INTRODUCTION

Retinoids, including vitamin A and its derivatives, have long been studied for their role in embryonic development because excess retinoids induce neural tube defects. Emerging evidence now indicates that retinoid signaling pathways also affect the functioning of the adult brain (for a review, see Mey and McCaffery, 2004; Lane and Bailey, 2005). Retinoids exert their effects by binding to nuclear retinoic acid receptors (RAR), which function as heterodimers with retinoid X receptors (RXR) (for a review, see Lane and Bailey, 2005). The RAR/RXR heterodimer regulates gene transcription by binding to specific DNA sequences termed retinoic acid response elements (RARE) found in the promoter regions of retinoid responsive genes. The adult brain expresses many of the components of the retinoid signaling machinery necessary to control retinoid metabolism and transport. For example, retinal dehydrogenases, which synthesize active all-trans-retinoic acid from dietary vitamin A (McCaffery and Drager, 1994; Thompson Haskell et al, 2002; Wagner et al, 2002), cellular retinoid binding proteins (Zetterstrom et al, 1994, 1999), and RARα and RARβ (Krezel et al, 1999; Zetterstrom et al, 1999) are all expressed in the adult brain. Given the role of retinoids in neuronal differentiation, it is not surprising that a large number of neuronal genes have been shown to be retinoid-responsive (for a review, see Lane and Bailey, 2005). However, only a small subset of neuronal genes have identified, functional RARE in their promoter regions: dopamine D2 receptor (Valdenaire et al, 1994; Samad et al, 1997; Valdenaire et al, 1998), neurogranin (RC3) (Iniguez et al, 1994; Husson et al, 2004), gonadotrophin-releasing hormone (Cho et al, 2001a, 2001b), and oxytocin (Richard and Zingg, 1991). Consequently, retinoid-responsive gene transcription in the central nervous system could have a significant impact on the function of the adult brain. Indeed, such a role for retinoids has been demonstrated in learning and memory behaviors (Chiang et al, 1998; Etchamendy et al, 2001; Cocco et al, 2002; Etchamendy et al, 2003; Crandall et al, 2004).

13-cis-RA (or isotretinoin) is a synthetic retinoid and the active ingredient in Accutane, a widely prescribed oral acne medication (Shalita et al, 1988). Originally intended to treat severe nodular acne that was unresponsive to conventional treatments, the use of 13-cis-RA for milder forms of acne is increasing (Hull and D'Arcy, 2003). Since the mid-1980s, reports in the literature suggest a relationship between 13-cis-RA use and the onset of psychological symptoms including depression, suicidal ideation, and psychosis (Hazen et al, 1983; Bigby and Stern, 1988; Scheinman et al, 1990; Byrne and Hnatko, 1995; Wysowski et al, 2001; Ng and Schweitzer, 2003; O'Connell et al, 2003; Barak et al, 2005). However, there are also studies suggesting that there is no evidence of such a link (Jick et al, 2000; Jacobs et al, 2001; Ferahbas et al, 2004; Chai et al, 2005). One recent study used positron emission tomography to examine brain function in patients with treatment-resistant acne (Bremner et al, 2005). In patients treated with 13-cis-RA, but not antibiotic, there was a significant decrease in orbitofrontal cortical metabolism. However, there was no difference in depressive symptoms between the two treatment groups (Bremner et al, 2005).

Patients with disfiguring acne exhibit problems with self-esteem, self-image, depression, and anger (Koo, 1995) that could confound depressive symptoms arising due to 13-cis-RA treatment. The use of an animal model allows for the separation of the effects of acne on self-image and mood from those of 13-cis-RA itself on depression-related behavior. However, there are only a few reports concerning the effects of 13-cis-RA in animal models of depression-related behavior. Crandall et al (2004) have shown that long-term administration of 13-cis RA in adult mice suppresses hippocampal neurogenesis (Crandall et al, 2004). Deficits in adult neurogenesis in the hippocampus have been proposed to underlie depression (Jacobs et al, 2000), especially as they are sensitive to reversal by antidepressant administration (Malberg and Schechter, 2005). Crandall et al (2004) reported impairments in spatial learning but did not examine depression-related behaviors in mice. More recently, Ferguson et al (2005) reported that high doses of 13-cis-RA do not affect depression-related behavior in adult rats. The objective of the present study was to determine whether 13-cis-RA, administered in a dose identical to that prescribed to human patients, could alter depression- or anxiety-related behaviors in young adult mice. Here, we report for the first time that chronic administration of 13-cis-RA to mice significantly increases the time spent immobile in both the forced swim and the tail suspension tests. Such increases in immobility in these behavioral despair paradigms are consistent with increased depression-related behavior.

METHODS

Animals

Young, adult male DBA/2J (Jackson Laboratories, Bar Harbor, ME) mice were used in this study. Animals were 3 weeks old at arrival and 4 weeks old at the start of treatment. Mice were group housed four per cage and maintained under a 12 : 12 h light/dark cycle. Food and water were provided ad libitum. All procedures and tests performed on animals were approved by the University of Texas IACUC, protocol number 04100403, and according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Treatment

Animals were allowed to adapt to the University of Texas animal facility for 1 week prior to initial injection. Animals were handled during this week to reduce the stress of receiving daily injections. A 2 mg/ml stock solution of 13-cis-RA in DMSO was stored for up to 1 week at −20°C. The stability of 13-cis-RA after 1 week of storage was confirmed by comparing the absorbance at 354 nm to that of a freshly prepared sample (Crandall et al, 2004). Sterile injection samples were prepared in the dark. Injections were performed 1 h prior to the start of the dark cycle in the same room where the animals were housed and one person performed all injections. Treated animals (n=12) received daily intraperitoneal (i.p.) injections of 1 mg/kg/day 13-cis-RA (Sigma, St Louis, MO), dissolved in sterile 50% DMSO/50% saline (0.9%) to a final volume of 200 μl. Control animals (n=12) were administered 200 μl of vehicle (50% DMSO/50% saline), i.p., daily. Animals were injected for 6 weeks prior to behavioral testing. Daily injections continued throughout the testing period and all animals received injections 14 h before any behavioral testing to eliminate any acute effects of injection. Only one behavioral test was performed each day and approximately 24 h elapsed between tests. The behavioral experiments were conducted in the following order: tail suspension test, forced swim test, open field test, and rotarod test.

13-Cis-RA Plasma Level Determination

After behavioral testing, blood samples were collected from treated animals immediately post mortem. 13-Cis-RA was administered to the mice approximately 16 h before killing. To prevent retinoid degradation, all steps were performed in the dark. The blood was allowed to coagulate for 30 min and then centrifuged at 1000g for 20 min. The plasma was then removed and stored at −80°C until the retinoid extraction was performed. Samples were extracted as described (Lane et al, 1999). In total, 0.7 volumes of acetonitrile : butanol (1 : 1) solution was added to the sample and vortexed for 45 s, then 0.6 volumes of saturated K2HPO4 was added and vortexed for 10 s. The samples were then centrifuged at 13 000g for 10 min. The upper layer was removed and the sample volume brought up to 100 μl with 100% acetonitrile. The absorbance of the samples was determined at 354 nm and 13-cis-RA concentration was determined by comparison to a standard curve. For the standard curve, a known amount of 13-cis-RA was added to plasma from mice not treated with Accutane and extracted as described above. All concentration values were corrected for initial and extracted sample volumes.

Tail Suspension Test

To perform the tail suspension test, animals were suspended by the tail on a horizontal beam 33 cm high. Mice were taped to the bar using adhesive tape placed approximately 1 cm from the tip of the tail. A 6-min test period was used and videotaped for subsequent analysis The animals engaged in several escape-related behaviors interspersed with periods of immobility. The duration of immobility was measured in seconds. The mice were tested in a randomized, double-blinded manner.

Porsolt Forced Swim Test

The forced swim test is a pharmacologically validated model of depression-related behavior in mice (Lucki et al, 2001). Mice were placed in a Plexiglas, cylindrical tank, 20 × 51 cm, filled to a depth of 35 cm with room temperature water (25±0.5°C) for a total of 6 min and the test was videotaped for subsequent analysis. Mice were scored for each of three behaviors: swimming, climbing, and immobility. Immobility was defined as the absence of escape-related behavior, such as swimming, climbing, or rearing. Swimming behavior was assigned when the animal was moving around the container with all the paws, while climbing behavior was assigned when the animal became vertical and clawed at the cylinder walls with its forepaws. The duration of each of these behaviors was measured in seconds over a 6-min period. The mice were tested in a randomized, double-blinded manner.

Open Field Test

The open field test was used to evaluate anxiety-related behaviors (Mathis et al, 1994, Prut and Belzung, 2003). Animals were placed in a 16 × 16 cm open field apparatus (Med Associates, St Albans, VT). Ambulatory and stereotypic activity was measured by infrared motion detectors spaced 2.5 cm apart. Rearing and jumping activity was measured by another set of infrared motion detectors 4.6 cm above the floor. The animal activity was monitored by Activity Monitor, version 5.10 (Med Associates, St Albans, VT) over the course of 5 min. The ambulatory time was determined in seconds for each minute spent in the open field as were the total ambulatory, vertical, and stereotypic times. The time spent and distance traveled in the center (38% of the total field area) and the number of entries into the center of the field were also calculated. Motor function was also assessed by considering the total distance traveled in the open field during the 5-min observation period.

Rotarod Test

In order to determine the effect of retinoids on motor coordination, the rotarod test was performed. The rotarod test consisted of placing the animal on a rotating rod, 4 cm in diameter, 35 cm high (Columbia Instruments, Columbus, OH). Low rotation speeds were chosen for this test, since DBA/2J mice perform this task relatively poorly at higher speeds (Brooks et al, 2004). The animals were first placed on a rod rotating at 5 rpm for 2 min to acclimate to this novel motor task. Each time an animal fell from the rod during this time period, the animal was placed back onto the rod. After the 2-min acclimation period, the animal was allowed to rest for 1 min and then placed back on the rod rotating at 5 rpm. The latency to fall was then measured in seconds. The animal was then placed on the rod for a second trial at a rotation speed of 10 rpm. The latency to fall was again measured in seconds. In all trials, if the mouse did not fall from the rod, it was removed from the rod after 2 min.

Behavioral Evaluation and Statistics

All behavioral tests were double-blinded and all were scored prior to decoding any of the tests. Immobility data from the last 4 min of both the tail suspension and forced swim tests were analyzed with a 2 × 2 × 4 (Group × Test × Minute) repeated measures ANOVA, with the two tests and 4 min serving as repeated measures. We analyzed the two immobility tests in this way because they both examine the same construct (depressive behavior) using the same units of measurement (seconds of immobility). This omnibus approach has been successfully applied to other behavioral measures meeting these criteria (Shumake et al, 2005). Only the last 4 min of each test were used in the analysis because of widespread zero values during the first 2 min of each test. The open-field and rotorod data were analyzed by two-tailed t-tests. Values reported are mean±SEM of n=12 observations unless otherwise indicated.

RESULTS

Confirmation of 13-Cis-RA Plasma Levels in Treated Mice

The human dose of 13-cis-RA used to treat acne ranges from 0.5 to 2 mg/kg/day. The mice used in this study received an amount of 13-cis-RA within this range and identical to the amount administered to mice in the study by Crandall et al (2004). The plasma level of 13-cis-RA in six mice chronically treated with 13-cis-RA was 1.5±0.4 μg/ml. This value is higher than that previously observed by Crandall et al (2004), potentially due to the use of different retinoid extraction techniques. However, the plasma levels obtained in the present study are approximately twice that displayed by humans administered 0.5 mg/kg/day of 13-cis-RA, one-half the dose used in the present study. Plasma 13-cis-RA concentrations in these patients reached 0.74 μg/ml (Kerr et al, 1982). In addition, the plasma concentration obtained in the present study is similar to that reported in the Accutane capsules package insert were the peak plasma concentration of 13-cis-RA in healthy patients aged 12–15 years receiving multiple oral doses of 13-cis-RA for the treatment of severe acne ranged from 0.37 to 1.09 μg/ml. Therefore, administration of 1 mg/kg/day of 13-cis-RA to young, adult male mice results in plasma 13-cis-RA levels similar to those observed in human patients.

13-Cis-RA Treatment had no Effect on Weight Gain

Weight gain over the 6 weeks of injections was followed to ensure that the animals were not showing overt signs of distress or discomfort. Weekly weighing of the mice showed that overall weight gain across the experiment was normal and that there was no difference in body weight between the 13-cis-RA-treated group and the control group. At the start of injections, the 13-cis-RA-treated group weighed 15.5±0.4 g and the control group weighed 15.1±0.3 g. At the beginning of the behavioral testing, the treatment group weighed 23.8±0.5 g and the control group weighed 23.1±0.5 g. The behavioral tests had no effect on the weight of the animals. 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. The injections were verified to be i.p. and not subcutaneous by feeling the skin on the underside of the animals daily for bubbles or scar tissue.

Chronic Administration of 13-Cis-RA Increases Immobility in the Tail Suspension and Forced Swim Tests

The effect of chronic administration of 13-cis-RA on young adult mouse behavior in the tail suspension (a and b) and forced swim (c and d) tests is shown in Figure 1. Both tests are pharmacologically validated models of behavioral despair in which the time spent immobile decreases with antidepressant treatment (Cryan et al, 2005; Lucki, 1997). The 2 × 2 × 4 (Group × Test × Minute) repeated measures ANOVA of the immobility data showed no significant interactions and no significant effect of test, supporting the validity of analyzing the two tests in this way. There was a significant main effect of group, F(1,21)=4.8, p=0.039, with 13-cis-RA-treated mice showing significantly more immobility than vehicle controls. The estimated marginal means and standard errors for the two groups were 19±2.1 s of immobility per minute for the 13-cis-RA-treated group and 12±2.0 s of immobility per minute for the vehicle controls. The increase in immobility due to 13-cis-RA treatment observed in the forced swim test appears to be due to a decrease in swimming behavior (Figure 1c). There was no difference between the groups for the amount of time the animals spent engaged in climbing activity. These data show that chronic 13-cis-RA administration increases immobility in both the tail suspension and forced swim tests, which is indicative of increased depression-related behavior.

Figure 1
figure 1

13-Cis-RA increases immobility in the tail suspension and forced swim tests. Total time spent immobile in the tail suspension (a) and forced swim (c) tests are expressed as the mean±SEM for the last 4 min of the observation period. The development of immobility during the tail suspension (b) and forced swim (d) tests are shown as mean±SEM seconds per each minute of the 6-min observation period. Overall, immobility time from both tests (as assessed by repeated measures ANOVA) showed a significant main effect of group, F(1,21)=4.8, p=0.039, with 13-cis-RA-treated mice showing significantly more immobility than vehicle controls. n=11, 13-cis-RA-treated mice and n=12, control mice.

Full size image

Chronic Administration of 13-Cis-RA has no Effect on Anxiety-Related and Motor Behavior in the Open Field Test

Approximately 25–50% of depressed youth also exhibit anxiety. Selective serotonin reuptake inhibitors have been used successfully to treat both depression and anxiety, potentially indicating a common etiology (for a review, see Axelson and Birmaher, 2001). Also, in mice, anxiety may be one factor that contributes to susceptibility to depression-related behaviors (Ducottet and Belzung, 2005). The open field test is an anxiety-related behavioral paradigm that is sensitive to anxiolytic drugs (Mathis et al, 1994; Prut and Belzung, 2003). An animal exhibiting anxiety-related behavior will spend less time in the center of the open field. There was no difference between the 13-cis-RA-treated group and the control group in the number of entries into the inner zone of the field between the two groups (p=0.61) (Figure 2a). 13-Cis-RA-treated mice made 21±4 entries into the inner field and control animals made 24±2 entries.

Figure 2
figure 2

13-Cis-RA does not affect performance in the open field test. The number of entries into the center (a), the total ambulatory time (b), and the total ambulatory distance (c) measured during a 5-min observation period in the open field test. In each case, there was no difference between 13-cis-RA-treated mice and vehicle-treated controls. Values are mean±SEM with n=12 for each group.

Full size image

Retinoid signaling affects locomotion. For example, RARβ-RXRβ, RARβ-RXRγ, and RXRβ-RXRγ double null mutant mice (but not the corresponding single mutants) exhibit reduced forward locomotion and rearing frequency in the open field test (Krezel et al, 1998). Also, retinoic acid-mediated effects have been implicated in Parkinson's (Eichele, 1997) and Huntington's disease (Luthi-Carter et al., 2000) (for a review, see Lane and Bailey, 2005). Therefore, to assure that the effects of 13-cis-RA on behavior in the tail suspension and forced swim tests were due to a specific increase in despair-related behavior and not to an overall decrease in locomotion or coordination, we conducted the open field and rotarod tests, respectively. There was no difference in total ambulatory time (p=0.68) or the ambulatory time in the inner zone of the field (p=0.68) between the two groups (Figure 2b). The total ambulatory time of mice treated with 13-cis-RA was 55±3 s vs 57±2 s for the control mice. The ambulatory time for the 13-cis-RA-treated mice in the inner zone of the field was 7±1 s, while control mice had 8±1 s of ambulatory time in the inner zone of the open field. Additionally, there was no difference in the ambulatory distance traveled over the entire field (p=0.78) or the inner zone of the field (p=0.93) (Figure 2c). 13-Cis-RA-treated mice traveled 1426±98 cm while control mice traveled 1451±57 cm over the entire field. 13-Cis-RA-treated mice traveled 211±30 cm in the inner zone and the control mice traveled 216±19 cm. As 13-cis-RA treatment did not alter the number of entries into, or the total distance traveled in, the center zone of the open field, it is unlikely that 13-cis-RA has an effect on anxiety-related behavior. The lack of a difference between groups with respect to total distance traveled or total ambulatory time indicates that 13-cis-RA has no effect on overall locomotion compared with vehicle-treated controls.

Chronic 13-Cis-RA Administration Increases Motor Coordination on the Rotarod

There was no impairment of motor coordination in mice treated with 13-cis-RA (Figure 3). There was a trend for mice treated with 13-cis-RA to remain on the rotating rod longer than control mice when tested at a rotational speed of 5 rpm (p=0.07). Mice administered 13-cis-RA were able to stay on the rod for 105±10 s, whereas control mice were only able to remain on the rod for 75±13 s. However, when tested at 10 rpm, there was no difference between the latency to fall time for the treated and control mice (p=0.91). Retinoids have been implicated in the regulation of locomotion because RARβ-RXRβ, RARβ-RXRγ, and RXRβ-RXRγ double null mutant mice were shown to have locomotor deficits (Krezel et al, 1998). In contrast, chronic 13-cis-RA treatment tends to improve motor coordination at low rotational speeds in the rotarod test without an effect on total locomotion (Figure 2).

Figure 3
figure 3

13-Cis-RA does not impair motor performance in the rotarod test. The mean latency to fall from a rod rotating at 5 or at 10 rpm is shown during a 2-min test period. The 13-cis-RA-treated mice tended to remain on the rotarod longer at 5 rpm. Treatment had no effect on rotarod performance at 10 rpm. Values are mean±SEM with n=12 for each group.

Full size image

DISCUSSION

Here we demonstrate that chronic administration of 13-cis-RA to young adult male mice increased the time spent immobile in both the tail suspension and forced swim tests. In the open field test, there was no evidence of a change in anxiety-related behavior. Furthermore, there was no impairment of locomotion or motor coordination in 13-cis-RA-treated animals. Taken together, these data indicate that the increased immobility in behavioral despair paradigms is not due to an effect on motor systems, but results from increased depression-related behavior. Increased immobility in the forced swim test has been exhibited by mice and rats in situations that increase vulnerability to depression such as estradiol deficiency due to aromatase knockout (Dalla et al, 2004), stress due to food restriction (Alcaro et al, 2002), magnesium depletion (Singewald et al, 2004), and choline administration (Carlezon et al, 2002). The behavior of the animals in the tail suspension test was not assessed in these studies. However, increased immobility in both the forced swim and tail suspension tests does occur in mice and rats due to the pro-depressant effects of withdrawal from chronic amphetamine administration (Cryan et al, 2003), centrally administered urotensin-II (Do-Rego et al, 2005), and interleukin-1 and endotoxin injection (Dunn and Swiergiel, 2005). Interestingly, in the forced swim test, antidepressants that target the serotonergic system increase swimming behavior, thereby decreasing immobility (Detke et al, 1995). On the other hand, antidepressants that target noradrenergic systems increase climbing behavior. In our experiments, 13-cis-RA treatment leads to an increased immobility with a concomitant decrease in swimming behavior, without an effect on climbing behavior. This may indicate that chronic 13-cis-RA administration disrupts serotonergic systems, leading to increased depression-related behavior.

A recent study reported the effects of chronic 13-cis-RA administration on adult rats. In the forced swim test, Ferguson et al (2005) reported a marginally significant effect of 13-cis-RA treatment on immobility. However, in contrast to our observations, these authors observed less immobility following 13-cis-RA administration. This reduced immobility was observed for a successive 2-day test in which the immobility for the second day was found to be decreased in all test groups. This analysis was performed by averaging the data across the two separate test days. The fact that immobility was decreased on the second test day is surprising because the forced swim test, when conducted on consecutive days, generally results in increased immobility each day regardless of treatment (Bhatnagar et al, 2004). In addition, Ferguson et al (2005) reported that there was no change in voluntary saccharin solution intake in treated rats, reflecting an absence of the depressive symptom anhedonia. The authors concluded that 13-cis-RA treatment did not severely affect depression-like behaviors in rats. There are three important differences between our study and that of Ferguson et al (2005): (1) age and species of test subjects, (2) method of dosing, and (3) dose of drug administered. The adolescent mice were used in our study because Accutane is administered primarily to human adolescents. Our mice are at an earlier stage of development than the adult rats used by Ferguson et al (2005); we began dosing between the time of cessation of weaning and the onset of sexual maturity. The dose of 13-cis-RA we administered was 1 mg/kg via i.p. injection in comparison to the much higher doses (7.5 and 22.5 mg/kg) used by Ferguson et al (2005) administered by oral gavage. The dose we used is within the recommended range for treatment of patients (0.5–2.0 mg/kg/day) and we achieved serum levels in treated mice that were equivalent to those seen in patients. Repeated drug administration is a potential stressor and stress has been shown to influence depression-related behavior (Shumake and Gonzalez-Lima, 2003). Ferguson et al (2005) also examined behavior in the open field, and in one treatment group saw that 13-cis-RA-treated rats were significantly less active than same sex controls. This increase in freezing time in the open field increased with the duration of treatment even in control animals. One explanation for this is that all animals were exhibiting enhanced stress-responses that can manifest as increased freezing time in the open field test. It is therefore possible that Ferguson et al (2005) were not able to detect changes in depression-related behaviors in 13-cis-RA-treated animals because they were masked by elevated stress in all treatment groups.

Changes in adult behavior as a result of 13-cis-RA administration are perhaps not surprising given the evidence linking retinoid signaling to adult CNS function (Mey and McCaffery, 2004; Lane and Bailey, 2005). A large number of neuronal genes have been shown to be retinoid-responsive (for a review see Lane and Bailey, 2005), but it is not yet known whether 13-cis-RA can regulate gene transcription in the adult brain. 13-Cis-RA could be activating gene transcription by either isomerizing to all-trans-retinoic acid and binding to the RAR (Tsukada et al, 2000) or directly binding to the RAR itself (Idres et al, 2002). Although 13-cis-RA binds to the RAR with much lower affinity than all-trans-retinoic acid, once bound to the RAR, 13-cis-RA is very efficient at initiating transcription (Idres et al, 2002). Given that depression is a complex, multifactorial phenomenon involving dopaminergic, serotonergic, noradrenergic, and other transmitter pathways, altered regulation of a subset of target genes by 13-cis-RA could produce subtle changes in neuronal function that lead to enhanced depression-related behaviors. For example, a verified RARE has been found in the promoter region of the dopamine D2 receptor and all-trans-retinoic acid can upregulate dopamine D2 receptor expression (Samad et al, 1997). Increased dopamine D2 receptor abundance has been reported in patients with major depression (Klimek et al, 2002). Similarly, the 5-hydroxytryptamine 1A receptor, which has a role in regulating the firing of serotonin neurons and is associated with depression-related personality traits (Strobel et al, 2003), has been shown to be upregulated in response to all-trans-retinoic acid treatment in neuronal cells (Charest et al, 1993). Altered regulation of either the D2 receptor or the 5-HT 1A receptor genes by 13-cis-RA, as well as other neuronal genes, could provide a possible mechanism for the changes in depression-related behaviors seen in chronically treated mice.

An alternative mechanism by which 13-cis-RA could alter depression-related behaviors is by affecting adult neurogenesis. Apart from the well-documented effects of retinoids on neuronal differentiation during development (Maden, 2002), retinoid signaling has been shown to promote adult neurogenesis in the hippocampus and olfactory bulb (Takahashi et al, 1999; Thompson Haskell et al, 2002). In contrast, long-term administration of 13-cis-RA suppresses hippocampal neurogenesis and decreases hippocampal cell survival (Crandall et al, 2004). Reductions in hippocampal neurogenesis have been correlated with depression in human patients (Sheline et al, 1996; Bremner et al, 2000). In animal models, inescapable stress also reduces hippocampal neurogenesis, a phenomena that can be reversed with antidepressant treatment (Malberg and Duman, 2003). The ability of 13-cis-RA to increase depression-related behavior in the present study may be due to decreased hippocampal neurogenesis and cell survival and likely involves altered regulation of neuronal gene expression. Interestingly, adult neurogenesis requires functional serotonergic signaling (Brezun and Daszuta, 1999; Santarelli et al, 2003), further suggesting that 13-cis-RA mediates its effects by specifically regulating serotonergic gene expression and/or hippocampal neurogenesis.

13-Cis-RA is an effective treatment for severe recalcitrant acne (Shalita et al, 1988). However, 13-cis-RA use has been linked with depression, although this remains controversial (Jacobs et al, 2001; O'Connell et al, 2003; O'Donnell, 2003; Chai et al, 2005). Between 1982 and 2000, the US Food and Drug Administration received 431 reports of depression, suicidal ideation, suicidal attempts, and suicides in patients taking 13-cis-RA (Ng and Schweitzer, 2003). In addition, the Food and Drug Administration has received at least 24 case reports in which depression was resolved after 13-cis-RA use was discontinued, but recurred after drug rechallenge (Ault, 1998). Diagnosing depressive symptoms in acne patients following 13-cis-RA treatment is complicated by the psychological effects of severe acne itself. Clinically significant anxiety and depression symptoms have been reported in both adolescents and adults with chronic acne (Koo, 1995; Kellett and Gawkrodger, 1999). Additionally, adolescents under social stress may be on the border of clinical depression and 13-cis-RA treatment may be a precipitating event (Gerard and Buehler, 2004). Despite the possible confounding effects of existing conditions, there are some instances where patients with no previous history of psychiatric symptoms develop depression and suicidal ideation following 13-cis-RA treatment (O'Donnell, 2003). Use of a mouse model allowed us to examine the effects of 13-cis-RA on depression-related behaviors exclusive of pre-existing conditions, self-image, and other factors that confound human studies. This report is the first to demonstrate that 13-cis-RA administration enhances depression-related behaviors in mice. The mechanism by which this occurs is yet to be elucidated, although altered neuronal gene regulation and changes in hippocampal neurogenesis are likely to be involved.