Postnatal handling reverses social anxiety in ... - Wiley Online Library

Genes, Brain and Behavior (2010) 9: 26–32

© 2009 The Authors Journal compilation © 2009 Blackwell Publishing Ltd/International Behavioural and Neural Genetics Society

Postnatal handling reverses social anxiety in serotonin receptor 1A knockout mice C. Zanettini† , V. Carola‡,§ , L. Lo Iacono‡,∗∗ , A. Moles†,¶ , C. Gross‡ and F. R. D’Amato†,∗ † CNR Institute of Neuroscience, Via del Fosso di Fiorano 64, 00143 Roma, Italy, ‡ Mouse Biology Unit, EMBL, Via Ramarini 32, 00015 Monterotondo, Italy, § Santa Lucia Foundation,

European Centre for Brain Research (CERC), Via del Fosso di Fiorano 64/65, 00143 Rome, Italy, and ¶ Genomnia srl, Milano, Italy ** Present address: Drug Discovery and Development Department, Istituto Italiano di Tecnologia, via Morego 30, 16163 Genova, Italy *Corresponding author: F. R. D’Amato, CNR Institute of Neuroscience, Via del Fosso di Fiorano 64, 00143 Roma, Italy. E-mail: [email protected]

Mice lacking the serotonin receptor 1A (Htr1a knockout, Htr1aKO ) show increased innate and conditioned anxiety. This phenotype depends on functional receptor activity during the third through fifth weeks of life and thus appears to be the result of long-term changes in brain function as a consequence of an early deficit in serotonin signaling. To evaluate whether this phenotype can be influenced by early environmental factors, we subjected Htr1a knockout mice to postnatal handling, a procedure known to reduce anxiety-like behavior and stress responses in adulthood. Offspring of heterozygous Htr1a knockout mice were separated from their mother and exposed 15 min each day from postnatal day 1 (PD1) to PD14 to clean bedding. Control animals were left undisturbed. Maternal behavior was observed during the first 13 days of life. Adult male offspring were tested in the open field, social approach and resident–intruder tests and assessed for corticosterone response to restraint stress. Knockout mice showed increased anxiety in the open field and in the social approach test as well as an enhanced corticosterone response to stress. However, while no effect of postnatal handling was seen in wild-type mice, handling reduced anxiety-like behavior in the social interaction test and the corticosterone response to stress in knockout mice. These findings extend the anxiety phenotype of Htr1aKO mice to include social anxiety and demonstrate that this phenotype can be moderated by early environmental factors. Keywords: corticosterone, knockout, maternal behavior, open field test, postnatal handling, serotonin, social approach test, stress, ultrasonic calls

Received 24 April 2009, revised 13 July 2009, accepted for publication 2 August 2009

26

Introduction Early life events are important determinants of behavioral traits in adulthood and have been shown to moderate risk for mental illness. In rodents’ alterations in rearing environment including natural variations in maternal care between individuals and/or strains as well as experimentally induced changes in rearing environment such as separation from the mother, communal rearing and early weaning have been associated with changes in anxiety-like behavior and stress response (Holmes et al . 2005). Among these, daily brief separation from the mother (15 min/day) during the first 2 weeks of life has arguably been the most commonly used experimental manipulation of rearing environment. In the mouse this manipulation has been shown to elicit increased maternal licking and grooming of pups upon reunion (Moles et al . 2008; Pryce and Feldon 2003) and this increase in maternal care is crucial for the long-term effects of the procedure (D’Amato et al . 2005; Zimmerberg et al . 2003). In the mouse, repeated brief manipulations during development lead to reduced responsiveness to stress and decreased anxiety-like behavior in adulthood (Cabib et al . 1993; D’Amato et al . 1998; Parfitt et al . 2004, 2007); however, a similar procedure has not been effective in several laboratory inbred strains (Millstein & Holmes 2007) suggesting that genetic background modifiers play an important role in determining susceptibility to maternal care. The aim of the present study was two-fold. First, we investigated whether Htr1aKO mice showed increased anxiety-like behavior under conditions of social interaction. Htr1aKO mice exhibit increased avoidance and decreased exploration in the open field, elevated plus maze, elevated zero maze and novelty-suppressed feeding tests (Gross et al . 2000, 2002; Heisler et al . 1998; Parks et al . 1998; Ramboz et al . 1998), all tests of innate anxiety. Furthermore, Htr1aKO mice demonstrate enhanced context and cue-dependent fear conditioning to ambiguous stimuli (Klemenhagen et al . 2006; Tsetsenis et al . 2007), suggesting a bias in their processing of aversive cues. However, social anxiety has not been systematically examined in this strain. One study measured ultrasound vocalizations (USVs) in response to maternal separation and found reduced calls in Htr1aKO pups (Weller et al . 2003). However, no tests measuring social anxietylike behavior in adulthood have been reported. To test social anxiety, we examined USVs in response to maternal separation in postnatal day 6 (PND 6) pups and measured preference in the social approach test in adulthood. doi: 10.1111/j.1601-183X.2009.00531.x

Postnatal handling in Htr1aKO mice

Second, we assessed the effect of brief daily handling during the neonatal period on anxiety-like behavior and stress response in Htr1aKO mice and their wild-type littermates. We showed previously that the increased innate anxiety seen in the open field and elevated plus maze by Htr1aKO mice is not further moderated by the level of maternal care (Carola et al . 2006). However, in this study we examined only non-social anxiety-like behavior and, moreover, the study was restricted to heterozygous knockout mice. Our present findings show that Htr1aKO mice exhibit increased social anxiety-like behavior both during early life and adulthood and that this phenotype can be reversed by exposing the pups to daily maternal separation during the neonatal period. These data demonstrate that serotonin plays a critical role in determining social behavior and that early handling can selectively reprogram social behavior in mice.

to daily separation from the mother on PNDs 1–14 for 15 min. Each day, the whole litter was transferred to a new cage with clean bedding placed on a hot plate at a temperature of 35◦ C. A second group of mother–infant pairs (N = 10, mean litter size: 5.6 + 0.34) left undisturbed in their cages was chosen as the control group. The manipulation sessions were conducted at a different time each day between 11:30 and 18:30 h. The home cages of the two experimental groups were cleaned on PND 10, and a small amount of soiled bedding was scattered into the clean cages before introducing the mice. A total of 132 pups were used in this experiment, 66 males and 66 females. Genotype distribution was as follows: wild − type = 38, heterozygous = 60, knockout = 34.

Maternal behavior Maternal behavior was observed twice a day (at 10:30 and 15:30 h) for 30 min from PND 1–12 using a sampling points procedure; the behavior of each female was recorded every 2 min, for a total of 30 observations/day. The following behaviors were scored: arched-back nursing and blanket nursing, licking/grooming pups and out of nest. Data were analyzed in 3-day blocks.

Methods Open field test Animal husbandry Heterozygous Htr1a null mice (Htr1atetO , Gross et al . 2002; Lo Iacono and Gross 2008) on a mixed 129S6/SvEvTac;C57BL/6;CBA mixed background were generated at EMBL and shipped to CNR for breeding. One week after arrival, 2-month-old animals were housed: two females and one male in a cage. Starting at day 15 post-coitus, females were weighed every 2 days to identify pregnant mice that were then isolated in cages (40 × 23 × 15 cm). Cages were inspected twice a day for pups and the day of birth was considered PND 0. Genotyping was conducted as previously described (Gross et al . 2002). Mice were marked by ear notching, weighed and weaned at PND 28 and housed in groups of four age- and sex-matched mice of the same experimental group taken from different litters. Male mice were used for all adult behavioral experiments. The colony room was kept at constant temperature (21◦ C ± 1) and maintained on a 12-h light/12-h dark cycle with free access to food and water. C57BL/6J (EMMA, Monterotondo, Italy) and NMRI (Harlan, Correzzana, Italy) male mice (6 weeks of age) housed in groups of four were used as partners in the social tests. All experiments were conducted under license from the Italian Department of Health and in accordance with Italian regulations on the use of animals for research (legislation DL 116/92) and NIH guidelines on animal care.

Ultrasound measurement Ultrasonic calls emitted by pups were recorded at PND 6. Cages were transferred to the experimental room at 14:30 h and after 1 h of acclimatization the mother was removed and transferred to a clean cage while the pups were left in their cage on a hot plate set at 35◦ C. After 5-min vocalizations uttered by pups under different experimental conditions were recorded using Ultravox 2.0 (Noldus Information Technology, Wageningen, The Netherlands) set to a frequency of 70 ± 4 kHz. Recording continued for 5 min after placing the pup in either (i) a beaker containing home cage bedding, or (ii) a beaker containing clean bedding. After the test, pup’s body weight and sex were recorded and a small piece of tail was cut for genotyping before returning them to the home cage. When all the pups had been tested, a small amount of the home cage bedding was scattered over them and the mother was reintroduced into the home cage.

Postnatal handling A second batch of Htr1aKO heterozygous pairs was used for testing the effects of postnatal handling. A total of 23 dams with their pups were selected for this experiment with four to seven pups per litter. Thirteen litters (mean litter size: 5.9 + 0.25) were submitted Genes, Brain and Behavior (2010) 9: 26–32

Open field test was performed at PND 90. The apparatus consisted of a square arena (60 × 60 × 25 cm) enclosed by opaque Plexiglass walls. We defined the region within 12 cm of the walls as the arena periphery and the remaining area as the arena center. Mice were transferred to the experimental room at 15:30 h and left for 1 h. Mice were placed into the arena periphery and wholebody locomotion recorded using a video-tracking system (TSE, Bad Homburg, Germany). The arena was carefully cleaned between sessions with 10% ethanol.

Social approach test Mice were tested at PND 110 in a rectangular Plexiglas box (35 × 20 × 15 cm) consisting of three interconnected chambers of equal size with clean bedding (Brodkin et al . 2004). Each lateral chamber contained a transparent Plexiglas cylinder (8 cm in diameter) with small holes (0.5 cm in diameter) to allow auditory, visual and olfactory investigation of the stimulus mouse. Mice were transported to the experimental room 1 h before the test. The test was divided in two sessions of each 5 min. During the first session the mouse was placed in the central chamber and allowed to freely explore the apparatus in the absence of a stimulus mouse. The mouse was briefly removed and returned to a holding cage during which time a 60-day-old C57BL/6J male mouse was introduced into one cylinder (pseudo-randomly chosen), and the test mouse was again placed in the central chamber and allowed to freely explore the apparatus. After each test, the entire apparatus was carefully cleaned with 10% ethanol. Both sessions were recorded by a whole-body videotracking system (Smart 1.1, San Diego Instruments, San Diego, CA) and the percentage of time spent in the chamber with the stimulus animal (time social/(time social + non − social)x 100) was calculated. Selected anxiety-related behaviors were manually scored from videotape (stretch attend and grooming, Carola et al . 2002) and the total duration of these behaviors calculated.

Resident–intruder test At PND 115, mice were individually housed. After 1-day habituation, an NMRI male intruder of approximately the same age and weight was introduced into the cage for 5 min. Social interaction was recorded by a video camera and the behavior of mice scored by an observer unaware of the experimental group using Observer software (Noldus Information Technology). The following behaviors were recorded, aggressive grooming, chase and attack, and summed to compute an aggressivity index. At the end of the test each subject was reunited with its cage mates.

27

Zanettini et al.

Stress response

Statistical analysis Statistical significance of differences in behavioral and biochemical measures were assessed by two- and three-way ANOVA as appropriate, followed by Tukey-HSD post hoc testing in case of significance. To reduce differences in variability between groups, anxiety-related behavior in the social approach test was subjected to square-root transformation prior to statistical testing.

Results

Clean bedding

Nest bedding

∗

200

number of USVs

At PND 120, mice were transferred to the experimental room 1 h before testing and pseudo-randomly assigned to the experimental or control group. Experimental subjects were immobilized in restraint tubes for 15 min after which they were moved to a new room and blood was collected by decapitation. Control subjects were removed from their home cage, moved to the new room and blood collected by decapitated. Blood was centrifuged at 4◦ C for 20 min and serum stored at −20◦ C for later radioimmunoassay (ICN, Costa Mesa, CA, USA). All blood samples were collected between 15:30 and 17:30 h.

∗

160 120 80 40 0

+/+

+/–

–/–

Figure 1: Increased USVs to clean bedding in Htr1aKO mice. Number of USVs emitted during 5 min by 6-day-old pups of wildtype (+/+), heterozygous (+/–), and homozygous knockout (–/–) littermates exposed to clean or nest bedding (mean ± SE; WT: N = 60, HET: N = 57, KO: N = 28; * P < 0.05)

Increased USV to separation in Htr1aKO mice We have validated the brief daily handling procedure in outbred laboratory mice as a technique to induce increases in maternal care and long-term behavioral changes in offspring (D’Amato et al . 1987, 1998). However, the use of this procedure in inbred mice has been equivocal (Millstein & Holmes 2007). To determine whether offspring of the genetic background used to maintain the Htr1aKO allele responded to maternal separation, we examined USVs of pups separated from their mothers and exposed to either familiar or fresh bedding at PND 6. In outbred mice, exposure to fresh, but not familiar bedding induced increased USV (D’Amato et al . 1987, 1990). A three-way ANOVA of the total number of USVs emitted during the 5-min test revealed a significant effect of bedding (F1,129 = 4.53, P < 0.05) with exposure to clean bedding eliciting more calls that familiar bedding (Fig. 1). Although no significant main effect of genotype or sex was observed, a significant genotype × sex × bedding interaction emerged (F2,129 = 3.42, P < 0.05). Post hoc testing only indicated that knockout males exposed to clean bedding emitted more USVs than those exposed to familiar bedding and than their wild-type controls (P < 0.05, Fig. 1). No effect of bedding was seen in wild-type littermates. These findings indicate that Htr1aKO male pups are more sensitive to the aversive effects of maternal separation than their wild-type littermates and that in the appropriate genetic background brief maternal separation can be effective in non-outbred mice.

Increased maternal care following separation of Htr1aKO pups We have shown that in outbred mice brief daily handling induces an increase in maternal care that persists during the early postnatal period and is required for the longterm changes in behavior and stress response associated with handling (D’Amato et al . 1987, 1998). To examine if a similar increase in maternal care was induced in our animals, we quantified maternal behaviors at two points during the day in mice either subjected to daily separation and

28

exposed to clean bedding (handled) or non-handled control animals. No difference in nursing, licking/grooming pups or time in nest was observed in handled vs. non-handled litters (Fig. 2a–c), although there was significant decreased in maternal care over time (F3,63 = 13.44, P < 0.0001; F3,63 = 3.79, P < 0.05; F3,63 = 22.09, P < 0.0001, respectively). No significant handling × time effect emerged. However, there was a trend for increased licking/grooming pups in handled litters during the first three PNDs (Fig. 2b). To determine whether maternal care induced by handling might depend on the genotype of the offspring involved, we compared licking/grooming pups in those litters with either more than 1 or no homozygous knockout male offspring. A two-way ANOVA of cumulative maternal care during PNDs 1–3 revealed a significant interaction between the presence or absence of knockout pups and handling (Fig. 2d; F1,19 = 5.35, P < 0.05). While no effect of handling on licking/grooming pups was seen in litters without knockout pups, significantly increased maternal care was seen in litters with at least one knockout. These data suggest that increased maternal care is elicited selectively in litters that have knockout male pups, possibly in response to their increased USVs following separation (Fig. 1).

Handling reverses social anxiety in Htr1aKO mice Htr1aKO mice show increased innate anxiety-like behavior in tests measuring avoidance of novel exposed spaces. To test whether Htr1aKO mice also exhibit increased avoidance of social stimuli and, furthermore, to test whether this behavior could be moderated by early handling, we examined social avoidance behavior in adult handled and non-handled mice. Consistent with previous studies, knockout mice showed increased avoidance of the center and decreased locomotion in the open field test (Fig. 3a,b; F2,129 = 7.61, P < 0.001 and F2,129 = 6.51, P < 0.005, respectively). However, no significant main effect of handling or interaction between genotype and handling was seen. In the social approach test, on the other hand, while groups did not differ during the Genes, Brain and Behavior (2010) 9: 26–32


c

1-3

70 60 50 40 30 20 10 1-3

4-6 7-9 10-12 postnatal day

Stress response in handled mice Postnatal handling has been consistently associated with decreased stress hormone responses in adulthood (Anisman et al . 1998; Parfitt et al . 2007). To determine whether stress hormone responses were altered under our experimental conditions, we measured control and restraint stress-induced

Licking/grooming pups

3.5 3 2.5 2 1.5 1 0.5

d

Out of nest

0

b


habituation phase, a significant interaction between handling and genotype emerged for relative time spent in the chamber with the stimulus animal (Fig. 3c; F2,71 = 8.40, P < 0.001). Post hoc analysis demonstrated significantly decreased social approach in non-handled knockout mice vs. wild-type littermates, an effect that was absent in handled animals. Moreover, an analysis of global anxiety-related behaviors (cumulative stretch attend and grooming ) in the social approach test revealed a significant genotype effect and a significant interaction between handling and genotype (Fig. 3d; F2,71 = 3.25, P < 0.05 and F2,71 = 5.21, P < 0.01, respectively). Post hoc analysis indicated significantly increased anxiety-related behaviors in non-handled knockout mice vs. wild-type littermates that was absent in handled animals. Taken together these data demonstrate that Htr1aKO mice show increased avoidance to novel social as well as spatial stimuli and that this phenotype can be reprogrammed by altering their early maternal environment. Finally, we conducted a resident–intruder test to determine whether the increased social avoidance of Htr1aKO mice could be a secondary consequence of a change in aggressive behavior. Aggressive behaviors (as assessed by an aggressivity index) exhibited by male residents in this test were not affected by handling or genotype (data not shown) suggesting that aggressive motivation did not differ in knockout mice.


non-handled handled

Fraction of observations


Figure 2: Increased maternal care toward litters with male Htr1aKO mice. Fraction of observations when mother was observed (a) nursing pups, (b) licking/grooming pups or (c) out of the nest from PND1 to PND 12. (d) Comparison of the amount of licking/grooming performed by mothers of non-handled and handled litters, according to the presence of at least one knockout male pup (including only litters with four to seven pups; non-handled: N = 10, handled: N = 13; *P < 0.05)

Nursing 80 70 60 50 40 30 20 10 0



a

0

1-3


Licking/grooming pups

∗

7 6

without KO with KO

5 4 3 2 1 0

non-handled

handled

levels of serum corticosterone. A three-way ANOVA revealed a significant main effect of stress (F1,101 = 179.53, P < 0.0001) with restrained animals showing higher levels of corticosterone, a significant main effect of handling (F1,101 = 8.95, P < 0.01) with handled mice showing lower hormone levels, and a significant interaction between handling and genotype (F2,101 = 3.10, P < 0.05). Post hoc testing indicated that handling significantly reduced the corticosterone response to restraint selectively in Htr1aKO animals (P < 0.05). These findings confirm that the effects of handling on both behavior and stress hormone response are more pronounced in Htr1aKO mice.

Discussion Two major findings emerged from the present study. First, Htr1aKO mice exhibited increased ultrasound calls to maternal separation and increased avoidance of an unfamiliar social stimulus. These findings expand the anxiety phenotype of Htr1aKO mice to include social anxiety. Second, the increased social anxiety-like behavior (but not non-social anxiety) of knockout mice was reversed by postnatal handling (Table 1). This result argues that the neural circuits controlling social anxiety in these mice are susceptible to reprogramming by early experience. Although we have previously reported reductions of anxiety-like behavior and response to stress in outbred NMRI mice subjected to brief daily handling (Cabib et al . 1993; D’Amato et al . 1987, 1998), the effect of this procedure in inbred mice has been equivocal. Anisman et al . (1998) reported similar effects in BALB/c, but not C57BL/6 mice, whereas Parfitt et al . (2004) reported that

29

Zanettini et al.

∗∗

Time in center (sec)

200

120

non-handled handled

150 100 50 0

+/+

∗

75

100 80 60 40 20

∗

50

25

+/+

d

0

+/+

+/–

0

–/–

Anxiety-related behaviors

% Time in social side

c

+/–

∗∗

b Distance traveled (m)

a

4

∗

–/–

∗∗

∗

+/–

–/–

Figure 3: Increased social avoidance in Htr1aKO mice is reversed by handling. Behavioral performance of wild-type (+/+), heterozygous (+/–) and knockout (–/–) littermates in (a, b) open field test (non-handled: N = 91, handled: N = 44; ** P < 0.01) and (c, d) social approach test (non-handled: N = 40, handled: N = 37; * P < 0.05; ** P < 0.01). Anxiety-related behaviors are expressed as the square-root of cumulative stretch attend and grooming

3 2 1 0

–/–

+/–

+/+

Table 1: Summary table describing behavioral effects of postnatal handling

Genotype WT KO

Open field anxiety Non-handled Handled = ⇑

= ⇑

Social anxiety Non-handled Handled = ⇑

brief daily handling affected the development of anxiety and hypothalamus–pituitary–adrenal axis functionality in C57BL/6. On the other hand, Millstein and Holmes (2007) did not observe any effect of brief daily handling on either maternal behavior or offspring emotionality in eight inbred strains. Due to these conflicting reports, we first examined ultrasound calls in pups during brief maternal separation. These calls are affected by olfactory, tactile and thermal manipulations (for a review see Costantini and D’Amato 2006) and have been suggested to measure primitive separation anxiety-like behavior and serve as a useful predictor of adult emotionality (Brunelli 2005; Burgdorf et al . 2005; Dichter et al . 1996; Winslow et al . 2000). To examine the importance of these olfactory cues in the response to maternal separation in our mice, we exposed pups to either clean or familiar nest bedding during the separation procedure. Our analysis of separation-induced ultrasound calls revealed no difference in calls between Htr1aKO and wild-type littermate mice when exposed to familiar nest bedding (Fig. 1). However, when mice were exposed to clean bedding, male Htr1aKO mice emitted significantly more calls than wild-type littermates with heterozygous mice showing intermediate levels. Importantly, in knockout mice familiar bedding significantly reduced ultrasound calls. These findings confirm our earlier studies demonstrating the importance of olfactory cues to separation-induced ultrasound calling in

30

= =

Stress response Non-handled Handled = =

= =

mice. These data argue that Htr1aKO mice show increased separation anxiety-like behavior in infancy. Our data are in contrast with those reported by Weller et al . (2003) who showed that Htr1aKO pups emitted fewer ultrasound calls than their wild-type littermates. Although the origin of this discrepancy is not entirely clear at this moment, the sex of pups was not considered in the Weller et al . study and a sex bias may have contributed to their result. Next, we examined maternal behavior in litters subjected to daily brief handling. The frequency of nursing, licking/grooming of pups and time out of the nest did not differ between litters subjected to separation and those left undisturbed (Fig. 2a–c). Maternal behavior was analyzed far from daily handling sessions, and it is possible that a more prolonged or aimed at data collection could show greater handling effects. This finding was not altogether unexpected as ultrasound calls were selectively increased only in knockout males following separation (Fig. 1) and handled litters contained pups of mixed sex and genotype. It has been reported that increased maternal care following handling depends on increased ultrasound calling of separated pups (D’Amato et al . 2005; Zimmerberg et al . 2003). However, a trend for an increase in maternal licking/grooming of pups was seen on the first three PNDs (Fig. 2b). To investigate whether this increase might derive from the presence of knockout males in the litter, we stratified data by whether or not the Genes, Brain and Behavior (2010) 9: 26–32


Corticosterone (ng/ml)

500

∗

non-handled handled

400 300 200 100 0

Baseline Stress

+/+

Baseline Stress

+/–

Baseline Stress

–/–

Figure 4: Stress response in Htr1aKO mice is reduced by handling. Serum corticosterone measured under either control conditions or after 15 min of restraint stress in wild-type (+/+), heterozygous (+/–) and knockout (–/–) littermates (non-handled: N = 70, handled: N = 43; * P < 0.05)

litter contained at least one knockout male pup. This analysis revealed a significant increase in licking/grooming of pups in litters with at least one male knockout (Fig. 2d) and supports the hypothesis that increased ultrasound calls underlie increased maternal care in handled mice. Although it was not technical feasible to directly quantify maternal care provided to individuals pups, we speculate that male knockouts pups received more maternal care than their wild-type littermates. When tested in adulthood, male Htr1aKO mice showed increased avoidance of the center and decreased total locomotion in an open field, confirming the increased innate anxiety-like behavior reported in these mice (Fig. 3a,b; Heisler et al . 1998; Parks et al . 1998; Ramboz et al . 1998 ). However, handling did not moderate innate anxiety in this test. This finding is consistent with earlier data showing that differences in maternal care did not moderate the anxiety phenotype of heterozygous Htr1aKO mice in the open field or elevated plus maze (Carola et al . 2006). In the social approach test, Htr1aKO mice showed significantly increased avoidance of the compartment containing the stimulus animal when compared to wild-type littermates (Fig. 3c). Wild-type mice showed a preference for the compartment containing the stimulus animal, while knockout mice did not. Unlike in the open field test, however, handling significantly reduced social avoidance among knockout mice in this test. A similar pattern emerged for anxiety-related behaviors, as measured by cumulative stretch attend and grooming (Fig. 3d) suggesting that social avoidance was due to increased anxiety rather than decreased interest in the stimulus. These data demonstrate that the avoidant phenotype of Htr1aKO mice extends to social stimuli. At the same time, the differential sensitivity of these phenotypes to handling argues that avoidance of social and non-social cues in these mice are likely to be controlled by different neural circuits. Finally, although there was only a trend for increased corticosterone response to restraint stress in knockout mice, handling selectively reduced corticosterone responses in knockout mice only (Fig. 4). Thus, again, knockout mice appeared particularly susceptible to the long-term effects of Genes, Brain and Behavior (2010) 9: 26–32

postnatal handling. However, handling did have a significant overall effect on corticosterone levels with the greatest decreases seen in unstressed mice. These findings are consistent with an earlier study examining the corticosterone response of Htr1aKO mice to open field exposure (Gross et al . 2000). In summary, our study demonstrates that Htr1aKO mice show increased anxiety-like behavior associated with separation from their mother and exposure to an unfamiliar mouse in adulthood. At the same time, our data show that postnatal handling reduces social anxiety-like behavior and stress hormone responses selectively in Htr1aKO mice. The apparent increase in sensitivity of Htr1aKO mice to handling could be explained by two mechanisms. First, Htr1aKO mice could be more sensitive because their anxiety is above a certain threshold required to engage the handling effect. According to this hypothesis wild-type mice of this strain are not sensitive to handling because their anxiety does not reach this threshold. Second, Htr1aKO mice could be more sensitive to handling because the absence of Htr1a facilitates the physiological impact of handling. This could happen if, for example, the set point for serotonin homeostasis (increased in the knockout due to the absence of the Htr1a autoreceptor) was critical to determine the outcome of handling. This hypothesis is supported by data in rats where maternal licking and grooming increase levels of serotonin in the developing cortex (Mitchell et al . 1990), and this increase has been proposed to mediate the long-term gene expression changes associated with increased maternal care (Meaney et al . 2000). Thus, our work raises the possibility that alterations in serotonin homeostasis might mask the effects of handling in at least some laboratory inbred strains (Millstein & Holmes 2007). Finally, our findings demonstrate that genetic variation can simultaneously influence the development of trait anxiety and its reprogramming by environmental adversity during the early postnatal period.

References Anisman, H., Zaharia M.D., Meaney, M.J. & Merali, Z. (1998) Do early-life events permanently alter behavioral and hormonal responses to stressors? Int J Dev Neurosci 16, 149–164. Brodkin, E.S., Hagemann, A., Nemetski, S.M. & Silver, L.M. (2004) Social approach-avoidance behavior of inbred mouse strains towards DBA/2 mice. Brain Res 1002, 151–157. Brunelli, S.A. (2005) Selective breeding for an infant phenotype: rat pup ultrasonic vocalization (USV). Behav Genet 35, 53–65. Burgdorf, J., Panksepp, J., Brudzynski S.M., Kroes, R. & Moskal, J.R. (2005) Breeding for 50-kHz positive affective vocalization in rats. Behav Genet 35, 67–72. Cabib, S., Puglisi-Allegra, S. & D’Amato, F.R. (1993) Effects of postnatal stress on dopamine mesolimbic system responses to aversive experiences in adult life. Brain Res 604, 232–239. Carola, V., D’Olimpio F., Brunamonti, E., Mangia, F. & Renzi, P. (2002) Evaluation of the elevated plus-maze and open-field tests for the assessment of anxiety-related behaviour in inbred mice. Behav Brain Res 134, 49–57. Carola, V., Frazzetto, G. & Gross, C. (2006) Identifying interactions between genes and early environment in the mouse. Genes Brain Behav 5, 189–199.

31

Zanettini et al. Costantini, F. & D’Amato, F.R. (2006) Ultrasonic vocalizations in mice and rats: social context and functions. Acta Zoolog Sin 52, 619–633. D’Amato, F.R. & Cabib, S. (1987) Chronic exposure to a novel odor increases pups’ vocalizations, maternal care, and alters dopaminergic functioning in developing mice. Behav Neural Biol 48, 197–205. D’Amato, F.R. & Cabib, S. (2000) Behavioral effects of manipulations of the olfactory environment in developing mice: Involvement of the dopaminergic system. In Oliverio, A. & PuglisiAllegra, S. (eds), Psychobiology of Stress. Kluwer Academic Publishers, Dordrecht, pp. 59–71. D’Amato, F.R., Cabib, S., Ventura, R. & Orsini, C. (1998) Long-term effects of post-natal manipulation on emotionality are prevented by maternal anxiolytic treatment in mice. Develop Psychobiol 32, 225–234. D’Amato, F.R., Scalera, E., Sarli, C. & Moles, A. (2005) Pups call, mothers rush: does maternal responsiveness affect the amount of ultrasonic vocalizations in mouse pups? Behav Genet 35, 103–112. (Erratum: (2006) Behav Genet 36, 471). Dichter, G.S., Brunelli, S.A. & Hofer, M.A. (1996) Elevated plus-maze behavior in adult offspring of selectively bred rats. Physiol Behav 60, 299–304. Gross, C., Santarelli, L., Brunner, D., Zhuang, X. & Hen, R. (2000) Altered fear circuits in 5-HT(1A) receptor KO mice. Biol Psychiatry 48, 1157–1163. Gross, C., Zhuang, X., Stark, K., Ramboz, S., Oosting, R., Kirby, L., Santarelli, L., Beck, S. & Hen, R. (2002) Serotonin1A receptor acts during development to establish normal anxiety-like behaviour in the adult. Nature 416, 396–400. Heisler, L.K., Chu, H.M., Brennan, T.J., Danao, J.A., Bajwa, P., Parsons, L.H. & Tecott, L.H. (1998) Elevated anxiety and antidepressant-like responses in serotonin 5-HT1A receptor mutant mice. Proc Natl Acad Sci USA 95, 15049–15054. Holmes, A., le Guisquet, A.M., Vogel, E., Millstein, R.A., Leman, S. & Belzung, C. (2005) Early life genetic, epigenetic and environmental factors shaping emotionality in rodents. Neurosci Biobehav Rev 29, 1335–1346. Klemenhagen, K.C., Gordon, J.A., David, D.J., Hen, R. & Gross, C.T. (2006) Increased fear response to contextual cues in mice lacking the 5-HT1A receptor. Neuropsychopharmacology 31, 101–111. Lo Iacono, L. & Gross, C. (2008) Alpha-Ca2+/calmodulin-dependent protein kinase II contributes to the developmental programming of anxiety in serotonin receptor 1A knock-out mice. J Neurosci 28, 6250–6257. Meaney, M.J., Diorio, J., Francis, D., Weaver, S., Yau, J., Chapman, K. & Seckl, J.R. (2000) Postnatal handling increases the expression of cAMP-inducible transcription factors in the rat hippocampus: the effects of thyroid hormones and serotonin. J Neurosci 20, 3926–3935. Millstein, R.A. & Holmes, A. (2007) Effects of repeated maternal separation on anxiety- and depression-related phenotypes in different mouse strains. Neurosci Biobehav Rev 31, 3–17.

32

Mitchell, J.B., Iny, L.J. & Meaney, M.J. (1990) The role of serotonin in the development and environmental regulation of type II corticosteroid receptor binding in rat hippocampus. Brain Res Dev Brain Res 55, 231–235. Moles, A., Sarli, C., Bartolomucci, A. & D’Amato, F.R. (2008) Interaction with stressed mothers affects corticosterone levels in pups after reunion and impairs the response to dexamethasone in adult mice. Psychoneuroendocrinology 33, 462–470. Parfitt, D.B., Levin, J.K., Saltstein, K.P., Klayman, A.S., Greer, L.M. & Helmreich, D.L. (2004) Differential early rearing environments can accentuate or attenuate the responses to stress in male C57BL/6 mice. Brain Res 1016, 111–118. Parfitt, D.B., Walton, J.R., Corriveau, E.A. & Helmreich, D.L. (2007) Early life stress effects on adult stress-induced corticosterone secretion and anxiety-like behavior in the C57BL/6 mouse are not as robust as initially thought. Horm Behav 52, 417–426. Parks, C.L., Robinson, P.S., Sibille, E., Shenk, T. & Toth, M. (1998) Increased anxiety of mice lacking the serotonin1A receptor. Proc Natl Acad Sci USA 95, 10734–10739. Pryce, C.R. & Feldon, J. (2003) Long-term neurobehavioral impact of the postnatal environment in rats: manipulations, effects and mediating mechanisms. Neurosci Biobehav Rev 27, 57–71. Ramboz, S., Oosting, R., Amara, D.A., Kung, H.F., Blier, P., Mendelsohn, M., Mann, J.J., Brunner, D. & Hen, R. (1998) Serotonin receptor 1A knockout: an animal model of anxietyrelated disorder. Proc Natl Acad Sci USA 95, 14476–14481. Tsetsenis, T., Ma, X.H., Lo Iacono, L., Beck, S.G. & Gross, C. (2007) Suppression of conditioning to ambiguous cues by pharmacogenetic inhibition of the dentate gyrus. Nat Neurosci 10, 896–902. Weller, A., Leguisamo, A.C., Towns, L., Ramboz, S., Bagiella, E., Hofer, M., Hen, R. & Brunner, D. (2003) Maternal effects in infant and adult phenotypes of 5HT1A and 5HT1B receptor knockout mice. Dev Psychobiol 42, 194–205. Winslow, J.T., Hearn, E.F., Ferguson, J., Young, L.J., Matzuk, M.M. & Insel, T.R. (2000) Infant vocalization, adult aggression, and fear behavior of an oxytocin null mutant mouse. Horm Behav 37, 145–155. Zimmerberg, B., Kim, J.H., Davidson, A.N. & Rosenthal, A.J. (2003) Early deprivation alters the vocalization behavior of neonates directing maternal attention in a rat model of child neglect. Ann NY Acad Sci 1008, 308–313.

Acknowledgments F.D., A.M. and C.Z. were supported by Regione Lazio funds for ‘‘Sviluppo della Ricerca sul Cervello’’ and by Telethon, Italy (Grant no. GGP05220). C.G., V.C. and L.L. were supported by funds from EMBL, NARSAD and the Fritz Thyssen Stiftung.
