The Effect of Chronic Antipsychotic Drug on Hypothalamic Expression of Neural Nitric Oxide Synthase and Dopamine D2 Receptor in the Male Rat

Antipsychotic-induced sexual dysfunction is a common and serious clinical side effect. It has been demonstrated that both neuronal nitric oxide (nNOS) and dopamine D2 receptor (DRD2) in the medial preoptic area (MPOA) and the paraventricular nucleus (PVN) of the hypothalamus have important roles in the regulation of sexual behaviour. We investigated the influences of 21 days’ antipsychotic drug administration on expression of nNOS and DRD2 in the rat hypothalamus. Haloperidol (0.5 mg/kg/day i.p.) significantly decreased nNOS integrated optical density in a sub-nucleus of the MPOA, medial preoptic nucleus (MPN), and decreased the nNOS integrated optical density and cell density in another sub-nucleus of the MPOA, anterodorsal preoptic nucleus (ADP). Risperidone (0.25 mg/kg) inhibited the nNOS integrated optical density in the ADP. nNOS mRNA and protein in the MPOA but not the PVN was also significantly decreased by haloperidol. Haloperidol and risperidone increased DRD2 mRNA and protein expression in both the MPOA and the PVN. Quetiapine (20 mg/kg/day i.p.) did not influence the expression of nNOS and DRD2 in either the MPOA or the PVN. These findings indicate that hypothalamic nNOS and DRD2 are affected to different extents by chronic administration of risperidone and haloperidol, but are unaffected by quetiapine. These central effects might play a role in sexual dysfunction induced by certain antipsychotic drugs.


Introduction
Antipsychotic-induced sexual dysfunction is a common and serious clinical side effect, which is gaining increasing attention within the past decade. Sexual dysfunction has important implications for satisfaction with sexual life and overall quality of life [1,2]. It is one of the major reasons for treatment noncompliance and inevitably affects overall clinical outcome and treatment success [3,4,5].
The mechanism of antipsychotic drug-induced sexual dysfunction is complex and remains unclear. Generally, the sensory information related to sexual behaviour is processed at various brain nuclei, which balance the inhibitory and excitatory influence on spinal sympathetic and parasympathetic centres, and then determine the functional state of the sexual effector organ [6]. Antipsychotics might act both centrally and peripherally to induce sexual dysfunction. We have shown that some antipsychotics change NOS activity and expression in penile tissues [7], as well as demonstrating functional effects on male sexual behaviour in the rat [8,9]. Research directly investigating the central mechanisms of antipsychotic-induced sexual dysfunction is particularly rare. Understanding of these mechanisms is mostly theoretical, deriving from general knowledge of sex physiology and psychopharmacology, and is generally unverified by basic or clinical investigation.
The medial preoptic area (MPOA) and the paraventricular nucleus (PVN) are two critical brain structures for male sexual behaviour. These nuclei receive direct and indirect input from every sensory modality and send projections to extra-hypothalamic brain areas for the initiation and patterning of copulation [10,11]. The MPOA and PVN also have mutual connectivity [12,13]. Neuroanatomical studies further indicated that the MPOA could be divided into several sub-regions, which have different roles in the regulation of sexual function. Lesion studies have inferred that the caudal MPOA could impair copulation more severely than the rostral MPOA while the dorsal MPOA (anterodorsal preoptic nucleus, ADP) may be more important than the medial and other MPOA regions for copulatory behaviour [14,15].
Previous studies have indicated that dopamine and nitric oxide (NO) might be two of the most important neuromodulators with facilitative effects on sexual function in both the MPOA and the PVN [16]. A direct dopamine D2 receptor antagonist effect has been proposed as the primary underlying mechanism of sexual dysfunction after antipsychotic drug administration [17]. Halo-peridol, a dopamine D2 receptor antagonist, has been found to impair sexual behaviour after acute microinjection into the MPOA [18] and the PVN [11]. However, the role of dopamine D2 antagonism has not been tested with other antipsychotic drugs, notably the newer atypical drugs.
The importance of NO in sexual function has been demonstrated by the observation that a NO precursor (L-arginine) facilitates male sexual function, while a NOS inhibitor (Lnitroarginine methyl ester, L-NAME) injected into either the MPOA [10] or the PVN [19], reduces it. An early study suggested that haloperidol could inhibit neuronal nitric oxide synthase (NOS) activity by preventing electron transfer [20], while apomorphine, a mixed D1/D2 agonist, increased NO production in the PVN, which was correlated with penile erection. Acute haloperidol (0.5 mg/kg i.p.) prevented apomorphine's effect on both NO 2 2 concentration and penile erection [21]. Therefore, the NO pathway might also be involved in the development of antipsychotic-induced sexual dysfunction.
The current study investigated the expression of neural NOS (nNOS) and the dopamine D2 receptor in the MPOA and PVN after chronic systemic administration of the typical antipsychotic haloperidol, and the atypical antipsychotics risperidone and quetiapine, in which haloperidol and risperidone, but not quetiapine are associated with a high incidence of sexual dysfunction in both humans [22,23] and rats [9].

3-month-old male Sprague-Dawley rats (Shanghai Laboratory
Animal Centre, China) were used. After being quarantined for 1 week, animals were housed five or six to a cage with free access to food and water, at an ambient temperature of 18-22uC and 60% humidity with a 12-h light/dark cycle (lights on 7.00AM -7.00PM). The animals were handled daily. All procedures in animal experiments were approved by the Animal Study Committee of Southeast University, China (ethics approval number 2006ZDLL009.0).

Drug Administration and Experimental Design
Four groups of 3-month-old male Sprague-Dawley rats (n = 7 in each group) received intraperitoneal injection of haloperidol (0.5 mg/kg), risperidone (0.25 mg/kg), quetiapine (20 mg/kg) or vehicle (saline) once daily for 21 days. The haloperidol dose was chosen to obtain effective receptor occupancy (Wadenberg et al, 2001); does for risperidone and quetiapine were chosen on the basis of their relative clinical potency (Jibson and Tandon, 1998). Risperidone (Sigma, Poole, UK), haloperidol (Seranace liquid, 2 mg/ml) and quetiapine (AstraZeneca, Macclesfield, UK) were dissolved in a few drops of glacial acetic acid, diluted with distilled water and buffered back with NaOH to pH 6.0. The vehicle was saline previously acidified to pH 6.0. The experiment was performed three times for immunocytochemistry, gene expression and protein expression.

Immunohistochemistry of nNOS
At the end of the drug administration, the rats were sacrificed by deep anaesthesia with 30 mg/kg pentobarbital i.p. followed by transcardial perfusion with 4% paraformaldehyde (PFA) in 0.1 M phosphate buffered saline (PBS). Brains were removed and fixed in 4% PFA for 4 days before being embedded in wax. Three sections (7 mm) from MPOA between Bregma 20.30 and 20.40 mm, determined using the rat brain stereotaxic atlas [24], were mounted onto slides coated with 3-aminopropyltriethoxysilane (APES) (Sigma). PVN sections were taken from Bregma 21.60 to -1.88 mm. Avidin-Biotin Complex (ABC) method was used to stain for nNOS immunoreactivity. Briefly, following clearing in xylene and rehydration in graded alcohol, sections were incubated in hydrogen peroxide solution to inhibit endogenous peroxidase activity and a microwave treatment protocol was employed to aid antigen retrieval. Sections were then blocked with 10% normal horse serum diluted in 0.01 M PBS. The sections were incubated with a mouse anti-rat polyclonal primary antibody raised against the C-terminal of nNOS protein (1095-1289) (Product NO 610308, BD Transduction Laboratories, USA) at a dilution of 1:6000 in antibody dilution buffer (10% normal serum in 0.01 M PBS) overnight at 4uC, followed by incubation with biotinylated secondary antibody (anti-mouse IgG) (Vector Laboratories, Burlingame, CA) at a dilution of 1:200 for 2 h at room temperature. Sections were then incubated for 2 h at room temperature with ABC solution (Vectastain elite kit, Vector Laboratories, Burlingame, CA) and peroxidase was stained by incubation with 3,39-diaminobenzidine (DAB) solution containing nickel chloride. No immunoreactivity could be detected in control sections, in which the primary antibody was omitted from the staining protocol.
The composite images containing total left and right MPOA and PVN were captured using a RGB JVC solid-state camera connected to an Olympus BH2 microscope at 10-fold objective magnification fitted with a motorized stage. Then two immunohistochemical parameters, the neural positive cell density and integrated optical density (IOD), were analysed using Image-pro plus software (version 4.1, Diagnostic, Inc.) in three horizontal sections (28 mm apart) in each structure. The quantification processes were set blind to section code until the analysis was finished. Two sub-nuclei, anterodorsal preoptic nucleus (ADP) and medial preoptic nucleus (MPN), were investigated in the MPOA, with ADP and MPN defined by a 0.2 6 0.3 mm 2 and 0.4 6 0.6 mm 2 rectangle area respectively [25]. The PVN field area was defined by the area within its clear border after DAB staining. Stained neurons were counted using a manual counting option, and then calculated as neuronal density (cells/mm 2 ). Integrated optical density (IOD) is the summation of total pixel optical density in the sub-nuclei [26,27,28,29].

nNOS and DRD2 mRNA Measurement
Brains from each animal were immediately frozen in liquid nitrogen and then allowed to thaw slightly on ice. The MPOA slice was made by cutting between the anterior and posterior edges of the optic chiasm. The tissue containing the MPOA was isolated by cutting laterally to the lateral sulcus and to the depth of the anterior commissure [30]. The PVN slice was made by cutting rostral to the optic chiasm and again 2 mm posterior to the first cut. The tissue containing the PVN was dissected out by cutting 1.5 mm lateral to the median line and undercutting at a depth of 3 mm [31]. The tissues were finally stored at 280uC until analysis.
Total RNA was isolated from the tissue after homogenization using phenol-chloroform TRIzol (Sigma, USA). Genomic DNA contaminants were removed by treatment of the tissue extract with RNase-free DNase (Promega, USA). First-strand cDNA was then synthesized from 2 mg total RNA by incubating with Oligo(dT) 15 Primer and M-MLV Reverse Transcriptase according to the manufacturer's protocol (Promega, USA).
The expression level of nNOS and DRD2 mRNA was examined by a quantitative real-time PCR method using an ABI Prism 7000 Sequence Detector (Applied Biosystems, USA). b-actin was used as the endogenous control. Rat-specific oligonucleotide primers were designed and synthesized by Invitrogen (China) as following: nNOS: FWD 59-TCA AAG CCA TCC AGC GCA  TA-39, REV 59-GCG GTT GGT CAC TTC ATA CGT TC-39;  DRD2: FWD 59-AGA CGA TGA GCC GCA GAA AG -39, REV  59-GCA GCC AGC AGA TGA TGA AC-39; b-actin: FWD 59-GGA GAT TAC TGC CCT GGC TCC TA-39, REV 59-GAC  TCA TCG TAC TCC TGC TTG CTG-39. A final 20 ml reaction was prepared with 10 ml SYBR Green Premix Ex Taq (TaKaRa, Japan), 0.4 ml ROX Reference Dye, 0.8 ml primer pairs, 6.8 ml sterile water and 2 ml cDNA sample. Samples were analysed in triplicate, while a negative control using water as template was included for each primer pair. The PCR reaction was started with an initial enzyme activation step at 95uC for 10 seconds followed by 40 cycles of denaturation at 95uC for 5 seconds and annealing/ extension at 62uC for 34 seconds. Data were evaluated by the relative quantification method with efficiency calibrated mathematical model using the control group as a calibrator for comparison of every unknown sample gene expression level [32]. The conversion between Ct and relative gene expression level is n fold induction 2-Ct relative to b-actin.

Western Blotting Measurement of nNOS and DRD2
The isolated samples were homogenized in lysis buffer (containing 50 mM Tris-HCl, 150 mM NaCl, 0.02 NaN2, 1 mM PMSF, 1 mg/mL Aprotitin and 1% Triton X-100) in the presence of protease inhibitors (5 mL/mL of protease inhibitor cocktail, Sigma). After centrifugation (12,000g, 25 min, 4uC), protein concentration of the supernatant was estimated by the method of Bradford. About 50 mg total protein was then boiled in sample buffer for 5 min and separated by SDS-PAGE using 10% polyacrylamide gels. After electrophoresis, the proteins were transferred to PVDF membranes (Millipore) for 120 min. The membranes were treated with a blocking buffer (5% nonfat dry milk with 0.1% Tween-20) and incubated with primary nNOS and DRD2 antibody respectively (mouse monoclonal antibody against nNOS 1:500, Transduction Laboratories, Lexington, KY; Rabbit polyclonal antibody against DRD2 1:200, Chemicon International Inc., Temecula, CA) at 4uC overnight. The blots were incubated with a peroxidase-conjugated anti-mouse IgG (Santa Cruz). The reactive bands were detected with an enhanced chemiluminescence method (ECL; Amersham Biosciences). Quantitation of protein bands was carried out by optical densitometry. b-tubulin was used as a control.

Statistical Analysis
Positive cell density, integrated optical density and mRNA expression were expressed as mean 6 s.e.means and were analysed by one-way ANOVA followed by post-hoc Dunnett's t-test. p,0.05 was considered statistically significant. Integrated optical density of nNOS staining. The effects of antipsychotic drugs on nNOS integrated optical density (IOD) in the MPOA and PVN are shown in Figure 1. One-way ANOVA revealed a significant between groups effect in the MPN [F (3,24) = 3.125, p = 0.045]. Individual comparisons with the vehicle group revealed that IOD of nNOS staining was significantly reduced by three weeks' administration of haloperidol (p = 0.015), but not by risperidone (p = 0.212) or quetiapine (p = 0.320). A significant effect was also seen in the ADP [F (3,24)  DRD2. The effects of antipsychotic drugs on DRD2 mRNA expression in the MPOA and the PVN were shown in Figure 3. One-way ANOVA revealed a significant effect of antipsychotic administration on DRD2 mRNA expression [F (3,24) = 23.691, p,0.001] in the MPOA. Individual comparisons with vehicle group revealed that DRD2 mRNA expression in the MPOA was significantly increased by three weeks administration of haloperidol (p,0.001) and risperidone (p = 0.001), but not by quetiapine (p = 0.929).

Immunohistochemistry of nNOS
One-way ANOVA also revealed a significant effect of antipsychotic administration on DRD2 mRNA expression [F (3,24) = 6.947, p = 0.002] in the PVN. Individual comparisons with vehicle group revealed that DRD2 mRNA expression in the PVN was significantly increased by administration of haloperidol (p = 0.001) and risperidone (p = 0.013), but not by quetiapine (p = 0.412).

Protein Expression of nNOS and DRD2
nNOS. The nNOS monomers display sizes corresponding to approximately 155 kDa, whileb-tubulin has a size of approximately 55 kDa. One-way ANOVA revealed a significant effect of antipsychotic administration on nNOS expression in the MPOA [F (3,24) = 18.115, p,0.001] but not the PVN [F (3,24) = 0.027, p = 0.994]. Individual comparisons with vehicle group revealed that nNOS protein expression in the MPOA was significantly decreased by administration of haloperidol (p,0.001) but not risperidone (p = 0.999) or quetiapine (p = 0.619) (Figure 4).
DRD2. The DRD2 immunoreactivity identified a protein band corresponding to approximately 48 kDa. One-way ANOVA revealed a significant effect of antipsychotic administration on DRD2 expression in both the MPOA [F (3,24) (Figure 4).

Discussion
The present study evaluated a possible relationship between chronic antipsychotic drug administration and the expression of nNOS in regions of the hypothalamus implicated in sexual function. nNOS catalyzes the synthesis of neuronal NO from L-arginine. Behavioural and neurochemical studies indicate that neuronal NO plays a key role in the central mechanism of sexual function. Originally nNOS was considered a constitutively expressed enzyme but accumulating evidence demonstrates that nNOS expression is highly susceptible to regulation by various physiological or pathological conditions [33]. Previous studies indicated that nNOS expression could also be affected by antipsychotic medication, while the reported effects varied widely with different antipsychotic drugs and different brain regions (as summarised in [7]). Our current study is consistent with those previous findings in that haloperidol and risperidone, but not quetiapine, decreased immunopositive neuron density and IOD of nNOS in the ADP and the MPN to different extents.
Along with the inhibitory effect of haloperidol on nNOS immunoreactivity, chronic treatment with haloperidol decreased levels of nNOS mRNA and protein in the MPOA. These findings indicate that haloperidol might suppress nNOS gene transcription and protein expression, which then reduces the production of neuronal NO in the MPOA and finally lead to sexual dysfunction. The mechanism underlying down-regulation of nNOS synthesis following haloperidol administration remains unclear. nNOS is regulated by Ca 2+ influx through L-type voltage-sensitive Ca 2+ channels [33], while haloperidol has been demonstrated to inhibit the L-type Ca 2+ channel at micromolar concentrations [34]. Wu et al., (2000) reported that haloperidol but not risperidone (both 10 uM) could significantly reduce the amplitude of L-type Ca 2+ current in vitro [35]. Collectively, these studies provide a possible mechanism underlying the effect of haloperidol on nNOS expression. However, these inhibitory effects occurred at relatively high concentrations, and extrapolation to the in vivo situation should be undertaken with caution.
Risperidone also showed an effect on nNOS expression in the MPOA, indicated by a suppression of nNOS IOD in the ADP, although nNOS-immunopositive neuron density was not significantly changed. Risperidone did not significantly influence nNOS mRNA and protein levels in the MPOA; however differences restricted to individual sub-nuclei may not be detectable in this analysis as they may not be robust enough to influence the result in the whole nucleus.
In the hypothalamus, dopamine D2 receptor density is rich in both the MPOA and the PVN [36,37]. Sexual function was noticeably inhibited after infusions of haloperidol to the MPOA or the PVN [11,18]. D2 receptor expression in the brain is increased after chronic treatment with antipsychotic drugs [38,39], while continuous infusion of the dopamine D2 agonist quinpirole can cause a significant down-regulation of striatal D2 dopamine receptors [40]. The alterations of D2 receptor expression also parallel dopamine-mediated behaviours, such as stereotypy and locomotor behaviour [40]. Although the increase in D2 receptors is not always accompanied by an increase in D2 mRNA [41], higher doses and more prolonged exposure to antipsychotic drugs do elevate D2 mRNA [42,43]. We find that haloperidol and risperidone increased the DRD2 mRNA and protein expression in the MPOA and the PVN. This is the first study providing direct evidence for dopamine D2 receptor alteration induced by antipsychotic drugs in those two neural sites. Our study also provides strong support for the hypothalamic dopamine D2 receptor antagonism hypothesis of antipsychotic drug-induced sexual dysfunction. This hypothesis is also supported by our previous observation of an association of D2 receptor polymorphisms with sexual dysfunction in male subjects receiving antipsychotic treatment [44].
While the current study is limited by using only single dose regimes, these were chosen to be equivalent to clinically-effective doses on the basis of inducing substantial and effective dopamine  D2 receptor occupancy. Previous studies from this laboratory have investigated the effects of a range of doses of these three drugs in rat sexual behavior; the doses chosen here were at (haloperidol) or below (risperidone and quetiapine) the threshold dose that impaired aspects of male sexual function [9].
As reviewed by Hull and Dominguez [10], chemosensory information activates the MPOA neurons, resulting in Ca 2+ influx and NOS activation. The resultant NO production may increase dopamine neurotransmission to facilitate sexual function via D2like receptors. The incerto-hypothalamic dopaminergic neuron terminals from the MPOA release dopamine in the PVN and then activate NOS production in the oxytocin-containing neurons of the PVN, which in turn facilitates penile erection [11,19]. According to our current findings, antipsychotic drugs might impair sexual function through the following various mechanisms: 1) down-regulation of nNOS expression, thereby decreasing production of NO and consequently synaptic dopamine in the MPOA and/or 2) blocking the postsynaptic D2 receptor directly in the PVN and consequently inhibiting sexual behaviour. These central effects might be involved in sexual dysfunction induced by certain antipsychotic drugs. These differences of sexual dysfunction induced by various antipsychotic drugs observed in clinical medication are likely because of the different pharmacological properties of these antipsychotics.