/novel synonymous or intronic variants identified in ABCB11 heterozygotes presenting as progressive intrahepatic cholestasis with low γ-glutamyltransferase,” Hepatology Research, vol. 48, no. 7, pp. 574–584, 2018.
View at: Publisher Site | Google Scholar
Y. Yang, J. Zhang, L. T. Li et al., “Whole-genome sequencing reveals large ATP8B1 deletion/duplications as second mutations missed by exome-based sequencing,” The Journal of Molecular Diagnostics, vol. 23, no. 11, pp. 1491–1499, 2021.
View at: Publisher Site | Google Scholar
W. Q. Xu, R. M. Wang, Y. Dong, and Z. Y. Wu, “Pathogenicity of intronic and synonymous variants of ATP7B in Wilson disease,” The Journal of Molecular Diagnostics, vol. 25, no. 1, pp. 57–67, 2023.
View at: Publisher Site | Google Scholar
H. A. Wai, J. Lord, M. Lyon et al., “Blood RNA analysis can increase clinical diagnostic rate and resolve variants of uncertain significance,” Genetics in Medicine, vol. 22, no. 6, pp. 1005–1014, 2020.
View at: Publisher Site | Google Scholar
D. Baralle and M. Baralle, “Splicing in action: assessing disease causing sequence changes,” Journal of Medical Genetics, vol. 42, no. 10, pp. 737–748, 2005.
View at: Publisher Site | Google Scholar
F. Schmid, E. Glaus, D. Barthelmes et al., “U1 snRNA-mediated gene therapeutic correction of splice defects caused by an exceptionally mild BBS mutation,” Human Mutation, vol. 32, no. 7, pp. 815–824, 2011.
View at: Publisher Site | Google Scholar
Y. Zheng, Y. Peng, S. Zhang et al., “Case report: MYO5B homozygous variant c.2090+3A>T causes intron retention related to chronic cholestasis and diarrhea,” Frontiers in Genetics, vol. 13, article 872836, 2022.
View at: Publisher Site | Google Scholar
H. Y. Xiong, B. Alipanahi, L. J. Lee et al., “RNA splicing. The Human splicing Code Reveals New Insights into the Genetic Determinants of Disease,” Science, vol. 347, no. 6218, article 1254806, 2015.
View at: Publisher Site | Google Scholar
R. Vaz-Drago, N. Custódio, and M. Carmo-Fonseca, “Deep intronic mutations and human disease,” Human Genetics, vol. 136, no. 9, pp. 1093–1111, 2017.
View at: Publisher Site | Google Scholar
B. B. Cummings, J. L. Marshall, T. Tukiainen et al., “Improving genetic diagnosis in Mendelian disease with transcriptome sequencing,” Science Translational Medicine, vol. 9, no. 386, article eaal5209, 2017.
View at: Publisher Site | Google Scholar
A. Moles-Fernández, J. Domènech-Vivó, A. Tenés, J. Balmaña, O. Diez, and S. Gutiérrez-Enríquez, “Role of splicing regulatory elements and in silico tools usage in the identification of deep intronic splicing variants in hereditary breast/ovarian cancer genes,” Cancers, vol. 13, no. 13, p. 3341, 2021.
View at: Publisher Site | Google Scholar
T. V. Riepe, M. Khan, S. Roosing, F. P. M. Cremers, and P. A. 't Hoen, “Benchmarking deep learning splice prediction tools using functional splice assays,” Human Mutation, vol. 42, no. 7, pp. 799–810, 2021.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2023 Li Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies