Dentate nNOS accounts for stress‐induced 5‐HT1A receptor deficiency: Implication in anxiety behaviors

Abstract Background Anxiety is a common disorder with high social burden worldwide. Dysfunction of serotonin‐1A receptor (5‐HT1A receptor) in the dentate gyrus (DG) of the hippocampus has been predominantly implicated in the anxiety behavior. However, the molecular mechanism underlying the deficiency of postsynaptic 5‐HT1A receptor in regulating anxiety behavior remains unclear. Methods Using pharmacological and genetic methods, we investigated the role of detate nNOS in 5‐HT1A receptor decline and anxiety behavior induced by chronic mild stress (CMS) in mice. Results Here we showed that local elevation of glucocorticoids in the DG accounted for chronic stress‐induced anxiety behavior. Neuronal nitric oxide synthase (nNOS) mediated chronic stress‐induced downregulation of 5‐HT1A receptor in the DG through peroxynitrite anion (ONOO•) pathway but not cyclic guanosine monophosphate (cGMP) pathway. By using pharmacological tool drugs and nNOS knockout mice, we found that nNOS in the DG played a key role in chronic stress‐induced anxiety behavior. Conclusions These findings uncovered an important role of nNOS‐5‐HT1A receptor pathway in the DG of the hippocampus in chronic stress‐induced anxiety. Accordingly, we developed a “dentate nNOS‐5‐HT1A receptor closed‐loop” theory (stress‐glucocorticoids‐nNOS‐Nitric oxide‐ONOO•‐5‐HT1A receptor ‐nNOS) of stress‐related anxiety.


| INTRODUC TI ON
Anxiety disorders are among the most disabling of all medical disorders. 1 It has been well demonstrated that serotonin (5-HT) neurotransmission plays important roles in modulation of numerous diseases. 2,3 These functions are mediated by at least 15 well identified different 5-HT receptors. [4][5][6] Among these receptors,  receptor, in particular, is proved to be implicated in the etiology of anxiety disorders. 7 Several lines of evidence supported that 5-HT 1A receptor level in the hippocampus of patients with anxiety disorders is in deficit and that partial 5-HT 1A receptor agonists are anxiolytics. [8][9][10][11] Moreover, 5-HT 1A receptor knockout mice represent a genetic animal model of anxiety. 7,12 Reportedly, stress is closely correlated with depression and anxiety. 13 Chronic stress increases corticosterone (CORT, glucocorticoids in rodents), decreases 5-HT 1A receptor in the dentate gyrus (DG), and gives rise to anxiety-related behaviors. 14,15 Hence, glucocorticoids may be the predominant substance that mediates the impact of stress on the hippocampus in the development of anxiety.
However, the molecular mechanisms underlying stress and glucocorticoids-induced decrease in expression of hippocampal 5-HT 1A receptor are still unclear.
Neuronal nitric oxide synthase (nNOS) is enriched in the hippocampus. 16 Our previous study revealed that chronic stress upregulates the level and activity of hippocampal nNOS and thus contributes to stress-induced depressive behaviors. 17 There are some reports that nNOS inhibitors produce anxiolytic-like effects, and 5-HT system is dysfunctional in nNOS knockout mice. [18][19][20] However, it remains unknown whether nNOS-NO pathway contributes to 5-HT 1A receptor deficiency in the development of anxiety.
In physiological condition, NO functions through the soluble guanylate cyclase/cyclic guanosine monophosphate/protein kinase G (sGC/cGMP/PKG) pathway. 17,21 Besides, excessive NO, synthesized by increased nNOS, may react with O2·and thereby form peroxynitrite anion (ONOO•). Both NO itself and ONOO• can regulate a variety of molecular expression and function. 17,[21][22][23] Importantly, it was found that sodium nitroprusside (SNP), a type of NO donors, causes anxiogenic effect at high doses but exerts anxiolytic property at low doses, implicating ONOO• in the induction of anxiety behaviors. [24][25][26] Here, using genetic and pharmacological methods, we investigated the link among glucocorticoids, nNOS, and 5-HT 1A receptor. We report that nNOS-NO-ONOO• pathway accounts for the high level of glucocorticoids-induced decline in 5-HT 1A receptor in the DG of the hippocampus and probably contributes to anxiety behavior, uncovering a novel molecular mechanism for anxiety pathogenesis.

| Animals
Young adult (6-to 7-week-old) male homozygous nNOS-deficient mice (B6.129S4-Nos1 tm1Plh /J, KO, stock number: 002986) and their wild-type controls (C57BL/6J, WT, stock number: 000664) (both from The Jackson Laboratory; maintained at Model Animal Research Center of Nanjing University, Nanjing, China), young adult (6-to 8-week-old) ICR mice were used for behavioral analysis. Littermates were used in a same individual experiment in this study. Animals were housed in an air-conditioned room (20 ± 2°C), with food and water ad libitum, except when specified otherwise. Five mice were housed in each cage with 12 hours light/dark cycle. All procedures involving the use of animals were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University (protocol number: IACUC-1704010).
Every effort was made to minimize the number of animals used and their suffering.

| Chronic unpredicted mild stress
The procedure of chronic unpredictable mild stress (CUMS) was designed as described previously. 28 Briefly, the CUMS protocol consists of the sequential application of a variety of mild stressors, including restraint in tubes, forced swimming in cold water, water and/or food deprivation, and pairing with another stressed animal in wet sawdust, reversal of the light/dark cycle, housing in wet sawdust, housing in constant illumination or darkness each for a period ranging from 10 minutes to 24 hours, in a schedule that lasts for 3 weeks. The schedule of stressors in the protocol is unpredicted to mice.

| Corticosterone level measurement
For measurement of CORT level, all mice were decapitated between 9:00 and 11:00 AM. Blood from angulus oculi vessels was collected in heparinized tubes, and CORT in plasma was measured with a CORT ELISA Kit according to the instructions of the manufacturer (Cayman Chemical Company). The detection limit of this kit is approximately 40 pg/mL. Mice were euthanized for plasma collection at least 15 hours after the end of behavior test. 29

| Western blot analysis
Western bolt analysis of samples from cultured hippocampal neurons and hippocampal tissues of animals was performed as described previously. 17 In this study, we dissected the DG from the hippocampus for Western blot. The primary antibodies were as follows: rabbit anti-nNOS (1:1000; Millipore Bioscience Research Reagents), mouse anti-nitrotyrosine (1:3000; Millipore Bioscience Research Reagents), and rabbit anti-5-HT 1A receptor (1:500; Santa Cruz Biotechnology).
Appropriate horseradish peroxidase-linked secondary antibodies were used for detection by enhanced chemilumi-nescence (Pierce; Thermo Fisher Scientific). When we prepared the samples, all the hippocampus was checked to find the needle track of infusion. Only the samples with obvious needle track of Alzet osmotic minipumps or infusion needles were remained for further measurement.

| RNA extraction and reverse transcription-PCR
Total RNA was extracted from the DG of the hippocampus using TRIzol reagent according to the manufacture's instructions (Sigma-Aldrich).

| Microinjection
Adult mice were anesthetized with 0.07 mL of a mixture of ketamine (90.9 mg/mL) and xylazine (9.1 mg/mL) and placed in a stereotaxic apparatus (Stoelting). The drug solutions in 2 μL volume were microinjected into the DGs (0.2 μL/min) at coordinates 2.3 mm posterior to bregma, 1.3 mm lateral to the midline, and 2.0 mm below dura. Osmotic pump was used for continuous delivery of drug. For the osmotic pump implantation, two 7d or 28d Alzet osmotic minipumps (DURECT Corporation) containing cPTIO solution were placed subcutaneously in the back of mice, and two brain infusion cannulas connected to the pump were positioned at the following coordinates: 2.3 mm posterior to bregma, 1.3 mm lateral to the midline, and 2.0 mm below dura. The infusion rate of the osmotic pump was 0.25 μL/h. We anesthetized mice with a mixture of ketamine (90.9 mg/mL) and xylazine (9.1 mg/ mL) and removed the osmotic pumps 1 day before behavioral tests. shown on the right. Mean ± SEM, *P < .05, **P < .01, ***P < .001, compared with Vehicle, # P < .05, ## P < .01, ### P < .001, compared with CUMS + Vehicle in B and G, one-way ANOVA for B and G. Student's t test for C, D, and E multiple comparison corrections. Comparisons between two groups were made with two-tail Student's t test. To compare two independent groups with respect to abnormal distribution, Mann-Whitney U tests were used. For all results, differences were considered significant when P < .05. To know the role of glucocorticoids in stress-induced anxiety-like behaviors, we used metyrapone (at daily dose of 100 mg/kg, s.c. during 21-day CUMS exposure, Figure 1A), a corticosteroids synthesis inhibitor, 17 to block CORT synthesis and to test behavioral changes in the OFT. In consistent with our previous study, the mice, which under-

| Hippocampal nNOS mediates the glucocorticoids-induced 5-HT 1A receptor expression decline
Downregulation of 5-HT 1A receptor in DG granule cells of the hippocampus is essential for induction of anxiety. 33 By Western blot and RT-PCR methods, we found that a high dose of CORT (20 mg/kg, s.c.,

| nNOS downregulates 5-HT 1A receptor expression via ONOO•
NO mediates the main biological function of nNOS. 22 To determine the role of NO in the regulation of 5-HT 1A receptor by CORT, we cleared endogenous NO by cPTIO, a tool drug that directly extinguishes NO generated by NO synthase (NOS) without affecting NOS activity. 17 The cPTIO was infused into the DGs via osmotic minipumps (Alzet, 20 μM, 0.25 μL/h) at day 1, followed by 5 days CORT administration (20 mg/kg, s.c., once per day). Western blots analysis showed that hippocampal 5-HT 1A receptor expression remained
To directly investigate whether the local nNOS in the DG account for glucocorticoids-induced anxiety behavior, we infused 7-NI (10 μM,

| D ISCUSS I ON
Extensive evidences demonstrate that 5-HT 1A receptor deficiency is implicated in the pathogenesis of anxiety and that agonists of 5-HT 1A receptor have anxiolytic effect in clinical use. 8,11,29 However, it remains an important question unresolved that how 5-HT 1A receptor become deficiency during the development of anxiety?
Previous studies from our laboratory found that chronic stress and F I G U R E 4 Dentate nNOS inhibition rescues glucocorticoids-induced anxiety-related behaviors. (A) Central distance of the mice exposed to CUMS for 21 d and the mice treated with 7-NI accompanied with CUMS for 21 d in open-field test. (B) Central time of the mice exposed to CUMS for 21 d and the mice treated with 7-NI accompanied with CUMS for 21 d in open-field test. (C) CUMS and 7-NI have no effect on locomotor activity. (D) Central distance of the mice exposed to CUMS for 21 d and the mice treated with 7-NI infusion into hippocampus followed by CUMS for 21 d in open-field test. (E) Central time of the mice exposed to CUMS for 21 d and the mice treated with 7-NI infusion into hippocampus followed by CUMS for 21 d. (F) CUMS and 7-NI infused into hippocampus have no effect on locomotor activity. Parameters assessed were the number of square crossings (horizontal) and the times of standing (vertical). Mean ± SEM, *P < .05, compared with control or vehichel; #P < .05, compared with CUMS or CORT. One-way ANOVA for A-I, two-way ANOVA for J-L glucocorticoids exposure increased the expression of nNOS in the DG. Here, we report a causal link between the nNOS and the deficiency of the 5-HT 1A receptor in the DG of the hippocampus after high dosage of glucocorticoids (corticosterone in rodent) exposure, which account for the generation of anxiety behavior after chronic stress. Interestingly, we found that ONOO• pathway but not cGMP pathway of NO was involved in the downregulation of 5-HT 1A receptor expression.
The serotonergic system is a main pathway accounting for anxiety disease. Postsynaptic 5-HT 1A receptor is expressed at a high density in the hippocampus. 34 Decreased content of 5-HT 1A receptor is found in the hippocampus of patients with anxiety and depression disorder. 11 Re-expression of 5-HT 1A receptor primarily in the hippocampus and cortex is sufficient to rescue the anxiety phenotype of the 5-HT 1A receptor knockout mice. 7 The hippocampal CA1 and DG densely expresses 5-HT 1A receptor. 35 In the DG, 5-HT 1A receptor majorly expresses on dentate granular cells. 35,36 Both transgenic and pharmacological evidences showed that 5-HT1A receptor on dentate granular cells is engaged for depression and anxiety pathology, and antidepressant response. 29,33,37 However, the molecular mechanism responsible for the key pathophysiological process, the dysfunction of 5-HT 1A receptor on dentate granular cells, remains obscure. It was reported that glucocorticoids have a profound influence on the function of the hippocampus via regulating several lines of genes. 38 Our previous study found that glucocorticoids modulate HPA activity through nNOS-GR (glucocorticoid receptors) pathway. 17 24 It has been demonstrated that 5-HT 1A receptor on the hippocampal DG granule cells is crucial for anxiety-related behavior. 33 Thus, excessive ONOO• eliminates the content of 5-HT 1A receptor in granule cells, which may finally induce anxiety phenotype. The OFT is a common measure of general locomotor activity levels and anxiety in rodent.
Beside OFT, EPM and LDT are another two widely used behavior tests for assessment of anxiety. In this study, the anxiety behavior only assessed in OFT. Thus, more evidence for confirmation of the role of glucocorticoids-nNOS-Nitric oxide-ONOO•-5-HT1A receptor pathway in stress-related anxiety is needed.
Benzodiazepines, a type of typical anxiolytics, exert anxiolytic effects by regulating 5-HT release in the ventral hippocampus. 39 Recently, it is fully demonstrated that 5-HT 1A receptor on mature dentate gyrus granule cells are critical for anxiety-and depression-related behaviors in rodents. 33 Stress potently modulates anxiety-and depression-related behaviors. 13,28,40 For instance, it has been demonstrated that foot shock stress as well as restraint stress can induce anxiety behavior in mice and rats. 41,42 Furthermore, 5-HT 1A receptor expression decrease is strongly implicated in the pathology of anxiety and 5-HT 1A receptor reactivation is critical for the anxiolytic effect of SSRIs. 43,44 In recent years, the function of adult hippocampal neurogenesis in mediating the effects of antidepressants has been realized. 8,45 Treatment of antidepressants increased hippocampal neurogenesis in human and rodents. 46,47 Disruption of hippocampal neurogenesis by X-ray irradiation blocked the anxiolytic action of fluoxetine. 8 Long-term antidepressant treatment activated 5-HT 1A receptor in the forebrain and 5-HT 1A receptor activation in the hippocampus was required for the treatment of F I G U R E 5 A proposed model of nNOS-ONOO•-5-HT 1A receptor pathway in stress-induced anxiety. High level of glucocorticoids-induced dentate nNOS overexpression accounts for 5-HT 1A receptor expression decline in the DG after stress. Our previous study demonstrated that hippocampal 5-HT 1A receptor expression reduction results in nNOS overexpression. Accordingly, the reduction of hippocampal 5-HT 1A receptor after stress, in turn, further increases the expression of nNOS, which is named as "nNOS-5-HT 1A receptor closed-loop." ONOO• mediates the downregulation of 5-HT 1A receptor by nNOS. Overall, the glucocorticoids-triggered "dentate nNOS-5-HT 1A receptor closed-loop" plays a critical role in the pathology development of stress-induced anxiety anxiety-associated behaviors. 29,48 Recently, we discovered that hippocampal neurogenesis is crucial for 5-HT 1A receptor in modulating anxiety behaviors. 37,49 Here, we found that a single intrahippocampal injection of 7-NI into the DG counteracted the anxiety behavior of chronic stress or systemic corticosterone administration. Impaired hippocampal neurogenesis contributed to chronic stress and elevated glucocorticoids-induced depression and anxiety behavior. 27,28 Moreover, nNOS knockout and 7-NI significantly enhanced hippocampal neurogenesis. 28 Therefore, the 5-HT 1A receptor reactivation-induced enhancement of hippocampal neurogenesis might contribute to the anxiety behavior rescue after a single infusion of 7-NI.
In our previous study, we found that blockade of 5-HT 1A receptor increased nNOS expression 29 ( Figure 5, demonstrated previously part). Downregulation of hippocampal nNOS expression mediated the anxiolytic effects of fluoxetine and 5-HT 1A receptor agonists. 29 These findings suggest that nNOS works as a downstream molecule in the 5-HT 1A receptor cascade, explaining the molecular basis of anxiolytic effects of selective serotonin reuptake inhibitor.
However, the function of hippocampal nNOS in stress-induced anxiety remains unclear. Here, we went one more step to understand the pathology of stress-induced anxiety. We found that hippocampal nNOS accounted for chronic stress-induced decrease in 5-HT 1A receptor expression via ONOO• pathway ( Figure 5). Previously, we demonstrated that high concentration of glucocorticoids after stress upregulated nNOS via mineralocorticoid receptor (MR) in the DG of the hippocampus. 17 Altogether, these evidences support a novel hypothesis of anxiety that chronic stress-induced nNOS overexpression reduce the expression of 5-HT 1A receptor in the DG, causing a higher level of nNOS, which in turn lead to more severe deficiency of 5-HT 1A receptor in the DG. We, thus, name this undesirable interaction between nNOS and 5-HT 1A receptor as "nNOS-5-HT 1A receptor loop" (stress-glucocorticoids-nNOS-Nitric oxides-ONOO•-5-HT 1A receptor -nNOS), which functions importantly in the pathology of stress-associated anxiety behavior ( Figure 5). We also illustrated that ONOO• plays an important role in the "nNOS-5-HT 1A receptor loop" under chronic stress state. Additionally, we clarified the mechanism by which 5-HT 1A receptor modulated nNOS. 29 The coordinate used for microinjection in this study specially target the DG of the ventral hippocampus. Therefore, we suggested a glucocorticoids-nNOS-NO-ONOO•-5-HT 1A receptor pathway, in the ventral hippocampal DG, implicated in the pathology of stress-related anxiety.

ACK N OWLED G M ENT
This work was supported by grants from the National Natural

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.