Polymorphisms in GRIK4, HTR2A, and FKBP5 Show Interactive Effects in Predicting Remission to Antidepressant Treatment

Neuropsychopharmacology (2010) 35, 727740& 2010 Nature Publishing Group All rights reserved 0893-133X/10 $32.00

http://www.neuropsychopharmacology.org

Web End =www.neuropsychopharmacology.org

Polymorphisms in GRIK4, HTR2A, and FKBP5 Show Interactive Effects in Predicting Remission to Antidepressant Treatment

Sonja Horstmann*,1, Susanne Lucae1, Andreas Menke1, Johannes M Hennings1, Marcus Ising1, Darina Roeske1, Bertram Mller-Myhsok1, Florian Holsboer1 and Elisabeth B Binder1

1Max Planck Institute of Psychiatry, Munich, Germany

Single-nucleotide polymorphisms (SNPs) in the FKBP5, GRIK4, and HTR2A genes have been shown to be associated with response to citalopram treatment in the STAR*D sample, but only associations with FKBP5 have so far been tested in the Munich Antidepressant Response Signature (MARS) project. Response and remission of depressive symptoms after 5 weeks of antidepressant treatment were tested against 82 GRIK4 and 37 HTR2A SNPs. Association analysis was conducted in about 300 depressed patients from the MARS project, 10% of whom had bipolar disorder. The most predictive SNPs from these two genes and rs1360780 in FKBP5 were then genotyped in a total of 387 German depressed in-patients to analyze potential additive and interactive effects of these variants. We could not replicate previous findings of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study in our sample. Although not statistically significant, the effect for the best GRIK4 SNP of STAR*D (rs1954787, p 0.076, p

corrected

0.98) seemed to be in the same direction. On the other hand, the nominally significant association with the top HTR2A SNPs of STAR*D (rs7997012, allelic, p 0.043, p

corrected

0.62) was with the opposite risk allele. The GRIK4 SNP (rs12800734, genotypic, p 0.0019, p corrected

0.12) and

0.02), which showed the strongest association with remission in our sample, had not been reported previously. Associations across all genetic markers within the GRIK4 (genotypic, p 0.022) or HTR2A

(genotypic, p 0.012) locus using the Fishers product method (FPM) were also significant. In all 374 patients, the best predictive model

included a main effect for GRIK4 rs12800734 and two significant interactions between GRIK4 rs12800734 and FKBP5 rs1360780, and GRIK4 rs12800734 and HTR2A rs17288723. This three SNP model explained 13.1% of the variance for remission after 5 weeks (p 0.00051 for the model). Analyzing a sub-sample of 194 patients, plasma ACTH (p 0.002) and cortisol (p 0.021) responses of

rs12800734 GG (GRIK4) carriers, who also showed...