Detection of three closely located single nucleotide polymorphisms in the EAAT2 promoter: comparison of single-strand...

Single-strand conformational polymorphism (SSCP) is still a frequently used genotyping method across different fields for the detection of single nucleotide polymorphisms (SNPs) due to its simplicity, requirement for basic equipment accessible in most laboratories and low cost. This technique was previously used to detect rs4354668:A > C (g.-181A > C) SNP in the promoter of astroglial glutamate transporter (EAAT2) and the same approach was initially used here to investigate this promoter region in a cohort of newborns.

Unexpectedly, four distinct DNA migration patterns were identified by SSCP. Sanger sequencing revealed two additional SNPs: g.-200C > A and g.-168C > T giving a rise to a total of ten EAAT2 promoter variants. SSCP failed to distinguish these variants reliably and thus pyrosequencing assays were developed. g.-168C > T was found in heterozygous form in one infant only with minor allele frequency (MAF) of 0.0023. In contrast, g.-200C > A and -181A > C were more common (with MAF of 0.46 and 0.49, respectively) and showed string evidence of linkage disequilibrium (LD). In a systematic comparison, 16% of samples were miss-classified by SSCP with 25-31% errors in the identification of the wild-type and homozygote mutant genotypes compared to pyrosequencing or Sanger sequencing. In contrast, SSCP and pyrosequencing of an unrelated single SNP (rs1835740:C > T), showed 94% concordance.

Our data suggest that SSCP cannot always detect reliably several closely located SNPs. Furthermore, caution is needed in the interpretation of the association studies linking only one of the co-inherited SNPs in the EAAT2 promoter to human diseases.