Chronic all-trans retinoic acid administration induced hyperactivity of HPA axis and behavioral changes in young rats

European Neuropsychopharmacology (2010) 20, 839–847

www.elsevier.com/locate/euroneuro

Chronic all-trans retinoic acid administration induced hyperactivity of HPA axis and behavioral changes in young rats Li Cai a,b , Xue-Bo Yan a , Xiao-Ning Chen a , Qing-Yuan Meng a , Jiang-Ning Zhou a,⁎ a

Key Laboratory of Brain Function and Diseases, School of Life Science, University of Science and Technology of China, Chinese Academy of Sciences, Hefei 230027, Anhui, PR China b Department of Pathology, School of Basic Medicine, Anhui Medical University, Hefei 230032, Anhui, PR China

Received 15 December 2009; received in revised form 14 June 2010; accepted 24 June 2010

KEYWORDS Affective disorders; All-trans retinoic acid; Hypothalamus–pituitary– adrenal axis; Corticotropin release factor

Abstract Although clinical reports suggest a possible relationship between excess retinoids and the development of depression, the effect of retinoids on mood-related behavior remains controversial. Hyperactivity of the hypothalamus–pituitary–adrenal (HPA) axis plays a key role in the development of affective disorders. The present study aimed to elucidate the effect of retinoid on the activity of HPA axis in rat and whether this goes together with behavioral changes. All-trans retinoic acid (ATRA) was administered to juvenile male rats by daily intraperitoneal injection for 6 weeks. ATRA treatment increased basal serum corticosterone concentration as well as the thickness of adrenal cortex in young rat. Furthermore, the mRNA expression of corticotropin release factor (CRF) and retinoic acid receptor-α (RAR-α) in the hypothalamus was both markedly increased in ATRA-treated rats compared with vehicle. Some behavioral alterations were also observed. ATRA-treated rats showed anxiety-like behavior in elevatedplus maze and decreased spontaneous exploratory activities in novel open field. However, in the sucrose preference test chronic ATRA treatment did not modify behavior in the juvenile animals. Chronic administration of ATRA did not impair physical motor ability in either the prehensile traction or the beam balance/walk test. In conclusion, long-term ATRA administration resulted in hyperactivated HPA axis which was accompanied by several behavioral changes in young rat. © 2010 Elsevier B.V. and ECNP. All rights reserved.

1. Introduction ⁎ Corresponding author. Tel.: +86 551 3607658; fax: +86 551 3600408. E-mail address: [email protected] (J.-N. Zhou).

Retinoids (vitamin A and its derivatives), specifically alltrans retinoic acid (ATRA), have been extensively examined for their role in the development of central nervous system

0924-977X/$ - see front matter © 2010 Elsevier B.V. and ECNP. All rights reserved. doi:10.1016/j.euroneuro.2010.06.019

840 (Maden, 2007; McCaffery et al., 2003). Accumulating reports indicate that retinoid signaling pathways also affect the functioning of the adult brain including synaptic plasticity, learning and memory, sleep and mood (Lane and Bailey, 2005; O'Reilly et al., 2008; Tafti and Ghyselinck, 2007). Retinoic acids (RAs) have wide therapeutic applications in the field of dermatology and oncology (Chandraratna, 1998). One synthetic retinoid, 13-cis-retinoic acid (13-cis-RA, isotretinoin) has been often prescribed for the treatment of severe nodular acne in adolescence. Case reports have suggested that in 1–10% of patients 13-cis-RA treatment can cause mood disorders including anxiety, depression and increased suicide attempts (Hull and D'Arcy, 2005). In chronic 13-cis-RA administered acne patients, the metabolic activity in the orbitofrontal cortex, a brain area known to mediate symptoms of depression, was significantly changed (Bremner et al., 2005). Other retinoid treatments have been linked to psychiatric side effects in patients as well, such as the oral retinoid treatment for psoriasis (Starling and Koo, 2005). Animal study by de Oliveira et al. showed that 28 days' vitamin A supplementation in rat induced anxiety-like behavior in light–dark box and decreased locomotion and exploration in an open field without motor impairment (de Oliveira et al., 2007). Trent et al. recently reported that chronic 13-cis-RA administration (1 mg/kg/day/i.p., 7–14 days) in adult rats reduced aggression- and increasing flight-related behaviors in the resident-intruder paradigm (Trent et al., 2009). The hypothalamus–pituitary–adrenal (HPA) axis plays an important role in the pathogenesis and neurobiology of affective disorders. Hyperactivity of HPA axis has long been closely associated with the symptoms involved in anxiety, depression and psychosis (Stokes and Sikes, 1991). Recently our laboratory has found, in vitro, that ATRA could upregulate the gene expression of corticotropin release factor (CRF) by the mediation of retinoic acid receptor-α (RAR-α) (Chen et al., 2009). The effects of ATRA on the state of HPA axis have not been examined in the animal model yet. The objective of the current study was to investigate whether long-term ATRA treatment could affect the state of the HPA axis in young rat.

2. Experimental procedures 2.1. Animals Male Sprague–Dawley rats (Anhui Experimental Animal Center) were 3 weeks old at arrival and 4 weeks old at the start of treatment. Rats were housed in colony cages with food and water provided ad libitum and maintained under a 12 h light–dark cycle (lights on at 7:00 a.m.). The ambient temperature was maintained at 21–22 °C with 50–60% relative humidity. Animals were allowed to acclimate for one week during which time they were handled daily. All animals were weighed upon arrival with subsequent measurements taken weekly throughout the course of the experiments. All experiments and animal care procedures were reviewed and approved by the Animal Resource Center of the University of Science and Technology of China in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH publication no. 8023, revised 1978).

L. Cai et al. saline solution (0.9% w/v sodium chloride) with dimethyl sulphoxide (DMSO) at a ratio of 1:1 v/v. Drug treated group (n = 10) received 2 mg/kg all-trans retinoic acid (ATRA, Sigma, St Louis, MO) dissolved in 1:1 v/v DMSO:saline. ATRA solutions were prepared under red light and flasks containing the ATRA were covered with aluminum foil to avoid photoisomerization (O'Reilly et al., 2006). All animals were treated for 6 weeks before behavioral testing commenced and treatment continued until the end of behavioral testing. To eliminate any acute effect of the injections on behavioral testing, injection was always made between 17:00 and 18:00 h, whereas behavioral testing was conducted between 10:00 and 14:00 h. Only one behavioral test was performed each day and approximately 24 h elapsed between tests. The behavioral tests were conducted in the following order: sucrose preference test, elevated-plus maze (EPM), open-field test, prehensile traction, beam balance and beam walk. Individual rats were tested in a random order in each behavioral test. All behavioral scoring was carried out by a trained observer who was unaware of the animal treatment. Daily monitoring confirmed that the animals were not showing any signs of distress as a result of the repeated injections, which were given on alternating sides of the abdominal cavity.

2.3. Behavioral tests 2.3.1. Sucrose preference test Sucrose intake is as a measure of anhedonia, a core symptom of depression. All animals were first trained to drink 2% (w/v) sucrose solution for 24 h. After a 12 h period (between 10 and 12 h of the light phase) of food and water deprivation, animals were given free access to two bottles containing water and 2% sucrose solution, respectively. After 2 h the volumes of water and sucrose consumed were measured. The percentage of sucrose solution from the total liquid ingested was used as a measure for rats' sensitivity to reward (Wu et al., 2007). 2.3.2. Elevated-plus maze Elevated-plus maze was often used to combine with open-field test, evaluating exploratory and anxiety-like behavior (Carobrez and Bertoglio, 2005; Yan et al., 2007). The elevated-plus maze was made of black Plexiglas, consisting of two opposite open arms (50 cm × 10 cm × 0.5 cm), an open platform (10 cm × 10 cm) in the center and two opposite closed arms (50 cm × 10 cm × 40 cm). The apparatus was elevated 50 cm above the floor and illuminated with a dim overhead light. Rat was placed on the central platform facing one of the open arms and allowed to explore the apparatus for 5 min. The following indices were recorded: the total number of entries into open arm and closed arm and the total time spent in each type of arm. From these values, the percentages (%) of both open-arm entries and time spent in open arms were calculated. Arm entry was defined as placing all four paws on it. 2.3.3. Open-field test The open-field apparatus was applied here to analyze spontaneous exploratory activity and curiosity of animals to a novel environment (Chen et al., 2009; Ramos and Mormede, 1998). It was made of black wood and consisted of a floor (96 cm × 96 cm) with 50-cm walls. The box floor was painted with white lines (6 mm) to form 16 equal squares, and the central four squares were defined as the center area. The test was performed under bright ambient room light. Each rat was placed in the center of the open field and left free to explore the unfamiliar arena for 5 min. Total number of squares crossed, central squares crossed, rearing, grooming and fecal pellets was recorded. 2.3.4. Motor function assessment

2.2. Treatment All animals received daily intraperitoneal injections at a volume of 1 ml/kg body weight. Vehicle control group (n = 10) received sterile

2.3.4.1. Prehensile traction test. This test measures muscle strength and equilibrium (Combs and D'Alecy, 1987), when the rat's forepaws are placed on the center of a horizontal rope. Elevated

Retinoid induced hyperactivated HPA axis (50 cm) 0.3-cm-thick and 55-cm-long rope was used and the performance of rats was scored as follows: 0, hangs on 0 to 2 s; 1, hangs on 3 to 4 s; 2, hangs on 5 s, no third limb up to the rope; and 3, hangs on 5 s and brings hind limb up to the rope. The suspension time of each rat before it falls down in 60 s was also recorded. 2.3.4.2. Beam balance and beam walk. These two tasks were utilized to assess gross and fine motor function, respectively (Alexis et al., 1995; Feeney et al., 1982). The beam balance task consists of placing the animal perpendicularly on an elevated (50 cm) narrow wooden beam (2.3 cm × 120 cm) for a maximum of 60 s and assessing the performance with a score from 0 to 6 [0 = balances with steady posture; 1 = grasps side of beam; 2 = hugs the beam and one limb falls down from the beam; 3 = hugs the beam and two limbs fall down from the beam, or spins on beam; 4 = attempts to balance on the beam but falls off (N 40 s); 5 = attempts to balance on the beam but falls off (N20 s); 6 = falls off: no attempt to balance or hang on to the beam (b20 s)] (Chen et al., 2001). The beam walk task was performed on the same elevated beam with markings at 5.0-cm intervals. The right-hand end was blocked, illuminated with 60-W light bulb and the left was connected to a dark platform. Rat was allowed to walk from the right-hand end toward the left for 60 s and the number of segments crossed was recorded. If the rat reached the goal in less than 60 s, the distance traversed was divided by the actual time and normalized to the number of segments per 60 s. If the animal fell from the beam during the test, the distance traversed and the time spent before falling were recorded and normalized to 60 s. Each session consists of 3 trials and the scores were averaged.

2.4. Tissue preparation After the last behavioral measure, rats were transferred to a room adjacent to the laboratory 1 h prior to decapitation. Rats were decapitated between 9:00 a.m. and 11:00 a.m. and the member of the two groups was sacrificed randomly. Whole brains and trunk blood were collected immediately. Trunk blood was collected in foillined tubes, allowed to coagulate for 30 min and then centrifuged at 1000 g for 20 min. Serum was stored at − 80 °C until assay. The whole brains of the ten rats in each group (ATRA-treated group and vehicle group, respectively) were randomly assigned for later measurement: five for CRF immunohistochemistry and five for reverse transcription (RT)-PCR analysis. For immunohistochemistry analysis, the whole brains were fixed in 4% (w/v) paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4 °C for 24 h, then blocked and embedded in paraffin according to the standard histological techniques. For mRNA analysis, hypothalamic parts were quickly dissected from the brains with the following limits (Huang et al., 2008; Yasin et al., 1993): anterior border of the optic chiasm, anterior border of the mamillary bodies, and lateral hypothalamic sulci. The depth of dissection was approximately 3 mm. Tissues were quickly frozen in liquid nitrogen and preserved at − 80 °C.

841 2.6. Measurement of serum corticosterone The corticosterone serum concentration was measured using a commercially available Enzyme-Linked Immunosorbent Assay (ELISA) kit (RapidBio, CA) according to the respective protocols.

2.7. Index of adrenal gland and morphology of adrenal cortex The adrenal glands were removed and weighed immediately post mortem. The index of adrenal gland was expressed as the ratio of adrenal gland weight to body weight. Hematoxylin and eosin (HE) stained paraffin sections were prepared, examined under a light microscope (Nikon 80i, Japan) and representative photos were taken. The thickness of adrenal cortex was measured using a digital image analysis software program (Motic Virtual Microscope 1.0) and obtained by measuring the distance between the medulla and the adrenal capsule in a straight line, one measurement being taken in each quadrant of the adrenal cortex. Three sections were examined in each animal and the average value was used.

2.8. Immunohistochemistry Serial coronal sections (3 μm) of the hypothalamus were cut using a Leica microtome (Leica RM 2135) and collected at 150-μm intervals. Sections were dewaxed in xylene and rehydrated through a graded ethanol series; rinsed in phosphate buffered saline (PBS) for 10 min and were microwave-treated in 0.05 M citrate-buffered saline (pH 6.0) at 90 °C (2 × 10 min) for antigen retrieval. Then, sections were rinsed, incubated in 3% hydrogen peroxide in methanol (10 min) to quench endogenous peroxidase activity, rinsed, and incubated in 5% normal goat serum (Vector Laboratories, Burlingame, CA) in PBS for 30 min at 37 °C to block nonspecific staining. Subsequently, sections were incubated overnight at 4 °C with primary antibody (rabbit antiCRF, 1:2000, Bachem, CA) diluted in PBS. Then the sections were washed and respectively incubated with biotinylated IgG (1:200, Vector Laboratories) and avidin–biotin horseradish peroxidase complex (1:200, Vector Laboratories) for 30 min at 37 °C. Sections were visualized with 0.025% (w/v) DAB (Sigma Chemicals, St Louis, MO) in 0.05 M Tris–HCl buffer (pH 7.6) and 0.01% (v/v) H2O2 for 20 min, washed, then dehydrated, hyalinized, and sealed with neutral gum. The sections were examined and photographed using a computerized Nikon 80i light microscope equipped with a Canon Digital camera (PowerShotS 40). For each animal sections containing the paraventricular nucleus (PVN) were identified using a rat brain stereotaxic atlas (Paxinos and Watson, 1998). The number of CRFimmunoreactive cells within the PVN was counted bilaterally at 400× magnification by an investigator blind to the experimental treatment and the total number of CRF-immunoreactive cells throughout the PVN for each animal was semiquantatively estimated as described previously (Bao et al., 2005).

2.5. Measurement of serum ATRA levels 2.9. RNA isolation and real time quantitative PCR (Q-PCR) Serum ATRA levels were measured by reversed phase high performance liquid chromatography (RP-HPLC) (Le Douze et al., 2000). To prevent retinoid degradation, HPLC steps were performed in dark. Serum (200 μl) were mixed with 200 μl of acetonitrile, shaken, and centrifuged (4000 rpm for 10 min). An aliquot (50 μl) of the upper phase was injected directly into the HPLC system. The chromatographic conditions were: column, reversed phase C18 [15 cm × 4.6 mm; particle size, 5 μm] (Supelco, St. Quentin Fallavier, France); mobile phase, 57.5% acetonitrile/25% acetic acid (2% in water)/17.5%methanol; flow rate, 1.3 ml/min; UV detection wavelength, 354 nM. Standard curve was prepared by adding known amounts of ATRA in DMSO to serum blank samples (200 μl). Two quality controls (low and high) were tested to estimate the reproducibility, precision, and reliability of the method.

Total RNA was extracted from the frozen hypothalamus using the Trizol (Invitrogen, Carlsbad, CA) method. cDNA was synthesized using reverse transcriptase (Promega, Wisconsin, USA). Q-PCR was performed using SYBR Green PCR Kit (Applied Biosystems, USA) and an ABI Prism 7000 Sequence Detector system in 25 μl volume for 40 cycles (15 s at 95 °C; 60 s at 64 °C for rat CRF and CRABP1, 60 s at 66 °C for rat RAR-α). The following primers were used: rat CRF 5′-CAGAACAACAGTGCGGGCTCA-3′ and 5′-AAGGCAGACAGGGCGACAGAG-3′; rat CRABP1 5′-CTCCTCAAGGCTCTGGGTGT-3′ and 5′-CGCACCGTAGTGGATGTCTT-3′; rat RAR-α 5′-ACCATTGCCGACCAGATTACCC-3′ and 5′AAGGTCATTGTGTCTTGCTCAGGT-3′; rat beta-actin 5′-TTGCTGACAGGATGCAGAA-3′ and 5′-ACCAATCCACACAGAGTACTT-3′. The relative amount of target gene was calculated using the 2−ΔΔCt method (Livak

842 and Schmittgen, 2001). The relative amplification efficiencies of the primers were tested and shown to be similar.

2.10. Statistical analysis Statistical analysis was performed using the SPSS 11.5 software for Windows. All results were expressed as mean ± standard error of the mean (SEM). The differences of means between vehicle group and ATRA group were analyzed by Independent-Samples T test with 95% confidence interval limits. p b 0.05 was considered to be statistically significant.

3. Results 3.1. Body weight Weekly weighing of the rats showed that body weight during the experiment was normal and that there was no significant difference in body weight between the ATRA-treated group and the control group in any week. At the start of injections, the mean body weight of control and ATRA-treated rats was 78.84 ± 13.56 g and 76.91 ± 13.28 g, respectively (n= 10 per group). After 6 weeks of administration, the ATRA-treated group weighed 317.2 ± 17.9 g and the control group weighed 316.6 ± 29.5 g. The behavioral tests had no effect on the weight of the animals.

3.2. Sucrose preference test There were no significant differences in sucrose preference between ATRA-treated rats and control rats (p N 0.05, Fig. 1a).

L. Cai et al.

3.3. Elevated-plus maze Fig. 1b and c shows the differences between the two groups in the EPM performance. In the chronic ATRA-treated rats, there was a significant reduction in the open:total arm entry ratio in the maze (p b 0.05, Fig. 1b). Furthermore, a significant reduction in the % time spent in the open arms of the maze was also detected in the ATRA-treated group compared with vehicle (p b 0.01, Fig. 1c).

3.4. Open-field test Fig. 1d–f illustrates open-field activity by chronic ATRA administration. ATRA-treated rats were less active than control rats in the unfamiliar open field. The total number of squares crossed of ATRA group was markedly lower than that of control group (p b 0.05, Fig. 1d). Long-term ATRA treatment also decreased the number of rearing in rats in the novel environment (p b 0.01, Fig. 1e). There were no statistical differences between the two groups in the number of center square crossing (Fig. 1f), grooming and fecal pellets.

3.5. Motor function assessment To determine whether any behavioral effects of ATRA administration could be attributed to individual motor deficits, an overall observation on physical motor ability or coordination was conducted by the prehensile traction and beam balance/walk tests. In the prehensile traction test, ATRA-treated rats showed similar performance to vehicle-treated rats and there was no difference in the suspension time between the two groups. There was also no significant difference in beam balance score or

Figure 1 Effects of chronic ATRA treatment on the behavioral changes. The presented data are mean ± SEM with n = 10 for each group. *p b 0.05, **p b 0.01 versus vehicle-treated controls. a: Effects of chronic ATRA treatment on the percentage of sucrose consumption from the total liquid in 2 h in sucrose preference test. b and c: Effects of chronic treatment with ATRA on performance in the elevated-plus maze. Data show the percentage of entries in open arms (b) and the percentage of time spent in open arms (c). An open-arm entry was defined as all four of the paws being placed in the open arm. d to f: Effects of chronic ATRA treatment on the openfield activity. Data show the number of total squares crossed (d), rearing (e) and central activity (f) in 5 min in open field.

Retinoid induced hyperactivated HPA axis

843 caused a significant elevation of basal corticosterone concentration in young rats compared with vehicle (p b 0.05, Fig. 2).

3.8. Index of adrenal gland and thickness of adrenal cortex

Figure 2 Serum corticosterone levels in basal conditions in ATRA and vehicle-treated groups. Data are mean ± SEM with 10 rats in each group. *p b 0.05 versus vehicle-treated controls. segments traversed per min between the two groups. The above results indicate that ATRA did not affect overall physical motor function in rats compared with vehicle.

Chronic ATRA treatment significantly increased both the index and the cortex thickness of adrenal gland in young rats relative to vehicle (p b 0.05, Fig. 3a; p b 0.01, Fig. 3b). Moreover, ATRA induced morphological changes in the adrenal cortex of rats and representative pictures are listed in Fig. 3c–f. Although the adrenal cortexes of ATRA-treated animals had a general architecture similar to that of the control group, they were more voluminous, especially in ZF, compared with those of the animals in the control group and had an intensely vascularized ZR (Fig. 3c and d). In addition, in spite of the fact that big lipid vacuoles were scarcely seen in the inner ZF- and outer ZR-cells of controls, the same cells of ATRA-treated rats were crowded with large lipid vacuoles (Fig. 3e and f).

3.6. ATRA serum levels in treated rats ATRA serum concentration was examined. The serum level of ATRA in ten rats chronically treated with ATRA was 1.86± 0.85 μg/ml. In the vehicle-treated rats, however, serum samples contained essentially no measurable ATRA.

3.7. Serum corticosterone levels Fig. 2 shows serum corticosterone levels in basal conditions in vehicle and ATRA-treated rats. Chronic ATRA administration

3.9. The expression of CRF-immunoreactive neurons in the PVN Chronic ATRA administration resulted in a significant augmentation of CRF-immunoreactive neurons in the paraventricular nucleus (PVN) as compared to that of vehicle group (p b 0.01, Fig. 4a), and representative photomicrographs of CRF-immunoreactive neurons in the PVN taken from adjacent series of sections from vehicle and ATRA-treated rats are listed (Fig. 4b–e). There were pronounced cluster of CRF-expressing cells in the hypotha-

Figure 3 Effects of chronic ATRA treatment on index of adrenal gland and thickness of adrenal cortex (a, b) and representative photomicrographs of the cortex of adrenal gland from control group (c, e) and chronic ATRA-treated group (d, f). ZG: zona glomerulosa; ZF: zona fasciculata; ZR: zona reticularis. Line segment shows the thickness of adrenal cortex. Solid arrows indicate large lipid vacuoles. Dotted arrows indicate vascularization in ZR. HE staining, original magnification: × 100 (c, d); × 400 (e, f). Bar = 100 μm in c and d, 50 μm in e and f. n = 10 for each group, *p b 0.05, **p b 0.01 versus vehicle-treated controls.

844

L. Cai et al.

Figure 4 Effects of chronic ATRA treatment on CRF expression in the PVN of hypothalamus and median eminence (vehicle group: b, d, f; ATRA group: c, e, g). The CRF was detected by immunohistochemical technique (see Experimental procedures). Solid arrows indicate PVN location. Dotted arrows indicate median eminence. In f and g, nuclei are stained with hematoxylin. 3V: third ventricle. Original magnification: ×100 (b, c, f, g); ×200 (d, e). Bar = 100 μm in b, c, d, e; 200 μm in f and g. n = 5 for each group, **p b 0.01 versus vehicle group. lamic PVN in ATRA-treated rats (Fig. 4c, amplified in Fig. 4e). In contrast, there were only a few of the CRF-immunoreactive cells in the PVN in vehicle-treated rats (Fig. 4b, amplified in Fig. 4d). Long-term ATRA administration also resulted in an enhancement of CRF immunoreactivity in fibers in the external lamina of the median eminence in rats compared with that in vehicle rats (Fig. 4f and g).

3.10. mRNA expression of CRF, RAR-α and CRABP1 in the hypothalamus Fig. 5 shows the mRNA expression of CRF, RAR-α and cellular retinoic acid binding protein 1 (CRABP1) in the hypothalamus of the two groups. Real time quantitative PCR revealed that CRF mRNA expression in the hypothalamus was significantly higher

Figure 5 Effects of chronic ATRA administration on the mRNA expression of CRF (a), RAR-α (b) and CRABP1 (c) in the hypothalamus, detected by Q-PCR. Data represent mean ± SEM with n = 5 in each group, *p b 0.05 versus vehicle group.

Retinoid induced hyperactivated HPA axis in chronic ATRA-treated group compared to the vehicle-treated group (p b 0.05, Fig. 5a). The RAR-α mRNA level was also markedly increased in the hypothalamus of ATRA-treated rats (p b 0.05, Fig. 5b). There was no statistical difference in the CRABP1 mRNA expression between the two groups (p = 0.75, Fig. 5c).

4. Discussion Given the possible relationship between retinoid and mood disorders, the present study demonstrated, for the first time, that chronic exposure to ATRA in young rat induced hyperactivity of the HPA axis and accompanied behavioral alterations. ATRA-treated rats showed anxiety-like behavior in elevatedplus maze and decreased exploratory activities in novel open field. The role of retinoid in mood disorder has recently attracted an increasing attention due to the negative psychiatric side effects in clinic (Hull and D'Arcy, 2005). Animal models are beginning to be employed to elucidate the possible link between retinoid treatment and affective disorders. O'Reilly et al. reported chronic 13-cis-RA treatment induced depression-related behavior in juvenile mice (O'Reilly et al., 2006). However, Ferguson et al. found that chronic oral treatment with 13-cis-RA or ATRA in adult rats did not have a prodepressive effect in the forced swim test or sucrose anhedonia paradigm (Ferguson et al., 2005, 2007). More recently Trent et al. showed that chronic 13-cis-RA administration changed the resident-intruder paradigm in rats: significantly reduced aggressive behavior towards intruder rats and increased submissive behavior including flight-submit and flight-escape behavior (Trent et al., 2009). There are many important differences among all these research (including ours): (1) different doses of different drug. ATRA is the bioactive form and exerts the most potent effects of any retinoid (Lane and Bailey, 2005). 13-cis-RA's pharmacologic actions are thought to be exerted via isomerization to ATRA (Crettaz et al., 1990; Geiger et al., 1996). Here we administered ATRA at the same dose with the upper limit of clinical acne treat (2 mg/kg/day) and achieved serum levels of 1.86 ± 0.85 μg/ml: (2) different routes of administration (i.p. versus oral gavage). Oral administration in rodents may result in a gavage-associated trauma which makes it difficult to measure exact concentration of medication. Ferguson et al. reported two animals euthanized or found dead due to gavage injuries during oral treatment (Ferguson et al., 2007). Considering the variance of absorption and bioavailability in vivo, the administration route might be important in determining behavioral consequences associated with RA treatment: (3) different ages of test subjects. ATRA and 13-cis-RA are prescribed primarily to human adolescents clinically (Cambier et al., 2001; Zouboulis and Piquero-Martin, 2003), so in our rat model we began treatment at 4 weeks of age, which can be considered equivalent to human adolescence (Spear, 2000). Adolescence and puberty are times of significant developmental changes in the brain, particularly in relation to motivational and emotional behaviors, that could make this age group particularly vulnerable to the effects of retinoids (Spear, 2000). Finally, Trent et al. suggested that the sensitivity of different behavioral tests might also influence results (Trent et al., 2009).

845 The aim of the current study was to evaluate the effects of long-term ATRA administration on HPA axis in young rats. The HPA axis is the core neuroendocrine system involved in regulation of the ‘flight, fight, fright’ response to stressful stimuli. It is an integrated multilevel axis, responsive to environmental as well as endogenous events, and the HPA physiology is complex. Dysregulation of HPA axis has been reported in various psychiatric illnesses and is believed to take an important part in the pathogenesis of affective disorders (Stokes and Sikes, 1991; Swaab et al., 2005). However, the effect of retinoid on the state of HPA axis has not been examined in the animal model till now. As ATRA is a lipophilic molecule, it easily crosses the blood–brain barrier (Le Douze et al., 2000). After 6 weeks of ATRA administration, we found that CRF, the initiating factor of HPA axis, was enhanced at both protein and mRNA levels in the hypothalamus of ATRAtreated animals. Following secretion from nerve terminals in the median eminence, CRF activates adrenocorticotrophic hormone (ACTH) secretion from anterior pituitary, which finally stimulates the secretion of corticosterone from the adrenal cortex. Here in the present study, we found that the adrenal index as well as the thickness of adrenal cortex in ATRA group significantly increased and many large lipid vacuoles were observed in the zona fasciculata cells, suggesting the hyperplasia of adrenal cortex and an active anabolic metabolism. We also noticed that the adrenal cortex of ATRA group had an intensely vascularized ZR which was not mentioned in stressed rats before. The vascularization observed here might be an indicator of vigorous functioning status since ZR-cells could produce and secrete glucocorticosteroid. Consistently, the basal corticosterone serum concentration was markedly up-regulated in ATRA group. Although we did not examine the level of plasma ACTH, the enhancement of CRF immunoreactivity in fibers in the external lamina of the median eminence indicated an increased input to the pituitary in chronic ATRAtreated rats. These above results indicate that long-term ATRA administration induced an over activity of HPA axis at multiple levels in young rat. Accompanied with hyperactivated HPA axis, several behavioral changes were observed. Chronic ATRA-treated young rats presented decreased open-arm entries as well as time spent in open arms in EPM test. The exploratory activity (total squares crossed and rearing) in a brightly illuminated novel open field was also decreased in ATRA group. The behavioral changes might be related to the hyperactivity of the HPA axis, since it has been demonstrated that excessive CRF and corticosterone could induce anxietylike behavior and decrease the spontaneous locomotor and exploratory activity of rats in an aversive novel environment (Stokes and Sikes, 1991; Owens and Nemeroff, 1991). However, chronic ATRA administration did not induce anhedonia, the main symptom of depression, in juvenile rats in sucrose preference test, which was also shown by others using 13-cis-RA (Ferguson et al., 2005, 2007; Trent et al., 2009). There was no impairment of physical motor ability or coordination in ATRA-treated animals, suggesting that the decreased total activity in open field is not due to an effect on motor systems. In order to preclude the possible interference of acute ATRA effect during the experiment, we also examined the above behaviors in rats received one time ATRA intraperitoneal injection (2 mg/kg) in the same time point, i.e. injection was made between 17:00 and 18:00 h and behavioral testing was conducted between 10:00 and 14:00 h

846 the next day. We did not observe significant behavioral changes in these rats. The presence of RAR-α (Krezel et al., 1999; Zetterstrom et al., 1999) together with the existence of local retinoic acid synthesis (Luo et al., 2004) and abundant cellular retinoid binding proteins CRBP and CRABP expression (Zetterstrom et al., 1999), suggests the possibility of active RA signaling in the rodent hypothalamus. Recently our laboratory has found over-activated RA metabolism and signaling in the hypothalamus of depressed patients as well as rats (Chen et al., 2009). Moreover, recruitment of RAR-α by the CRF promoter was observed in rat hypothalamus and studies in vitro demonstrated that RAR-α mediated an upregulation of CRF gene expression (Chen et al., 2009). Here in the present study, after long-term ATRA administration, we also observed increased levels of RAR-α in the hypothalamus in young rats, which was accompanied by up-regulated CRF expression. Since ATRA is the final output of cellular RA synthesis, we did not examine the upstream members of RA signaling pathway such as RALDH 3 (Lane and Bailey, 2005). ATRA exerts its effects by activating the nuclear RAR-α which further regulates the transcription of a large number of retinoid-responsive neuronal genes (Lane and Bailey, 2005). Considering the intricate neuronal network in the brain, the increased expression of CRF in the hypothalamus observed here might be either a direct transcriptional-regulating result by ATRA or an indirect downstream effect of RA signaling, which needs more refined work to confirm in the future. In our previous study we found an increased CRABP1 in chronic unpredictable mild stress rat model (Chen et al., 2009) whereas there was no significant difference in the CRABP1 mRNA expression in the present study. CRABP1 is the main cellular RA binding protein functioning to decrease cellular responses to ATRA by catalyzing its degradation (Fiorella et al., 1993). This inconsistency may indicate that ATRA is catalyzed differently under different conditions of rat model which deserves for the further study. Combined with the recent findings in our laboratory, these results suggest that the increased CRF in the hypothalamus might be mediated by the RA-RAR-α-CRF pathway. In conclusion, long-term over exposure of ATRA resulted in up-regulated activity of HPA axis accompanied by behavioral changes in young rat. Even though it is difficult to extrapolate the results presented here to humans, these results may provide a possible explanation for the adverse psychiatric effects of retinoids and indicate that RA signaling pathway is involved in the pathogenesis of affective disorders. More investigations remain to elucidate whether intervention of RA signaling pathway may influence HPA axis in a specific way.

Role of the funding source This study was supported by the Nature Science Foundation of China (30530310). The funding source had no further role in study design; in the collection, analysis and interpretation of data; in the writing the report; and in the decision to submit the paper for publication.

Contributors Li Cai contributed to the study design, experimental work, data analysis and preparation of the manuscript. Xue-Bo Yan, Xiao-Ning

L. Cai et al. Chen and Qing-Yuan Meng contributed to the technical help of experiment and data analysis. Jiang-Ning Zhou contributed to the study design, instruction of experiment and preparation of the manuscript. All authors contributed to and have approved the final manuscript.

Conflict of interest None of the authors has any conflict of interest related to this study.

Acknowledgements We would like to thank Mr. Feng Yang and Dr. Rong Li for their technical assistance.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.euroneuro.2010.06.019.

References Alexis, N.E., Dietrich, W.D., Green, E.J., Prado, R., Watson, B.D., 1995. Nonocclusive common carotid artery thrombosis in the rat results in reversible sensorimotor and cognitive behavioral deficits. Stroke 26, 2338–2346. Bao, A.M., Hestiantoro, A., Van Someren, E.J., Swaab, D.F., Zhou, J.N., 2005. Colocalization of corticotropin-releasing hormone and oestrogen receptor-alpha in the paraventricular nucleus of the hypothalamus in mood disorders. Brain 128, 1301–1313. Bremner, J.D., Fani, N., Ashraf, A., Votaw, J.R., Brummer, M.E., Cummins, T., Vaccarino, V., Goodman, M.M., Reed, L., Siddiq, S., Nemeroff, C.B., 2005. Functional brain imaging alterations in acne patients treated with isotretinoin. Am. J. Psychiatry 162, 983–991. Cambier, N., Menot, M.L., Schlageter, M.H., Balitrand, N., Leblanc, T., Bordigoni, P., Rohrlich, P., Lamagnère, J.P., Donadieu, J., Herbelin, C., Puissant, C., Gourand, F., Baruchel, A., Chomienne, C., 2001. All trans retinoic acid abrogates spontaneous monocytic growth in juvenile chronic myelomonocytic leukaemia. Hematol. J. 2, 97–102. Carobrez, A.P., Bertoglio, L.J., 2005. Ethological and temporal analyses of anxiety-like behavior: the elevated plus-maze model 20 years on. Neurosci. Biobehav. Rev. 29, 1193–1205. Chandraratna, R.A., 1998. Future trends: a new generation of retinoids. J. Am. Acad. Dermatol. 39, S149–S152. Chen, J., Li, Y., Wang, L., Zhang, Z., Lu, D., Lu, M., Chopp, M., 2001. Therapeutic benefit of intravenous administration of bone marrow stromal cells after cerebral ischemia in rats. Stroke 32, 1005–1011. Chen, X.N., Meng, Q.Y., Bao, A.M., Swaab, D.F., Wang, G.H., Zhou, J.N., 2009. The involvement of retinoic acid receptor-alpha in corticotropin-releasing hormone gene expression and affective disorders. Biol. Psychiatry 66, 832–839. Combs, D.J., D'Alecy, L.G., 1987. Motor performance in rats exposed to severe forebrain ischemia: effect of fasting and 1,3-butanediol. Stroke 18, 503–511. Crettaz, M., Baron, A., Siegenthaler, G., Hunziker, W., 1990. Ligand specificities of recombinant retinoic acid receptors RAR alpha and RAR beta. Biochem. J. 272, 391–397. de Oliveira, M.R., Silvestrin, R.B., Mello, E.S.T., Moreira, J.C., 2007. Oxidative stress in the hippocampus, anxiety-like behavior and decreased locomotory and exploratory activity of adult rats:

Retinoid induced hyperactivated HPA axis effects of sub acute vitamin A supplementation at therapeutic doses. Neurotoxicology 28, 1191–1199. Feeney, D.M., Gonzalez, A., Law, W.A., 1982. Amphetamine, haloperidol, and experience interact to affect rate of recovery after motor cortex injury. Science 217, 855–857. Ferguson, S.A., Cisneros, F.J., Gough, B., Hanig, J.P., Berry, K.J., 2005. Chronic oral treatment with 13-cis-retinoic acid (isotretinoin) or all-trans-retinoic acid does not alter depression-like behaviors in rats. Toxicol. Sci. 87, 451–459. Ferguson, S.A., Cisneros, F.J., Hanig, J.P., Berry, K.J., 2007. Oral treatment with ACCUTANE does not increase measures of anhedonia or depression in rats. Neurotoxicol. Teratol. 29, 642–651. Fiorella, P.D., Giguere, V., Napoli, J.L., 1993. Expression of cellular retinoic acid-binding protein (type II) in Escherichia coli. Characterization and comparison to cellular retinoic acid-binding protein (type I). J. Biol. Chem. 268, 21545–21552. Geiger, J.M., Hommel, L., Harms, M., Saurat, J.H., 1996. Oral 13-cis retinoic acid is superior to 9-cis retinoic acid in sebosuppression in human beings. J. Am. Acad. Dermatol. 34, 513–515. Huang, Q., Zhu, H., Fischer, D.F., Zhou, J.N., 2008. An estrogenic effect of 5alpha-androstane-3beta, 17beta-diol on the behavioral response to stress and on CRH regulation. Neuropharmacology 54, 1233–1238. Hull, P.R., D'Arcy, C., 2005. Acne, depression, and suicide. Dermatol. Clin. 23, 665–674. Krezel, W., Kastner, P., Chambon, P., 1999. Differential expression of retinoid receptors in the adult mouse central nervous system. Neuroscience 89, 1291–1300. Lane, M.A., Bailey, S.J., 2005. Role of retinoid signalling in the adult brain. Prog. Neurobiol. 75, 275–293. Le Douze, F., Debruyne, D., Albessard, F., Barre, L., Defer, G.L., 2000. Pharmacokinetics of all-trans retinoic acid, 13-cis retinoic acid, and fenretinide in plasma and brain of rat. Drug Metab. Dispos. 28, 205–208. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2 (− Delta Delta C(T)) Method. Methods 25, 402–408. Luo, T., Wagner, E., Grun, F., Drager, U.C., 2004. Retinoic acid signaling in the brain marks formation of optic projections, maturation of the dorsal telencephalon, and function of limbic sites. J. Comp. Neurol. 470, 297–316. Maden, M., 2007. Retinoic acid in the development, regeneration and maintenance of the nervous system. Nat. Rev. Neurosci. 8, 755–765. McCaffery, P.J., Adams, J., Maden, M., Rosa-Molinar, E., 2003. Too much of a good thing: retinoic acid as an endogenous regulator of neural differentiation and exogenous teratogen. Eur. J. Neurosci. 18, 457–472.

847 O'Reilly, K.C., Shumake, J., Gonzalez-Lima, F., Lane, M.A., Bailey, S.J., 2006. Chronic administration of 13-cis-retinoic acid increases depression-related behavior in mice. Neuropsychopharmacology 31, 1919–1927. O'Reilly, K., Bailey, S.J., Lane, M.A., 2008. Retinoid-mediated regulation of mood: possible cellular mechanisms. Exp. Biol. Med. 233, 251–258. Owens, M.J., Nemeroff, C.B., 1991. Physiology and pharmacology of corticotropin-releasing factor. Pharmacol. Rev. 43, 425–473. Paxinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates, Fourth Edition. Academic Press, Spiral Bound, New York. Ramos, A., Mormede, P., 1998. Stress and emotionality: a multidimensional and genetic approach. Neurosci. Biobehav. Rev. 22, 33–57. Spear, L.P., 2000. The adolescent brain and age-related behavioral manifestations. Neurosci. Biobehav. Rev. 24, 417–463. Starling III, J., Koo, J., 2005. Evidence based or theoretical concern? Pseudotumor cerebri and depression as acitretin side effects. J. Drugs Dermatol. 4, 690–696. Stokes, P.E., Sikes, C.R., 1991. Hypothalamic–pituitary–adrenal axis in psychiatric disorders. Annu. Rev. Med. 42, 519–531. Swaab, D.F., Bao, A.M., Lucassen, P.J., 2005. The stress system in the human brain in depression and neurodegeneration. Ageing Res. Rev. 4, 141–194. Tafti, M., Ghyselinck, N.B., 2007. Functional implication of the vitamin A signaling pathway in the brain. Arch. Neurol. 64, 1706–1711. Trent, S., Drew, C.J., Mitchell, P.J., Bailey, S.J., 2009. Chronic treatment with 13-cis-retinoic acid changes aggressive behaviours in the resident-intruder paradigm in rats. Eur. Neuropsychopharmacol. 19, 876–886. Wu, L.M., Han, H., Wang, Q.N., Hou, H.L., Tong, H., Yan, X.B., Zhou, J.N., 2007. Mifepristone repairs region-dependent alteration of synapsin I in hippocampus in rat model of depression. Neuropsychopharmacology 32, 2500–2510. Yan, X.B., Wang, S.S., Hou, H.L., Ji, R., Zhou, J.N., 2007. Lithium improves the behavioral disorder in rats subjected to transient global cerebral ischemia. Behav. Brain. Res. 177, 282–289. Yasin, S.A., Costa, A., Besser, G.M., Hucks, D., Grossman, A., Forsling, M.L., 1993. Melatonin and its analogs inhibit the basal and stimulated release of hypothalamic vasopressin and oxytocin in vitro. Endocrinology 132, 1329–1336. Zetterstrom, R.H., Lindqvist, E., Mata de Urquiza, A., Tomac, A., Eriksson, U., Perlmann, T., Olson, L., 1999. Role of retinoids in the CNS: differential expression of retinoid binding proteins and receptors and evidence for presence of retinoic acid. Eur. J. Neurosci. 11, 407–416. Zouboulis, C.C., Piquero-Martin, J., 2003. Update and future of systemic acne treatment. Dermatology 206, 37–53.