Anxiety and panic responses to a predator in male and female Ts65Dn mice, a model for Down syndrome

Authors

  • C. Martínez-Cué,

    Corresponding author
    1. Laboratory of Developmental Neurobiology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Cantabria, Santander, Spain
    Search for more papers by this author
  • N. Rueda,

    1. Laboratory of Developmental Neurobiology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Cantabria, Santander, Spain
    Search for more papers by this author
  • E. García,

    1. Laboratory of Developmental Neurobiology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Cantabria, Santander, Spain
    Search for more papers by this author
  • J. Flórez

    1. Laboratory of Developmental Neurobiology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Cantabria, Santander, Spain
    Search for more papers by this author

C. Martínez-Cué, Laboratory of Developmental Neurobiology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Cantabria, Santander 39011, Spain. E-mail: martinec@unican.es

Abstract

Hyperactivity is a feature frequently reported in behavioral studies on the Ts65Dn (TS) mouse, the most widely accepted model of Down syndrome, when tested in anxiety-provoking situations such as the plus-maze and the open-field tests. Although this behavior could be considered as an expression of reduced anxiety, it has been considered as a consequence of a lack of behavioral inhibition and/or reduced attention. This study addressed anxiety and panic behavior of male and female TS mice by evaluating serum biochemical parameters and behavioral responses to a predator in the Mouse Defense Test Battery. Flight, risk assessment, defensive threat/attack and escape attempts were measured during and after rat confrontation. When confronted to a rat, male TS mice showed similar biochemical and behavioral responses as control mice. However, female control and TS mice presented lower serum adrenocorticotropic hormone (ACTH) levels under basal conditions and higher corticosterone levels after predator exposure than male mice. Thus, there was a larger increase in ACTH and corticosterone levels after predator exposure with respect to the undisturbed condition in females than in males. In addition, TS females showed some alterations in defensive behaviors after predator exposure. The results emphasize the need to consider gender as a confounding factor in the behavioral assessment of TS mice.

Individuals with Down syndrome (DS) are at a higher risk of significant psychopathology, as this condition is more often accompanied by depressive disorders and other anxiety-associated disturbances than the general population. Individuals with DS show higher prevalence of depression (11.3% compared with 4.3% of controls) (Collacott et al. 1992; McCarthy & Boyd 2001). They also may present anxiety disorders including phobias (e.g. agoraphobia, fear of loud noises or certain situations in daily life), repetitive behaviors and obsessional slowness (Charlot et al. 2002; Myers & Pueschel 1991; Nicham et al. 2003).

The Ts65Dn (TS) mouse is the most robust and widely used model of DS. It contains an extra chromosome spanning most of the region of MMU16 that is homologous to HSA21. The trisomic region extends from Mrp139 to Znf295 and contains roughly 136 genes that are orthologous to human genes on HSA21 (Kahlem et al. 2004). It possesses many physical, neurological and behavioral features that are reminiscent of those seen in people with DS (Baxter et al. 2000; Dierssen et al. 1996, 1997; Galdzicki et al. 2001; Galdzicki & Siarey 2003; Granholm et al. 2000; Holtzman et al. 1996; Hyde & Crnic 2001; Insausti et al. 1998; Kurt et al. 2000; Martínez-Cuéet al. 1999, 2002; Paz-Miguel et al. 1999; Reeves et al. 1995; Ruiz de Azúa et al. 2001; Siarey et al. 1997; 1999; Stasko & Costa 2004).

The behavioral characterization of TS mice has been focused mainly on cognitive processes. TS mice showed developmental delay, reduced capacity of attention and alterations in cognitive flexibility, spatial learning and working and reference memory (Coussons-Read & Crnic 1996; Escorihuela et al. 1995; 1998; Holtzman et al. 1996; Stasko & Costa 2004). Gender differences in learning responses to environmental enrichment have been reported (Martínez-Cuéet al. 2002). However, features related to other dimensions of behavior are less documented. Anxiety in TS mice was assessed in the plus-maze and open-field tests (Coussons-Read & Crnic 1996; Escorihuela et al. 1995; 1998). In both paradigms, only male mice were studied. In these tests, TS mice showed behavior that was rated as less anxious, as they displayed more active exploratory behavior in the open arms of the plus maze as well as in the open field. However, this hyperactivity in anxiety-provoking situations has been rather interpreted as a lack of behavioral inhibition and/or reduced attention to relevant stimuli (see Crnic & Pennington 2000; Escorihuela et al. 1998).

It has been suggested that defensive behaviors of rodents to predators may constitute a reliable model for understanding anxiety disorders. The Mouse Defense Test Battery (MDTB) measures a full range of specific defensive behaviors either to a present threat source (a predator) or to situations closely associated with it. A full range of antipredator defensive behaviors such as escape attempts, flight, risk assessment, immobility and defensive attack are evaluated. This battery of tests allows discrimination between the motor and cognitive aspects of anxiety and between anxiety and panic-like responses (see Blanchard et al. 1997; Griebel et al. 1995b; 1996a; 1996b; 1996c; 1998).

The present work was aimed at characterizing and comparing anxiety and panic responses in male and female TS mice. This was accomplished by: (a) analyzing several behavioral responses in ethological anxiety-provoking situations, such as the presence of a predator and (b) characterizing the hormonal response of the hypothalamic-pituitary-adrenal (HPA) axis following exposure to an acute stressor, the predator.

Materials and methods

All procedures herein described were in compliance with ethical principles and guidelines for scientific experiments on animals of the European Communities directive 86/609/EEC regulating animal research and were approved by the Local Ethical Committee of the University of Cantabria.

All the behavioral procedures and blood sampling were performed during the light phase of the light/dark cycle between 1000 and 1500 . To avoid circadian hormone variation effects on the results, we equally distributed animals of the four different groups [male and female TS and control (CO) mice] throughout the experimental period. During all the biochemical and behavioral assessments, the experimenters were blind to gender and genotype of the animals.

Animals

TS mice were bred in the Faculty of Medicine colony, from TS females and B6EiC3HF1 male breeders provided by the Robertsonian Chromosome Resources (The Jackson Laboratory, Bar Harbor, ME). CO mice were the non-trisomic littermates. A total of 80 male and female TS and CO mice (20 animals in each group) of 3–5 months of age were used. Ten mice from each group were used to determine the baseline levels of ACTH, corticosterone and testosterone, and the rest were used to evaluate their behavior in the MDTB and ACTH, corticosterone and testosterone levels after predator exposure. All animals were housed in standard laboratory conditions, and mixed groups of two to three mice of both genotypes from the same litters and the same age and gender were housed together in macrolon cages with a temperature of 22 ± 2 °C, a light cycle of 12:12 h and ad lib access to food and water.

MDTB

Apparatus

The experimental apparatus was a black wooden oval runway, 0.4 m wide, 0.3 m high and 4.8 m in total length, consisting of two 2-m straight segments joined by two 0.4-m curved segments and separated by a median wall (2.0 × 0.3 × 0.06). The apparatus was elevated 0.8 m from the floor to enable the experimenter to easily hold the rat, while minimizing the mouse's visual contact. The floor was marked every 20 cm to facilitate distance measurement. Activity was recorded with video cameras mounted above the apparatus. Experiments were performed in the dark under red light.

Procedure

The experimental procedure described by Griebel et al. (1995a) was consistently followed.

Pre-test: motor activity before exposure to a predator. Subjects were placed in the oval runway for a 3-min familiarization period, during which line crossings and wall rearings (an index of escape attempts) were recorded.

Predator avoidance test. Immediately after the familiarization period, a hand-held dead rat (killed by CO2 inhalation) was introduced into the runway and brought up to the subject at a speed of approximately 0.5 m/second. Approach was terminated when contact with the animal was made or the subject ran away from the approaching rat. If the subject fled, avoidance distance (the distance from the rat to the subject at the point of flight) was recorded. This was repeated five times.

Chase/flight test. The hand-held rat was brought up to the subject at a speed of approximately 2.0 m/second. During a 1-min period, the following parameters were measured: speed and distance traveled escaping from the predator, number of stops (pause in movement), orientations (subject stops, then orients the head toward the rat) and reversals (subject stops, then runs in the opposite direction). The latter three responses are considered risk assessment activities.

Straight alley. The runway was converted into a straight alley by the closing of two doors. For 1 min, the hand-held rat remained at a constant distance of 40 cm from the subject, and freezing time and the number of approaches/withdrawals (subjects move more than 0.2 m toward the rat and then return back) were recorded. The latter response is considered to be a risk assessment activity.

Forced contact. The experimenter brought the rat up to contact the subject for 30 seconds, and vocalization, biting and upright postures (fighting behavior) done by the subject were noted.

Post-test: contextual defense.  Immediately after the forced contact test, the predator was removed and doors were opened. Line crossings and wall rearings were recorded during a 3-min period.

Determination of biochemical parameters

In order to assess serum ACTH, corticosterone and testosterone levels under basal conditions, we extracted blood from the venous sinus of the animals to 40 undisturbed male and female TS and CO mice. To evaluate these hormone levels after exposure to the predator, we extracted blood to the 40 mice subjected to the MDTB immediately after finishing the behavioral tests (i.e. 15 min after the first exposure to the predator).

ACTH

Serum ACTH concentration was determined by an ICN Biomedicals Inc. (Costa Mesa, CA) radioimmunoassay detection kit (catalog number: 07-106101) that uses I125 together with ACTH and a specific antiserum against ACTH. Blood samples were collected in plastic tubes with ethylenediaminetetraacetic acid. Standard and samples were duplicated (200 µl), antibody against ACTH marked with I125 was added (100 µl). A second antibody was then added and allowed to incubate for 16 h at 4 °C; later on, the supernatant was decanted and the radioactivity present was determined in a gamma counter.

Corticosterone

To determine serum corticosterone concentration, we used ICN Biomedicals Inc. radioimmunoassay detection kit (catalog number: 07-120102 and 07-120103) that uses I125 together with corticosterone and a specific antiserum against corticosterone. After the addition of I125-corticosterone and antiserum against corticosterone to the blood samples and standards, tubes were incubated at room temperature during 2 h. Precipitant solution was added, the tubes were centrifuged, the supernatant was decanted and radioactivity present was determined by a gamma counter.

Testosterone

Serum concentration of testosterone was determined by ICN Biomedicals Inc. radioimmunoassay detection kit (catalog number: 07-189102) that uses I125 together with testosterone and a specific antibody against testosterone. I125-testosterone and the antibody antitestosterone were added to standards and blood samples, and tubes were incubated at 37 °C during 2 h. A second antibody was then added; tubes were vortexed and incubated again for 1 h. Finally, they were centrifuged, the supernatant was decanted and radioactivity present was determined by a gamma counter.

Statistics

Results from the concentration curves of ACTH, corticosterone and testosterone were analyzed by non-lineal regression analysis using the software package graphpad inplot (GraphPad Software Inc, San Diego, CA). Biochemical and behavioral data were analyzed by anova (ACTH, corticosterone, testosterone, crossings, rearings, number of avoidances, avoidance distance, flight distance and speed, stops, orientations and immobility time) or the non-parametric Kruskal–Wallis anova for some infrequently occurring or highly variable behaviors (reversals, approaches/withdrawals, bites, vocalizations and boxing). Post hoc comparisons were made using Tukey tests or the non-parametric Mann–Whitney U-test. Comparisons between plasma ACTH, corticosterone, testosterone and the different behaviors assessed in the MDTB were examined using Spearman's Rank-Order correlations. The analyses were done with the statistical package spss for windows version 11.0. Differences between groups were considered statistically significant when P < 0.05.

Results

Biochemical measures

The biochemical responses of male and female TS and CO mice are shown in Fig. 1. Under basal conditions, ACTH levels were lower in female than in male mice, and this difference reached statistical significance in controls (anova genotype × gender: F3,37 = 9.61, P < 0.001; Fig. 1a). However, after exposure to the predator, no differences between both genotypes or genders were found (F3,37 = 0.81, NS; Fig. 1a), because ACTH levels increased in females (F3,37 = 24.81, P < 0.001) but not in males (F3,37 = 2.66, NS) with respect to the basal condition.

Figure 1.

Mean ± SEM of plasma concentration of(a) ACTH,(b) corticosterone and(c) testosterone in male and female control and Ts65Dn mice.*P < 0.05, †P < 0.01; ‡P < 0.001 male vs. female mice; §P < 0.05, ¶P < 0.01, **P < 0.001 basal vs. post-predator exposure levels; Tukey tests after significant anovas.

No significant differences were found in corticosterone levels between genders or genotypes when they were assessed in undisturbed mice (F3,37 = 1.37, NS; Fig. 1b). However, after the MDTB, female TS and CO mice showed higher corticosterone levels than male mice (F3,39 = 9.19, P < 0.001) due to the large increase that occurred in female [F3,37 = 11.29, P < 0.001] but not in male mice (F3,37 = 0.52, NS).

As expected, female CO and TS mice presented much lower levels of testosterone than males both under basal conditions (F3,38 = 4.51, P < 0.05; Fig. 1c) and after been exposed to the predator (F3,38 = 5.68, P < 0.01). However, no significant difference was found between CO and TS mice in any gender condition. and no significant difference was found in testosterone levels between undisturbed or predator-exposed female (F3,38 = 0.76, NS) or male (F3,38 = 0.49, NS) mice.

MDTB

The assessment of contextual fear in the four groups of mice is shown in Fig. 2. No significant differences were found either in the number of crossings (Fig. 2a) made in the pre-test (F3,39 = 0.54, NS) and in the post-test (F3,39 = 1.76, NS) or in the number of escape attempts (i.e. wall rearings) (Fig. 2b) done before (F3,39 = 0.88, NS) or after (F3,39 = 1.01, NS) exposure to a predator. Figure 2a also shows that exposure to a predator produced a larger increase in locomotor activity in male and female TS mice than in controls, although this difference did not reach statistical significance (F3,39 = 0.47, NS).

Figure 2.

Mean ± SEM of the number of crossings in pre- and post-test periods and the difference between them(a) and of the number of rearings in the pre- and post-test periods(b) in male and female control (CO) and Ts65Dn (TS) mice.

The expression of predator avoidance is shown in Fig. 3. Female TS mice performed a lower number of avoidances to the predator than CO mice (F3,39 = 4.49, P < 0.01), but no differences were found in males (Fig. 3a). Although the avoidance distances were smaller in both male and female TS mice than in controls, the difference did not reach statistical significance (F3,39 = 1.58, NS; Fig. 3b).

Figure 3.

Mean ± SEM of the number of avoidances(a) and of the avoidance distance(b) during the predator avoidance test in male and female control (CO) and Ts65Dn (TS) mice.*P < 0.01 TS vs. CO mice; Tukey tests after significant anovas.

When being chased by a predator, male and female TS and CO mice made a similar number of risk assessment behaviors (Fig. 4a; stops: F3,39 = 1.10, NS; orientations: F3,39 = 0.78, NS; reversals: K = 0.83, NS). Flight characteristics of the chase test showed no differences among the four groups of mice either in the flight distance (F3,39 = 0.62, NS; Fig. 4b) or in the speed (F3,39 = 0.62, NS; Fig. 4c).

Figure 4.

Mean ± SEM of the number of risk assessment behaviors (stops, orientations and reversals),(a) the distance traveled(b) and speed(c) during the chase/flight test by male and female control (CO) and Ts65Dn (TS) mice.

In the straight alley test, TS female mice displayed a longer immobility time than the rest of the mice (F3,39 = 4.03, P < 0.05; Fig. 5a) and performed a smaller number of approaches/withdrawals to the rat (K = 10.59, P < 0.05; Fig. 5b).

Figure 5.

Mean ± SEM of the(a) freezing time,(b) number of approaches/withdrawals and(c) vocalizations, upright postures and bitings in the forced contact test done by male and female Ts65Dn (TS) and control (CO) mice.*P < 0.05, TS vs. CO mice; †P < 0.05, male vs. female mice; Tukey tests after significant anovas. ‡P < 0.01 TS vs. CO mice; §P < 0.01, ¶P < 0.01 male vs. female mice. Mann–Whitney U-tests after significant Kruskal–Wallis anova.

In the forced contact test, female TS and CO mice performed a higher number of vocalizations than male mice (K = 20.54, P < 0.001; Fig. 5c). However, no differences were found in the other attack parameters among the four groups of mice (upright postures: K = 2.57, NS; defensive biting: K = 4.46, NS).

Correlation between biochemical and behavioral measures

Table 1 summarizes the relationship between the biochemical and behavioral measures obtained. ACTH did no correlate with any of the behavioral or biochemical parameters assessed. However, freezing time in the straight alley was positively correlated to plasma corticosterone (R = 0.32, P < 0.05) and negatively correlated to plasma testosterone levels (R = −0.42, P < 0.01). Vocalizations during the forced contact test were also positively correlated to plasma corticosterone levels (R = 0.47, P < 0.01) and negatively correlated to plasma testosterone levels (R = −0.59, P < 0.001). Plasma corticosterone and testosterone levels were negatively correlated (R = −0.64, P < 0.001).

Table 1.  Correlations between the behavioral measures obtained in the The Mouse Defense Test Battery and the plasma ACTH, corticosterone and testosterone levels in male and female Ts65Dn and control mice
 Spearmans' R
 ACTHCorticosteroneTestosterone
Habituation/ Contextual fear
 Crossings during the pre-tests0.49, NS−0.26, NS0.01, NS
 Crossings during the post-test−0.03, NS−0.04, NS0.15, NS
 Rearings during the pre-tests0.03, NS−0.28, NS0.04, NS
 Rearings during the post-test0.01, NS−0.04, NS0.02, NS
Predator avoidance test
 Number of avoidances−0.11, NS0.17, NS0.01, NS
 Mean avoidance distance−0.21, NS0.18, NS−0.13, NS
Chase/flight test
 Risk assessment
  Stops0.22, NS0.06, NS0.15, NS
  Orientations0.16, NS0.22, NS−0.13, NS
  Reversals−0.01, NS0.08, NS−0.15, NS
 Distance0.16, NS0.12, NS−0.05, NS
 Speed0.16, NS0.12, NS−0.05, NS
Straight alley test
 Freezing time−0.17, NS0.32, P < 0.05−0.42, P < 0.01
 Approaches/withdrawals0.04, NS−0.24, NS0.20, NS
Forced contact
 Vocalizations−0.30, NS0.47, P < 0.01−0.59, P < 0.001
 Upright postures−0.02, NS0.09, NS0.03, NS
 Biting0.02, NS−0.01, NS0.15, NS
Biochemical parameters
 ACTH1.0−0.23, NS0.23, NS
 Corticosterone−0.23, NS1.0−0.64, P < 0.001
 Testosterone0.23, NS−0.64, P < 0.0011.0

Discussion

The present study documents the existence of several alterations in the response of TS mice to an ethological anxiety-provoking situation. The response, however, was gender dependent: whilst male TS mice behaved similarly to CO when confronted to a rat and showed similar HPA activation, all female mice presented lower ACTH levels and higher corticosterone levels after predator exposure. Additionally, TS females displayed alterations in some behavioral responses.

Previous work from this and other laboratories suggested that TS mice could display less anxiety than their CO littermates in the plus-maze and open-field tasks, as they showed hyperactivity in situations that usually provoke caution and suppress activity in normal mice (Coussons-Read & Crnic 1996; Demas et al. 1996; Escorihuela et al. 1995; 1998). Male TS mice were more active in the light/dark open field and ventured more frequently into the center of the open field than CO mice. Furthermore, in the elevated plus maze, male TS mice showed less initial freezing and made more arm entries and visited the open arms more frequently than CO. The number of arm entries and the amount of time spent in the open arms have traditionally been estimated as an index of anxiety. However, the hyperactivity seen in TS mice in the anxiety-provoking situations in these tests has not been interpreted as a decrease in anxiety but rather as a failure to inhibit activity or as a deficit in the ability to attend to relevant stimuli (see Crnic & Pennington 2000; Escorihuela et al. 1998).

The need for a better characterization of anxiety and other emotional responses (Stasko & Costa 2004) was accomplished by using the MDTB. This battery of tests elicits and measures defensive reactions of mice to both present (fear) and anticipated (anxiety) threat (a predator). In addition, it allows discrimination between motor and cognitive aspects of anxiety and between anxiety and panic-like responses (see Blanchard et al. 1997; Griebel et al. 1995b; 1996a; 1996b; 1996c; 1998).

Regarding motor activity in basal conditions and after exposure to the predator, TS mice were more active in all conditions than CO mice, but the difference did not reach statistical significance. Although hyperactivity of TS mice has been consistently reported in the open field and plus maze (Coussons-Read & Crnic 1996; Demas et al. 1996; Escorihuela et al. 1995; 1998), the effect was milder in the oval runway, probably due to the conditions of the equipment that is a dimly lighted dark, and offers less-aversive environment.

All groups of mice showed increased contextual fear evaluated by the number of escape attempts (wall rearings) after exposure to the predator; however, no differences between the different groups of mice were found in the number of escape attempts performed before or after exposure to the predator, indicating that this kind of expression of the predator-induced anxiety was similar in all groups. However, exposure to a predator produced a larger, although non-significant, increase in locomotor activity in male and female TS mice than in controls which is consistent with the reduced behavioral inhibition described in male TS mice in other experimental paradigms (Escorihuela et al. 1995; 1998; Coussons-Read & Crnic 1996), and that is also a characteristic of individuals with DS (see Crnic & Pennington 2000).

In male mice, no differences were found between TS and CO for any anxiety or panic measurements. This supports the interpretation that the hyperactivity found in male TS mice on the plus maze and open field, previously reported, was not accounted for by differences in anxiety but rather by other behavioral and/or cognitive processes (Crnic & Pennington 2000; Escorihuela et al. 1998).

By contrast, TS female mice, which have never been evaluated in anxiety tests so far, showed alterations in some anxiety measures of the test battery in relation to CO females. In the straight alley test, where mice are constrained in one part of the runway and the predator is presented to them at a closer distance, mice display freezing followed by predator assessment consisting of approaches followed by withdrawals. In this situation, TS female mice presented longer immobility time than the rest of the mice, indicating enhanced anxiety (Fig. 5). Furthermore, TS females performed less predator assessment activity as they performed a smaller number of approaches/withdrawals to the rat, which is consistent with the increased freezing, and does not reflect a reduction in fear to the source of threat. Risk assessment was also assessed during the chase test. When mice escape from a predator, they often show an abrupt movement arrest followed by orientation to the source of danger, i.e. predator assessment, as well as a reversal of movement to approach the predator. These activities are associated with a gathering of information concerning potential threat sources and facilitate acquisition of information leading to intensified defensiveness and defensive behavior (see Blanchard et al. 1997). During the chase test whilst male TS and CO mice made a similar number of risk assessment activities, no gender effect was found in this cognitive aspect of anxiety. However, there was a non-significant decrease in the number of stops in TS females, consistent with the decreased predator assessment found in the straight alley in this group of mice.

In addition, when the contact between the rat and the mouse is forced, defensive threat and attack behaviors occur. These defensive behaviors, including vocalization, biting and upright postures, reflect a more ‘affective’-oriented anxiety-related defense (Griebel et al. 1996c). In the forced contact test (Fig. 5c), no significant difference between TS and CO mice was found in defensive threat and attack behaviors. However, sexual dimorphism was found in the magnitude of a defensive behavior such as vocalization. Both TS and CO females vocalized much more than males, and female TS vocalized more than CO mice.

In the rat avoidance test, female TS mice avoided the predator to a lower extent than CO mice. Number of avoidances and avoidance distances are considered a measure of panic behavior, a lower distance being indicative of reduced panic response (Blanchard et al. 1993; Griebel et al. 1995b; 1996c). However, this result should be interpreted cautiously, because it can be contaminated by the concurrent enhanced freezing behavior observed in the female TS mice.

In the chase test, no differences were found among the four groups of mice in the flight distance or speed. Overall, our findings show a more variable anxiety profile in female TS compared with male TS mice.

A gender-related difference was also evident in the biochemical measures. Under basal conditions, ACTH levels were lower in female than in male mice; however, after exposure to the predator, no differences between both genotypes or genders were found as ACTH levels increased in females but not in males with respect to the basal condition. Furthermore, although no significant differences were found in corticosterone levels of the undisturbed mice, female TS and CO mice showed higher corticosterone levels than male mice after the MDTB, due to the large increase that occurred in female mice. Therefore, ACTH and corticosterone levels were higher in female mice subjected to predator exposure than in undisturbed mice.

Corticosterone levels were positively correlated with some of the altered behaviors in female mice after being confronted to a predator, such as freezing time in the straight alley and vocalizations during the forced contact test (the former altered in female TS mice and the later in both groups of females). By contrast, ACTH levels after predator exposure was not correlated to any behavioral measure. As expected, testosterone values were much lower in females than in males. Their hormone levels were negatively correlated to corticosterone and to those behaviors enhanced in females (i.e. freezing in the straight alley and vocalizations during rat confrontation). However, predator exposure did not affect testosterone levels in TS and CO mice. Other studies performed in rats have similarly reported that the predator exposure increased the corticosterone response but did not modify testosterone levels (Blanchard et al. 1998).

In the present study, blood testosterone concentration was evaluated, because this hormone is known to be implicated in aggressive behavior (Birger et al. 2003; Kalin 1999), and aggression is an important component of the mice defense/attack repertoire, specially of the behaviors provoked when been attacked by a predator (i.e. in the forced contact test). Due to the gender effect found in some of the analyzed behaviors, future studies should also assess other hormone levels such as estrogen and progesterone on anxiety and panic behavior performance.

Gender differences can be found in almost every aspect of physiological functioning, and the stress-reactivity literature shows that this is also the case for stress. It is known that male and female stress responses are different. The pattern of stress-dependent corticosterone secretion also appears different between the sexes. Females generally have higher basal levels of circulating corticosterone as well as higher levels following exposure to a stressor than males (Atkinson & Waddell 1997; Beck & Luine 1999; 2002; Bowman et al. 2001; 2002; Galea et al. 1997; Haleem et al. 1988; Handa et al. 1994; Luine et al. 1994; Rivier 1993; 1999). A greater inhibition of food intake is produced in female than in male rats after stress (Kuriyama & Shibasaki 2004), and females are also more sensitive to the effects of prenatal stress (McCormic et al. 1995). Sex differences in the stress response are probably due to differences in the circulating levels of estradiol between the sexes, because estrogens may alter or interact with the HPA axis in regulating corticosterone release (Atkinson & Waddell 1997; Rivier 1993; Viau & Meaney 1991). In humans too, gender differences in the HPA axis is related to circulating estradiol (see Luine 2002). It is therefore possible that gender-specific changes in the HPA-axis response to a stressor may sensitize female mice, make them more vulnerable and facilitate the expression of anxious behavior.

Although less frequently than in other developmental disorders, behavioral problems of several types may appear in DS population. Anxiety disorders have been reported with prevalence estimates ranging from 1 to 7% (Barba 2005; Myers 1992). With the exception of autistic spectrum disorders (Capone et al. 2005), no gender-related prevalences have been observed.

In conclusion, the present study highlights the importance of considering female behavior when studying this mouse model of DS. TS females are not often used in behavioral research either because they are thought to be too variable or because they are used for reproduction. However, this work and previous work from our laboratory (Martínez-Cuéet al. 2002) suggest that there are important differences between male and female TS behavior and that using only males would limit the information obtained in these studies.

Acknowledgments

We are very grateful to Dr G. Griebel for helping us to set up the MDTB experiments. This work was supported by the Spanish Ministry of Education (SAF-2002-02178) and The Jérôme Lejeune Foundation.

Ancillary