Nova2 Interacts with a Cis-Acting Polymorphism to Influence the Proportions of Drug-Responsive Splice Variants of SCN1A
2007; Elsevier BV; Volume: 80; Issue: 5 Linguagem: Inglês
10.1086/516650
ISSN1537-6605
AutoresErin L. Heinzen, Woohyun Yoon, Sarah K. Tate, Arjune Sen, Nicholas Wood, Sanjay M. Sisodiya, David B. Goldstein,
Tópico(s)Drug Transport and Resistance Mechanisms
ResumoAn intronic polymorphism in the SCN1A gene, which encodes a neuronal sodium-channel α subunit, has been previously associated with the dosing of two commonly used antiepileptic drugs that elicit their pharmacologic action primarily at this ion-channel subunit. This study sought to characterize the functional effects of this polymorphism on alternative splicing of SCN1A and to explore the potential for modulating the drug response in the pharmacologically unfavorable genotype by identification of a splice modifier acting on SCN1A. The effects of the genotype at the SCN1A IVS5N+5 G→A polymorphism on SCN1A splice-variant proportions and the consequences of increased expression of splice modifiers were investigated both in human temporal neocortex tissue and in a cellular minigene expression system. Quantitative real-time polymerase chain reaction was used to quantify the amounts of SCN1A transcripts forms. We show that the polymorphism has a dramatic effect on the proportions of neonate and adult alternative transcripts of SCN1A in adult brain tissue and that the effect of the polymorphism also appears to be modified by Nova2 expression levels. A minigene expression system confirms both the effect of the polymorphism on transcript proportions and the role of Nova2 in the regulation of splicing, with higher Nova2 expression increasing the proportion of the neonate form. A larger Nova2-mediated effect was detected in the AA genotype that is associated with increased dose requirements. The effects of Nova2 on modulation of the alternative splicing of 17 other neuronally expressed genes were investigated, and no effect was observed. These findings emphasize the emerging role of genetic polymorphisms in modulation of drug effect and illustrate both alternative splicing as a potential therapeutic target and the importance of considering the activity of compounds at alternative splice forms of drug targets in screening programs. An intronic polymorphism in the SCN1A gene, which encodes a neuronal sodium-channel α subunit, has been previously associated with the dosing of two commonly used antiepileptic drugs that elicit their pharmacologic action primarily at this ion-channel subunit. This study sought to characterize the functional effects of this polymorphism on alternative splicing of SCN1A and to explore the potential for modulating the drug response in the pharmacologically unfavorable genotype by identification of a splice modifier acting on SCN1A. The effects of the genotype at the SCN1A IVS5N+5 G→A polymorphism on SCN1A splice-variant proportions and the consequences of increased expression of splice modifiers were investigated both in human temporal neocortex tissue and in a cellular minigene expression system. Quantitative real-time polymerase chain reaction was used to quantify the amounts of SCN1A transcripts forms. We show that the polymorphism has a dramatic effect on the proportions of neonate and adult alternative transcripts of SCN1A in adult brain tissue and that the effect of the polymorphism also appears to be modified by Nova2 expression levels. A minigene expression system confirms both the effect of the polymorphism on transcript proportions and the role of Nova2 in the regulation of splicing, with higher Nova2 expression increasing the proportion of the neonate form. A larger Nova2-mediated effect was detected in the AA genotype that is associated with increased dose requirements. The effects of Nova2 on modulation of the alternative splicing of 17 other neuronally expressed genes were investigated, and no effect was observed. These findings emphasize the emerging role of genetic polymorphisms in modulation of drug effect and illustrate both alternative splicing as a potential therapeutic target and the importance of considering the activity of compounds at alternative splice forms of drug targets in screening programs. The study of pharmacogenetics has been confined predominantly to the elucidation of genetic variation in processes that control drug exposure, including drug-metabolizing enzymes and transporters. However, there is a growing consensus that genetic variability in drug absorption, distribution, metabolism, and elimination plays only a modest role in variability in treatment response and that variation related to the mode of action of a drug may often be as or more important. Such genes, however, have not been thoroughly studied, and, to date, there remain relatively few examples of polymorphisms in a pharmacologic target or downstream pathway that contribute to variable drug response in humans. Perhaps the best documented example is the VKORC1 gene (MIM 608547), which encodes a subunit of the vitamin K epoxide reductase complex and the pharmacological target for the commonly used anticoagulant warfarin. A common polymorphism in this gene has been significantly associated with reduced warfarin dose.1D’Andrea G D’Ambrosio RL Di Perna P Chetta M Santacroce R Brancaccio V Grandone E Margaglione M A polymorphism in the VKORC1 gene is associated with an interindividual variability in the dose-anticoagulant effect of warfarin.Blood. 2005; 105: 645-649Crossref PubMed Scopus (621) Google Scholar Although not yet definitive, it now appears that the molecular mechanism of this variation in warfarin response is regulated at the transcriptional level.2Rieder MJ Reiner AP Gage BF Nickerson DA Eby CS McLeod HL Blough DK Thummel KE Veenstra DL Rettie AE Effect of VKORC1 haplotypes on transcriptional regulation and warfarin dose.N Engl J Med. 2005; 352: 2285-2293Crossref PubMed Scopus (1209) Google Scholar Another example of a common polymorphism that influences drug response is one in SCN1A (MIM 182389), the gene encoding the target of certain antiepileptic drugs. The SCN1A IVS5N+5 G→A polymorphism (rs3812718, formerly SCN1A IVS4-91G→A) was shown to be significantly associated with maximum dose of both phenytoin and carbamazepine in a cohort of patients with various forms of epilepsy.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar Dosing in epilepsy, as well as in many other therapeutic areas, often involves long periods of trial and error and could benefit from genetic predictors of dosing requirements. In that study, Tate et al.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar found antiepileptic drug dose to be lowest in patients with the GG genotype, intermediate in those with the GA genotype, and highest in those with the AA genotype. In a separate study by the same authors that involved patients of Chinese ancestry, an association was found between SCN1A IVS5N+5 G→A and phenytoin serum concentrations at maintenance dose, again in the direction AA→GA→GG for highest→lowest dose.4Tate S Singh R Hung C Tai J Depondt C Cavalleri G Sisodiya S Goldstein D Horng-Huei L A common polymorphism in the SCN1A gene associates with phenytoin serum levels at maintenance dose.Pharmacogenet Genomics. 2006; 16: 721-726Crossref PubMed Scopus (85) Google Scholar The phenotype of serum concentration at maintenance dose was chosen, to eliminate the variation between drug doses and serum levels (which are likely to be caused by pharmacokinetic factors) and therefore increase the chance of detecting the effect of pharmacodynamic variation in the drug target. Interestingly, SCN1A IVS5N+5 G→A is located in the consensus-site sequence located after the neonatal form of exon 5 (exon 5N) of the SCN1A gene (fig. 1).3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar Collectively, this G→A change, which notably is a frequent mutation associated with mammalian disease,6Nakai K Sakamoto H Construction of a novel database containing aberrant splicing mutations of mammalian genes.Gene. 1994; 141: 171-177Crossref PubMed Scopus (252) Google Scholar results in the disruption of the conserved consensus-site sequence, which would theoretically result in a possibly weaker 5′ splice site.7Cartegni L Chew SL Krainer AR Listening to silence and understanding nonsense: exonic mutations that affect splicing.Nat Rev Genet. 2002; 3: 285-298Crossref PubMed Scopus (1646) Google Scholar Consistent with this theory, preliminary data reported by Tate et al.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar suggest that the SCN1A IVS5N+5 G→A polymorphism may control the proportion of the two transcript forms in adults. This exon 5 alternative splicing event is conserved across all neuronally expressed sodium channels.8Copley RR Evolutionary convergence of alternative splicing in ion channels.Trends Genet. 2004; 20: 171-176Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar, 9Plummer NW Meisler MH Evolution and diversity of mammalian sodium channel genes.Genomics. 1999; 57: 323-331Crossref PubMed Scopus (169) Google Scholar Although not well studied for SCN1A, there is extensive evidence of the developmental regulation of this splicing event at other neuronally expressed sodium channels.8Copley RR Evolutionary convergence of alternative splicing in ion channels.Trends Genet. 2004; 20: 171-176Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar, 9Plummer NW Meisler MH Evolution and diversity of mammalian sodium channel genes.Genomics. 1999; 57: 323-331Crossref PubMed Scopus (169) Google Scholar The sequence conservation across neuronally expressed sodium channels suggests that this also may occur for SCN1A, as indicated by semiquantitative assays reported by Tate et al.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar The molecular control of developmentally regulated sodium-channel alternative splicing is unknown. Recently, several neuronally expressed splice-modifier proteins have been identified, including neuro-oncological ventral antigen 1 and 2 (Nova1 and Nova2, respectively) and sodium channel modifier 1 (Scnm1).10Buchner DA Trudeau M Meisler MH SCNM1, a putative RNA splicing factor that modifies disease severity in mice.Science. 2003; 301: 967-969Crossref PubMed Scopus (97) Google Scholar, 11Ule J Darnell RB RNA binding proteins and the regulation of neuronal synaptic plasticity.Curr Opin Neurobiol. 2006; 16: 102-110Crossref PubMed Scopus (133) Google Scholar, 12Ule J Jensen KB Ruggiu M Mele A Ule A Darnell RB CLIP identifies Nova-regulated RNA networks in the brain.Science. 2003; 302: 1212-1215Crossref PubMed Scopus (762) Google Scholar It is possible that these neuronally expressed splice-modifier proteins may contribute to the alternative splicing of exon 5 of SCN1A. Importantly, the genomic region in and around exon 5 contains several putative Nova-binding sites (fig. 1), suggesting a possible role of Nova proteins in regulation of transcript splicing.5Ule J Stefani G Mele A Ruggiu M Wang X Taneri B Gaasterland T Blencowe BJ Darnell RB An RNA map predicting Nova-dependent splicing regulation.Nature. 2006; 444: 580-586Crossref PubMed Scopus (379) Google Scholar Given the association of this SCN1A polymorphism with altered doses of antiepileptic drugs, it is possible that drug doses found clinically may be influenced by the proportion of SCN1A transcripts in the neonatal form. Elucidating the genetic and molecular process that regulates this event not only may help in the prediction of how patients may respond to particular antiepileptic drugs but also may reveal a means for driving the splice-variant transcript proportion to the drug-responsive state in patients with the unfavorable (AA) genotype. Therefore, this study sought to quantify the effect of SCN1A IVS5N+5 G→A genotype on exon 5 splice-variant proportions and to determine the effects of neuronally expressed splice-modifier proteins on the controlling of this splice event. This study was approved by the Joint Research Ethics Committee of the Institute of Neurology and the National Hospital for Neurology and Neurosurgery and by the Duke University Institutional Review Board. All patients gave written informed consent for use of resected brain tissue for research. Forty-three temporal neocortical tissue samples were obtained from material resected from patients who underwent therapeutic surgery for drug-resistant mesial temporal lobe epilepsy, in accordance with routine clinical protocols. All tissue used for research was surplus to diagnostic requirements. Subjects were aged 18–60 years, and the group consisted of 15 men and 28 women. The population of patients with mesial temporal lobe epilepsy was the same as in the 2005 study by Tate et al.,3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar although random subjects were selected for both studies (thus it is possible that there may be an overlap of subjects in this study and subjects in the study by Tate et al.). All subjects had received a diagnosis of hippocampal sclerosis that was confirmed histologically. The hippocampal tissue was excluded from analysis because of potential confounding issues of neuronal loss. All tissue was flash frozen in liquid nitrogen and was stored at −80°C until use. Twenty-nine nonepileptic control temporal cortex tissue samples were obtained through the Kathleen Price Bryan Alzheimer’s Disease Brain Bank at Duke University. All patients gave written informed consent for use of the brain tissue for research. The control tissue was collected postmortem (time to collection ranged from 1 to 30 h postmortem, with controlling for agonal state). Subjects were aged 56–90 years, and the group consisted of 14 men and 15 women. Ethical permission for the use of this tissue was provided by Duke University Medical Center Institutional Review Board. DNA was extracted using the Trizol method (Invitrogen), per the manufacturer’s instructions. DNA quantity and quality were assessed spectrophotometrically. Samples were genotyped for the SCN1A IVS5N+5 G→A polymorphism (rs3812718) by use of a custom designed TaqMan-based allelic discrimination assay (Applied Biosystems). Assay primer and fluorescently labeled probe sequences were as follows (all are 5′→3′): forward primer, TTTCAGAGTCTTGAGAGCTTTGAAAA; reverse primer, TGTGACGTACCTGTAATAGGGAGTTC; G probe, FAM-AATTCCAGGTAAGAAGTGA-MGB; and A probe, VIC-AATTCCAGGTAAAAAGTGA-MGB. Assay conditions were in accordance with manufacturers' standard protocols. Fluorescence outputs were quantified in real time by use of a 7900HT Fast Real Time PCR System, and the data were analyzed using SDS software, version 2.2.2 (Applied Biosystems). Total RNA was extracted from brain tissue by use of the RNeasy Lipid Tissue Purification Kit (Qiagen), in accordance with the manufacturer’s instructions. RNA from cells was extracted using the RNeasy Purification Kit (Qiagen). The resulting RNA was quantified spectrophotometrically, and 1 μg of total RNA was reverse transcribed into cDNA by use of a High Capacity cDNA Synthesis Kit (Applied Biosystems), in accordance with product instructions. An Exon Trapping system (Gibco) was used to assess the genetic and molecular control of the alternative splicing of exon 5N and 5A in SCN1A. A genomic DNA fragment of 696 bp containing exon 5N and 5A of SCN1A was PCR amplified using a primer set—forward primer, GATCGAATTCCCAGAGTGACAACAAGGGTGT, and reverse primer, GATCGGATCCGGTCACCTTGACCTCAAATTACA—and was subcloned into a pSPL3 vector. The A and G forms of the minigene were obtained by separate amplification of homozygous individuals. SCNm1 (MIM 608095), NOVA1 (MIM 602157), and NOVA2 (MIM 601991) cDNA with a full-length ORF was PCR amplified with cDNAs from neuroblastoma cells, SK-N-BE(2), by use of primer sets (all are 5′→3′): SCNm1 forward primer, GATCAGATCTATGTCTTTCAAGAGGGAAGGA; SCNm1 reverse primer, GATCGAATTCTCAGTCCAAGGGGAGATCAGG; NOVA1 forward primer, ATTACTCGAGAACATGATGGCGGCAGCTCC; NOVA1 reverse primer, ATTAGGTACCTCAACCCACTTTCTGAGGATT; NOVA2 forward primer, TATTACTCGAGATGGAGCCCGAGGCCCCGGAT; and NOVA2 reverse primer, ATTAGGTACCCAGGCCTCATCCCACTTTCT. Total RNA was extracted from the SK-N-BE(2) cells and was converted to cDNA by use of a High-Capacity cDNA Archive Kit (Applied Biosystems) for NOVA2 cloning. PCR was performed with a high-fidelity DNA polymerase, Platinum Pfx (Invitrogen). The PCR product was subcloned into pcDNA3.1 (Invitrogen), with an N-terminal hemagglutinin (HA) tag. All the sequences were verified by DNA sequencing. To confirm the causal polymorphism of the observed effects in the minigene system, the SCN1A IVS5N+5 G→A polymorphism (rs3812718) was introduced in the G construct by use of a QuikChange XL Site-Directed Mutagenesis Kit (Stratagene). HEK293 cells were grown to 80% confluence in 6-well plates. The cells were transfected with 0.1 μg of SCN1A minigene (A, G, or equal quantities of A and G) and varying amounts of NOVA2 plasmids by use of Lipofectamine 2000 (Invitrogen), in accordance with the manufacturer’s guide. After 48 h, cells were harvested for RNA extraction and western-blot analyses. The same experimental design was employed for transfection of SCNm1 and NOVA1 plasmids; however, this was limited to a single transfection of 1 μg. TaqMan-based real-time PCR was used to determine the percentage of SCN1A transcripts containing exon 5N and 5A. Assays were custom designed through Applied Biosystems, and primers or probe sequences were used in these assays as follows (all are 5′→3′). For SCN1A 5A transcript quantification in human brain tissue, forward primer AACTGGCTCGATTTCACTGTCATTA, reverse primer ATTGCCCAGGTCCACAAACT, and probe FAM-CTGTGACGTACGCAAATG-NFQ were used. For SCN1A 5N transcript quantification, forward primer TGGCTCGATTTCACTGTCATTACAT, reverse primer TCTGAAAGTGCGAAGAGCTGAAAA, and probe FAM-CCTAGGTTTACAAATTCTG-NFQ were used. For SCN1A 5A transcript quantification in HEK293 cells, forward primer GTGAACTGCACTGTGACAAGCTGC, reverse primer CACCTGAGGAGTGAATTGGTCG, and probe FAM-CCGAGCATTGAAGACGA-NFQ were used. For SCN1A 5N transcript quantification, forward primer GTGAACTGCACTGTGACAAGCTGC, reverse primer TCTGAAAGTGCGAAGAGCTGAAAA, and probe FAM-CCTAGGTTTACAAATTCTG-NFQ were used. SCNm1, NOVA1, NOVA2, and β-actin mRNA expression levels were quantified using commercially available TaqMan assays (Applied Biosystems). Fluorescence outputs were quantified in real time by use of a 7900HT Fast Real Time PCR System, and the data were analyzed using SDS software, version 2.2.2 (Applied Biosystems). In subjects with the AA genotype, the output data often were below the limit of detection for the 5N assay. In these cases, the percentage of SCN1A transcripts containing exon 5N was considered to be zero. The effect of Nova2 on alternative splicing events in 17 ion-channel genes (splicing events are defined in table 1) was investigated in cells transfected with 0.1 μg of SCN1A minigene with and without 1 μg NOVA2 cDNA transfection (the same preparation was used for investigation of the effects of Nova2 on SCN1A exon 5 splicing). TaqMan-based real-time PCR assays were used to quantify splice-variant ratios (variant:common transcript forms, as defined in table 1) by targeting unique exon-exon boundaries.Table 1Alternative Splicing Events in Ion-Channel Genes That Were Screened for Nova2-Mediated RegulationGenBank Accession Number for Splice FormGeneCommonVariantCACNA1ANM_023035BE972738CACNA1BNM_000718M94173CACNA1GNM_018896NM_198376CACNB1NM_000723NM_199247CACNB4NM_000726AY054985CLCN7NM_001287AK096963FXYD6NM_022003BP372334GRIA1NM_000827A46050GRIA3NM_007325NM_000828HTR3ANM_000869BG341613KCNJ15NM_170736CQ732921KCNK1NM_002245AV733795KCNN1NM_002248BM718136KCNN2NM_021614BG769522KCNQ2NM_004518AY358189MCOLN1NM_020533CA489568SCN1ANM_006920CQ727958 Open table in a new tab Protein was extracted with 2% SDS from cell pellet. Protein was separated on an NuPAGE 4%–12% Bis-Tris gel (Invitrogen). The gel was blotted to nitrocellulose membrane (BioRad). The membrane was blocked with 5% skim milk. Then, the membrane was incubated with rabbit anti-HA antibody (1:500 [Abcam]) overnight at 4°C and subsequently was incubated with horseradish peroxidase–conjugated anti-rabbit antibody (1:5,000 [Amersham]) for 1 h at room temperature. The signal was detected by Amersham ECL kit. Rabbit anti–β-actin antibody (1:2,000 [Abcam]) was used for a loading control. ImageJ software was used to perform densitometric analysis. One-way analysis of variance (ANOVA) was used to determine the statistical significance of the SCN1A exon 5N splice variant compared with SCN1A IVS5N+5 G→A genotype correlation in human brain tissue. Statistical analyses were performed separately in control and mesial temporal lobe epilepsy samples, in addition to a combined subject analysis. Tukey post hoc tests were used to determine the statistically significant differences between splicing proportions in the three genotype groups. The effects of NOVA2 mRNA expression (NOVA2) on the percentage of SCN1A mRNA transcripts containing exon 5N (SCN1A 5N%) were fit to a sigmoidal Emax model defined as follows:SCN1A 5N%=5N%0+Emax×E(NOVA250γ)E(NOVA250γ)+NOVA2γ ,(1) where Emax is the maximum NOVA2-mediated change in SCN1A 5N%, E(NOVA2γ50) is amount of NOVA2 needed to elicit one-half the maximum effect on SCN1A 5N%, γ is the shape factor governing the relationship between SCN1A 5N% and NOVA2 mRNA expression, and 5N%0 is the percentage of transcripts containing exon 5N in the presence of no or very low quantities of NOVA2. Linear regression was used to assess the effects of splice-modifier protein mRNA expression on the proportion of SCN1A transcripts for all subjects and for the individual genotypes. Student’s t test was used to compare splice-variant ratios of ion-channel genes in HEK293 cells transfected with SCN1A minigene and in the presence and absence of 1-μg NOVA2 cDNA transfection (n=3 independent replications per group). In human brain tissue, the SCN1A IVS5N+5 G→A polymorphism had a substantial effect on the percentage of transcripts containing exon 5N (neonatal form) of SCN1A (fig. 2A). Individuals with the AA genotype had a mean (±SEM) of 0.7%±1% of SCN1A transcripts in the neonatal form, whereas subjects with the GG genotype had 41%±9% of transcripts containing exon 5N. The G allele elicited a dominant effect, with those with the AG genotype having 28%±4% of transcripts in the neonatal form. The presence or absence of a history of temporal lobe epilepsy did not affect the correlation between genotype and the alternative splicing event. However, the magnitude of the proportion of SCN1A neonatal transcripts was slightly altered in the disease state, with a trend toward a lower percentage of transcripts containing exon 5N in patients with mesial temporal lobe epilepsy (fig. 2B). We also investigated a possible role for neuronally expressed splice-modifier proteins—Nova1, Nova2, and Scnm1—in controlling the SCN1A exon 5 splicing event, by assessing the correlation between the percentage of transcripts containing exon 5N and the mRNA expression within individual genotypes. We found no effect of Nova1 and Scnm1 on modulation of the SCN1A exon 5 splicing event, collectively or for individual genotypes. Although no NOVA2-mediated modulation of SCN1A exon 5 splicing was detected when all subjects were evaluated, a statistically significant correlation (uncorrected for multiple testing) was observed in human brain tissue between NOVA2 mRNA expression and the percentage of SCN1A transcripts containing exon 5N for the AA genotype (fig. 3). NOVA2 mRNA expression was equivalent in both mesial temporal lobe epilepsy and control brain tissue, and a similar correlation was observed in both experimental groups (data not shown). We used a minigene expression system to confirm a causal relationship between these correlations observed in human adult brain tissue. First, we tested the role of the SCN1A IVS5N+5 G→A polymorphism. In this cellular system, the experimentally simulated GG genotype resulted in 51% of transcripts containing exon 5N, whereas the percentage was significantly reduced for the AA genotype, to 7% (fig. 4). The AG genotype resulted in 40% of transcripts being in the neonatal form (fig. 4); however, we note that we did not quantify the true copy numbers of the A and G alleles in this study. Because of the presence of other SNPs in the minigene construct (fig. 1), we also performed mutagenesis studies to confirm that the SCN1A IVS5N+5 G→A polymorphism was eliciting the observed splicing changes and that the alternations were not due, in whole or in part, to other polymorphisms in the minigene construct. Direct sequencing of the G minigene construct revealed the reference sequence indicated in figure 1. The A construct had the alternate sequence at all the SNP locations (rs2195144 [A→G], rs2217199 [T→C], rs3812719 [G→T], and rs3812718 [G→A]). To confirm that the SCN1A IVS5N+5 G→A polymorphism is the causal SNP in the minigene system, we mutated the SCN1A IVS5N+5 G→A polymorphism in the G construct to the A form (at the IVS5N+5 site) and tested the effects of this change on the percentage of transcripts containing exon 5N. This experiment revealed no statistical difference in transcripts containing exon 5N between the original AA minigene system and the GG-mutated system (data not shown). These findings confirm that the SCN1A IVS5N+5 G→A polymorphism is, in fact, the casual variant in the modulation of the alterations in SCN1A exon 5 splicing. Next, we tested how overexpression of the common neuronally expressed splice-modifier proteins Scnm1, Nova1, and Nova2 influences SCN1A splicing. Overexpression of neither Scnm1 nor Nova1 showed any effects on SCN1A exon 5 alternative splicing in the minigene cellular system investigated here (data not shown). However, increasing Nova2 decreased the quantity of SCN1A transcripts containing the adult form of exon 5 while it simultaneously increased the number of transcripts containing the neonatal form. Overall, this translated into Nova2 increasing the percentage of SCN1A transcripts containing exon 5N (fig. 5). However, the effects of Nova2 varied for the different SCN1A IVS5N+5 G→A genotypes. For the GG genotype, NOVA2 mRNA expression at the observed maximum elicited an ∼2.5-fold increase in the proportion of SCN1A transcripts containing exon 5N, whereas, for the AA genotype, the expression change produced an ∼20-fold change in the splicing event. Similar to the observations in human brain tissue, these findings support a greater effect of Nova2 in the AA genotype. We note that, in these experiments, the maximum changes in splicing elicited by Nova2 in the AA genotype were predicted only and were not directly observed (i.e., an experimentally derived maximum plateau of effect was not obtained in these experiments) (fig. 5). Despite the failure to reach the plateau, a much larger change from baseline in SCN1A alternative splicing was observed, despite a lower maximum-observed increase in NOVA2 mRNA expression in the AA minigene cellular system. Increases in NOVA2 mRNA expression were associated with increases in protein expression for both genotypes (fig. 5), which is consistent with the hypothesis that expression of the splice-modifier protein is modulating the observed effects on SCN1A alternative splicing. Importantly, we emphasize that, with modulation of Nova2 expression, the SCN1A exon 5 splice-variant percentage in the AA system approached that observed for the GG minigene expressing endogenous levels of NOVA2 mRNA (fig. 5). Finally, we tested the specificity of Nova2-mediated alternative splicing by evaluating the effects of increases in expression of this protein on modulation of splicing events in 17 neuronally expressed ion-channels genes. We found that Nova2 exhibited no effect on splice-variant ratios for any of the 17 alternative splicing events defined in table 1 (data not shown). The SCN1A IVS5N+5 G→A polymorphism, located in the splice-donor site of exon 5N, had a striking effect on the percentage of transcripts in the neonatal form of the SCN1A gene in adults with and without epilepsy (fig. 2A). In those individuals with the GG genotype, up to 50% of the transcripts include the neonatal version of exon 5, compared with an often undetectable level of the neonatal version in some subjects with the AA genotype. This correlation reveals a statistically significant increase in exon 5 splicing between the AG and GG genotypes; however, the magnitude of the change is less between the AG and GG genotypes compared with that between the AA and AG genotypes both in vivo and in vitro (figs. 2A and 4, respectively). It is unclear why the G allele appears to elicit a dominant effect on exon 5 splicing (figs. 2A and 4). The association was observed clearly both in autopsy samples from individuals who do not have known neurological conditions and in resection tissue samples from patients with a history of refractory temporal lobe epilepsy (fig. 2A), providing evidence that this splice variation is likely to be largely genetically controlled, rather than a seizure-induced event. Tate et al.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar reported a very modest effect of genotype on SCN1A exon 5 alternative splicing in patient temporal neocortical tissue only, not in control brain tissue.3Tate SK Depondt C Sisodiya SM Cavalleri GL Schorge S Soranzo N Thom M Sen A Shorvon SD Sander JW et al.Genetic predictors of the maximum doses patients receive during clinical use of the anti-epileptic drugs carbamazepine and phenytoin.Proc Natl Acad Sci USA. 2005; 102: 5507-5512Crossref PubMed Scopus (255) Google Scholar The discrepancy could be attributed to the lack of quantitative real-time PCR in the original study, coupled with the comparatively smaller sample sizes evaluated. It is also possible that the difference may simply be due to a different population of control subjects used in this study. We note that animal models of epilepsy also indicate an association between increased exon 5N levels and seizures,13Gastaldi M Bartolomei F Massacrier A Planells R Robaglia-Schlupp A Cau P Increase in mRNAs encoding neonatal II and III sodium channel α-isoforms during kainate-induced seizures in adult rat hippocampus.Brain Res Mol Brain Res. 1997; 44: 179-190Crossref PubMed Scopus (48) Google Scholar, 14Aronica E Yankaya B Troost D van Vliet EA Lopes da Silva FH Gorter JA Induction of neonatal sodium channel II and III α-isoform mRNAs in neurons and microglia after status epilepticus in the rat hippocampus.Eur J Neurosci. 2001; 13: 1261-1266Crossref PubMed Google Scholar which is not apparent in our samples by the quantitative assays used here. In fact, in our samples, there was a slightly reduced percentage of neonatal transcripts in the brain tissue collected from mesial temporal lobe epilepsy for each genotype group (fig. 2B). This opposing finding may simply be due to differences between the animal models and humans. Although additional work is needed to establish the role of seizures in mediating SCN1A exon 5 alternative splicing, our data clearly show that the genotype predominantly drives this alternative splicing event, regardless of disease status. Consistent with the correlation between SCN1A IVS5N+5 G→A genotype and alternative transcript splicing in human brain tissue, the exon 5N and 5A minigene system also demonstrated that the expression level of exon 5N was directly affected by the genotype (fig. 4). The proportion found is strikingly similar to the data from human samples. These data reveal virtually no change in the level of exon 5A expression, indicating that the effect of the SCN1A IVS5N+5 G→A polymorphism appears to be mainly the result of increased exon 5N expression, a finding that also is consistent with the observations in humans. The SCN1A IVS5N+5 G→A polymorphism is located 6 bases into the intron in the 5′ splice site following exon 5N (fig. 1). The guanine base at this nucleotide position is highly conserved, with an estimated presence of 75%–86% in human introns, whereas the adenine nucleotide at this base position is far less common, with a presence of ∼5%–11%.15Zhang MQ Statistical features of human exons and their flanking regions.Hum Mol Genet. 1998; 7: 919-932Crossref PubMed Scopus (319) Google Scholar Given that the SCN1A IVS5N+5 G→A polymorphism causes a change from a highly conserved sequence to a lesser conserved sequence, it was not unexpected that a functionally weaker 5′ splice site at exon 5N was seen in the presence of this change. This unambiguous evidence of genetic control of a splicing event in the SCN1A gene, in combination with the new data further supporting genotype dependency of phenytoin dose in patients with epilepsy,4Tate S Singh R Hung C Tai J Depondt C Cavalleri G Sisodiya S Goldstein D Horng-Huei L A common polymorphism in the SCN1A gene associates with phenytoin serum levels at maintenance dose.Pharmacogenet Genomics. 2006; 16: 721-726Crossref PubMed Scopus (85) Google Scholar provides an example of a functionally significant genetic polymorphism that appears to directly influence the therapeutic dose of antiepileptic medications and could have other currently unknown phenotypic consequences. The genetically controlled splicing variation and, consequently, the altered dose requirements of carbamazepine and phenytoin in humans could be the result of the product of transcript variant of the SCN1A gene producing altered electrical signaling in patients with epilepsy (i.e., epilepsy is worse in patients with the AA genotype, therefore requiring higher doses of antiepileptic medication to effectively control the seizures). However, our observations also could be the result of altered pharmacology of the antiepileptic drugs in the presence of the different genotypes (i.e., altered potency and/or efficacy of the drugs at the SCN1A gene product). Consistent with the latter possibility, an analogous functional splicing alteration in the sodium-channel gene of the German cockroach has been reported to cause altered sensitivity to deltamethrin.16Tan J Liu Z Nomura Y Goldin AL Dong K Alternative splicing of an insect sodium channel gene generates pharmacologically distinct sodium channels.J Neurosci. 2002; 22: 5300-5309Crossref PubMed Google Scholar, 17Du Y Liu Z Nomura Y Khambay B Dong K An alanine in segment 3 of domain III (IIIS3) of the cockroach sodium channel contributes to the low pyrethroid sensitivity of an alternative splice variant.Insect Biochem Mol Biol. 2006; 36: 161-168Crossref PubMed Scopus (15) Google Scholar Furthermore, we cannot rule out the possibility that the electrical properties of SCN1A at baseline and in the presence of drugs differ in the presence of altered transcript-variant quantities. To address this important question, future work will focus on identifying the functional differences between the products of the SCN1A exon 5N and 5A transcripts in the absence and presence of drugs. In addition to the effects of the SCN1A polymorphism, we also have documented an important role of Nova2 in modulating the proportion of SCN1A alternative transcripts. Because the two transcripts appear to result in different pharmacological properties, this work clearly highlights the potential therapeutic importance of control of splice variation. For example, because the neonate form appears to be more responsive to some antiepileptic drugs, it would seem possible, in principle, to make some forms of epilepsy more easily treatable by increasing the proportion of the neonate transcript through manipulation of NOVA2 expression or other modifier genes. Given these findings, Nova2 may, in the future, prove to be a drug target for improvement of pharmacologic responsiveness in a population of patients with refractory epilepsy. However, we note that Nova2 has been shown to modulate numerous alternative splicing events in vivo,18Ule J Ule A Spencer J Williams A Hu JS Cline M Wang H Clark T Fraser C Ruggiu M et al.Nova regulates brain-specific splicing to shape the synapse.Nat Genet. 2005; 37: 844-852Crossref PubMed Scopus (373) Google Scholar and it could be argued that such a strategy may suffer from extensive nonspecific effects. As a preliminary step toward evaluating the generality of NOVA2 effects on neuronal splicing, we looked at 17 alternative splicing events in ion-channel genes in the presence and absence of increases in Nova2 expression. Of these selected alternative splicing events (table 1), none exhibited Nova2-mediated alternative splicing, although additional work is needed to evaluate comprehensively the specificity of Nova2 in modulating neuronal splicing in humans. These data highlight a specific regulatory role of Nova2 in controlling a clinically significant alternative splicing of SCN1A and provide novel avenues to pursue in the attempt to understand and regulate genetic response to medications. The work was supported by the Institute for Genome Sciences and Policy, the UK National Society for Epilepsy (to S.K.T.), the UK Medical Research Council (to S.M.S. and D.B.G.), and the Guarantors of Brain (to A.S.). Plasmid pSPL3 was kindly provided by Dr. Douglas Marchuk (Duke University). We also thank Kenneth Cronin for his technical assistance with this project.
Referência(s)