Increased Behavioral and Neuronal Responses to a Hallucinogenic Drug in PACAP Heterozygous Mutant Mice

Accumulating evidence from human genetic studies implicates the pituitary adenylate cyclase-activating polypeptide (PACAP) gene as a risk factor for psychiatric disorders, including schizophrenia and stress-related diseases. Mice with homozygous disruption of the PACAP gene display profound behavioral and neurological abnormalities that are ameliorated with the atypical antipsychotic and dopamine D2 and serotonin (5-HT)2 antagonist risperidone and the 5-HT2 receptor antagonist ritanserin; however, the underlying mechanisms remain unknown. Here, we investigated if PACAP heterozygous mutant (PACAP+/−) mice, which appear behaviorally normal, are vulnerable to aversive stimuli. PACAP+/− mice were administered a 5-HT2 receptor agonist, (±)-2,5-dimethoxy-4-iodoamphetamine (DOI), a hallucinogenic drug, and their responses were compared with the littermate wild-type mice. After DOI injection, PACAP+/− mice showed increased head-twitch responses, while their behavior was normal after saline. DOI induced deficits in sensorimotor gating, as determined by prepulse inhibition, specifically in PACAP+/− mice. However, other 5-HT2 receptor-dependent responses, such as corticosterone release and hypothermia, were similarly observed in PACAP+/− and wild-type mice. c-Fos expression analysis, performed in various brain regions, revealed that the DOI-induced increase in the number of c-Fos-positive cells was more pronounced in 5-HT2A receptor-negative cells in the somatosensory cortex in PACAP+/− mice compared with wild-type mice. These results indicate that PACAP+/− mice exhibit specific vulnerability to DOI-induced deficits in cortical sensory function, such as exaggerated head-twitch responses and sensorimotor gating deficits. Our findings provide insight into the neural mechanisms underlying impaired behavioral responses in which 5-HT2 receptors are implicated.


Introduction
Pituitary adenylate cyclase-activating polypeptide (PACAP) is a neuropeptide with multiple roles, including neurotransmitter, neuromodulator and neurotrophic factor [1]. Our recent studies have suggested that PACAP is associated with psychiatric disorders, including schizophrenia. Genetic variants of the PACAP gene that are overrepresented in schizophrenia patients are associated with reduced hippocampal volume and impaired memory performance [2]. A copy number gain of the PACAP gene due to a partial trisomy has been shown to cause severe mental retardation [3]. PACAP-deficient mice exhibit remarkable behavioral changes related to psychosis, memory impairment and depression that can be treated with the atypical antipsychotic and mixed D 2 and serotonin (5-HT) 2 antagonist risperidone and the 5-HT 2 receptor antagonist ritanserin [4][5][6][7][8][9]. In addition, Vacic et al. [10] found a significant association of copy number gains at chromosome 7q36.3 with schizophrenia, which results in increased expression of the common VIP and PACAP receptor VPAC2 in cultured lymphocytes. Furthermore, Ressler et al. [11,12] demonstrated a sex-specific association of post-traumatic stress disorder (PTSD) with PACAP and the PACAP-selective receptor PAC 1 in females. These studies provide convergent evidence for psychiatric implications of the PACAP signaling system; however, the underlying mechanisms remain unknown.
There is a great deal of evidence implicating 5-HT 2 receptors in various neurological and psychiatric conditions. Hallucinogenic drug-induced activation of 5-HT 2 receptors is closely related to their reinforcing and/or aversive effects [13]. Impaired 5-HT 2A receptor signaling plays a major role in schizophrenic episodes. Almost all currently available atypical antipsychotic drugs are 5-HT 2A receptor inverse agonists, as well as dopamine D 2 receptor antagonists or partial agonists [14]. The attenuation of extrapyramidal symptoms by atypical antipsychotic drugs has been proposed to be mediated by dopamine release in the striatum induced by 5-HT 2A antagonism [15]. Furthermore, in a study using 5-HT 2A conditional knockout mice, a specific role for cortical 5-HT 2A receptor function in the modulation of conflict anxiety was observed, consistent with the hypothesized ''topdown'' control model of anxiety-related processes [16].
To provide insight into the neural mechanisms underlying impaired behavioral responses in which 5-HT 2 receptors are implicated, we examined if PACAP mutant (PACAP +/2 ) mice, which appear behaviorally normal, are vulnerable to aversive stimuli; in this case, injection of the hallucinogenic 5-HT 2 receptor agonist (6)-2,5-dimethoxy-4-iodoamphetamine (DOI). We also performed c-Fos expression analysis to identify brain regions with an altered response to DOI in PACAP +/2 mice.

Animals
All animal care and handling procedures were performed according to the Guidelines for the Care and Use of Laboratory Animals approved by the Japanese Pharmacological Society, and were approved by the Animal Care and Use Committee of the Graduate School of Pharmaceutical Sciences, Osaka University. All efforts were made to minimize the number of animals used.

Behavioral Analysis
Each behavioral study was performed using a separate cohort of mice. For assessment of the head-twitch response, mice were individually placed in observation cages (19610611 cm) for a 60min habituation period. They were then intraperitoneally injected with either saline or DOI (Sigma-Aldrich, Tokyo, Japan), which were prepared just before use, and recordings were made for a duration of 60 min. Scoring began immediately after injection by trained blind observers. The head-twitch response is a distinctive paroxysmal head-twitching behavior that is easily distinguished from head-bobbing, lateral movements of the head and grooming.
Acoustic startle responses for the prepulse inhibition (PPI) experiment were measured in a startle chamber (SR-LAB; San Diego Instruments, San Diego, CA) using standard methods described previously [17]. Animals were placed in the startle chamber 5 min after intraperitoneal injection of DOI (1.0 mg/kg). The testing session started with 5 min of acclimatization to the startle chamber in the presence of 65 dB background broadband (white) noise. Testing consisted of forty 120 dB pulses alone and 10   pulses preceded (100 ms) by a prepulse of 68, 71 or 74 dB. Pulses were randomly presented with an average interval of 15 s between pulses. Twelve no-stimulus trials were included to assess spontaneous activity during testing. PPI was calculated as a percentage score: PPI (%) = (1 2 (startle response for pulse with prepulse)/ (startle response for pulse alone))6100.

Statistics
All data are expressed as the mean 6 standard error of the mean (S.E.M.). Student's t-test, one-way analysis of variance (ANOVA) followed by Dunnett's test, or two-way ANOVA followed by the Tukey-Kramer test were used to assess statistical significance as appropriate. Data for open field test and headtwitch response were analyzed using two-way ANOVA for genotype as the intersubject factor and repeated measures with time as the intrasubject factor. Data for PPI were analyzed using three-way ANOVAs (genotype and treatment as the intersubject factors, and prepulse intensity as the intrasubject factors). Multiple comparisons were performed using the Student-Newman-Keuls test. A p-value lower than 0.05 was considered statistically significant. The statistical analyses were performed using a software package (StatViewH 5.0 for Windows; SAS Institute, Cary, NC).

Results
Behavioral Abnormalities in PACAP +/2 Offspring from Maternal PACAP +/2 Mating, not from Maternal PACAP +/+ Mating Because genetic background may account for behavioral variation [19], we first investigated locomotor activity in PA-CAP +/2 mice on the C57BL/6J-129S6/SvEvTac F 1 mixed genetic background (B6-129S6) in the open field test. To obtain B6-129S6 F 1 mice, male PACAP +/2 mice on the 129S6/SvEvTac background were mated with female PACAP +/+ or PACAP +/2 mice on the C57BL/6J background. F 1 PACAP +/2 mice from PACAP +/+ dams did not exhibit hyperactivity or differences in rearing time or time spent in the center of the field in the open field ( Fig. S2A, C, E). Repeated two-way ANOVA revealed no significant main effect of genotype [F (1, 295) = 0.005, p = 0.95]. In contrast, F 1 PACAP +/2 mice from PACAP +/2 dams showed significant increases in locomotor activity compared with PA-CAP +/+ littermates (Fig. S2B). Repeated two-way ANOVA revealed a significant main effect of genotype [F (1, 170) = 5.93, p,0.05]. Concomitant with the increase in locomotor activity, F 1 PACAP +/2 mice from PACAP +/2 dams showed significant increases in rearing time and time spent in the center of the field compared with PACAP +/+ littermates (Fig. S2D, F). The regions are indicated in Figure S1. Data shows mean 6 SEM of 4-6 mice. Statistically significant differences vs. saline were assessed with the t-test. doi:10.1371/journal.pone.0089153.t001 To investigate the mechanisms by which susceptibility genes alter neural responses, we used PACAP +/2 mice obtained from PACAP +/+ dams, which display behaviors undistinguishable from their PACAP +/+ littermates, for the experiments that follow. To examine whether the behavioral response to external stimuli is altered in PACAP +/2 mice, we examined the DOI-induced headtwitch response, which is a characteristic head-shaking movement induced by the hallucinogenic drug via stimulation of 5-HT 2 receptors [20]. The total number of head-twitch responses induced by DOI during a 60-min period was significantly increased in both genotypes in a dose-dependent manner, but the response in PACAP +/2 mice occurred significantly more frequently than in their PACAP +/+ littermates with doses of DOI (1.0 or 3.0 mg/kg) (Fig. 1A). Two-way ANOVA revealed a significant main effect of genotype [F (1,28) = 39.6, p,0.001] and DOI dose [F (4, 28) = 55.8, p,0.001], and there was a significant interaction between genotype and treatment [F (4, 28) = 14.1, p,0.001]. The maximal head-twitch response induced by 1.0 mg/kg DOI was elicited within approximately 20 minutes in both PACAP +/+ and PA-CAP +/2 mice (Fig. 1B).
PPI is a reliable, robust quantitative phenotype that is useful for probing the neurobiology and genetics of gating deficits in schizophrenia across species [21]. DOI is known to disrupt PPI via 5-HT 2A receptors in rodents [22]. We therefore examined PPI in PACAP +/+ and PACAP +/2 mice treated with or without DOI. There was no difference in PPI levels between the two genotypes of mice following saline injection ( Fig. 2A). Although 1.0 mg/kg DOI had no effect in wild-type mice, it evoked PPI deficits in PACAP +/2 mice ( Fig. 2A). Three-way ANOVA revealed a significant main effect of genotype [F (1,33)  (1 mg/kg) did not affect PPI in PACAP +/+ mice, whereas it significantly disrupted PPI compared with saline in PACAP +/2 mice ( Fig. 2A). There were no significant differences in startle amplitudes between the four groups (PACAP +/+ and PACAP +/2 injected with saline or DOI) (Fig. 2B).
As it has been reported that stimulation of 5-HT 2A receptors activates hypothalamic neurons to increase the secretion of several hormones, such as corticosterone, and induces hypothermia [23,24], we examined the effect of DOI on plasma corticosterone and body temperature in both genotypes. Although DOI dosedependently increased the level of plasma corticosterone, there was no difference between the two genotypes (Fig. S3A). Repeated two-way ANOVA revealed a significant main effect of treatment [F (3,32)

Identification of Brain Regions with an Altered Response to DOI in PACAP +/2 Mice
To identify the brain regions that may underlie the differences in behavioral response to DOI, we first examined DOI-induced changes in c-Fos protein expression in 16 brain regions known to be involved in the regulation of the head-twitch response, in PPI, in the release of corticosterone and the regulation of body temperature [25][26][27][28] in PACAP +/+ mice on the C57BL/ 6J6129S6/SvEvTac hybrid background (Fig. S1). Among the brain regions examined, significant differences in c-Fos expression were observed only in six; the mPFC, SSCx, VL-CPu, MD, BLA and PVN (Table 1). Therefore, we subsequently examined DOIinduced c-Fos expression in these six regions in PACAP +/2 mice. Representative photomicrographs of c-Fos staining in these regions are shown in Fig. S4. Interestingly, the number of c-Fospositive cells was significantly increased specifically in the SSCx in PACAP +/2 mice compared with PACAP +/+ mice in response to DOI (Fig. 3 and S4). Two-way ANOVA revealed a significant main effect of treatment [F (1,18) = 54.2, p,0.001] and genotype [F (1,18) = 6.47, p,0.05], and a significant interaction between genotype and treatment for the SSCx [F (1,18) = 6.46, p,0.05]. In the remaining five regions (mPFC, VL-CPu, MD, BLA and PVN), there was no statistically significant difference in the number of c-Fos-positive cells after DOI injection between the two genotypes (Fig. 3).
Increase in the Number of c-Fos-positive Cells Among 5-HT 2A Receptor-negative Cells in the SSCx of PACAP +/2 Mice Injected with DOI We next examined whether the difference between PACAP +/+ and PACAP +/2 mice in the molecular response of SSCx neurons to DOI was dependent on the expression levels of the 5-HT 2A receptor. 5-HT 2A receptor protein levels in the SSCx were similar in both genotypes as measured by Western blot analysis (Fig. S5).
The types of neurons activated by DOI in the SSCx were then examined using double immunostaining for c-Fos and the 5-HT 2A receptor. The number of c-Fos-positive cells among 5-HT 2A receptor-positive cells was not changed in either genotype. However, the number of c-Fos-positive/5-HT 2A receptor-negative cells was significantly increased in the SSCx in PACAP +/2 mice compared with PACAP +/+ mice (Fig. 4). Two-way ANOVA revealed a significant main effect of genotype [F (1,12)

Discussion
In this study, we obtained four major findings. First, PACAP +/2 mice injected with DOI exhibit exaggerated head-twitch responses and PPI deficits, while PACAP +/+ mice with the same treatment show significantly reduced head-twitch responses and normal levels of PPI. Second, other 5-HT-induced responses mediated by 5-HT 2A receptor stimulation, such as increased corticosterone levels and hypothermia, did not significantly differ between PACAP +/2 and PACAP +/+ mice. Third, the DOI-induced increase in the number of c-Fos-positive cells was more pronounced in 5-HT 2A receptor-negative cells in the SSCx in PACAP +/2 mice compared with PACAP +/+ mice. Fourth, F 1 PACAP +/2 mice from PACAP +/2 dams, but not from PACAP +/+ dams, show significant increases in locomotor activity, rearing time and time spent in the center of the open field compared with PACAP +/+ littermates.
We previously reported that PACAP 2/2 mice exhibit an increased head-twitch response to an intermediate dose of DOI, 0.25 mg/kg body weight [4]. In addition, they display PPI deficits without receiving DOI injection [8]. In PACAP +/2 mice, we have not yet observed any appreciable behavioral changes. However, impaired long-term potentiation has been observed in the dentate gyrus in PACAP +/2 mice and in mice with altered expression of the PAC 1 receptor [7], which have approximately 25% of the [ 125 I]-PACAP27 binding density (due to a targeted deletion of exon 2) in their brain [29]. In addition, Ohtaki et al. [30] demonstrated that PACAP +/2 mice have an increased vulnerability to ischemic neuronal cell death, which is associated with decreased signal transducer and activator of transcription (STAT) 3 and extracellular signal-regulated kinase (ERK) activities. These observations imply that, given that the PACAP signaling system is a risk factor for psychiatric disorders, PACAP +/2 mice could serve as a useful model to assess vulnerability to non-genetic risk factors in subjects with genetic susceptibility.
Our present findings that PACAP +/2 mice display exaggerated head-twitch responses and PPI deficits, but not alterations in other parameters (such as corticosterone release or hypothermia) after DOI injection, suggest that these mice have a selective vulnerability to the hallucinogenic drug that specifically affects cortical sensory function. Indeed, among the brain regions examined (the mPFC, SSCx, VL-CPu, MD, BLA and PVN), the DOI-induced increase in the number of c-Fos-positive/5-HT 2A receptornegative cells was more pronounced in the SSCx in PACAP +/2 mice compared with their wild-type littermates. Scruggs et al. [31] demonstrated that DOI activates 5-HT 2A receptors on thalamocortical neurons and thereby increases glutamate release, which in turn drives c-Fos expression in cortical glutamatergic neurons through a mechanism dependent on the ionotropic glutamatergic AMPA receptor. It may be possible that endogenious PACAP influences on the 5-HT 2A -glutamate interactions. In our preliminary study, we have observed that PACAP modulates cell surface expression of 5-HT 2A receptors in heterologous cells. This effect of PACAP may explain susceptibility to hallucinogenic drugs in PACAP +/2 mice, although further studies (e.g. in neuronal cultures or in vivo brain via viral delivery) are necessary.
Hallucinogenic drugs have been shown to concomitantly induce both the head-twitch response and the expression of specific genes, including egr-1 and egr-2, in the SSCx in a 5-HT 2A receptordependent manner [32]. The suppression of this system by activation of the metabotropic glutamate receptor 2 has been shown to abolish the hallucinogen-specific signaling and behavioral responses [33]. In support of this, we recently observed that a metabotropic glutamate 2/3 receptor agonist can reverse psychomotor abnormalities and recognition memory deficits in PA-CAP 2/2 mice [34]. In a mouse model of maternal influenza viral infection, which is a risk factor for schizophrenia, an adult-onset abnormal response to DOI is observed, with an exaggerated headtwitch response and expression of the genes c-fos, egr-1 and egr-2 in cortical neurons [35]. It might be worth examining whether PACAP signaling pathways are altered in the mouse viral infection model [36].
The present observation that F 1 PACAP +/2 mice from PACAP +/2 dams show behavioral abnormalities in the open field test may reflect genetic and environmental vulnerability in PACAP +/2 offspring. However, the potential genotypic impact of PACAP +/2 dams on the offspring might not be related to the changes in the head-twitch response, PPI deficits or SSCx activation observed in F 1 PACAP +/2 mice from PACAP +/+ dams. Nonetheless, it will be interesting to investigate the neural and molecular mechanisms underlying the abnormalities in F 1 PA-CAP +/2 mice obtained from PACAP +/2 dams. A future experiment, for example, using cross-fostering, should provide insight into the interactions between genetic and environmental risk factors.
In conclusion, we demonstrate that PACAP +/2 mice show specific vulnerability to the hallucinogenic drug DOI, which impacts cortical sensory function and results in exaggerated headtwitch responses and sensorimotor gating deficits. These findings suggest that the PACAP signaling pathway is critically involved in 5-HT 2 receptor-dependent cortical processing. PACAP +/2 mice provide a promising model to investigate the neural and molecular mechanisms underlying impaired behavioral responses in which 5-HT 2 receptors are involved.