No association between polymorphisms in the glutamate transporter SLC1A2 and Parkinson disease

Essential tremor (ET) and Parkinson disease (PD) are clearly distinct disorders. However, there is considerable clinical and pathophysiological overlap.¹ ET patients are at an approximately fourfold risk to develop PD compared to population controls. Both, ET and PD, have a genetic component. For PD numerous monogenic forms are known and genome-wide-association-studies (GWAS) in sporadic patients have uncovered a number of disease associated genetic variants. For ET only two GWAS studies have been performed up to date.^{2, 3} In one of them we discovered an association between variants in the major glutamate reuptake transporter of the brain (EAAT2 encoded by the SLC1A gene) and ET.³ The strongest association was found for the intronic single nucleotide polymorphism (SNP) rs3794087. To investigate whether SNPs in or around the SLC1A2 gene are also associated with PD we undertook a two-part study. The study was approved by the local ethics committees and written informed consent was obtained from all participants. Firstly, we investigated the association between our ET lead SNP rs3794087 and PD in three previously described large German PD case/control samples (1798 PD patients, 1482 controls, a detailed description of the samples can be found in reference ⁴). Genotyping was performed using a TaqMan^® assay. Association was assessed using PLINK.⁵ We calculated Hardy-Weinberg equilibrium (HWE, p-exclusion <0.001) and association in the allelic model. Power calculations can be found in supporting figure 1. The genotyping rate of rs3794087 was >95% and the SNP was in HWE. No significant p-values were obtained for rs3794087, neither in the complete sample (p=0.775), nor in the subsamples from Kiel (p=0.623), Lübeck (p=0.998) or Tübingen (p=0.766)(Table 1). Secondly, we analysed all 45 SNPs in the region of the SLC1A gene typed in the first stage sample of a large GWAS performed for PD (1713 PD patients, 3978 controls, a detailed description of the samples can be found in reference ⁶). All SNPs were in HWE. Two SNPs (rs12294045, p=0.034; rs3847618, p=0.035) yielded nominally significant p-values. Both p-values do not withstand Bonferroni correction for multiple testing (Bonferroni corrected alpha-level p<0.001). The ET associated lead SNP rs3794087 was excluded from the GWAS analysis because the samples are partially overlapping with the Tübingen samples from the first analysis. A list of all SNPs with allele frequencies and p-values can be found in supporting table 2. In summary, we conclude that we didn’t find any evidence for association between SNPs in the SLC1A2 gene and PD neither in our three German samples nor in the GWAS-sample which contains American and German samples. A previous GWAS identified association between SNPs in the LINGO1 gene and ET.² Also in this case most groups, including our own, could not find association of these variants with PD.⁷ SLC1A2 encodes the predominant glutamate reuptake transporter EAAT2 in the brain which removes glutamate from the synaptic cleft. Interestingly EAAT2 is strongly expressed in the inferior olive, a structure implicated in rhythm generation in ET but not in the substantia nigra which is subject to neurodegeneration in PD. Therefore we hypothesize that SLC1A2 is an ET specific gene, not involved in the pathogenesis of PD.

Table 1.

Association results between SNP rs3794087 and PD

German samples (MAF – minor allele frequency)
SNP rs3794087	MAF cases	MAF controls	p- value
whole sample	0.252	0.249	0.775
subs. Kiel	0.238	0.229	0.623
subs. Lübeck	0.273	0.273	0.998
subs. Tübingen	0.251	0.256	0.766