Influence of Sex and Corticotropin-Releasing Factor Pathways as Determinants in Serotonin Sensitivity

Abstract

Stress sensitivity and sex are predictive factors in affective disorder susceptibility. Serotonin (5-HT) pathway recruitment by corticotropin-releasing factor (CRF) during stress is necessary in adaptive coping behaviors, but sex differences in such responses have not been investigated. Using selective 5-HT reuptake inhibitor (SSRI) administration to acutely elevate 5-HT in a genetic model of stress sensitivity, we examined behavioral and physiological responses in male and female stress-sensitive CRF receptor-2-deficient (R2KO) mice. Chronic SSRI treatment was used to confirm outcomes were specific to acute 5-HT elevation and not antidepressant efficacy. We hypothesized that R2KO mice would show a greater sensitivity to acute changes in 5-HT and that, because females typically are more stress sensitive, R2KO females would be the most responsive. Our results supported this hypothesis because females of both genotypes and R2KO males showed a greater sensitivity to an acute 10 mg/kg dose of citalopram in a tail suspension test, displaying decreased immobile time and increased latency to immobility. Furthermore, acute citalopram promoted significant anxiogenic-like effects that were specific to R2KO females in the elevated plus maze and light-dark box tests. Chronic citalopram did not produce these behavioral changes, supporting specificity to acute 5-HT modulation. Mechanistically, females had decreased hippocampal 5-HT transporter (SERT) levels, whereas R2KO mice showed reduced SERT in the prefrontal cortex, supporting a possible intersection of sex and genotype where R2KO females would have the lowest SERT to be blocked by the SSRI. This sensitivity to 5-HT-mediated anxiety in females may underlie a heightened vulnerability to stress-related affective disorders.

An increased sensitivity to acute selective serotonin re-uptake inhibitor administration may be a predictive marker of an underlying vulnerability to disease.

Affective disorders such as depression and anxiety comprise a complex pattern of neurocircuitry dysregulation (1,2,3). Females present with depression at twice the frequency of males (3,4,5,6,7), suggesting an important sex-dependant susceptibility to disease risk factors, including exposure to stress. Although a link between stress sensitivity and disease susceptibility has been observed in clinical settings, the full implications of sex differences have not yet been examined. Furthermore, use of intact, healthy organisms in research has not provided the necessary assessment of the maladaptive stress responses that likely contribute to disease onset.

The recruitment of the serotonin (5-HT) system provides a mechanism by which stress alters neurotransmitter signaling critical in mood regulation and in the prevention of affective disorders (8,9,10,11,12,13). One mediator in the interaction between stress and 5-HT is the neuropeptide corticotropin-releasing factor (CRF), which is induced by stress exposure and acts on its two receptors, CRF receptor-1 (CRFR1) and -2 (CRFR2) (14,15,16,17). In the raphe, stress exposure results in increased gene expression related to 5-HT production (18) and alterations in excitatory and inhibitory drives on 5-HT producing raphe neurons (19,20,21,22,23,24,25). Additionally, sex hormones have been implicated as modulators of 5-HT transmission in raphe projection regions (26,27). Despite the known complex modulation of 5-HT by stress, sex differences in these midbrain dorsal and median raphe nuclei producing the main ascending 5-HT projections have not been determined.

To examine the contribution of stress pathway dysregulation and sex on behavioral responses to acute increases in 5-HT, we have used a mouse model of stress sensitivity deficient for CRFR2 [CRFR2 knockout (R2KO)]. R2KO mice display maladaptive physiological and behavioral responses to stress perturbations (28,29) and fail to produce adaptive and homeostatic responses in the raphe after stress (30). We hypothesize that because CRF modulates 5-HT pathways, mice with a genetic disruption in CRF signaling will show a greater sensitivity to changes in local 5-HT in behavioral and physiological outcomes after acute citalopram treatment. Chronic citalopram treatment was assessed as a control to ascertain the specificity of these effects to acute elevations in 5-HT rather than antidepressant efficacy. Because intrinsic differences in the 5-HT system may contribute to altered responses to citalopram, sex and genotypic comparisons for 5-HT transporter (SERT) levels were conducted in brain regions where stress-mediated signaling interact and modulate 5-HT pathways, including the dorsal and ventral hippocampus, prefrontal cortex, and amygdala.

Materials and Methods

Animals

All mice used were generated in-house from CRFR2 heterozygous breeding on a mixed C57BL/6:129J background as previously described (28). Male and female wild-type (WT) and CRFR2-deficient (R2KO) littermates (9–12 wk, genotyped at 3 wk) were group housed under a 12-h light, 12-h dark cycle (lights on at 0700 h), with food and water ad libitum. All studies were in accordance with experimental protocols approved by the University of Pennsylvania Institutional Animal Care and Use Committee, and all procedures were conducted in accordance with institutional guidelines.

Citalopram administration

Acute

To examine the physiological and behavioral effects of acute increases in 5-HT, WT and R2KO male (n = 13–14) and female (n = 18) mice were treated with 10 mg/kg of the selective 5-HT reuptake inhibitor (SSRI) citalopram (Anawa, Zurich, Switzerland) dissolved in saline and administered via ip injection (100 μl) or saline vehicle control (100 μl). Citalopram was administered at 10 mg/kg as a moderate dose shown to produce behavioral responses in mice (31,32,33). Behavioral measures were examined 30 min after administration of citalopram or vehicle (34).

Chronic

To distinguish between antidepressant efficacy and effects of acute elevations in 5-HT in these studies, a separate cohort of WT and R2KO male (n = 17–18) and female (n = 19–24) mice were exposed to chronic citalopram dissolved in their drinking water for 4 wk as previously described (35). For controls in chronic administration groups, male and female WT and R2KO littermates were given water without drug. Body weights were taken at weekly intervals, and water consumption of mice was measured and water volumes refreshed with new drug solutions at 3-d intervals (see supplemental Table 1). Citalopram levels in drinking water were adjusted based on recent consumption and weight data to ensure continued delivery of a physiologically relevant dosage of 10 mg/kg. In behavioral testing, responses were measured in mice after 4 wk citalopram (35). Mice continued to receive citalopram during behavioral testing periods.

Behavioral responses after citalopram administration

Behavioral stress responses were examined in male and female mice 30 min after acute citalopram administration. A separate cohort of mice was examined during chronic citalopram treatment. Testing was completed in a designated room within the mouse home environment. Behavioral tests were separated by 5–7 d and occurred in the following order: elevated plus maze, tail suspension test, and light-dark box. At the end of each test, female mice were vaginally lavaged with warm saline to attain cellular samples for determining estrous cycle stage. All studies were conducted by an investigator blinded to genotype and treatment groups.

Tail suspension test

Testing was performed as previously described (28). Briefly, the 6-min behavioral test was performed between 1100 and 1300 h. Immobility was hand scored by an experimenter blinded to experimental details. Mice were excluded from statistical analysis if they climbed their tails during the test. Data analysis was aided by the use of AnyMaze software (Stoelting, Wood Dale, IL) to measure time spent immobile and latency to first bout of immobility.

Elevated plus maze

Testing was conducted as previously described (36). Briefly, the 5-min test occurred during the light cycle between 1100 and 1300 h. Light intensity in open arms was 6 lux. Analysis used AnyMaze software to measure open and closed arm time and entries and total distance traveled. Mice were excluded from statistical analysis if they fell off the open arms of the maze.

Light-dark box

Testing was performed as previously described (28). Briefly, light intensity was 5 lux in the dark compartment and 300 lux in the light compartment. Test duration was 10 min and occurred 2 h into the dark cycle. AnyMaze software was used to measure transitions between light and dark sides of the test and time spent in the light side.

Estrous cycle

To control for peak estrogen levels during the time of testing, estrous cycle stages were examined by vaginally lavaging each female with warm saline immediately after completion of each test. The smear was air dried on a slide and hematoxylin stained. Estrous cycle stages were grouped according to hormone levels into proestrus, estrus, and diestrus categories.

Hypothalamic-pituitary-adrenal (HPA) axis assessment

To examine sex and genotypic effects of acute citalopram administration on the physiological HPA axis stress response, corticosterone levels were examined after a 15-min restraint stress in a separate cohort of male and female mice. Testing was administered between 0900–1100 h by placing mice in a 50-ml conical tube. In acute treatment groups, restraint onset began 30 min after administration of citalopram to male (n = 7–11) and female (n = 7–10) WT and R2KO mice. Tail blood samples (∼15 μl per sample) were taken at onset and completion of restraint (0 and 15 min, respectively). For female mice, vaginal lavage saline samples were also collected immediately after the last blood sample collection for estrous cycle determination. To examine the HPA axis stress recovery phase, additional samples were collected at 15 min and 75 min after restraint end. Chronic citalopram male (n = 9–10) and female (n = 8–12) WT and R2KO mice were exposed to the same restraint and blood collection procedure. Corticosterone was determined by RIA (MP Biomedicals, Orangeburg, NY) using 3 μl plasma with a variance coefficient R² = 0.9985.

SERT autoradiography

To determine whether sex differences in response to acute citalopram were related to transporter levels, SERT binding was analyzed in male and female mice using [³H]citalopram autoradiography. Fresh-frozen brain sections (30 μm) were selected from experimentally naive male and female WT and R2KO mice (n = 6–8). Sections were atlas matched (37) and immersed in incubation buffer (4 C, 10 min) containing 50 mm Tris (pH 7.6), 120 mm NaCl, and 5 mm KCl. Sections were then placed in incubation buffer containing [³H]citalopram at room temperature (0.7 nm; Perkin-Elmer, Boston, MA) for 1 h. Slides were washed in incubation buffer and rapidly dipped in cold deionized H₂O, dried under a cold airstream, and apposed to Kodak Hyperfilm (Eastman Kodak, Rochester, NY) for 21 d. Images were normalized to background and analyzed by a predetermined region of interest tool in the basolateral amygdala, prefrontal cortex, and ventral and dorsal hippocampus. OD measurements of matched slides were conducted using IPlabs software (BD Biosciences/Scanalytics, Rockville, MD) and averaged across two consecutive sections for each animal. Analyses were conducted by an investigator blinded to sex and genotype.

Statistical analyses

Examination of sex differences in citalopram effects on behavioral tests and physiological measurements used multifactor ANOVA with factors for sex, genotype, and drug. Additional analysis of main effects by genotype was accomplished using ANOVA with factors for sex and drug treatment separated by WT and R2KO. For further distinction of differential effects of citalopram in males and females, data were analyzed by multifactor ANOVA, separated by sex, with factors for genotype and drug treatment. HPA measurements were analyzed using repeated-measures ANOVA over time with the same factors as above. Assessing estrous cycle effects used an ANOVA model for cycle effects in a behavioral measurement from each test (immobility time, open arm time, light-dark transitions). All acute and chronic groups were analyzed separately, and an additional test was run to assess acute-chronic differences where treatment duration was added as a factor to the tests described above. Analysis was carried out with JMP statistical software (SAS, Cary, NC).

Results

Behavioral tests

Tail suspension test

Analysis of male and female vehicle-treated controls by ANOVA revealed that R2KO mice displayed a main effect for basal elevation in immobile time compared with WT controls [F_{(1, 62)} = 2.6; P < 0.05; Fig. 1A]. Post hoc testing within sexes showed that this effect was significant in males [F_{(1, 15)} = 2.3; P < 0.05]. In measuring latency to become immobile within vehicle-treated mice, a main effect of genotype was again observed [F_{(1, 62)} = 2.6; P < 0.05; Fig. 1B] because R2KO mice displayed a decreased latency to become immobile compared with WT.

Figure 1.

Enhanced sensitivity in females and CRFR2-deficient mice after acute citalopram administration in the tail suspension test. Citalopram (CIT) treatment decreased immobile time (A) in male R2KO and female mice relative to vehicle-treated mice (VEH) and increased latency (B) to first bout of immobility in female R2KO. A main effect of drug was detected to decrease immobility (P < 0.01) and increase latency (P < 0.01). Data are presented as mean ± sem. *, Significant genotype effect, P < 0.05; #, significant citalopram effect, P < 0.05; ##, P < 0.01.

Acute citalopram administration resulted in main effects of drug for decreased immobility time [F_{(1, 60)} = 5.3; P < 0.01; Fig. 1A] and increased latency to first become immobile [F_{(1, 60)} = 3.8; P < 0.01; Fig. 1B]. Acute citalopram reduced time spent immobile for male and female R2KO mice [F_{(1, 16)} = 3.7 and 3.1, respectively; P < 0.01] as well as in WT females [F_{(1, 11)} = 2.5; P < 0.05] but not in WT male mice [F_{(1, 15)} = 1.3; P = 0.20].

Chronic citalopram did not alter time spent immobile [F_{(1, 66)} = 0.11; P = 0.91] or latency to become immobile [F_{(1, 66)} = 0.12; P = 0.91; see Table 1] in male or female WT or R2KO mice.

Table 1.

Behavioral effects of chronic citalopram administration

	Tail suspension test		Elevated plus maze				Light-dark box
	Time immobile (sec)	Immobility latency (sec)	Open arm time (sec)	Open arm entries (no.)	Total entries (no.)	Total distance (m)	Transitions (no.)	Light side time (sec)
Male
WT male VEH	95.8 ± 12.4	62.9 ± 13.9	33.6 ± 10.9	5.2 ± 1.3	17.8 ± 1.4	7.4 ± 0.5	18.1 ± 3.2	116.2 ± 21.8
WT male CIT	91.1 ± 9.9	62.8 ± 8.3	42.5 ± 9.3	4.4 ± 1.3	20.8 ± 2.3	6.0 ± 1.2	13.4 ± 2.7	231.0 ± 36.0
R2KO male VEH	132.8 ± 17.2	47.5 ± 5.4	31.2 ± 8.56	5.2 ± 1.2	19.2 ± 2.4	6.2 ± 0.6	14.2 ± 1.6	109.9 ± 18.3
R2KO male CIT	112.4 ± 12.4	58.9 ± 9.0	42.4 ± 11.4	4.5 ± 0.7	23.4 ± 3.1	5.9 ± 0.7	11.8 ± 1.2	109.4 ± 13.4
Female
WT female VEH	118.8 ± 20.8	61.4 ± 9.8	46.0 ± 12.8	5.1 ± 1.1	16.3 ± 1.7	5.6 ± 0.8	8.9 ± 2.8	152.5 ± 60.7
WT female CIT	144.7 ± 11.1	36.4 ± 4.8	38.9 ± 12.0	6.3 ± 1.7	12.9 ± 1.3	6.2 ± 0.9	11.1 ± 1.9	107.5 ± 22.0
R2KO female VEH	136.6 ± 9.4	42.4 ± 3.8	13.1 ± 6.2	2.7 ± 1.0	17.8 ± 1.6	4.7 ± 0.6	8.2 ± 2.4	77.8 ± 33.3
R2KO female CIT	135.4 ± 11.4	50.7 ± 6.0	41.0 ± 17.2	4.0 ± 1.0	13.0 ± 2.2	5.3 ± 0.6	6.9 ± 1.4	145.5 ± 46.4

CIT, Citalopram; VEH, vehicle. 

Elevated plus maze

In the elevated plus maze, analysis by ANOVA revealed a main effect of drug for acute citalopram to decrease open arm time [F_{(1, 62)} = 7.9; P < 0.01; Fig. 2A]. Testing within sexes revealed this effect to be specific to female mice [F_{(1, 23)} = 10.7; P < 0.01]. Male WT and R2KO groups displayed no change in open arm time after citalopram injection [F_{(1, 20)} = 0.9; P = 0.36]. Further analysis showed a main effect of sex where female mice displayed fewer open arm entries compared with males [F_{(1, 62)} = 6.5; P < 0.01; Fig. 2B], although this measure was not affected by citalopram. Females displayed a citalopram-mediated decrease in total number of entries in the maze [F_{(1, 23)} = 5.7; P < 0.05; Fig. 2C] and elevated distance traveled in the maze when compared with males [F_{(1, 62)} = 9.2; P < 0.01; Fig. 2D].

Figure 2.

Female mice display increased anxiety-like behavior after acute citalopram in the elevated plus maze. A, Administration of citalopram (CIT) produced decreased time in open arms relative to vehicle-treated controls (VEH) in WT and R2KO females but not in male mice. B and C, Females displayed a main effect for reduction in open arm entries (B) compared with male mice (P < 0.01) and a citalopram-mediated decrease in total number of entries in the maze (C) (P < 0.05). D, A main effect of sex was also detected where female mice displayed increased total distance traveled in the maze compared with males (P < 0.01). Data are presented as mean ± sem. #, Significant citalopram effect, P < 0.05; *, significant main effect of sex, P < 0.05.

No effects of chronic citalopram administration were observed in male or female WT or R2KO mice (Table 1).

Light-dark box

Acute citalopram produced a sex-specific drug effect on light side time [drug × sex interaction, F_{(1, 68)} = 2.1; P < 0. 05], where drug administration reduced time spent in the light side for females but not for males (Fig. 3A). This effect was significant in R2KO female mice [F_{(1, 15)} = 2.7; P < 0.05]. Acute citalopram administration also caused a reduction in light-dark transitions across all groups [F_{(1, 68)} = 2.6; P < 0.01; Fig. 3B]. This effect was significant only in female R2KO mice [F_{(1, 15)} = 2.5; P < 0.05]. In vehicle-treated mice, there was a main effect of sex for females to show increased transitions between dark and light sides [F_{(1, 68)} = 2.2; P < 0.05] and decreased time spent in the light side of the apparatus [F_{(1, 68)} = 2.7; P < 0.01; Fig. 3A].

Figure 3.

Female R2KO mice display enhanced sensitivity to acute citalopram in the light-dark box test. In R2KO females, citalopram (CIT) produced a reduction in time spent in the light side (A) and in the number of light-dark transitions (B) relative to wild-type (WT) mice in the light-dark box after acute administration. No sex differences were observed in vehicle-treated (VEH) mice. Data are presented as mean ± sem. #, Significant citalopram effect, P < 0.05.

Chronic citalopram administration resulted in a drug by genotype interaction effect in male mice [F_{(1, 33)} = 2.1; P < 0. 05; see Table 1], where only WT male mice displayed increased light side time after citalopram.

Estrous cycle

Estrous cycle analysis after behavior and physiological tests revealed no significant differences among acute or chronic treatment groups (supplemental Table 2).

HPA stress axis

In groups administered citalopram acutely before restraint stress, a main effect was observed where the HPA response was significantly elevated in both male and female mice [F_{(1, 81)} = 25.1; P < 0.0001; Fig. 4, A and B, respectively]. Vehicle-treated mice displayed a significant sex effect where female mice displayed an elevated corticosterone response across multiple testing time points compared with males [F_{(1, 81)} = 37.1; P < 0.0001]. In male mice, there was a trend for a genotype-specific effect of citalopram [drug × genotype interaction F_{(1, 27)} = 2.0; P = 0.06], where citalopram-treated R2KO mice displayed greater corticosterone levels at 30 min compared with WT males administered drug. In females, acute citalopram significantly increased basal (0 min) corticosterone levels in R2KO mice compared with WT mice [F_{(3, 59)} = 6.7; P = 0.0001]. Acute citalopram delayed HPA axis recovery 90 min after stress in females [F_{(1, 54)} = 2.5; P < 0.05] and males [F_{(1, 27)} = 3.3; P < 0.01]. Post hoc testing of the 90-min time point revealed that although all citalopram-treated groups showed apparent corticosterone elevation, this effect was significant in R2KO males and females [F_{(1, 11)} = 2.0–2.4; P < 0.05].

Figure 4.

HPA activation after acute citalopram is enhanced in female and R2KO mice. Administration of citalopram (CIT) produced enhanced corticosterone (CORT) levels in males (A) and females (B) relative to vehicle-treated mice (VEH). Female R2KO mice displayed a significantly enhanced citalopram effect compared with citalopram-treated WT females. Data are presented as mean values ± sem. *, Significant effect of genotype, P < 0.05; #, significant citalopram effect, P < 0.05.

Chronic administration of citalopram produced no significant effects on HPA axis stress response in male or female WT or R2KO mice (supplemental Table 3). Although an apparent reduction in corticosterone was seen among female mice at the 0-min time point, this did not reach significance [F_{(3, 40)} = 3.0; P = 0.09].

SERT autoradiography

To examine whether sex and genotypic differences in 5-HT reuptake in limbic and frontal areas might contribute to observed behavioral responses to acute citalopram, SERT levels were compared between male and female WT and R2KO mice in the amygdala, hippocampus, and prefrontal cortex. In the hippocampus, an interaction effect between sex and genotype was observed in dorsal dentate gyrus [DG; F_{(1, 24)} = 2.2; P < 0.05; Fig. 5A], where lower SERT levels relative to males were observed in female WT mice but not in R2KO females. A main effect of sex was also detected in the dorsal and ventral DG, where males displayed elevated SERT compared with females [F_{(1, 11)} = 3.8 and 2.1, respectively; P < 0.01; Fig. 5B]. A genotypic effect for reduced SERT was detected in the dorsal DG of R2KO males compared with WT male controls [F_{(1, 11)} = 2.1; P < 0.05]. An interaction between sex and genotype was also detected in dorsal area CA3 [F_{(1, 24)} = 2.6; P < 0.05; Fig. 5E], where female mice, but not males, displayed lower SERT levels. Males also displayed significantly elevated SERT compared with females in dorsal DG [F_{(1, 11)} = 3.9; P < 0.01], and a trend was detected where R2KO males showed reduced SERT compared with WT males in dorsal DG [F_{(1, 11)} = 2.0; P = 0.06]. No effects of sex or genotype were observed in the CA1 region of the dorsal or ventral hippocampus.

Figure 5.

Female and R2KO mice display reduced SERT levels in hippocampus and prefrontal cortex. SERT levels were examined by autoradiography in the DG (A and D), area CA1 (B and E), and area CA3 (C and F) in the dorsal and ventral hippocampus (dHpc and vHpc, respectively), the prefrontal cortex (PFC) (G), and basolateral amygdala (BLA) (H). Significant sex by genotype interaction effects were observed in the DG and CA3 areas in the dorsal hippocampus where WT females displayed decreased SERT compared with males. Significant effect of genotype was detected in the prefrontal cortex, where R2KO mice SERT was reduced compared with WT. Representative autoradiogram images of i) PFC, ii) dHPC, and iii) vHPC are displayed, outlining regions sampled for analysis, accompanied by cytoarchitectural schematics (adapted from Ref. 65). Data are presented as mean ± sem. *, Significant effect of genotype, P < 0.05; +, significant effect of sex, P < 0.05; ++, significant effect of sex, P < 0.01.

A significant main effect of genotype was observed in the prefrontal cortex [F_{(1, 23)} = 2.1; P < 0.05] with a decrease in SERT in R2KO mice (Fig. 5G). Additionally, a significant interaction between sex and genotype was observed in the basolateral amygdala [F_{(1, 23)} = 2.1; P < 0.05, see Fig. 5H], where females displayed decreased SERT in WT mice but not in R2KO groups.

Discussion

Examination of behavioral and physiological responses to acute SSRI exposure in a model of stress dysregulation may be a critical step toward understanding the neurobiology of affective disorders and improving therapeutic efficacy. The inability to effectively recruit and maintain the homeostatic changes in 5-HT pathways mediated by CRF may be a factor in precipitating stress-induced disease onset (1,29,38,39,40). Furthermore, sex differences in affective disorder presentation necessitate the inclusion of females in these studies. Male and female stress-sensitive R2KO were examined for behavioral and physiological responses to acute or chronic citalopram treatment. In these studies, multiple behavioral paradigms were used to explore the interaction of stress pathway dysregulation and 5-HT in active and passive behavioral coping responses (41,42,43,44,45). Although acute testing required repeated SSRI administration that complicates interpretation from single tests, consistent effects across tests indicated a more robust support of 5-HT sensitivity.

In the tail suspension test, the predicted genotypic differences were detected where R2KO mice showed both increased immobile time and reduced latency to immobility compared with WT littermates. Acute citalopram administration at a dose of 10 mg/kg resulted in significant reductions in immobility in females of both genotypes. Interestingly, reduced immobility in males was only detected in R2KO mice, which may suggest that males are less sensitive to behavioral effects of acute 5-HT manipulation compared with female mice, which display significant drug effects across genotypes. The observed increase in latency to immobility and decrease in immobile time support our hypothesis that a stress-sensitive phenotype correlates with increased responsivity to acute SSRI administration. We hypothesize that both basal sex differences and a dysregulation of upstream stress pathways likely produces compensatory changes in the 5-HT system that may be responsible for the observed increased sensitivity to acute SSRI administration (38,41). As predicted, chronic citalopram administration did not produce these behavioral effects, supporting their dependence on acute increases in 5-HT and not antidepressant efficacy (46,47).

In an elevated plus maze, females displayed increased sensitivity to the anxiogenic effects of acute citalopram as exhibited by significant reductions in open arm time and entries after treatment. Interestingly, these outcomes were specific to females because male mice showed no significant behavioral changes in this test after acute citalopram. Similar anxiogenic responses to acute citalopram were observed in the light-dark box where female R2KO mice showed decreased time in the light side of the box. Chronic citalopram failed to exhibit effects on these anxiety measures for either sex or genotype in the elevated plus maze. Acute antidepressant administration has previously been associated with anxiogenic behavioral effects (47,48,49,50,51), and a subclass of patients also report increased anxiety during initial SSRI treatment (52,53). Interestingly, chronic citalopram produced opposing results in the light-dark box, with anxiolytic-like effects that were restricted to WT males, hypothesized to be the least stress-sensitive group in these studies. Such a finding may be dependent in part on the nature of the light-dark box paradigm where testing occurs during the dark cycle when recent water consumption may have resulted in a greater intake of drug before testing. Additionally, inspection of acute and chronic vehicle groups revealed an apparent reduction in light-dark transitions and light side time in chronic mice compared with acute vehicle groups, indicating that injection alone may induce a behavioral state more sensitive to increased 5-HT. The profile of acute responses across behavioral stress tests clearly delineates two additive factors in stress sensitivity: 1) sex and 2) CRF receptor dysregulation. Gonadal hormones may have an influence on synaptic plasticity and neurotransmission that mediates these sex differences. In females, differences in responsivity may arise from rapid actions between estrogen receptors and glutamatergic signaling in limbic brain regions (54), particularly within regions receiving stress-mediated 5-HT inputs. This enhanced excitatory input could be exacerbated by increased activity of raphe neurons in R2KO animals, ultimately resulting in modified responses in behavioral stress measurements (30).

SERT levels in male and female WT and R2KO mice present a possible contributing factor to increased SSRI sensitivity observed in behavioral tests. Autoradiographic analysis revealed a main effect of sex in the dorsal and ventral hippocampus where females had overall lower SERT levels. Reduced female SERT levels may indicate an altered ability to accommodate fluctuations in 5-HT levels after SSRI treatment, potentially sensitizing the 5-HT system and further enhancing drug effects. This finding, coupled with enhanced changes in behavioral stress measures within females, may contribute to the anxiety-like responses observed in our behavioral studies, supporting previous findings showing a link between serotonergic pathways and anxiety (55,56,57,58). Analysis of SERT in prefrontal cortex revealed a genotypic effect where both male and female R2KO mice showed significantly reduced SERT. Together, the decreased SERT among female mice in the hippocampus and in R2KO mice in the prefrontal cortex may provide a mechanism whereby female R2KO mice show the greatest sensitivity and behavioral responses to acute SSRI. Serotonergic influence on the cortex, including changes in SERT levels has been described as a point of genetic vulnerability in affective disorders (59,60,61,62). Although SERT is among multiple targets that modulates 5-HT activity, we have also recently reported increased expression of the rate-limiting synthesis enzyme for 5-HT, tryptophan hydroxylase-2, in the raphe of R2KO mice, further supporting a genetic basis for the increased sensitivity to acute citalopram in these mice (30).

The HPA stress axis is a physiological indicator of organismal stress state and is subject to sex-specific regulation (63). In our examination of corticosterone levels after an acute restraint stress, we detected the predicted sex differences where females displayed significantly elevated corticosterone compared with male mice. Acute citalopram treatment before the start of restraint increased baseline corticosterone for both sexes. Within females, this effect was significantly elevated in R2KO mice, again supporting the increased stress sensitivity of this group. In males, citalopram further increased corticosterone levels throughout the HPA time course. Both males and females treated with citalopram showed a main effect of drug to delay stress recovery, suggesting that acute pharmacological increases in 5-HT may override glucocorticoid-mediated negative feedback in the HPA stress axis. Although chronic citalopram failed to induce significant drug effects, trends in data suggesting a reduced HPA response after chronic SSRI exposure may indicate a potential blunting of the HPA and suggest a potential measure for therapeutic effects of chronic SSRI (53,64).

Overall, these studies have examined the hypothesis that a dysregulation of stress pathways produces alterations in 5-HT-mediated behavioral and physiological responses during stress. Furthermore, comparisons identifying the intersection between genotype and sex in these outcomes provided support for this hypothesis because the most stress-sensitive group, R2KO females, showed a consistent and significant response to acute citalopram treatment in behavioral tests. Sex and genotype effects on SERT levels that were brain region specific likely contribute to the complexity of affective disorder susceptibility and to the challenges in developing effective treatment strategies. We hypothesize that stress susceptibility may be produced by dysfunction of the stress-5-HT system arising from both basal sex differences as well as genetic dysregulation of the CRF system, leaving certain subpopulations at especially high risk of developing stress-induced affective disorders. These findings suggest that an increased sensitivity to acute SSRI administration may be a predictive marker of an underlying vulnerability to disease.