Sensorimotor gating depends on polymorphisms of the serotonin-2A receptor and catechol-O-methyltransferase, but not on neuregulin-1 Arg38Gln genotype: a replication study

Background: Prepulse inhibition (PPI) of the acoustic startle response (ASR) is an operational measure of sensorimotor gating and a promising endophenotype of schizophrenia. We have recently shown that the linked serotonin-2A receptor (5-HT 2A R) A-1438G and T102C polymorphisms modulate PPI in schizophrenia patients. Moreover, it was shown that genetic variation in the catechol-O -methyltransferase (COMT) and the neuregulin-1 (NRG-1) proteins influences PPI in schizophrenia patients and healthy volunteers. Therefore, we aimed to replicate these results and investigated the impact of the related polymorphisms on PPI in healthy human volunteers. Methods: We analyzed the 5-HT 2A R A-1438G/T102C (rs6311/rs6313), the COMT Val158Met (rs4680), and the NRG-1 Arg38Gln (rs3924999) polymorphisms, assessed startle reactivity, habituation, and PPI of ASR in 107 healthy Caucasian volunteers. Results: Subjects homozygous for the 5-HT 2A R T102C-T/A-1438G-A allele showed increased PPI levels. In particular, male subjects with the COMT Met158Met-genotype showed also elevated PPI. The NRG-1 Arg38Gln genotype did not have a significant impact on PPI. Startle reactivity was not affected by any of the investigated polymorphisms. Conclusions: We confirmed in an independent sample of healthy volunteers that PPI is influenced by genetic variation in the 5-HT 2A R gene. The influence of the COMT Val158Met genotype on PPI appears gender-specific. These results underscore the significance of the serotonin and dopamine systems in the modulation of sensorimotor gating.


Introduction
Prepulse inhibition (PPI) of the acoustic startle response (ASR) is widely used as an operational measure of sensorimotor gating (1). PPI refers to the reduction of ASR magnitude when a distinctive non-startling stimulus is presented 30 to 500 msec before a startling stimulus (2). It was proposed that the mechanism underlying PPI regulates sensory input by filtering out irrelevant or distracting stimuli to prevent sensory information overflow (1).
PPI was suggested as a promising endophenotype of schizophrenia for several reasons (3): I.) Schizophrenia patients consistently show PPI deficits (4), II.) Unaffected first-degree relatives of schizophrenia patients also exhibit decreased PPI levels (5,6), III.) The PPI deficit seems to be a trait marker of schizophrenia as it is already present in the prodromal phase of schizophrenia (7), IV.) Inbred rodent studies and human twin studies suggested that PPI is heritable (8)(9)(10) and V.) PPI is measurable in a wide range of species in which PPI deficits could be artificially induced by environmental or pharmacological manipulations. This offers the possibility to study the neurobiological basis of PPI in translational investigations (1).
Recently, several single nucleotide polymorphisms (SNPs) have been reported to strongly affect PPI thereby illuminating both the molecular mechanisms and genetic influences on sensorimotor gating.
Initially, Hong et al. (11) found that the neuregulin-1 (NRG-1) Arg38Gln SNP modulates PPI in healthy volunteers and schizophrenia patients. Roussos et al. (12) followed with their finding that the catechol-O-methyltransferase (COMT) Val158Met SNP affects PPI in healthy male volunteers, a discovery which we recently replicated in a mixed-sex sample of schizophrenia patients (13).
Additionally, we showed that the serotonin-2A receptor (5-HT 2A R) A-1438G and T102C SNPs (which are in complete linkage disequilibrium) are associated with PPI in schizophrenia patients (14). Most recently, it has been demonstrated that PPI also depends on the dopamine-D3 receptor Ser9Gly SNP (15). All of these SNPs have at some stage been suggested as susceptibility polymorphisms for schizophrenia (e.g., 16,17) but current meta-analyses do not support their role in aetiology of the disease itself, with the exception of the 5-HT 2A R A-1438G polymorphism which may have a small effect on the risk for schizophrenia (SchizophreniaGene: www.schizophreniaforum.org/res/sczgene) (18).
Replication is essential for establishing the credibility of genotype-phenotype associations (19). Therefore, we investigated the impact of four SNPs, which were previously linked to sensorimotor gating and schizophrenia, on PPI in a new and independent sample of healthy human volunteers: The linked 5-HT 2A R T102C/A-1438G, the COMT Val158Met, and the NRG-1 Arg38Gln polymorphisms.
Participants were screened for the exclusion criteria of DSM-IV Axis I disorders using the Structured Clinical Interview for DSM-IV Disorders (SCID-I). Additional exclusion criteria were a history of head injuries, any known neurological abnormalities or systemic illness with known neurological complication, a first-degree relative with psychosis or obsessive-compulsive disorder, and a history of substance abuse or dependence. Ethical approval of the local ethics committee was obtained and participants provided written informed consent.

Startle Response Measurement
Equipment, set up, PPI testing, and data acquisition and scoring procedures have been described in detail previously (22). Each examination began with a 4-min acclimation period of 70-dB background noise that was continued throughout the session. Participants received 49 white noise sound pulses at an intensity of 115 dB (duration of 40 ms) separated by variable inter-trial intervals between 9 and 23 s (mean 15 s). In 36 of the trials, the pulse was preceded by a 20-ms 85-dB white noise prepulse with stimulus-onset-asynchronies (SOA) of 30, 60 and 120 ms (12 trials each). The initial trial was a pulsealone (PA) trial, which was separated for further analysis. All following trials were presented in a pseudorandomized order. The entire test session lasted about 16 min. To ensure that PPI was not influenced by smoking withdrawal, smoking ad libitum was permitted before testing (23). Trial exclusion and scoring criteria were identical to those used in previous studies (24). Subjects with response rejections >50% were excluded from data analysis (n=4).

Statistical analysis
Startle reactivity was assessed by the mean amplitude of the first block of PA trials and the mean amplitude of all PA trials. For the assessment of startle habituation, PA trials were divided each in four blocks. The calculation of the mean percent PPI and the habituation measures (percent habituation and linear gradient coefficient b) have been described in detail elsewhere (24).
All demographical data were analyzed by analysis of variance (ANOVA) with exception of frequency data. Frequency data were analyzed using χ 2 -Tests. Given that gender (25) and smoking (23) could affect PPI, these variables were introduced as covariates in all analyses of covariance (ANCOVA) of the psychophysiological parameters independent of the statistical significance of the covariates. Based on significant main effects or interactions, Tukey-HSD post-hoc comparisons were performed. Given that we proposed directional hypotheses regarding the genotype effects on sensorimotor gating, statistical comparisons of the PPI data were carried out at a significance level set at p<.05 (2-tailed).
Considering that we investigated four SNPs, all other confirmatory statistical comparisons were carried out at a Bonferroni-corrected significance level of p<.0125 (2-tailed). Within the Pearson product-moment correlation analyses, the significance level was set at p<.01 (2-tailed) to avoid accumulation of α-error. Effect size calculations between two groups refer to Cohen's d. When posthoc tests of PPI data are reported, Cohen's d calculations based on pooled SOA-conditions.

c)*** COMT Val158Met polymorphism
The COMT Val158Met genotype frequencies were distributed in accordance to the HWE [Chi 2 (1)=1.9; p=.17]. The three genotype groups did not differ in demographic characteristics, pooled PPI scores, startle reactivity and habituation measures (see Table 2). Startle latency measures were also not influenced by COMT-genotype (data not shown).

NRG-1 Arg38Gln polymorphism
The NRG-1 Arg38Gln genotype frequencies were distributed in accordance with HWE [Chi 2 (1)=0.04; p=.84]. The three genotype groups did not differ regarding demographic characteristics, pooled PPI scores, startle reactivity and startle latency measures (see Table 3) but there was a trend for a different distribution of gender between the genotype groups. Moreover, there was a strong trend for a genotype effect regarding early habituation (see Table 3). The total habituation and the slope of habituation revealed a similar but also non-significant pattern of effect as observed for early habituation.

****Insert Table 3*** Correlation analysis and genotype interactions
Age and years of education did not correlate with any of the psychophysiological parameters. A 3x3x3 (SOA-condition x SNP x genotype) repeated measurement ANCOVA with the three SNPs under investigation and gender as a covariate did not show any significant interactions between SNPs.
Interaction analyses with two SNPs each did also not reveal any interactions. However, these analyses should be interpreted with caution because of the limited sample size.

Discussion
The present work aimed to replicate initial findings on the dependency of PPI on polymorphisms of the 5-HT 2A receptor, the COMT enzyme, and the NRG-1 signal protein in an independent sample of healthy human volunteers. First, we confirmed that the 5-HT 2A R T102C/A-1438G polymorphisms exerts the same impact on sensorimotor gating in healthy humans as was previously shown in schizophrenia patients (14). Carriers of the C102/G-1438-alleles exhibited a significantly lower PPI than subjects homozygous for the T102/A-1438-alleles. Interestingly, the PPI variance explained by these 5-HT 2A R SNPs was comparable between both studies (schizophrenia patients: 11%; healthy controls: 10%), although we used different PPI paradigms and recruited the subjects in different European countries. Secondly, we were able to replicate the finding that male COMT Val158Met Methomozygotes display elevated PPI levels (12), but we were unable to detect this effect in our total mixed-sex sample. Roussos et al. (12) reported their COMT effects on PPI from a sample of male students and we initially confirmed this effect for schizophrenia patients in a sample consisting of nearly 70% males (13). The fact that one study did not find an impact of that polymorphism on PPI in female subjects (27) further supports the notion that the COMT Val158Met genotype might affect PPI only in male subjects. Furthermore, the COMT Met158-allele has also been reported to have a greater impact on cognitive function in males than females, and this sex-specific effect results most likely via transcriptional regulation by estrogens (28). The explained PPI variance by COMT genotype for the males in the present study was 13%, which is situated between the effect in Roussos male sample (25%) and our mixed-gender schizophrenia sample (9%). Thirdly, we could not replicate the finding of Hong et al. (11) who reported a moderate impact of the NRG-1 Arg38Gln polymorphism on PPI previously (7.9% variance explanation by NRG-1 genotype). In contrast to the lowered PPI levels in the homozygous A-allele carriers in Hong's sample, we rather found slightly elevated PPI levels in this group. Thus, it is unlikely that our study was simply underpowered to generate the same effect.
The mixed sample of Hong et al. (11) was indeed larger than ours but was heterogeneous, consisting of Caucasian and African Americans, schizophrenia patients and healthy controls, whereas our sample includes exclusively Caucasian healthy controls. Finally, Hong et al. assessed PPI only at an SOA of 120 ms and they used a slightly weaker prepulse intensity (80 vs. 85 dB). However, a selective analysis of our 120 msec SOA-condition still revealed a different PPI pattern (AA>GG>AG), which could also not explained by the slightly different prepulse intensity. Thus, the discrepant results may be most likely caused by different sample compositions.
Importantly, the present finding on the 5-HT 2A R polymorphsisms match nearly all criteria for a replication of a genotype-phenotype association proposed by the National Cancer Institute-National

Human Genome Research Institute (NCI-NHGRI) Working Group on Replication in Association
Studies (19): I.) We used a larger size than reported in the initial report, II.) We generated an independent data set, III.) We assessed the same phenotype, IV.). We found similar effect sizes in the same SNPs, V.) We used the same statistical tests, VI.) We had a rationale for reassessing these SNPs (susceptibility genes of schizophrenia), and VII.) We provide at least the same level of detail for study design, analysis, and sample characteristics as reported in the initial study. Moreover, all analysed genotype frequencies were distributed in accordance to the HWE so that genetic inhomogeneity of the investigated population is unlikely.  (29,30). Thus, the C-allele of the T102C variation and/or the G-allele of the A-1438G variation may cause lower 5-HT 2A R densities in several brain areas that are involved in the processing of sensorimotor gating (14,31). Additionally, PPI deficits have also been reported in obsessive-compulsive disorder (32,33) and autism (34,35), in which both the 5-HT 2A R A-1438G polymorphism have been implicated (36,37). On the other hand, the increased PPI in male COMT Val158Met Met-homozygotes could not be plausibly explained by the assumed increase of dopamine in the prefrontal cortex, as initially suggested (12,13), because an activation of prefrontal dopamine receptors by a local infusion of apomorphine disrupted PPI in rats (38,39). Thus, the neurobiological basis of the COMT-genotype effect on PPI remains to be clarified.
In conclusion, the present findings support the view that sensorimotor gating is strongly modulated by 5-HT 2A R A-1438G/T102C genotype independent of gender, whereas the COMT Val158Met genotype only influences PPI in male subjects. In concert with previous human and animal findings showing that PPI is affected by multiple mutations, it is suggested that PPI (like schizophrenia) is modulated by polygenetic factors. Future studies with larger samples are needed to explore the multiple single and epistatic effects of different gene mutations on PPI, which may provide also windows into the polygenetic causation of schizophrenia. Table 1 Demographic data and psychophysiological parameters of healthy human volunteers grouped according to their 5-HT 2A A-1438G and T102C receptor genotype (means and standard error of means in parentheses, adjusted for gender and smoking; gender and smoking in frequency data). 5-HT 2A A-1438G and T102C receptor polymorphisms were in complete linkage disequilibrium.