Abstracts

RATIONALE: Twin and family studies show that liability to absence epilepsy (AE) is largely genetic, but susceptibility genes in humans have not been identified. Voltage-sensitive ion channels (especially T-type and so-called HVA Ca++ channels), neurotransmitter receptors (including GABAB and some GABAA subunits), and at least one other protein controlling transmembrane ion flux (i.e, Na+-H+ exchanger) have been implicated in genetic and pharmacologic studies of spike-wave discharges and AE in animal models. Here we report preliminary results of association studies of candidate genes in childhood and juvenile AE. METHODS: We identify bi-allelic, single-nucleotide polymorphisms (SNPs) within candidate genes by database search. We also screen for novel SNPs by PCR and conformation-sensitive gel electrophoresis in a panel of 30 unrelated AE probands and characterize variations by sequencing. SNPs are genotyped by single-strand conformation analysis in a separate sample of AE trios, each consisting of two parents and an AE child. We use the transmission-disequilibrium test (TDT) to detect association between candidate gene SNPs and AE. The TDT eliminates the problem of spurious associations that may arise in case-control studies due to inadequate matching for genetic background. RESULTS: SNPs identified thus far are silent. In our present collection of 30 AE trios we have analyzed SNPs at SLC9A1 (sodium-hydrogen ion exchanger) and CACNA1G (T-type Ca++ channel). We found no evidence for association between AE and two SNPS in SLC9A1. A novel SNP in CACNA1G showed 20 transmissions of allele B vs 9 transmissions of allele A by heterozygous (AB) parents (p=.041, chi-square=4.17, 1 df). This nominally significant result must be interpreted cautiously in light of the multiple candidate genes (10) that are under investigation. CONCLUSIONS: These preliminary findings (1) are consistent with the hypothesis that variation within or near CACNA1G influences AE susceptibility, (2) justify a thorough examination of CACNA1G for nonsilent SNPs (i.e., variations affecting protein sequence or expression) in AE patients, and (3) warrant testing for association between CACNA1G and AE in a larger sample of trios.