Fine Mapping of a Dravet Syndrome Modifier Locus on Mouse Chromosome 5 and Candidate Gene Analysis by RNA-Seq

Abstract

A substantial number of mutations have been identified in voltage-gated sodium channel genes that result in various forms of human epilepsy. SCN1A mutations result in a spectrum of severity ranging from mild febrile seizures to Dravet syndrome, an infant-onset epileptic encephalopathy. Dravet syndrome patients experience multiple seizures types that are often refractory to treatment, developmental delays, and elevated risk for SUDEP. The same sodium channel mutation can produce epilepsy phenotypes of varying clinical severity. This suggests that other factors, including genetic, modify the primary mutation and change disease severity. Mouse models provide a useful tool in studying the genetic basis of epilepsy. The mouse strain background can alter phenotype severity, supporting a contribution of genetic modifiers in epilepsy. The Scn1a+/- mouse model has a strain-dependent epilepsy phenotype. Scn1a+/- mice on the 129S6/SvEvTac (129) strain have a normal phenotype and lifespan, while [129xC57BL/6J]F1-Scn1a+/- mice experience spontaneous seizures, hyperthermia-induced seizures and high rates of premature death. We hypothesize the phenotypic differences are due to strain-specific genetic modifiers that influence expressivity of the Scn1a+/- phenotype. Low resolution mapping of Scn1a+/- identified several Dravet syndrome modifier (Dsm) loci responsible for the strain-dependent difference in survival. One locus of interest, Dsm1 located on chromosome 5, was fine mapped to a 9 Mb region using interval specific congenics. RNA-Seq was then utilized to identify candidate modifier genes within this narrowed region. Three genes with significant total gene expression differences between 129S6/SvEvTac and [129xC57BL/6J]F1 were identified, including the GABAA receptor subunit, Gabra2. Further analysis of Gabra2 demonstrated allele-specific expression. Pharmological manipulation by clobazam, a common anticonvulsant with preferential affinity for the GABRA2 receptor, revealed dose-dependent protection against hyperthermia-induced seizures in Scn1a+/- mice. These findings support Gabra2 as a genetic modifier of the Scn1a+/- mouse model of Dravet syndrome.

Fig 1. Fine mapping of Dsm1 with ISC strains.

A. Dsm1 ISC lines A-E carry varying 129 derived chromosome 5 segments (colors) on a congenic B6 background (grey). B. Survival of resulting F1.KO offspring was monitored to 8 weeks of age. Kaplan Meier survival plot is shown. Survival was significantly improved in mice with homozygous 129/129 alleles in ISC-A (p<0.01), ISC-B (p<0.0001), ISC-C (p<0.05), and ISC-D (p < 0.0002) compared to F1.KO heterozygous controls. P values were determined by LogRank Mantel-Cox test.

Fig 2. Evaluation of Dsm1 Candidate Genes by RNA-Seq.

A. The refined map interval is shown with mouse RefSeq gene annotations. The genes with strain-dependent differential expression are indicated by grey shading (Ensembl Mouse GRCm38.p4)[53]. B. Three genes within the Dsm1 interval showed significant differences in brain expression between wildtype 129 and [129xB6]F1 mice: Gabra2 (q < 0.0029**), Nsun7 (q < 0.02*), and Bend4 (q< 0.02*). Shown are RNA-Seq FPKM values and 95% confidence intervals calculated by String Tie. q-values are false discovery rate-adjusted p-values.

Fig 3. Allele specific Gabra2 transcript and protein expression.

A. Relative brain Gabra2 transcript levels between mice carrying 129/129, 129/B6 or B6/B6 alleles in ISC-D were determined by ddRT-PCR and expressed are as a ratio of Gabra2 to Tbp. Highest levels of Gabra2 transcript are observed with homozygous 129/129 alleles, intermediate levels with heterozygosity, and lowest levels with B6 homozygosity. B and C. Brain GABRA2 protein expression for mice carrying 129/129, 129/B6 or B6/B6 alleles in ISC-D was assayed by immunoblotting using a rabbit polyclonal antibody for GABRA2 (Phosphosolutions, 822-GA2C; 1:2000) and a mouse monoclonal antibody for mortalin (NeuroMab 75–127; 1:1000) as a loading control. Relative levels were determined by densitometry and expressed as a ratio of GABRA2 to mortalin. Relative GABRA2 protein expression followed the same trend as Gabra2 transcript. Scatter plot data points represent samples from individual mice (n = 5–6 biological replicates). Average values are depicted as a horizontal bar and error bars represent the SD. Statistical comparison between groups was made using ANOVA with Tukey’s post-hoc tests.

Fig 4. Pharmacological modulation of Gabra2 by clobazam.

F1.Scn1a^+/- mice underwent hyperthermia-induced seizure threshold testing after IP injections of vehicle or clobazam. The temperature at which generalized tonic-clonic seizure activity occurred is shown. Clobazam administration provided dose-dependent protection against hyperthermia-induced seizures. p values were determined by Mantel-Cox log-rank test (**p<0.01; ***p<0.001).