j ourna l homepage: www.e lsev ie r .com/ locate /gene
Review
First functional analysis of a novel splicing mutation in the B3GALTLgene by an ex vivo approach in Tunisian patients with typical Petersplus syndrome
Afif Ben Mahmoud a,⁎, Olfa Siala a, Riadh Ben Mansour b, Fatma Driss c, Siwar Baklouti-Gargouri a,Emna Mkaouar-Rebai a, Neila Belguith d, Faiza Fakhfakh a,⁎a Laboratoire de Génétique Moléculaire Humaine, Faculté de Médecine de Sfax, Université de Sfax, Tunisiab ISBS (Institut Supérieur de Biotechnologie de Sfax) Sfax, Tunisiac CBS (Centre de Biotechnologie de Sfax) Sfax, Tunisiad Laboratoire de Génétique Médicale, EPS Hédi Chaker, Sfax, Tunisia
Article history:Accepted 11 July 2013Available online 14 August 2013
Keywords:Peters plus syndromeFunctional analysisExon skippingB3GALTL gene
Peters plus syndrome is a rare recessive autosomal disorder comprising ocular anterior segment dysgenesis,short stature, hand abnormalities and distinctive facial features. It was related only to mutations in theB3GALTL gene in the 13q12.3 region. In this study, we undertook the first functional analysis of a novel c.597-2A N G splicing mutation within the B3GALTL gene using an ex-vivo approach. The results showed a complete skip-ping of exon 8 in the B3GALTL cDNA, which altered the open reading frame of themutant transcript and generateda PTC within exon 9. This finding potentially elicits the nonsense mRNA to degradation by NMD (nonsense-mediatedmRNAdecay). The theoretical consequences of splice sitemutations, predictedwith the bioinformaticstool Human Splice Finder, were investigated and evaluated in relation to ex-vivo results. The findings confirmedthe key role played by the B3GALTL gene in typical Peters-plus syndromes and the utility of mRNA analysis tounderstand the primary impacts of this mutation and the phenotype of the disease.
Peters plus syndrome (PPS) is a rare recessive autosomal disorder,with less than 70 cases reported throughout the world. It is character-ized by anterior chamber-eye anomalies, disproportionate short stature,developmental/intellectual disruptions, dysmorphic facial features,and cleft lip/palate (Maillette and Hennekam, 2002). This conditionis related to mutations in the B3GALTL gene, originally identified byHeinonen and collaborators (Heinonen et al., 2003) and later found toencode a glycosyltransferase (Kozma et al., 2006; Sato et al., 2006).The latter gene contains 15 coding exons and spans 132 kb of genomicDNA. It is transcribed in a wide range of human tissues in the form ofthree transcripts resulting in three different alternative polyA sites, allin exon 15 (Heinonen et al., 2003). The B3GTL protein spans 498amino acids and contains a short N-terminal tail (aa 1–4), a trans-membrane region (aa 5–28), a so-called stem region (aa 29–260), anda C-terminal catalytic domain (aa 261–498) (Heinonen et al., 2003).This protein is a glycosyltransferase: the β‐1,3-glycosyltransferase in-volved in the synthesis of the disaccharide Glc-β1,3-Fuc-O- that occurson thrombospondin type 1 repeats (TSRs) ofmany biologically importantproteins (Hess et al., 2008).
Up to now, only sevenmutations in the B3GALTL genewere identifiedin patients with PPS. The most frequent mutation occurring in 75% ofcases is the c.660 + 1G N A located at the donor splice site (5′ss) ofexon 8 and harbored in several populations (Oberstein et al., 2006).Other mutations were also reported including the c.230insT mutation inexon 4, the c.347 + 5 G N A variation in exon 5, the c.459 + 1G N Atransition in exon 6, the c.1178G N A mutation in exon 13 (Dassie-Ajdidet al., 2009), and a p.Tyr366X substitution in exon 13 (Aliferis et al.,2010). More recently, two Tunisian patients were found to have ac.597-2A N G mutation (Siala et al., 2012), whose effect in the acceptorsplice site of exon 8 of the B3GALTL gene on the mRNA splicing processusing an ex-vivo approach was studied in the present study.
2. Materials and methods
The study concerned two Tunisian patients belonging to twounrelated non-consanguineous families suspected to be affected byPPS. Both patients were underweight at birth, with disproportionateshort stature and microcephaly. They also showed facial dysmorphism,including dolichocephaly, round face, and bilateral corneal opaqueness.The progress of the latter wasmonitored by the determination of visualacuteness, followed by cornea transplantations in both patients.
2.1. DNA extraction
Total DNA extraction from blood leukocyteswas performed accordingto a previously described protocol (Kawasaki, 1990).
2.2. PCR amplification
The PCR amplification of exon 8 (64 bp) and its intron boundarieswas performed using the following primers F: 5′GATAAGGGGTCACCAAAGCTTATGACTTTTTTCC3′ containing amismatch to create a BsteII re-striction site and reverse primer R: 5′CCCATTCAGCTAGCTTAAAAGTAAAGAATCATG3′ containing a mismatch to create a NheI endonucleaserestriction site (restriction sites are underlined). These primers led toa 615 bp PCR product containing 296 bp of intron 7 and 255 bp of in-tron 8. Amplification was performed in a thermal cycler (Perkin Elmer
Gene A PCR System 9700) in a total volume of 25 μl containing0.05 μg of genomic DNA, 10 mM dNTP, 25 mM MgCl2, 20 pmol ofeach primer, 5 μl of 5× buffer, and 0.5 unit of Go Taq DNA polymerase(Go Taq, Promega). The PCR conditions were as follows : 2 min at95 °C followed by 35 cycles, each consisting of 45 s at 95 °C, 60 s at57 °C, 40 s at 72 °C, and final elongation at 72 °C for 7 min.
2.3. Cloning of wild type and mutant mini-genes
A splicing cassette (p(13,17)/cytomegalovirus [CMV]) was designed tocontain the 2 adjacent constitutive exons 13 and 17 of human 4.1 gene,with their downstream and upstream flanking intron sequences, respec-tively (Deguillien et al., 2001). Final PCRproductsweredigested and ligatedinto theBstEII/NheI sites of the cassette as 615 bp generated from the geno-mic fragment. After the subcloning of the mutant and wild type vectors,transformation was performed with TOP 10 competent cells (Invitrogen).The cells were grown to an optical density value of 0.8 at 37 °C. The cas-sette, as well as theWT andmutant inserts were fully sequenced to ensurethe absence of any additional mismatch resulting from PCR or cloning er-rors. The size of recombinant cassette was 1200 bp, a feature that wasused to distinguish the recombinant from the native plasmids.
2.4. Ex-vivo splicing assays: cell culture and transfection
Hela cells were cultured in RPMI (Roswell Park Memorial Institute)with glutamine in flask disk plates T-25. The cells were transfectedwith 2.5 μg of wild type or mutated minigene constructs using Fugen6 Transfection reagent (ROCHE) according to the protocol provided bythe manufacturer. Transfected cells were selected for 6–8 days in thesame medium containing 600–800 μg G418/ml (Geneticin invitrogen).
2.5. RNA extraction and RT-PCR
Total RNA was extracted from the cells using a Trizol reagent(invitrogen) and was then treated with recombinant DNase I/RNase-free (Takara). Briefly, 107 cells were lysed using 1 ml of Trizol reagent.200 μl of chloroform were added to the tubes and centrifuged at12,000 tr for 25 min at 4 °C. The aqueous phase was precipitated with1 vol 70% ethanol. The RNA was pelleted by centrifugation at 10,000 trfor 1 min with Rnase-free water, and the RNA preparation was thenstored at −80 °C. RNA content was measured at 260 nm using aNanoDrop ND1000 spectrophotometer (NanoDrop Technologies). Itwas treated with MMLV reverse transcriptase (Fermentas) in a 10 μlof reaction in accordance with the manufacturer's instructions. 10 μlof the cDNA was added to a final volume of 50 μl PCR reaction mixturecontaining 125 μM of dNTP, 25 μM of forward and reverse primers and1 unit of Go Taq DNA polymerase (promega). The forward primer (5′CGCCTAGATGCCTCTGCTAA3′) and the reverse primer (5′AAGCGCTTGTCCCACTCGCT3′) were located at each extremity of the splicing cas-sette. The splicing products were extracted from the agarose geland sequenced to confirm the identity of each product.
2.6. Bioinformatics prediction
The effect of the c.597-2A N G mutation on the splicing process ofthe B3GALTL gene was predicted by a new bioinformatics tool, thehuman splicing finder (HSF) software that is freely available online(http://www.umd.be/HSF/). This program includes new algorithms toevaluate the strength of 5′ss, 3′ss, and branch points (Beroud et al.,
Fig. 1. A—Overview of mutations described so far for the 15 exons of the B3GALTL gene. The B3GALTL protein, consisting of a transmembrane region (TMR); a stem region (SR); and acatalytic domain (CD). B—Gel electrophoresis of the RT-PCR amplification product of the region spanning UE upstream and DE downstream of the cassette, showing (2) a single bandin the wild type at the expected size of 349 bp, (3) a smaller single band of 285 bp in the mutant, and (1) a negative control without RNA. The size marker is a 100 bp DNA ladderfrom fermentas (M). Results of the ex-vivo splicing of the unaltered acceptor splicing site of intron 7, leading to the efficient and total inclusion of exon 8.
15A. Ben Mahmoud et al. / Gene 532 (2013) 13–17
2000, 2005). In addition, in order to identify cis-acting elements,it includes already-published algorithms, such as the RESCUE-ESE(Fairbrother et al., 2004) and ESE-Finder (Cartegni et al., 2003) aswell as new algorithms designed to use available or newly createdmatrices. HSF includes all genes and alternative transcripts as well asintronic sequences thatwere extracted from the Ensembl human genomedatabase (http://www.ensembl.org/) (Flicek et al., 2008).
Fig. 2. Sequencing of RT-PCR products performed on the total RNA from cells transfectedwith thmutant.
3. Results
3.1. The c.597-2A N G mutation leads to exon skipping and truncatedprotein
The effects of the c.597-2A N G mutation on RNA splicing were in-vestigated using a minigene approach. A genomic fragment containing
e control (A) andwithmutated construct (B) showing the total skipping of exon (8) in the
exon 8 and 296 bp and 255 bp of flanking intron 7 and intron 8, respec-tively, was cloned into an efficient splicing cassette. Control andmutat-ed minigene constructs were transfected in Hela-cells. The total RNAextracted from the cells transfected with the normal construct was sub-mitted to RT-PCR analyses, which revealed a 349 bp fragment corre-sponding to the expected splicing product containing the exon 8 ofthe B3GALTL gene. The RT-PCR performed on the total RNA from cellstransfected with the mutated construct, on the other hand, displayeda shorter band of 285 bp (Fig. 1). Direct sequencing revealed a totallack of exon 8 (Fig. 2B). These results confirmed that the B3GALTLmRNA underwent an aberrant splicing that resulted in the total skip-ping of exon 8.
3.2. Predicted effect of the c.597-2A N G mutation on the structure and thefunction of the B3GTL protein
The absence of exon 8 induced a reading frame shift with theintroduction of a downstream premature translation stop codon (PTC)in the mRNA at position +10 within exon 9, encoding the β1,3-glucosyltransferase protein. The truncated protein, if translated, wouldcontain only 212 residues over a total of 498. This loss of 286 amino-acids comprises the catalytic domain, a region which is crucial for awide range of conformational features pertaining to the GT31 familyof beta-3 glucosyltransferase. In addition, the MFOLD program resultsshowed several changes in the RNA secondary structure. In fact, thenormal sequence contained an external closing pair between G9C39with a hairpin loop between U16 and A34 positions containing the A atposition 597 of the B3GALTL mRNA. The c.597-2A N G mutation deletesthe 19 bp hairpin loop between U16 and A34 positions and replaces itby two external hairpin loops with different nucleotide composition andorientation compared to the normal sequence (Siala et al., 2012).
3.3. HSF prediction: the c.597-2A N G mutation has a deleterious effect onthe splicing process
In silico analyses of the putative effects of the c.597-2A N Gmutationon the splicing event and the strength (% values) of the acceptor anddonor splice sites of exon 8 were predicted by the HSF program version2.4. The strength of the splice site was indicated by the consensusvalue and the CV variation Δcv; the consensus sequences were (C/A)(AG/gt(a/g)agt) for the donor splice site and cag/G for the acceptorsite (Mount, 1982). The numerical score calculated for the altered se-quence tctatttcttgg/GC was 55.78% compared to 84.73% for the wildtype. Furthermore, this mutation was predicted to induce the activationof a cryptic splice site and to induce a slight increase in a putative acceptorsplice site (CV = 70.26 to CV = 70.45; Δcv + 0.27). Hence, the splicesite showed a CV of above 70 in combination with a Δcv reductionbelow 10%, indicating that this site was active. The activation of the cryp-tic splice site led to the exonic addition of 10 bp (ggg ctt acc aag/AG)(according toHSF predictions, if a cryptic sitewas in use, exon length var-iation = −10).
4. Discussion
The present work is the first attempt to perform a functional studyon a novel c.597-2A N G mutation in the acceptor splice site of exon 8of the B3GALTL gene. Ex-vivo splicing assays and subsequent RT-PCRanalyses were performed on mRNA isolated from the Hela-cellstransfected with normal and mutant expression vectors. The resultsshowed that the c.597-2A N G altered the acceptor site of exon 8, lead-ing to the total skipping of this exon. The skipping of exons 8 and 5with-in the B3GALTL cDNA was previously reported in Dutch patientsharboring respectively the c.660 + 1G N A and the c.437 + 5G N Amutations. Those patients showed the same phenotypic severity ob-served in our patients, including facial dysmorphism, disproportionateshort stature, and corneal opacity (Oberstein et al., 2006).
However, the skipping of exon 8 leads to a translation frameshift andthe introduction of a PTC at position +10 within exon 9, resulting in anonsense mRNA that, according to the position rule previously statedfor PTC, elicited NMD degradation (Nagy and Maquat, 1998). In all,about 78% of PTCs were located in more than 50 nucleotides upstreamof the last exon–exon junction, and were thus predicted to produce amarked proportion of NMD substrates (Maquat, 2004). The conse-quences for protein sequence and function alteration, as well as forthe triggering of theNMDpathway, have been previously demonstratedfor exon-skipping events in several studies (Baek and Green, 2005; Liuet al., 2001).
Defects in pre-mRNA splicing have been shown as a commondisease-causing mechanism in several studies (Cooper et al., 2009;Tazi et al., 2009). In the present case, the c.597-2A N G mutation waspredicted to induce the activation of a cryptic splice site and to slightlyincrease a putative acceptor splice site at position +10 of exon 8, thusgiving evidence for the activity of this site and leading to the exonicaddition of 10 bp (see Results). These results are different from thoseobserved by experimental assays showing the skipping of exon 8 show-ing the reliability of the predictions made by bioinformatics tools andthus, the importance of performing an important experimental analysisto confirm the findings from molecular diagnosis.
The remaining information required for splicing is contained withinrelatively short (~6 nucleotide) sequences located within both exonsand introns that either enhance or suppress splicing (Wang andBurge, 2008). For this reason, a systematic analysis of the cis-acting reg-ulatory sequences that reside within exon 8 and flanking introns wasperformed. The results showed that (1) the upstream intron of exon 8(296 bp) had a strong splicing enhancer element corresponding toSRp40 and a linked protein type Tra2-β. (2) The downstream intron(255 bp), on the other hand, showed the presence of at least a splicingenhancer element corresponding to SRp55 and a linked protein type9G8. (3) The exon itself contained a strong constitutive splicing silencer.The results also showed the occurrence of a splicing silencer motif.These results suggest that the interplay between the negative and pos-itive elementsmay determine the inclusion or exclusion of exon 8 of theB3GALTL gene. The further clarification of thesemechanismsmight opennew opportunities for the understanding of the phenotypic variabilityand genetic heterogeneity associated with PPS and the ultimate earlyprevention and/or treatment of this serious congenital disorder.
In this context, the present study provided qualitative support forthe effect of the c.597-2A N G mutation on the splicing event in thecell. It would be highly desirable to introduce a very detailed studyin vivo testing together withmethods for detailed quantitative analyses.This would facilitate a direct quantitative comparison of exceedinglydiverse approaches on at least the cellular level. This would allow forcomparison of different approaches with one another. In fact, the alter-native maturity of pre-mRNA would allow for both a qualitative andquantitative control of gene expression. It allows for the qualitative con-trol of gene expression bymodulating interaction properties, intracellularlocalization, enzymatic activity, and protein stability (Stamm et al.,2005). It also allows for the quantitative control of gene expression byalternative splicing, which may be attributed to the modification ofthe 5′ and 3′ untranslated regions of mRNAs, thereby guiding theirstability, localization or translatability, or even to the introduction ofpremature stop codons, and thereby directing the mRNA to the degra-dation mechanisms.
4.1. Conclusion
Splicing mutations that alter the highly conserved dinucleotide ateither the 5′or 3′ splice site, are commonly known to induce, in virtuallyall cases, a total exon skipping or intron inclusion, or/and use of crypticsplice sites. Mutations in the consensus sequences or in splicingenhancers or silencers, affect splicing with various degrees. The applica-tion of minigene ex-vivo assays offers a particularly powerful tool to
17A. Ben Mahmoud et al. / Gene 532 (2013) 13–17
address this issue. This approach exemplifies the importance of deter-mining the actual impacts of a genomicmutation onmRNAmetabolismto better understand the pathophysiology of this disease.
Conflict of interest
There is no conflict of interest.
Acknowledgments
The authors would like to express their gratitude to all colleaguesand technical staff at the Centre de Biotechnologie de Sfax (CBS) andInstitut Supérieur de Biotechnologie de Sfax (ISBS) for their valuablehelp. Thanks are also due to Mr Anouar Smaoui from Faculté desSciences de Sfax (FSS) for his constructive proofreading and languagepolishing services.
References
Aliferis, K., et al., 2010. A novel nonsense B3GALTL mutation confirms Peters plussyndrome in a patient with multiple malformations and Peters anomaly. OphthalmicGenet. 31 (4), 205–208.
Baek, D., Green, P., 2005. Sequence conservation, relative isoform frequencies, andnonsense mediated decay in evolutionarily conserved alternative splicing. Proc.Natl. Acad. Sci. U. S. A. 102 (36), 12813–12818.
Beroud, C., Collod-Beroud, G., Boileau, C., Soussi, T., Junien, C., 2000. Universal mutationdatabase, a generic software to build and analyze locus-specific databases. Hum.Mutat. 15, 86–94.
Beroud, C., Hamroun, D., Collod-Beroud, G., Boileau, C., Soussi, T., Claustres, M., 2005.Universal mutation database. Update. Hum. Mutat. 26, 184–191.
Cooper, T.A., Wan, L., Dreyfuss, G., 2009. RNA and disease. Cell 136, 777–793.Dassie-Ajdid, J., et al., 2009. Novel B3GALTL mutation in Peters-plus syndrome. Clin.
Genet. 76 (5), 490–492.
Deguillien, M., Huang, S.C., Morinière, M., Dreumont, N., Benz, E.J., Baklouti, F., 2001.Multiple cis elements regulate an alternative splicing event at 4.1R pre-mRNA duringerythroid differentiation. Blood 98 (13), 3809–3816.
Flicek, P., et al., 2008. Ensembl. Nucleic Acids Res. 36, 707–714.Heinonen, T.Y.K., et al., 2003. A novel human glycosyltransferase: primary structure and
characterization of the gene and transcripts. Biochem. Biophys. Res. Commun. 309,166–174.
Hess, D., Keusch, J.J., Oberstein, S.A., Hennekam, R.C., Hofsteenge, J., 2008. Peters plussyndrome is a new congenital disorder of glycosylation and involves defectiveOmicronglycosylation of thrombospondin type 1 repeats. J. Biol. Chem. 283 (12),7354–7360.
Kawasaki, E., 1990. Sample Preparation from Blood Cells and Other Fluids. San DiegoAcademic Press.
Kozma, K., et al., 2006. Identification and characterization of a beta1,3-glucosyltransferasethat synthesizes the Glc-beta1,3-Fuc disaccharide on thrombospondin type 1 repeats.J. Biol. Chem. 281, 36742–36751.
Mount, S.M., 1982. A catalogue of splice junction sequences. Nucleic Acids Res. 10,459–472.
Nagy, E., Maquat, L.E., 1998. A rule for termination-codon position within intron-containinggenes. When nonsense affects RNA abundance. Trends Biochem. Sci. 23 (6).
Oberstein, S.A., et al., 2006. Peters plus syndrome is caused by mutations in B3GALTL, aputative glycosyltransferase. Am. J. Hum. Genet. 79 (3), 562–566.
Sato, T., et al., 2006. Molecular cloning and characterization of a novel human b1,3-glucosyltransferase, which is localized at the endoplasmic reticulum and glucosylatesO-linked fucosylglycan on thrombospondin type 1 repeat domain. Glycobiology 16,1194–1206.
Siala, O., et al., 2012. Two Tunisian patients with Peters plus syndrome harbouring a novelsplice site mutation in the B3GALTL gene that modulates the mRNA secondarystructure. Gene 507 (1), 68–73.
Stamm, S., et al., 2005. Function of alternative splicing. Gene 344, 1–20.Tazi, J., Bakkour, N., Stamm, S., 2009. Alternative splicing and disease. Biochim. Biophys.
Acta 1792, 14–26.Wang, Z., Burge, C.B., 2008. Splicing regulation: from a parts list of regulatory elements to
