- Email: [email protected]
Behavioral and hormonal effects of partner familiarity in periadolescent rat pairs upon novelty exposure
Psychoneuroendocrinology 24 (1999) 639 – 656
Behavioral and hormonal effects of partner familiarity in periadolescent rat pairs upon novelty exposure M. Livia Terranova a, Francesca Cirulli b,*, Giovanni Laviola b a
Department of Psychology, Uni6ersita` degli Studi di Roma ‘La Sapienza’, 6ia dei Marsi 78, I-00185 Rome, Italy b Section of Beha6ioural Pathophysiology, Labor. F. O. S., Istituto Superiore di Sanita`, 6iale Regina Elena 299, I-00161 Rome, Italy Received 19 March 1998; accepted 28 January 1999
Abstract In periadolescent rats, social interactions are typically characterized by elevated levels of playful and affiliative behavior. Aim of the present study was to assess the behavioral and hormonal effects of partner familiarity upon the separation and reunion in a novel environment of established pairs of periadolescent subjects. At weaning (post-natal day, PND 21), Sprague–Dawley rats were pair housed with a non-sibling subject of the same age and sex. On PND 35, the members of each pair were separated for a 24-h period, and randomly assigned to different experimental groups: (1) sacrificed before separation; (2) sacrificed immediately after the isolation period; (3– 4) placed individually in a novel cage for 30 min either in low-light or in high-light conditions; (5 – 6) reunited for 30 min in a novel cage either with their previous cagemate (familiar, FAM); or (7 – 8) with an unfamiliar rat (UNF) of the same age and sex, in either light conditions. During reunion, the occurrence of social and non-social behaviors was scored. Blood samples were collected at the end of the session from all groups and assayed for corticosterone (CORT). The separation of the two members of an established pair did not affect baseline CORT levels. Upon reunion, the presence of a conspecific exerted a significant buffering effect on the novelty-induced increase in CORT levels. Such an effect of the social companion appeared more marked in males than in females, and in FAM compared to UNF pairs. Interestingly, FAM rats also expressed a
* Corresponding author. Tel.: +39-06-4990-2105; fax: + 39-06-4957-821. E-mail address: [email protected] (F. Cirulli) 0306-4530/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 3 0 6 - 4 5 3 0 ( 9 9 ) 0 0 0 1 9 - 0
640
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
significantly higher amount of social investigation and play-soliciting behavior compared to UNF animals. Behavioral results, together with previous data, suggest that periadolescent rats housed in established pairs develop a sort of amicable relationship. The overall CORT output measured at the end of the session is also in line with this interpretation. As a whole, these findings indicate that periadolescence is a time period during rat development, during which social variables play a very important role in modulating both behavioral and physiological responses to novelty in a fashion that does not completely overlap with data on adult subjects. © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Periadolescence; Social interactions; Partner familiarity; Environmental novelty; Corticosterone; Rat
1. Introduction During periadolescence — which has been defined as the ontogenetic period that encompasses in the rat the 7 – 10 days preceding the onset of puberty (at about 40 days of age) and the first few days thereafter (Spear and Brake, 1983)—rats and mice differ markedly from both younger and older conspecifics in several aspects of behavior and physiology (Spear and Brake, 1983; File, 1986; Witt, 1994; Laviola et al., 1995; Adriani et al., 1998; Terranova et al., 1998). In particular, animals around this age are characterized by elevated levels of affiliative and playful social interactions among peers, which mainly consist of rough-and-tumble play (Meaney and Stewart, 1981; Panksepp et al., 1984; Thor and Holloway, 1984, 1986; see also Terranova et al., 1993 for a comparison with the mouse). However, the specific characteristics of the social relationship underlying such a high tendency to socialize have not yet been clarified. In this frame, surprisingly few and contrasting results are available on the effects of partner familiarity on rats social interactions around periadolescence. On the one hand, Latane´ et al. (1971) could not find any difference between familiar (FAM) and unfamiliar (UNF) animals in measures of social proximity. On the other hand, both Barefoot et al. (1975) and Monroe and Milner (1977) found that partner’s novelty increased juvenile social interactions (see also Thor et al., 1988). Furthermore, while the importance of social variables in affecting the response to stressful stimuli has been characterized in adult rodents, few data are available in young subjects. A previous work by our group (Cirulli et al., 1996), designed to address such an issue, was the first, to our knowledge, to assess, along with the behavioral outcomes, also the adrenocortical response to both separation and reunion of periadolescent rat pairs with different degrees of familiarity. From a behavioral point of view, the above study showed that an elevated amount of investigative and playful behaviors was exhibited by UNF subjects. Such a finding is probably amenable to a ‘curiosity’ factor related to the attraction properties of novel conspecifics (Barefoot et al. 1975; Monroe and Milner, 1977; File and Peet, 1980; Armario et al., 1983a,b; Thor et al., 1988; File, 1990). On the other hand, FAM rats were found to have established some kind of a preferential relationship. These
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
641
animals actually expressed a higher amount of proximity-maintaining affiliative interactions which are reminiscent of those characterizing attachment-like bonds (for such kind of relationships in primates and in monogamous rodents such as the prairie vole, see Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al., 1990; Williams et al., 1994). However, hormonal results strongly suggested the absence both of a buffering effect of the interactions between UNF subjects towards novelty-induced neuroendocrine activation, and of an attachment-like social relationship between FAM animals. In fact, the corticosterone (CORT) levels that were measured in the two members of each pair at the end of the social encounter were found to be totally unaffected by partner familiarity. However, it should be taken into account that in the previous study (Cirulli et al., 1996) social interactions during reunion took place in a cage of a different size from the home cage. We need to consider that exposure to novelty is one of the most powerful stimuli capable of activating the hypothalamic-pituitary-adrenal (HPA) system (e.g. to a novel testing apparatus; see Hennessy and Levine, 1977, 1978; File and Vellucci, 1979; File and Peet, 1980; File, 1990). Moreover, familiarity with the environment is known to be one of the critical determinants of gregariousness among laboratory rats. Both fear and curiosity about novel environments have been shown to compete with social attraction, apparently representing a stronger drive state than the social one (File and Hyde, 1978; File and Vellucci, 1979; File and Peet, 1980; Armario et al., 1983a,b; File, 1986; Lister and Hilakivi, 1988; Thor et al., 1988; File, 1990). Since, in turn, social variables, such as the availability of a FAM or UNF conspecific, have been shown to affect the response of the HPA system to stressful stimuli (Armario et al., 1983a,b; Hennessy, 1986; Levine, 1993; Hennessy, 1997; see discussion in Cirulli et al., 1996), it is reasonable to hypothesize that, in our previous experiment, the novelty of the test cage might have affected both hormonal and behavioral responses. Specifically, the strong novelty-induced activation mi0ght have resulted in a ‘ceiling effect’ (see File and Peet, 1980), with the high values reached by CORT secretion masking the potential differences between the FAM and the UNF groups. Therefore, in order to at least partially reduce the novelty of the testing environment, in the present study rats were placed in a cage of the same type and size as the home-cage. Moreover, with the aim of separating the effects of partner familiarity on CORT secretion from those of environmental novelty, an additional control group of animals (NOV group), placed individually in the novel cage, was included. We also chose to manipulate the illumination level in the experimental room so as to expose the animals to a dimmer light compared to the previous experiment. In fact, exposure of nocturnal rodents, such as the rat, to bright light has been found to be another important stress-inducing stimulus from both a hormonal and a behavioral standpoint (File and Hyde, 1978; File and Peet, 1980; Lister and Hilakivi, 1988). In summary, since previous work with adult rodents and primates has indicated an important role of social variables on both behavioral and neuroendocrine responses to stressful stimuli, the present study was aimed at investigating this issue in a developmental phase (periadolescence) when rats show a particularly strong
642
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
tendency to be social. Partner familiarity was the specific variable manipulated upon the separation and reunion of established pairs, with the additional aim of evaluating whether periadolescents high sociality is amenable either to an aspecific social attraction among peers or to the presence of specific attachment-like relationships. Since the elevation of CORT levels upon separation is usually viewed as an important marker of attachment bonds (Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al., 1990; Williams et al., 1994; Hennessy, 1997), data on CORT levels were also collected at the end of the 24-h separation period.
2. Methods
2.1. Animals, breeding and rearing conditions Sprague – Dawley pregnant rats were obtained from Charles River Italia (Calco, Italy). Upon arrival, animals were housed separately in 40× 25×20 cm Plexiglas boxes, with sawdust as bedding and a metal top, in an air-conditioned room at 2191°C (relative humidity: 50 910%). They were maintained in a 12:12 h light (35.00 scotopic lux)/dark (0.02 scotopic lux) cycle, with lights on at 0930 h. Food (enriched standard diet purchased from Piccioni, I-25100, Brescia, Italy) and water were freely available. All animal handling and experimental procedures were performed in accordance with the EEC guidelines (EEC Council Directive 86/609, 1987) and with the Italian legislation (Decreto L.vo 116/92) on animal experimentation. From weaning (PND 21) until the time of testing, each pup was kept together with a non-sibling, same-age and same-sex conspecific, in a 40× 25× 20 cm Plexiglas box. On PND 34, all subjects were weighed and rehoused individually with their own sawdust in identical cages for a 24-h period. In both sexes, each rat was randomly assigned to one of eight different experimental conditions:NT (non-treated): animals blood sampled before separation on PND 34 (N= 8 subjects);24-h ISOL: animals blood sampled following the 24-h period of isolation, i.e. on PND 35 (N =8 subjects);NOV-LL (novelty-low light): following the 24-h isolation, subjects were put individually for 30 min in a cage identical to the home cage with new sawdust as bedding, under low light conditions (28.00 scotopic lux), and blood sampled at the end of the session (N= 8 subjects);NOV-HL (noveltyhigh light): same as above but under high light conditions (65.00 scotopic lux) (N= 8 subjects);FAM-LL (familiar-low light): the two members of a pair were reunited following the 24-h period of separation for 30 min in a cage identical to the home cage with new sawdust as bedding, under LL conditions, and blood sampled at the end of the session (N = 7 pairs);FAM-HL (familiar-high light): same as above but under HL conditions (N= 8 pairs);UNF-LL (unfamiliar-low light): following the 24-h period of isolation, two subjects from different pairs were placed together for 30 min in a cage identical to the home cage with new sawdust as bedding, under LL conditions, and blood sampled at the end of the session (N= 8
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
643
pairs);UNF-HL (unfamiliar-high light): same as above but under HL conditions (N =8 pairs). Age of testing was chosen taking into account literature on periadolescent rats reporting that the most playful and intense social interactions are shown by animals around PND 35 (see Section 1).
2.2. Blood samples Animals were killed by decapitation and trunk blood collected in heparinized tubes. Blood samples were centrifuged at 2000 rpm for 20 min to obtain cell-free plasma, and then frozen at − 80°C until they were assayed for CORT with a radioimmunoassay kit (ICN Biochemicals, Costa Mesa, CA; sensitivity9 0.125 mg/dl). Inter- and intra-assay coefficients of variation were 18 and 11.5%, respectively. Blood was collected from ten of the 16 subjects belonging to the eight couples in each treatment group.
2.3. Beha6ioral obser6ations Behavioral testing took place in an experimental room maintained in the same temperature and humidity conditions as the housing room, from 1000 to 1400 h, and the time of testing was counterbalanced between experimental groups. The behavior of each pair was videorecorded by means of two videocameras connected to a professional Sony videocassette recorders V0-5800PS. A software system for collection and analysis of observational data was used for scoring duration and frequency of each response. Only the last 10 min of the 30 min sessions were scored and statistically analyzed, since preliminary observations carried out in three sampled pairs of subjects for each experimental group confirmed that, as found in a previous work (see Cirulli et al., 1996), no behavioral differences were detectable during the first portion of the social encounters. A ‘focal animal—all occurrences’ sampling method was used (see Altmann, 1974), with each animal being continuously observed for randomly-chosen 5 min. Since the behavior of one of the two subjects in a pair cannot be considered independent from that of the other, pair means were used for further analysis. The social and non-social behavioral categories listed below, and their classification, are mainly based upon the ethological profiles of rat behavior described by Poole and Fish (1976), Meaney and Stewart (1981), Panksepp et al. (1984), Thor and Holloway (1984) and Thor and Holloway (1986) (see also Terranova et al., 1993, 1994; Terranova and Laviola, 1995; Cirulli et al., 1996). Unless otherwise specified, both frequency (f) and duration (d) were measured for each behavioral item.
2.4. Social acti6ities
Investigative elements: (1) Social sniff: self-explanatory (s.e.); (2) Follow: s.e.; (3) Mutual circle: partners are mutually sniffing each other’s anogenital region, while describing tight circles with their reciprocal following movements.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
644
Affiliative elements: (1) Social inacti6e: a rat is lying flat or standing still (eyes closed or open) while maintaining close physical contact with the partner; (2) Social rest: is being groomed by the partner; (3) Allogroom: s.e. Play-soliciting: (1) Crawl o6er (f): crawls over the partner’s body, crossing it transversely from one side to the other. Rough-and-tumble play: The two rats exhibit sudden darting movements, which are associated with pouncing on the partner back, chasing, wrestling and pinning (the so-called ‘on-top’ and ‘on-back’ postures). The play bout is often anticipated by a prolonged and intense bout of allogrooming, often accompanied by gross movements of the whole body of the performer and by vigorous pulling of the fur of the partner. The bout is considered concluded when both participants either turn their attention away from each other, or shift to amicable non-playful interactions (e.g. sniffing).
2.5. Non-social acti6ities
Activity and exploration/escape behaviors: (1) Inacti6e: s.e.; (2) Explore: s.e.; (3) Dig: is digging in the sawdust, pushing and kicking it around using the snout and/or both the forepaws and hindpaws, mostly moving around the cage and sometimes changing the whole arrangement of the substrate material. Maintenance activities: (1) Selfgroom: s.e.
2.6. Design and statistical analysis Behavioral data were analysed by means of a 2× 2× 2 ANOVA, with familiarity (FAM, UNF), light (LL, HL) and sex (male, female) as between-subject factors. Mean comparisons were carried out by means of the Superanova package (version 1.1, 1989 – 90, Abacus Concepts) wherever significant interaction effects were detected by the ANOVA. See below for corticosterone data analysis.
3. Results
3.1. Corticosterone Corticosterone data (see Fig. 1) were first analyzed by means of an 8 (treatment: NT, 24-h ISOL, NOV-LL, NOV-HL, FAM-LL, FAM-HL, UNF-LL, UNFHL)×2 (sex: male, female) ANOVA. Post-hoc comparisons were performed using the Tukey’s HSD test. Corticosterone levels differed significantly among the different treatment groups (treatment: F[1,7]= 13,48, PB .001). In particular, hormonal levels at the end of the 24-h of social isolation did not differ from pre-separation levels and were significantly lower than those found at the end of the 30 min reunion, independently from the degree of familiarity and the light level (P B.01). The ANOVA revealed a main effect of sex (F[1,7]= 8.79, PB .01), with females showing overall higher corticosterone levels than males. A second analysis
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
645
was thus performed for each sex separately to prevent small but consistent differences in the male group from being masked by female values. To evaluate simultaneously the effects of the presence/familiarity of another subject and those of different light conditions in the same analysis, we performed for each sex a 3 (treatment: NOV, FAM, UNF)× 2 (light: LL, HL) factorial ANOVA, excluding the NT and 24-h ISOL groups. In males, a main effect of treatment was found (F[1,2]=11.06, P B.0001], and post hoc comparisons indicated that, when exposed to novelty in pairs, they showed CORT levels that were significantly lower than those reached when placed alone in the test cage, irrespectively of the degree of partner familiarity (NOV vs. UNF: P B .01; NOV vs. FAM: PB .01). A significant interaction between treatment and light (F[2,50]= 3.79, PB .05) also appeared, and post hoc comparisons further clarified that such a buffering effect of the social companion was only evident when animals where tested in LL conditions (NOV-LL vs. FAM-LL: P B .01; NOV-LL vs. UNF-LL: PB .01). In addition, as it can be seen in Fig. 1, there was a tendency for CORT levels in the FAM group to be lower than those in the UNF one.
Fig. 1. Mean (9 SEM) corticosterone levels of male periadolescent rats in the different experimental conditions. NT, subjects were sampled before the isolation period; ISOL, subjects were isolated for 24 h and blood sampled at the end of the isolation period; NOV, subjects were placed in a test cage alone for 30 min after 24-h of isolation and sampled at the end of this time period; UNF, subjects were reunited for 30 min with an unfamiliar conspecific after 24-h of isolation and sampled at the end of the reunion; FAM, subjects were reunited for 30 min with their previous cagemate after 24-h of isolation and sampled at the end of the reunion (N =8 subjects in the NT, ISOL and NOV groups, N =10 subjects in the FAM and UNF groups). **NOV-LL vs. UNF-LL and NOV-LL vs. FAM-LL, P B .01.
646
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Fig. 2. Mean duration ( 9 SEM) of non-social behaviors recorded during the last 10 min of the 30 min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N = 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples).
As for females (see Table 1), light affected CORT secretion significantly (light: F[1,2]= 4.29, P B.05), with overall higher levels in the LL than in the HL condition. Although no significant effects of treatment were found, female CORT profile was rather similar to that of males.
3.2. Social and non-social beha6ior The ANOVA did not yield significant results for explore, dig, sniff and roughand-tumble play behaviors. When both frequency and duration were scored, the results for the two measures were consistent with each other and, for the sake of space, only the results regarding the latter are reported.
NT
1.1 (9 0.1) a
ISOL
2.4 (90.5)
NOV
UNF
FAM
LL
HL
LL
HL
LL
HL
28.2 ( 9 6.1)
19.1 (95.9)
26.7 (9 4.1)
25.3 (9 3.1)
23.8 (9 5.5)
10.6 (9 3.1)
NT, subjects were sampled before the isolation period; ISOL, subjects were isolated for 24 h and blood sampled at the end of the isolation period; NOV, subjects were placed in a test cage alone for 30 min after 24 h of isolation and sampled at the end of this time period; UNF, subjects were reunited for 30 min with an unfamiliar conspecific after 24-h of isolation and sampled at the end of the reunion; FAM, subjects were reunited for 30 min with their previous cagemate after 24-h of isolation and sampled at the end of the reunion (N =8 subjects in the NT, ISOL and NOV groups, N = 10 subjects in the FAM and UNF groups).
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Table 1 Mean ( 9 SEM) corticosterone levels (mg/dl) of female periadolescent rats in the different experimental conditionsa
647
648
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
In general, periadolescent rats allocated longer portions of their time to inacti6e (Fig. 2) and social inacti6e (an affiliative behavior, see Fig. 3) when tested in HL compared to LL conditions, while exhibiting significantly less mutual circle (an item of social investigation; Fig. 4) (light: F[1,55]= 3.93, 6.12, 12.29, PB .05 or less, respectively). Females tended to be more involved than males in the affiliative social rest behavior (32.502 ( 94.121) s and 22.954 ( 9 3.321) s, respectively; sex: F[1,55]= 3.52, P=.06), and, within the FAM group, expressed a higher amount of mutual circle (Fig. 4; sex× familiarity: F[1,55]= 7.39, PB .01; means comparison= P B .01). In addition, as shown in Fig. 3, females showed a tendency towards a decreased duration of allogroom affiliative behavior when tested in HL compared to LL conditions (Fig. 4; sex× light: F[1,55]= 3.66, P= .06).
Fig. 3. Mean duration ( 9SEM) of affiliative social behaviors recorded the last 10 min of the 30 min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N= 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples). Data refer to the same animals of Figs. 1 and 2.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
649
Fig. 4. Mean duration ( 9 SEM) of investigative social behaviors recorded during the last 10 min of the 30-min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N= 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples). Data refer to the same animals of Fig. 1.
FAM rats appeared to be more involved than UNF animals in the social investigative behaviors follow and mutual circle (Fig. 4) and in the play-soliciting item crawl o6er (FAM: 0.84690.211; UNF: 0.3929 0.085) (familiarity: F[1,55]= 3.49, 6.85, 3.79, P = .06, P B.01 and .05, respectively). The effects of partner familiarity were particularly marked within the female group, with FAM rats spending less time staying inacti6e (Fig. 2) and much more in mutual circle (Fig. 4) than UNF animals (sex×familiarity: F[1,55]= 5.20, 7.39, PB .05 or less, respectively; means comparisons= P = .06 and PB .001, respectively). Light conditions also interacted with partner familiarity, in that, compared to UNF rats, FAM subjects tended to show an increased duration of both mutual circle (Fig. 4) and allogroom (Fig. 3; means comparison= PB .01) behaviors when
650
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
tested in LL but not in HL conditions (light× familiarity: F[1,55]=3.65, 7.95, P = .06 and P B .01, respectively). Moreover, compared to UNF animals whose behavior was apparently not affected by light conditions, FAM rats showed a greater involvement in allogroom (means comparison= PB .01) and, as a tendency, in mutual circle when tested in LL compared to HL conditions (light× familiarity: F[1,55]=7.95, 3.65, P B.01 and P = .06, respectively). Among FAM animals, some interesting interactions between the sex and the light variables were also detected. As shown in Figs. 2 and 4, FAM females tended to devote longer periods of time to selfgroom, and spent significantly shorter periods of time in mutual circle, when tested in HL compared to LL conditions (sex × light×familiarity: F[1,55] =3.45, 3.55, P= .06 for both items), whereas in the same conditions FAM males significantly decreased their crawl o6er frequency (LL: 1.56290.521; HL: 0.5009 0.231; sex× light× familiarity: F[1,55]= 4.28, PB .04; means comparison= P B.01). Interestingly, within the male group tested in LL conditions, crawl o6er was exhibited significantly more often by FAM than by UNF rats (FAM: 1.5629 0.521; UNF: 0.3649 0.152; sex× light× familiarity: F[1,55]= 4.28, PB .04; means comparison= P B .001). Finally, the sex variable also seemed to affect the behavioral profile within the FAM group tested in LL conditions. In such a group, females tended to overcome males as for the duration of mutual circle behavior (Fig. 4), whereas the opposite was found for the frequency of crawl o6er (males: 1.562 9 0.521; females: 0.500 9 0.289; means comparisons=P B.04) and for selfgroom duration (Fig. 2) (sex× light× familiarity: F[1,55] = 3.55, 4.28, 3.45, P= .06, .04 and .06, respectively).
4. Discussion The main findings of the present study can be summarized as follows: 1. A 24-h period of separation of the members of periadolescent rat pairs did not affect baseline corticosterone (CORT) levels. 2. As expected, the exposure of animals individually (NOV group) to environmental novelty produced a significant increase in CORT levels. 3. Male subjects reunited with a conspecific following separation, and thus confronted with novelty in pairs, showed a significantly lower increase in CORT levels compared to NOV males. 4. There was only a tendency for such a buffering effect of the social companion on CORT secretion to be more marked in familiar (FAM) compared to unfamiliar (UNF) pairs. 5. FAM rats also expressed a significantly higher amount of social investigation and play-soliciting behavior compared to UNF animals, and this was particularly evident in the female group. A strong activation of the HPA axis upon separation is usually considered as a main indicator of an attachment relationship, such as that characterizing mother– infant dyads in non-human primates or heterosexual pairs in monogamous rodents (Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
651
al., 1985; Levine et al., 1989; Carter et al., 1990; Williams et al., 1994; Hennessy, 1997). In studies investigating gregariousness or affiliative peer interactions, it has been reported that adult rats do not show changes in CORT upon separation from social companions (File and Peet, 1980; Armario et al., 1983a,b; File, 1990). In this experiment we investigated whether this might be true in periadolescent rats, which are specifically characterized by high levels of social behavior (see Section 1). No hormonal changes were found following social separation of the two members of an established pair, a result which is apparently in line with the notion that the high levels of social behavior characterizing periadolescent rats are not amenable to specific attachment-like bonds. Moreover, when male subjects were confronted in pairs with a novel environment—which was per se responsible for a marked CORT elevation — a buffering effect of the social companion on CORT secretion was to some extent observed independently from the degree of partner familiarity. Thus, our data seem to confirm, although indirectly, the hypothesis that periadolescent rats high tendency to sociality is mainly the result of an aspecific social attraction among peers (see also Cirulli et al., 1996). The finding of a buffering effect of the social companion on CORT secretion has to be viewed in the frame of previous studies, carried out in both rats and non-human primates, showing the major role played by social variables in modulating the response of the HPA system to stress (Armario et al., 1983a,b; Hennessy, 1986; Levine et al., 1989; Levine, 1993; Hennessy, 1997; see discussion in Cirulli et al., 1996). Studies conducted in primates have shown that social variables can modulate HPA activity, both in terms of increased hormone release and/or the inhibition of endocrine responses to environmental perturbations (Levine et al., 1989; Levine, 1993). For example, the presence of a social group can ameliorate the neuroendocrine response to aversive stimuli that normally evoke a marked adrenal response in monkeys housed alone, such as a snake or a snake-like object (Mineka et al., 1980; Vogt et al., 1981). An unfamiliar conspecific can also be effective in reducing the pituitary-adrenal activation in response to stressful stimuli (Armario et al., 1983a,b; Hennessy, 1986). Nonetheless, a fine-grain behavioral analysis confirmed previous data (Cirulli et al., 1996) showing that the degree of familiarity affects the pattern of social interactions of periadolescent rat pairs. Although differences were not detected in the expression of affiliative behaviors, FAM animals were significantly more involved than UNF subjects in both social investigation and play-soliciting interactions. Such a finding does not completely agree with literature data on both adult (Armario et al., 1983a,b) and adolescent rats (Barefoot et al. 1975; Monroe and Milner, 1977; Thor et al., 1988; Cirulli et al., 1996), which indicate a soliciting effect of partner novelty on active social interactions, and associate such attraction properties of novel conspecifics to a ‘curiosity’ factor (see also File and Peet, 1980; Hennessy, 1986; File, 1990). It also remains to be ascertained why in the present study, FAM rats expressed elevated amounts of investigative and play-soliciting behaviors, instead of devoting increasing amount of time to affiliative interactions, as it would be expected on the basis of both the attachment-related literature (Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al.,
652
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
1990; Williams et al., 1994) and our previous results (Cirulli et al., 1996). Anyhow, the present data seem to indicate that periadolescent rats do recognize and interact preferentially with FAM rather than with UNF conspecifics. Although neuroendocrine responses were not significantly affected by the degree of familiarity of the two partners, they suggested that, consistently with the above behavioral results, some amicable relationship develops and can be recalled by FAM periadolescent rats. Whereas the previous study (Cirulli et al., 1996) found similar levels of CORT secretion upon social reunion, the present data provide some indication that FAM subjects might exert a greater buffering effect than UNF animals on novelty-induced CORT release. The possibility that a social variable such as partner familiarity might exert in the rat an influence on important biological processes such as CORT secretion is worth further experimentation. Nonetheless, it is important to stress here that a direct correlation between hormonal and behavioral data cannot be drawn since CORT data represent the overall hormonal output of the session, while behavioral data only refer to the last 10 min of the observation period. Overall, these findings suggest some interesting differences in physiological and behavioral regulation between periadolescent and adult rats. When compared to young animals, adults are reported to show a different response pattern, namely a higher amount of social interactions and lower CORT levels when placed with UNF compared to FAM conspecifics (Armario et al., 1983a,b). When pair-housed in the same home cage early on during development rather than in adulthood, rats may experience a different quality of social interactions, which in turn might well influence later recognition in the test box. Moreover, it seems quite likely that both the behavioral and the hormonal outcomes of affiliative social experiences are more readily seen at a time, such as periadolescence, when amicable social behavior is expressed at high levels. One must also take into account that different experimental paradigms could be responsible for the differences between the data presented in this paper and previous reports on adult subjects. Also the discrepancy between the pattern of results of the present and the previous (Cirulli et al., 1996) study, both carried out in periadolescent rats, should be attributed to differences in methodology. In fact, in this experiment, environmental novelty was reduced by placing animals in a test cage of the same type and size as the home cage, whereas in the previous study the two cages differed in size and shape. However, the present study seems to be in agreement with the previous one (Cirulli et al., 1996; see also Shors and Wood, 1995) in that females appeared to be more sensitive than males to partner familiarity. Sex differences in exploratory and emotional behavior have been documented in a variety of mammals including man (Archer, 1975; Johnston and File, 1991), and, more specifically, sex has been shown to interact with stimulus complexity (e.g. Russel, 1977). In general, females show a greater preference for novelty than males, and thus are more inclined to explore novel environments. In addition, as for the physiological response to exogenous stressors, females have been reported to exhibit higher basal levels of glucocorticoids, as well as a greater range of CORT response to a stressor (Kitay, 1961; Critchlow et al., 1963), and such a functional dimorphism has been implicated in the regulation of various behaviors.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
653
As for the manipulation of the illumination level of the testing environment, which was here accomplished by comparing a high light (HL) condition with the condition of low light (LL) uniformly used in the previous study (Cirulli et al., 1996), results are not straightforward. As expected, a decrease in immobility and an increase in the expression of social behavior in LL was observed, most likely resulting from the reduction of the illuminance-related stress and of the associated behavioral inhibition (see e.g. File and Hyde, 1978). Surprisingly, from a hormonal standpoint, the LL condition did not result in lower cort levels as hypothesized on the basis of few but consistent literature reports (File and Vellucci, 1979; File and Peet, 1980), especially in the NOV and UNF groups. However, CORT levels in the two lighting conditions were differentially affected by the social setting. Under LL, as we pass through different conditions, CORT levels appear to mirror different arousal levels, being higher in NOV, reaching intermediate levels in the presence of an UNF conspecific, while tending to diminish in the presence of a familiar subject. Instead, in the HL condition, CORT levels were more resilient to change as a function of the presence of another subject, being it UNF or FAM. Three hypotheses can be put forward to explain the ineffectiveness of high light to arouse the experimental subjects above the LL levels. From a behavioral point of view, although we have no data on the NOV group, we found that in LL conditions the behavior of UNF and FAM subjects was less inhibited, and this, in turn, could well have influenced CORT levels in a direction opposite to that expected. A possible additional explanation is that, since there is a strong relationship between increasing stimulus intensity and magnitude of ACTH and CORT secretion, subjects exposed to a stronger stimulus, such as the HL condition, might have reached a secretory peak earlier compared to LL animals (see Hennessy and Levine, 1978; Dallman et al., 1987). Thus, at the time of testing, i.e. at the end of a 30 min exposure to the different experimental conditions (in line with our previous study, see Cirulli et al., 1996), CORT levels of subjects exposed to HL condition might have already reflected the pattern of return to resting levels, while in the LL subjects CORT secretion was still reaching its peak, thus resulting higher than the other group. Running a time course of the CORT response to the different experimental conditions could reveal whether this might indeed be true. The intriguing possibility finally exists that such a complex pattern of CORT response to lighting might be due to ontogenetic factors underlying the development of neuroendocrine mechanisms of anxiety. Appropriate behavioral and neuroendocrine responses shown by adult rats are in part due to changes occurring during puberty (see Primus and Kellogg, 1989). Guillet and Kellogg (1988) reported that the CORT response to the stress of a novel environment matures in the rat as early as 28 days, i.e. during the prepubertal phase. However, the behavioral responses to these anxiogenic conditions, as assessed by means of the social interaction test, appear slightly later, i.e. as rats enter puberty (PND 30-32; see also Primus and Kellogg, 1989). Thus, hormonal and behavioral responses seem to be dissociated in their developmental onset. The differential effect exerted in the present experiment by light intensity on behavioral and hormonal measures might be another example of a separate emergence of these two responses to novelty during periadolescence.
654
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
It can be concluded that periadolescence is a peculiar time period during rat development. During this period, social variables appear to play a very important role in modulating both behavioral and physiological responses to novelty, in a fashion that does not completely overlap with data on adult subjects.
Acknowledgements This research was supported as part of the Nervous and Mental Disorders Research Project on ‘Psychobiological risk or protection factors for behavioural disorders and vulnerability to recreational substances abuse during development’ intramural grant to G.L., Instituto Superiore di Sanita`, Roma, Italy, and by the Ministry of Environment Project, PR22/IS, A79/L2, grant ‘Animal models of neurobehavioural risk following exposure to gaseous pollutants’. We acknowledge the help of Dr F. Chiarotti in the statistical analysis of the data and Angelina Valanzano for expert technical suggestions and help.
References Adriani, W., Chiarotti, F., Laviola, G., 1998. Elevated novelty seeking and peculiar d-amphetamine sensitization in periadolescent mice. Behav. Neurosci. 112, 1152 – 1166. Altmann, J., 1974. Observational study of behavior: sampling methods. Behavior 49, 227 – 265. Archer, J., 1975. Rodent sex differences in emotional and related behavior. Behav. Biol. 14, 451 – 479. Armario, A., Luna, G., Balasch, J., 1983a. The effect of conspecifics on corticoadrenal response of rats to a novel environment. Behav. Neural Biol. 37, 332 – 337. Armario, A., Ortiz, R., Balasch, J., 1983b. Corticoadrenal and behavioral response to open field in pairs of male rats either familiar or non-familiar to each other. Experientia 39, 1316 – 1317. Barefoot, J.C., Aspey, W.P., Olson, J.M., 1975. Effect of partner novelty on affiliation in the rat. Bull. Psychonomic Soc. 6, 655–657. Bowlby, J., 1975. Attachment theory, separation anxiety, and mourning. In: Hamburg, D.A., Brodie, K.H. (Eds.), American Handbook of Psychiatry. New Psychiatric Frontiers. Basic Books, New York, pp. 292–309. Carter, C.S., Witt, D.M., Thompson, E.G., Carlstead, K., 1990. Effects of hormonal, sexual, and social history on mating and pair bonding in prairie voles. Physiol. Behav. 44, 691 – 697. Cirulli, F., Terranova, M.L., Laviola, G., 1996. Affiliation in periadolescent rats: behavioral and corticosterone response to social reunion with familiar or unfamiliar partners. Pharmacol. Biochem. Behav. 54, 99–105. Coe, C.L., Wiener, S.G., Rosenberg, L.T., Levine, S., 1985. Physiological consequences of maternal separation and loss in the squirrel monkey. In: Rosenblum, L.A., Coe, C.L. (Eds.), Handbook of Squirrel Monkey Research. Plenum, New York, pp. 127 – 148. Critchlow, V., Liebelt, R.A., Bar-Sela, M., Mountcastle, W., Lipscomb, A.S., 1963. Sex differences in resting pituitary-adrenal function in the rat. Am. J. Physiol. 205, 807 – 815. Dallman, M.F., Akana, S.F., Cascio, C.S., Darlington, D.N., Jacobson, L., Levin, N., 1987. Regulation of ACTH secretion: variations on a theme of B. Recent Prog. Horm. Res. 43, 113 – 173. File, S.E., 1986. The effects of neonatal administration of clonazepam on passive avoidance and on social, aggressive and exploratory behavior of adolescent male rats. Neurobehav. Toxicol. Teratol. 8, 447 – 452. File, S.E., 1990. Interactions of anxiolytic and antidepressant drugs with hormones of the hypothalamicpituitary-adrenal axis. Pharmacol. Ther. 46, 357 – 375.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
655
File, S.E., Hyde, J.R.G., 1978. Can social interaction be used to measure anxiety? Br. J. Pharmacol. 62, 19 – 24. File, S.E., Vellucci, S.V., 1979. Behavioral and biochemical measures of stress in hooded rats from different sources. Physiol. Behav. 22, 31–35. File, S.E., Peet, L.A., 1980. The sensitivity of the rat corticosterone response to environmental manipulations and to chronic chlordiazepoxide treatment. Physiol. Behav. 25, 753 – 758. Guillet, R., Kellogg, C.K., 1988. Age-related effects of handling, diazepam exposure, and gonadectomy on corticosterone response. Soc. Neurosci. Abstracts 14, 112.5. Harlow, H.F., Harlow, M.K., 1965. The affectional system. In: Schrier, A., Harlow, H., Stollnitz, F. (Eds.), Behavior of Nonhuman Primates, vol. 2. Academic Press, New York. Hennessy, M.B., 1986. Effects of social partners on pituitary-adrenal activity during novelty exposure in adult female squirrel monkeys. Physiol. Behav. 38, 803 – 807. Hennessy, M.B., 1997. Hypothalamic-pituitary-adrenal responses to brief social separation. Neurosci. Biobehav. Rev. 21, 11–29. Hennessy, M.B., Levine, S., 1977. Effects of various habituation procedures on pituitary-adrenal responsiveness in the mouse. Physiol. Behav. 18, 799 – 802. Hennessy, M.B., Levine, S., 1978. Sensitive pituitary-adrenal responsiveness to varying intensities of psychological stimulation. Physiol. Behav. 21, 295 – 297. Johnston, A.L., File, S.E., 1991. Sex differences in animal tests of anxiety. Physiol. Behav. 49, 245 – 250. Kitay, J.I., 1961. Sex differences in adrenal cortical secretion in the rat. Endocrinology 68, 818 – 824. Latane´, B., Schneider, E., Warinf, P., Zweigenhaft, R., 1971. The specificity of social attraction in rats. Psychonom. Sci. 23, 28–29. Laviola, G., Wood, R.D., Kuhn, C., Francis, R., Spear, L.P., 1995. Cocaine sensitization in periadolescent and adult rats. J. Pharmacol. Exp. Ther. 275, 345 – 357. Levine, S., 1993. The influence of social factors on the response to stress. Psychother. Psychosom. 60, 33 – 38. Levine, S., Coe, C., Wiener, S.G., 1989. Psychoneuroendocrinology of stress: a psychobiological perspective. In: Brush, R.F., Levine, S. (Eds.), Psychoneuroendocrinology. Academic Press, New York, pp. 341–377. Lister, R.G., Hilakivi, L.A., 1988. The effects of novelty, isolation, light and ethanol on the social behavior of mice. Psychopharmacology 96, 181 – 187. Meaney, M.J., Stewart, J., 1981. A descriptive study of social development in the rat (Rattus nor6egicus). Anim. Behav. 29, 34–45. Mineka, S., Keir, R., Price, V., 1980. Fear of snakes in wild- and lab-reared rhesus monkeys. Anim. Learn. Behav. 8, 653–663. Monroe, B.D., Milner, J.S., 1977. Social deprivation and novelty effects on gregarious behavior in the rat. Bull. Psychonom. Soc. 9, 219–220. Panksepp, J., Siviy, S., Normansell, L., 1984. The psychobiology of play: theoretical and methodological perspectives. Neurosci. Biobehav. Rev. 8, 465 – 492. Poole, T.B., Fish, J., 1976. An investigation of individual, age and sexual differences in the play of Rattus norvegicus (Mammalia: Rodentia). J. Zool. Lond. 179, 249 – 260. Primus, R.J., Kellogg, C.K., 1989. Pubertal-related changes influence the development of environmentrelated social interaction in the male rat. Dev. Psychobiol. 22, 633 – 643. Russel, P.A., 1977. Sex differences in rats’ stationary exploration as a function of stimulus and environmental novelty. Anim. Learn. Behav. 5, 297 – 302. Shors, T.J., Wood, G.E., 1995. Contribution of stress and gender to exploratory preferences for familiar versus unfamiliar conspecifics. Physiol. Behav. 58, 995 – 1002. Spear, L., Brake, S.C., 1983. Periadolescence: age-dependent behavior and psychopharmacological responsivity in rats. Dev. Psychobiol. 16, 83 – 109. Spencer-Booth, Y., Hinde, R.A., 1971. Effects of 6 days separation from mother on 18- to 32-week-old rhesus monkeys. Anim. Behav. 19, 174–191. Terranova, M.L., Alleva, E., Laviola, G., 1994. Affiliation and neophobia in developing mice prenatally exposed to oxazepam. Behav. Pharmacol. 5, 52 – 60.
656
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Terranova, M.L., Laviola, G., 1995. Individual differences in mouse behavioral development: effects of precocious weaning and ongoing sexual segregation. Anim. Behav. 50, 1261 – 1271. Terranova, M.L., Laviola, G., Alleva, E., 1993. Ontogeny of amicable social behavior in the mouse: gender differences and on-going isolation outcomes. Dev. Psychobiol. 26, 467 – 481. Terranova, M.L., Laviola, G., De Acetis, L., Alleva, E., 1998. A description of the ontogeny of mouse agonistic behavior. J. Comp. Psychol. 112, 3 – 12. Thor, D.H., Harrison, R.J., Schneider, S.R., 1988. Sex differences in investigatory and grooming behaviors of laboratory rats (Rattus nor6egicus) following exposure to novelty. J. Comp. Psychol. 102, 188–192. Thor, D.H., Holloway, J.R., 1984. Developmental analyses of social play behavior in juvenile rats. Bull. Psychonom. Soc. 22, 587–590. Thor, D.H., Holloway, W.R., 1986. Social play soliciting by male and female juvenile rats: effect of neonatal androgenization and sex of cagemates. Behav. Neurosci. 100, 275 – 279. Vogt, J.L., Coe, C.L., Levine, S., 1981. Behavioral and adrenocorticoid responsiveness of squirrel monkeys to a live snake: is flight necessarily stressful? Behav. Neural Biol. 32, 391 – 405. Williams, J.R., Taymans, S., Hastings, N., Carter, C.S., 1994. Endocrine responses to mate separation and reunion in a monogamous mammal (Microtus ochrogaster). International Behavioral Neuroscience Society Meeting, abstract 23. Witt, E.D., 1994. Mechanisms of alcohol abuse and alcoholism in adolescents: a case for developing animal models. Behav. Neural Biol. 62, 168 – 177.
.
Behavioral and hormonal effects of partner familiarity in periadolescent rat pairs upon novelty exposure M. Livia Terranova a, Francesca Cirulli b,*, Giovanni Laviola b a
Department of Psychology, Uni6ersita` degli Studi di Roma ‘La Sapienza’, 6ia dei Marsi 78, I-00185 Rome, Italy b Section of Beha6ioural Pathophysiology, Labor. F. O. S., Istituto Superiore di Sanita`, 6iale Regina Elena 299, I-00161 Rome, Italy Received 19 March 1998; accepted 28 January 1999
Abstract In periadolescent rats, social interactions are typically characterized by elevated levels of playful and affiliative behavior. Aim of the present study was to assess the behavioral and hormonal effects of partner familiarity upon the separation and reunion in a novel environment of established pairs of periadolescent subjects. At weaning (post-natal day, PND 21), Sprague–Dawley rats were pair housed with a non-sibling subject of the same age and sex. On PND 35, the members of each pair were separated for a 24-h period, and randomly assigned to different experimental groups: (1) sacrificed before separation; (2) sacrificed immediately after the isolation period; (3– 4) placed individually in a novel cage for 30 min either in low-light or in high-light conditions; (5 – 6) reunited for 30 min in a novel cage either with their previous cagemate (familiar, FAM); or (7 – 8) with an unfamiliar rat (UNF) of the same age and sex, in either light conditions. During reunion, the occurrence of social and non-social behaviors was scored. Blood samples were collected at the end of the session from all groups and assayed for corticosterone (CORT). The separation of the two members of an established pair did not affect baseline CORT levels. Upon reunion, the presence of a conspecific exerted a significant buffering effect on the novelty-induced increase in CORT levels. Such an effect of the social companion appeared more marked in males than in females, and in FAM compared to UNF pairs. Interestingly, FAM rats also expressed a
* Corresponding author. Tel.: +39-06-4990-2105; fax: + 39-06-4957-821. E-mail address: [email protected] (F. Cirulli) 0306-4530/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 3 0 6 - 4 5 3 0 ( 9 9 ) 0 0 0 1 9 - 0
640
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
significantly higher amount of social investigation and play-soliciting behavior compared to UNF animals. Behavioral results, together with previous data, suggest that periadolescent rats housed in established pairs develop a sort of amicable relationship. The overall CORT output measured at the end of the session is also in line with this interpretation. As a whole, these findings indicate that periadolescence is a time period during rat development, during which social variables play a very important role in modulating both behavioral and physiological responses to novelty in a fashion that does not completely overlap with data on adult subjects. © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Periadolescence; Social interactions; Partner familiarity; Environmental novelty; Corticosterone; Rat
1. Introduction During periadolescence — which has been defined as the ontogenetic period that encompasses in the rat the 7 – 10 days preceding the onset of puberty (at about 40 days of age) and the first few days thereafter (Spear and Brake, 1983)—rats and mice differ markedly from both younger and older conspecifics in several aspects of behavior and physiology (Spear and Brake, 1983; File, 1986; Witt, 1994; Laviola et al., 1995; Adriani et al., 1998; Terranova et al., 1998). In particular, animals around this age are characterized by elevated levels of affiliative and playful social interactions among peers, which mainly consist of rough-and-tumble play (Meaney and Stewart, 1981; Panksepp et al., 1984; Thor and Holloway, 1984, 1986; see also Terranova et al., 1993 for a comparison with the mouse). However, the specific characteristics of the social relationship underlying such a high tendency to socialize have not yet been clarified. In this frame, surprisingly few and contrasting results are available on the effects of partner familiarity on rats social interactions around periadolescence. On the one hand, Latane´ et al. (1971) could not find any difference between familiar (FAM) and unfamiliar (UNF) animals in measures of social proximity. On the other hand, both Barefoot et al. (1975) and Monroe and Milner (1977) found that partner’s novelty increased juvenile social interactions (see also Thor et al., 1988). Furthermore, while the importance of social variables in affecting the response to stressful stimuli has been characterized in adult rodents, few data are available in young subjects. A previous work by our group (Cirulli et al., 1996), designed to address such an issue, was the first, to our knowledge, to assess, along with the behavioral outcomes, also the adrenocortical response to both separation and reunion of periadolescent rat pairs with different degrees of familiarity. From a behavioral point of view, the above study showed that an elevated amount of investigative and playful behaviors was exhibited by UNF subjects. Such a finding is probably amenable to a ‘curiosity’ factor related to the attraction properties of novel conspecifics (Barefoot et al. 1975; Monroe and Milner, 1977; File and Peet, 1980; Armario et al., 1983a,b; Thor et al., 1988; File, 1990). On the other hand, FAM rats were found to have established some kind of a preferential relationship. These
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
641
animals actually expressed a higher amount of proximity-maintaining affiliative interactions which are reminiscent of those characterizing attachment-like bonds (for such kind of relationships in primates and in monogamous rodents such as the prairie vole, see Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al., 1990; Williams et al., 1994). However, hormonal results strongly suggested the absence both of a buffering effect of the interactions between UNF subjects towards novelty-induced neuroendocrine activation, and of an attachment-like social relationship between FAM animals. In fact, the corticosterone (CORT) levels that were measured in the two members of each pair at the end of the social encounter were found to be totally unaffected by partner familiarity. However, it should be taken into account that in the previous study (Cirulli et al., 1996) social interactions during reunion took place in a cage of a different size from the home cage. We need to consider that exposure to novelty is one of the most powerful stimuli capable of activating the hypothalamic-pituitary-adrenal (HPA) system (e.g. to a novel testing apparatus; see Hennessy and Levine, 1977, 1978; File and Vellucci, 1979; File and Peet, 1980; File, 1990). Moreover, familiarity with the environment is known to be one of the critical determinants of gregariousness among laboratory rats. Both fear and curiosity about novel environments have been shown to compete with social attraction, apparently representing a stronger drive state than the social one (File and Hyde, 1978; File and Vellucci, 1979; File and Peet, 1980; Armario et al., 1983a,b; File, 1986; Lister and Hilakivi, 1988; Thor et al., 1988; File, 1990). Since, in turn, social variables, such as the availability of a FAM or UNF conspecific, have been shown to affect the response of the HPA system to stressful stimuli (Armario et al., 1983a,b; Hennessy, 1986; Levine, 1993; Hennessy, 1997; see discussion in Cirulli et al., 1996), it is reasonable to hypothesize that, in our previous experiment, the novelty of the test cage might have affected both hormonal and behavioral responses. Specifically, the strong novelty-induced activation mi0ght have resulted in a ‘ceiling effect’ (see File and Peet, 1980), with the high values reached by CORT secretion masking the potential differences between the FAM and the UNF groups. Therefore, in order to at least partially reduce the novelty of the testing environment, in the present study rats were placed in a cage of the same type and size as the home-cage. Moreover, with the aim of separating the effects of partner familiarity on CORT secretion from those of environmental novelty, an additional control group of animals (NOV group), placed individually in the novel cage, was included. We also chose to manipulate the illumination level in the experimental room so as to expose the animals to a dimmer light compared to the previous experiment. In fact, exposure of nocturnal rodents, such as the rat, to bright light has been found to be another important stress-inducing stimulus from both a hormonal and a behavioral standpoint (File and Hyde, 1978; File and Peet, 1980; Lister and Hilakivi, 1988). In summary, since previous work with adult rodents and primates has indicated an important role of social variables on both behavioral and neuroendocrine responses to stressful stimuli, the present study was aimed at investigating this issue in a developmental phase (periadolescence) when rats show a particularly strong
642
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
tendency to be social. Partner familiarity was the specific variable manipulated upon the separation and reunion of established pairs, with the additional aim of evaluating whether periadolescents high sociality is amenable either to an aspecific social attraction among peers or to the presence of specific attachment-like relationships. Since the elevation of CORT levels upon separation is usually viewed as an important marker of attachment bonds (Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al., 1990; Williams et al., 1994; Hennessy, 1997), data on CORT levels were also collected at the end of the 24-h separation period.
2. Methods
2.1. Animals, breeding and rearing conditions Sprague – Dawley pregnant rats were obtained from Charles River Italia (Calco, Italy). Upon arrival, animals were housed separately in 40× 25×20 cm Plexiglas boxes, with sawdust as bedding and a metal top, in an air-conditioned room at 2191°C (relative humidity: 50 910%). They were maintained in a 12:12 h light (35.00 scotopic lux)/dark (0.02 scotopic lux) cycle, with lights on at 0930 h. Food (enriched standard diet purchased from Piccioni, I-25100, Brescia, Italy) and water were freely available. All animal handling and experimental procedures were performed in accordance with the EEC guidelines (EEC Council Directive 86/609, 1987) and with the Italian legislation (Decreto L.vo 116/92) on animal experimentation. From weaning (PND 21) until the time of testing, each pup was kept together with a non-sibling, same-age and same-sex conspecific, in a 40× 25× 20 cm Plexiglas box. On PND 34, all subjects were weighed and rehoused individually with their own sawdust in identical cages for a 24-h period. In both sexes, each rat was randomly assigned to one of eight different experimental conditions:NT (non-treated): animals blood sampled before separation on PND 34 (N= 8 subjects);24-h ISOL: animals blood sampled following the 24-h period of isolation, i.e. on PND 35 (N =8 subjects);NOV-LL (novelty-low light): following the 24-h isolation, subjects were put individually for 30 min in a cage identical to the home cage with new sawdust as bedding, under low light conditions (28.00 scotopic lux), and blood sampled at the end of the session (N= 8 subjects);NOV-HL (noveltyhigh light): same as above but under high light conditions (65.00 scotopic lux) (N= 8 subjects);FAM-LL (familiar-low light): the two members of a pair were reunited following the 24-h period of separation for 30 min in a cage identical to the home cage with new sawdust as bedding, under LL conditions, and blood sampled at the end of the session (N = 7 pairs);FAM-HL (familiar-high light): same as above but under HL conditions (N= 8 pairs);UNF-LL (unfamiliar-low light): following the 24-h period of isolation, two subjects from different pairs were placed together for 30 min in a cage identical to the home cage with new sawdust as bedding, under LL conditions, and blood sampled at the end of the session (N= 8
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
643
pairs);UNF-HL (unfamiliar-high light): same as above but under HL conditions (N =8 pairs). Age of testing was chosen taking into account literature on periadolescent rats reporting that the most playful and intense social interactions are shown by animals around PND 35 (see Section 1).
2.2. Blood samples Animals were killed by decapitation and trunk blood collected in heparinized tubes. Blood samples were centrifuged at 2000 rpm for 20 min to obtain cell-free plasma, and then frozen at − 80°C until they were assayed for CORT with a radioimmunoassay kit (ICN Biochemicals, Costa Mesa, CA; sensitivity9 0.125 mg/dl). Inter- and intra-assay coefficients of variation were 18 and 11.5%, respectively. Blood was collected from ten of the 16 subjects belonging to the eight couples in each treatment group.
2.3. Beha6ioral obser6ations Behavioral testing took place in an experimental room maintained in the same temperature and humidity conditions as the housing room, from 1000 to 1400 h, and the time of testing was counterbalanced between experimental groups. The behavior of each pair was videorecorded by means of two videocameras connected to a professional Sony videocassette recorders V0-5800PS. A software system for collection and analysis of observational data was used for scoring duration and frequency of each response. Only the last 10 min of the 30 min sessions were scored and statistically analyzed, since preliminary observations carried out in three sampled pairs of subjects for each experimental group confirmed that, as found in a previous work (see Cirulli et al., 1996), no behavioral differences were detectable during the first portion of the social encounters. A ‘focal animal—all occurrences’ sampling method was used (see Altmann, 1974), with each animal being continuously observed for randomly-chosen 5 min. Since the behavior of one of the two subjects in a pair cannot be considered independent from that of the other, pair means were used for further analysis. The social and non-social behavioral categories listed below, and their classification, are mainly based upon the ethological profiles of rat behavior described by Poole and Fish (1976), Meaney and Stewart (1981), Panksepp et al. (1984), Thor and Holloway (1984) and Thor and Holloway (1986) (see also Terranova et al., 1993, 1994; Terranova and Laviola, 1995; Cirulli et al., 1996). Unless otherwise specified, both frequency (f) and duration (d) were measured for each behavioral item.
2.4. Social acti6ities
Investigative elements: (1) Social sniff: self-explanatory (s.e.); (2) Follow: s.e.; (3) Mutual circle: partners are mutually sniffing each other’s anogenital region, while describing tight circles with their reciprocal following movements.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
644
Affiliative elements: (1) Social inacti6e: a rat is lying flat or standing still (eyes closed or open) while maintaining close physical contact with the partner; (2) Social rest: is being groomed by the partner; (3) Allogroom: s.e. Play-soliciting: (1) Crawl o6er (f): crawls over the partner’s body, crossing it transversely from one side to the other. Rough-and-tumble play: The two rats exhibit sudden darting movements, which are associated with pouncing on the partner back, chasing, wrestling and pinning (the so-called ‘on-top’ and ‘on-back’ postures). The play bout is often anticipated by a prolonged and intense bout of allogrooming, often accompanied by gross movements of the whole body of the performer and by vigorous pulling of the fur of the partner. The bout is considered concluded when both participants either turn their attention away from each other, or shift to amicable non-playful interactions (e.g. sniffing).
2.5. Non-social acti6ities
Activity and exploration/escape behaviors: (1) Inacti6e: s.e.; (2) Explore: s.e.; (3) Dig: is digging in the sawdust, pushing and kicking it around using the snout and/or both the forepaws and hindpaws, mostly moving around the cage and sometimes changing the whole arrangement of the substrate material. Maintenance activities: (1) Selfgroom: s.e.
2.6. Design and statistical analysis Behavioral data were analysed by means of a 2× 2× 2 ANOVA, with familiarity (FAM, UNF), light (LL, HL) and sex (male, female) as between-subject factors. Mean comparisons were carried out by means of the Superanova package (version 1.1, 1989 – 90, Abacus Concepts) wherever significant interaction effects were detected by the ANOVA. See below for corticosterone data analysis.
3. Results
3.1. Corticosterone Corticosterone data (see Fig. 1) were first analyzed by means of an 8 (treatment: NT, 24-h ISOL, NOV-LL, NOV-HL, FAM-LL, FAM-HL, UNF-LL, UNFHL)×2 (sex: male, female) ANOVA. Post-hoc comparisons were performed using the Tukey’s HSD test. Corticosterone levels differed significantly among the different treatment groups (treatment: F[1,7]= 13,48, PB .001). In particular, hormonal levels at the end of the 24-h of social isolation did not differ from pre-separation levels and were significantly lower than those found at the end of the 30 min reunion, independently from the degree of familiarity and the light level (P B.01). The ANOVA revealed a main effect of sex (F[1,7]= 8.79, PB .01), with females showing overall higher corticosterone levels than males. A second analysis
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
645
was thus performed for each sex separately to prevent small but consistent differences in the male group from being masked by female values. To evaluate simultaneously the effects of the presence/familiarity of another subject and those of different light conditions in the same analysis, we performed for each sex a 3 (treatment: NOV, FAM, UNF)× 2 (light: LL, HL) factorial ANOVA, excluding the NT and 24-h ISOL groups. In males, a main effect of treatment was found (F[1,2]=11.06, P B.0001], and post hoc comparisons indicated that, when exposed to novelty in pairs, they showed CORT levels that were significantly lower than those reached when placed alone in the test cage, irrespectively of the degree of partner familiarity (NOV vs. UNF: P B .01; NOV vs. FAM: PB .01). A significant interaction between treatment and light (F[2,50]= 3.79, PB .05) also appeared, and post hoc comparisons further clarified that such a buffering effect of the social companion was only evident when animals where tested in LL conditions (NOV-LL vs. FAM-LL: P B .01; NOV-LL vs. UNF-LL: PB .01). In addition, as it can be seen in Fig. 1, there was a tendency for CORT levels in the FAM group to be lower than those in the UNF one.
Fig. 1. Mean (9 SEM) corticosterone levels of male periadolescent rats in the different experimental conditions. NT, subjects were sampled before the isolation period; ISOL, subjects were isolated for 24 h and blood sampled at the end of the isolation period; NOV, subjects were placed in a test cage alone for 30 min after 24-h of isolation and sampled at the end of this time period; UNF, subjects were reunited for 30 min with an unfamiliar conspecific after 24-h of isolation and sampled at the end of the reunion; FAM, subjects were reunited for 30 min with their previous cagemate after 24-h of isolation and sampled at the end of the reunion (N =8 subjects in the NT, ISOL and NOV groups, N =10 subjects in the FAM and UNF groups). **NOV-LL vs. UNF-LL and NOV-LL vs. FAM-LL, P B .01.
646
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Fig. 2. Mean duration ( 9 SEM) of non-social behaviors recorded during the last 10 min of the 30 min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N = 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples).
As for females (see Table 1), light affected CORT secretion significantly (light: F[1,2]= 4.29, P B.05), with overall higher levels in the LL than in the HL condition. Although no significant effects of treatment were found, female CORT profile was rather similar to that of males.
3.2. Social and non-social beha6ior The ANOVA did not yield significant results for explore, dig, sniff and roughand-tumble play behaviors. When both frequency and duration were scored, the results for the two measures were consistent with each other and, for the sake of space, only the results regarding the latter are reported.
NT
1.1 (9 0.1) a
ISOL
2.4 (90.5)
NOV
UNF
FAM
LL
HL
LL
HL
LL
HL
28.2 ( 9 6.1)
19.1 (95.9)
26.7 (9 4.1)
25.3 (9 3.1)
23.8 (9 5.5)
10.6 (9 3.1)
NT, subjects were sampled before the isolation period; ISOL, subjects were isolated for 24 h and blood sampled at the end of the isolation period; NOV, subjects were placed in a test cage alone for 30 min after 24 h of isolation and sampled at the end of this time period; UNF, subjects were reunited for 30 min with an unfamiliar conspecific after 24-h of isolation and sampled at the end of the reunion; FAM, subjects were reunited for 30 min with their previous cagemate after 24-h of isolation and sampled at the end of the reunion (N =8 subjects in the NT, ISOL and NOV groups, N = 10 subjects in the FAM and UNF groups).
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Table 1 Mean ( 9 SEM) corticosterone levels (mg/dl) of female periadolescent rats in the different experimental conditionsa
647
648
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
In general, periadolescent rats allocated longer portions of their time to inacti6e (Fig. 2) and social inacti6e (an affiliative behavior, see Fig. 3) when tested in HL compared to LL conditions, while exhibiting significantly less mutual circle (an item of social investigation; Fig. 4) (light: F[1,55]= 3.93, 6.12, 12.29, PB .05 or less, respectively). Females tended to be more involved than males in the affiliative social rest behavior (32.502 ( 94.121) s and 22.954 ( 9 3.321) s, respectively; sex: F[1,55]= 3.52, P=.06), and, within the FAM group, expressed a higher amount of mutual circle (Fig. 4; sex× familiarity: F[1,55]= 7.39, PB .01; means comparison= P B .01). In addition, as shown in Fig. 3, females showed a tendency towards a decreased duration of allogroom affiliative behavior when tested in HL compared to LL conditions (Fig. 4; sex× light: F[1,55]= 3.66, P= .06).
Fig. 3. Mean duration ( 9SEM) of affiliative social behaviors recorded the last 10 min of the 30 min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N= 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples). Data refer to the same animals of Figs. 1 and 2.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
649
Fig. 4. Mean duration ( 9 SEM) of investigative social behaviors recorded during the last 10 min of the 30-min social encounters (see Section 2) in same-sex and same-age pairs of periadolescent rats of the different groups (N= 8 couples of subjects scored in each group, except for FAM-LL group: N = 7 couples). Data refer to the same animals of Fig. 1.
FAM rats appeared to be more involved than UNF animals in the social investigative behaviors follow and mutual circle (Fig. 4) and in the play-soliciting item crawl o6er (FAM: 0.84690.211; UNF: 0.3929 0.085) (familiarity: F[1,55]= 3.49, 6.85, 3.79, P = .06, P B.01 and .05, respectively). The effects of partner familiarity were particularly marked within the female group, with FAM rats spending less time staying inacti6e (Fig. 2) and much more in mutual circle (Fig. 4) than UNF animals (sex×familiarity: F[1,55]= 5.20, 7.39, PB .05 or less, respectively; means comparisons= P = .06 and PB .001, respectively). Light conditions also interacted with partner familiarity, in that, compared to UNF rats, FAM subjects tended to show an increased duration of both mutual circle (Fig. 4) and allogroom (Fig. 3; means comparison= PB .01) behaviors when
650
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
tested in LL but not in HL conditions (light× familiarity: F[1,55]=3.65, 7.95, P = .06 and P B .01, respectively). Moreover, compared to UNF animals whose behavior was apparently not affected by light conditions, FAM rats showed a greater involvement in allogroom (means comparison= PB .01) and, as a tendency, in mutual circle when tested in LL compared to HL conditions (light× familiarity: F[1,55]=7.95, 3.65, P B.01 and P = .06, respectively). Among FAM animals, some interesting interactions between the sex and the light variables were also detected. As shown in Figs. 2 and 4, FAM females tended to devote longer periods of time to selfgroom, and spent significantly shorter periods of time in mutual circle, when tested in HL compared to LL conditions (sex × light×familiarity: F[1,55] =3.45, 3.55, P= .06 for both items), whereas in the same conditions FAM males significantly decreased their crawl o6er frequency (LL: 1.56290.521; HL: 0.5009 0.231; sex× light× familiarity: F[1,55]= 4.28, PB .04; means comparison= P B.01). Interestingly, within the male group tested in LL conditions, crawl o6er was exhibited significantly more often by FAM than by UNF rats (FAM: 1.5629 0.521; UNF: 0.3649 0.152; sex× light× familiarity: F[1,55]= 4.28, PB .04; means comparison= P B .001). Finally, the sex variable also seemed to affect the behavioral profile within the FAM group tested in LL conditions. In such a group, females tended to overcome males as for the duration of mutual circle behavior (Fig. 4), whereas the opposite was found for the frequency of crawl o6er (males: 1.562 9 0.521; females: 0.500 9 0.289; means comparisons=P B.04) and for selfgroom duration (Fig. 2) (sex× light× familiarity: F[1,55] = 3.55, 4.28, 3.45, P= .06, .04 and .06, respectively).
4. Discussion The main findings of the present study can be summarized as follows: 1. A 24-h period of separation of the members of periadolescent rat pairs did not affect baseline corticosterone (CORT) levels. 2. As expected, the exposure of animals individually (NOV group) to environmental novelty produced a significant increase in CORT levels. 3. Male subjects reunited with a conspecific following separation, and thus confronted with novelty in pairs, showed a significantly lower increase in CORT levels compared to NOV males. 4. There was only a tendency for such a buffering effect of the social companion on CORT secretion to be more marked in familiar (FAM) compared to unfamiliar (UNF) pairs. 5. FAM rats also expressed a significantly higher amount of social investigation and play-soliciting behavior compared to UNF animals, and this was particularly evident in the female group. A strong activation of the HPA axis upon separation is usually considered as a main indicator of an attachment relationship, such as that characterizing mother– infant dyads in non-human primates or heterosexual pairs in monogamous rodents (Harlow and Harlow, 1965; Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
651
al., 1985; Levine et al., 1989; Carter et al., 1990; Williams et al., 1994; Hennessy, 1997). In studies investigating gregariousness or affiliative peer interactions, it has been reported that adult rats do not show changes in CORT upon separation from social companions (File and Peet, 1980; Armario et al., 1983a,b; File, 1990). In this experiment we investigated whether this might be true in periadolescent rats, which are specifically characterized by high levels of social behavior (see Section 1). No hormonal changes were found following social separation of the two members of an established pair, a result which is apparently in line with the notion that the high levels of social behavior characterizing periadolescent rats are not amenable to specific attachment-like bonds. Moreover, when male subjects were confronted in pairs with a novel environment—which was per se responsible for a marked CORT elevation — a buffering effect of the social companion on CORT secretion was to some extent observed independently from the degree of partner familiarity. Thus, our data seem to confirm, although indirectly, the hypothesis that periadolescent rats high tendency to sociality is mainly the result of an aspecific social attraction among peers (see also Cirulli et al., 1996). The finding of a buffering effect of the social companion on CORT secretion has to be viewed in the frame of previous studies, carried out in both rats and non-human primates, showing the major role played by social variables in modulating the response of the HPA system to stress (Armario et al., 1983a,b; Hennessy, 1986; Levine et al., 1989; Levine, 1993; Hennessy, 1997; see discussion in Cirulli et al., 1996). Studies conducted in primates have shown that social variables can modulate HPA activity, both in terms of increased hormone release and/or the inhibition of endocrine responses to environmental perturbations (Levine et al., 1989; Levine, 1993). For example, the presence of a social group can ameliorate the neuroendocrine response to aversive stimuli that normally evoke a marked adrenal response in monkeys housed alone, such as a snake or a snake-like object (Mineka et al., 1980; Vogt et al., 1981). An unfamiliar conspecific can also be effective in reducing the pituitary-adrenal activation in response to stressful stimuli (Armario et al., 1983a,b; Hennessy, 1986). Nonetheless, a fine-grain behavioral analysis confirmed previous data (Cirulli et al., 1996) showing that the degree of familiarity affects the pattern of social interactions of periadolescent rat pairs. Although differences were not detected in the expression of affiliative behaviors, FAM animals were significantly more involved than UNF subjects in both social investigation and play-soliciting interactions. Such a finding does not completely agree with literature data on both adult (Armario et al., 1983a,b) and adolescent rats (Barefoot et al. 1975; Monroe and Milner, 1977; Thor et al., 1988; Cirulli et al., 1996), which indicate a soliciting effect of partner novelty on active social interactions, and associate such attraction properties of novel conspecifics to a ‘curiosity’ factor (see also File and Peet, 1980; Hennessy, 1986; File, 1990). It also remains to be ascertained why in the present study, FAM rats expressed elevated amounts of investigative and play-soliciting behaviors, instead of devoting increasing amount of time to affiliative interactions, as it would be expected on the basis of both the attachment-related literature (Spencer-Booth and Hinde, 1971; Bowlby, 1975; Coe et al., 1985; Carter et al.,
652
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
1990; Williams et al., 1994) and our previous results (Cirulli et al., 1996). Anyhow, the present data seem to indicate that periadolescent rats do recognize and interact preferentially with FAM rather than with UNF conspecifics. Although neuroendocrine responses were not significantly affected by the degree of familiarity of the two partners, they suggested that, consistently with the above behavioral results, some amicable relationship develops and can be recalled by FAM periadolescent rats. Whereas the previous study (Cirulli et al., 1996) found similar levels of CORT secretion upon social reunion, the present data provide some indication that FAM subjects might exert a greater buffering effect than UNF animals on novelty-induced CORT release. The possibility that a social variable such as partner familiarity might exert in the rat an influence on important biological processes such as CORT secretion is worth further experimentation. Nonetheless, it is important to stress here that a direct correlation between hormonal and behavioral data cannot be drawn since CORT data represent the overall hormonal output of the session, while behavioral data only refer to the last 10 min of the observation period. Overall, these findings suggest some interesting differences in physiological and behavioral regulation between periadolescent and adult rats. When compared to young animals, adults are reported to show a different response pattern, namely a higher amount of social interactions and lower CORT levels when placed with UNF compared to FAM conspecifics (Armario et al., 1983a,b). When pair-housed in the same home cage early on during development rather than in adulthood, rats may experience a different quality of social interactions, which in turn might well influence later recognition in the test box. Moreover, it seems quite likely that both the behavioral and the hormonal outcomes of affiliative social experiences are more readily seen at a time, such as periadolescence, when amicable social behavior is expressed at high levels. One must also take into account that different experimental paradigms could be responsible for the differences between the data presented in this paper and previous reports on adult subjects. Also the discrepancy between the pattern of results of the present and the previous (Cirulli et al., 1996) study, both carried out in periadolescent rats, should be attributed to differences in methodology. In fact, in this experiment, environmental novelty was reduced by placing animals in a test cage of the same type and size as the home cage, whereas in the previous study the two cages differed in size and shape. However, the present study seems to be in agreement with the previous one (Cirulli et al., 1996; see also Shors and Wood, 1995) in that females appeared to be more sensitive than males to partner familiarity. Sex differences in exploratory and emotional behavior have been documented in a variety of mammals including man (Archer, 1975; Johnston and File, 1991), and, more specifically, sex has been shown to interact with stimulus complexity (e.g. Russel, 1977). In general, females show a greater preference for novelty than males, and thus are more inclined to explore novel environments. In addition, as for the physiological response to exogenous stressors, females have been reported to exhibit higher basal levels of glucocorticoids, as well as a greater range of CORT response to a stressor (Kitay, 1961; Critchlow et al., 1963), and such a functional dimorphism has been implicated in the regulation of various behaviors.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
653
As for the manipulation of the illumination level of the testing environment, which was here accomplished by comparing a high light (HL) condition with the condition of low light (LL) uniformly used in the previous study (Cirulli et al., 1996), results are not straightforward. As expected, a decrease in immobility and an increase in the expression of social behavior in LL was observed, most likely resulting from the reduction of the illuminance-related stress and of the associated behavioral inhibition (see e.g. File and Hyde, 1978). Surprisingly, from a hormonal standpoint, the LL condition did not result in lower cort levels as hypothesized on the basis of few but consistent literature reports (File and Vellucci, 1979; File and Peet, 1980), especially in the NOV and UNF groups. However, CORT levels in the two lighting conditions were differentially affected by the social setting. Under LL, as we pass through different conditions, CORT levels appear to mirror different arousal levels, being higher in NOV, reaching intermediate levels in the presence of an UNF conspecific, while tending to diminish in the presence of a familiar subject. Instead, in the HL condition, CORT levels were more resilient to change as a function of the presence of another subject, being it UNF or FAM. Three hypotheses can be put forward to explain the ineffectiveness of high light to arouse the experimental subjects above the LL levels. From a behavioral point of view, although we have no data on the NOV group, we found that in LL conditions the behavior of UNF and FAM subjects was less inhibited, and this, in turn, could well have influenced CORT levels in a direction opposite to that expected. A possible additional explanation is that, since there is a strong relationship between increasing stimulus intensity and magnitude of ACTH and CORT secretion, subjects exposed to a stronger stimulus, such as the HL condition, might have reached a secretory peak earlier compared to LL animals (see Hennessy and Levine, 1978; Dallman et al., 1987). Thus, at the time of testing, i.e. at the end of a 30 min exposure to the different experimental conditions (in line with our previous study, see Cirulli et al., 1996), CORT levels of subjects exposed to HL condition might have already reflected the pattern of return to resting levels, while in the LL subjects CORT secretion was still reaching its peak, thus resulting higher than the other group. Running a time course of the CORT response to the different experimental conditions could reveal whether this might indeed be true. The intriguing possibility finally exists that such a complex pattern of CORT response to lighting might be due to ontogenetic factors underlying the development of neuroendocrine mechanisms of anxiety. Appropriate behavioral and neuroendocrine responses shown by adult rats are in part due to changes occurring during puberty (see Primus and Kellogg, 1989). Guillet and Kellogg (1988) reported that the CORT response to the stress of a novel environment matures in the rat as early as 28 days, i.e. during the prepubertal phase. However, the behavioral responses to these anxiogenic conditions, as assessed by means of the social interaction test, appear slightly later, i.e. as rats enter puberty (PND 30-32; see also Primus and Kellogg, 1989). Thus, hormonal and behavioral responses seem to be dissociated in their developmental onset. The differential effect exerted in the present experiment by light intensity on behavioral and hormonal measures might be another example of a separate emergence of these two responses to novelty during periadolescence.
654
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
It can be concluded that periadolescence is a peculiar time period during rat development. During this period, social variables appear to play a very important role in modulating both behavioral and physiological responses to novelty, in a fashion that does not completely overlap with data on adult subjects.
Acknowledgements This research was supported as part of the Nervous and Mental Disorders Research Project on ‘Psychobiological risk or protection factors for behavioural disorders and vulnerability to recreational substances abuse during development’ intramural grant to G.L., Instituto Superiore di Sanita`, Roma, Italy, and by the Ministry of Environment Project, PR22/IS, A79/L2, grant ‘Animal models of neurobehavioural risk following exposure to gaseous pollutants’. We acknowledge the help of Dr F. Chiarotti in the statistical analysis of the data and Angelina Valanzano for expert technical suggestions and help.
References Adriani, W., Chiarotti, F., Laviola, G., 1998. Elevated novelty seeking and peculiar d-amphetamine sensitization in periadolescent mice. Behav. Neurosci. 112, 1152 – 1166. Altmann, J., 1974. Observational study of behavior: sampling methods. Behavior 49, 227 – 265. Archer, J., 1975. Rodent sex differences in emotional and related behavior. Behav. Biol. 14, 451 – 479. Armario, A., Luna, G., Balasch, J., 1983a. The effect of conspecifics on corticoadrenal response of rats to a novel environment. Behav. Neural Biol. 37, 332 – 337. Armario, A., Ortiz, R., Balasch, J., 1983b. Corticoadrenal and behavioral response to open field in pairs of male rats either familiar or non-familiar to each other. Experientia 39, 1316 – 1317. Barefoot, J.C., Aspey, W.P., Olson, J.M., 1975. Effect of partner novelty on affiliation in the rat. Bull. Psychonomic Soc. 6, 655–657. Bowlby, J., 1975. Attachment theory, separation anxiety, and mourning. In: Hamburg, D.A., Brodie, K.H. (Eds.), American Handbook of Psychiatry. New Psychiatric Frontiers. Basic Books, New York, pp. 292–309. Carter, C.S., Witt, D.M., Thompson, E.G., Carlstead, K., 1990. Effects of hormonal, sexual, and social history on mating and pair bonding in prairie voles. Physiol. Behav. 44, 691 – 697. Cirulli, F., Terranova, M.L., Laviola, G., 1996. Affiliation in periadolescent rats: behavioral and corticosterone response to social reunion with familiar or unfamiliar partners. Pharmacol. Biochem. Behav. 54, 99–105. Coe, C.L., Wiener, S.G., Rosenberg, L.T., Levine, S., 1985. Physiological consequences of maternal separation and loss in the squirrel monkey. In: Rosenblum, L.A., Coe, C.L. (Eds.), Handbook of Squirrel Monkey Research. Plenum, New York, pp. 127 – 148. Critchlow, V., Liebelt, R.A., Bar-Sela, M., Mountcastle, W., Lipscomb, A.S., 1963. Sex differences in resting pituitary-adrenal function in the rat. Am. J. Physiol. 205, 807 – 815. Dallman, M.F., Akana, S.F., Cascio, C.S., Darlington, D.N., Jacobson, L., Levin, N., 1987. Regulation of ACTH secretion: variations on a theme of B. Recent Prog. Horm. Res. 43, 113 – 173. File, S.E., 1986. The effects of neonatal administration of clonazepam on passive avoidance and on social, aggressive and exploratory behavior of adolescent male rats. Neurobehav. Toxicol. Teratol. 8, 447 – 452. File, S.E., 1990. Interactions of anxiolytic and antidepressant drugs with hormones of the hypothalamicpituitary-adrenal axis. Pharmacol. Ther. 46, 357 – 375.
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
655
File, S.E., Hyde, J.R.G., 1978. Can social interaction be used to measure anxiety? Br. J. Pharmacol. 62, 19 – 24. File, S.E., Vellucci, S.V., 1979. Behavioral and biochemical measures of stress in hooded rats from different sources. Physiol. Behav. 22, 31–35. File, S.E., Peet, L.A., 1980. The sensitivity of the rat corticosterone response to environmental manipulations and to chronic chlordiazepoxide treatment. Physiol. Behav. 25, 753 – 758. Guillet, R., Kellogg, C.K., 1988. Age-related effects of handling, diazepam exposure, and gonadectomy on corticosterone response. Soc. Neurosci. Abstracts 14, 112.5. Harlow, H.F., Harlow, M.K., 1965. The affectional system. In: Schrier, A., Harlow, H., Stollnitz, F. (Eds.), Behavior of Nonhuman Primates, vol. 2. Academic Press, New York. Hennessy, M.B., 1986. Effects of social partners on pituitary-adrenal activity during novelty exposure in adult female squirrel monkeys. Physiol. Behav. 38, 803 – 807. Hennessy, M.B., 1997. Hypothalamic-pituitary-adrenal responses to brief social separation. Neurosci. Biobehav. Rev. 21, 11–29. Hennessy, M.B., Levine, S., 1977. Effects of various habituation procedures on pituitary-adrenal responsiveness in the mouse. Physiol. Behav. 18, 799 – 802. Hennessy, M.B., Levine, S., 1978. Sensitive pituitary-adrenal responsiveness to varying intensities of psychological stimulation. Physiol. Behav. 21, 295 – 297. Johnston, A.L., File, S.E., 1991. Sex differences in animal tests of anxiety. Physiol. Behav. 49, 245 – 250. Kitay, J.I., 1961. Sex differences in adrenal cortical secretion in the rat. Endocrinology 68, 818 – 824. Latane´, B., Schneider, E., Warinf, P., Zweigenhaft, R., 1971. The specificity of social attraction in rats. Psychonom. Sci. 23, 28–29. Laviola, G., Wood, R.D., Kuhn, C., Francis, R., Spear, L.P., 1995. Cocaine sensitization in periadolescent and adult rats. J. Pharmacol. Exp. Ther. 275, 345 – 357. Levine, S., 1993. The influence of social factors on the response to stress. Psychother. Psychosom. 60, 33 – 38. Levine, S., Coe, C., Wiener, S.G., 1989. Psychoneuroendocrinology of stress: a psychobiological perspective. In: Brush, R.F., Levine, S. (Eds.), Psychoneuroendocrinology. Academic Press, New York, pp. 341–377. Lister, R.G., Hilakivi, L.A., 1988. The effects of novelty, isolation, light and ethanol on the social behavior of mice. Psychopharmacology 96, 181 – 187. Meaney, M.J., Stewart, J., 1981. A descriptive study of social development in the rat (Rattus nor6egicus). Anim. Behav. 29, 34–45. Mineka, S., Keir, R., Price, V., 1980. Fear of snakes in wild- and lab-reared rhesus monkeys. Anim. Learn. Behav. 8, 653–663. Monroe, B.D., Milner, J.S., 1977. Social deprivation and novelty effects on gregarious behavior in the rat. Bull. Psychonom. Soc. 9, 219–220. Panksepp, J., Siviy, S., Normansell, L., 1984. The psychobiology of play: theoretical and methodological perspectives. Neurosci. Biobehav. Rev. 8, 465 – 492. Poole, T.B., Fish, J., 1976. An investigation of individual, age and sexual differences in the play of Rattus norvegicus (Mammalia: Rodentia). J. Zool. Lond. 179, 249 – 260. Primus, R.J., Kellogg, C.K., 1989. Pubertal-related changes influence the development of environmentrelated social interaction in the male rat. Dev. Psychobiol. 22, 633 – 643. Russel, P.A., 1977. Sex differences in rats’ stationary exploration as a function of stimulus and environmental novelty. Anim. Learn. Behav. 5, 297 – 302. Shors, T.J., Wood, G.E., 1995. Contribution of stress and gender to exploratory preferences for familiar versus unfamiliar conspecifics. Physiol. Behav. 58, 995 – 1002. Spear, L., Brake, S.C., 1983. Periadolescence: age-dependent behavior and psychopharmacological responsivity in rats. Dev. Psychobiol. 16, 83 – 109. Spencer-Booth, Y., Hinde, R.A., 1971. Effects of 6 days separation from mother on 18- to 32-week-old rhesus monkeys. Anim. Behav. 19, 174–191. Terranova, M.L., Alleva, E., Laviola, G., 1994. Affiliation and neophobia in developing mice prenatally exposed to oxazepam. Behav. Pharmacol. 5, 52 – 60.
656
M.L. Terrano6a et al. / Psychoneuroendocrinology 24 (1999) 639–656
Terranova, M.L., Laviola, G., 1995. Individual differences in mouse behavioral development: effects of precocious weaning and ongoing sexual segregation. Anim. Behav. 50, 1261 – 1271. Terranova, M.L., Laviola, G., Alleva, E., 1993. Ontogeny of amicable social behavior in the mouse: gender differences and on-going isolation outcomes. Dev. Psychobiol. 26, 467 – 481. Terranova, M.L., Laviola, G., De Acetis, L., Alleva, E., 1998. A description of the ontogeny of mouse agonistic behavior. J. Comp. Psychol. 112, 3 – 12. Thor, D.H., Harrison, R.J., Schneider, S.R., 1988. Sex differences in investigatory and grooming behaviors of laboratory rats (Rattus nor6egicus) following exposure to novelty. J. Comp. Psychol. 102, 188–192. Thor, D.H., Holloway, J.R., 1984. Developmental analyses of social play behavior in juvenile rats. Bull. Psychonom. Soc. 22, 587–590. Thor, D.H., Holloway, W.R., 1986. Social play soliciting by male and female juvenile rats: effect of neonatal androgenization and sex of cagemates. Behav. Neurosci. 100, 275 – 279. Vogt, J.L., Coe, C.L., Levine, S., 1981. Behavioral and adrenocorticoid responsiveness of squirrel monkeys to a live snake: is flight necessarily stressful? Behav. Neural Biol. 32, 391 – 405. Williams, J.R., Taymans, S., Hastings, N., Carter, C.S., 1994. Endocrine responses to mate separation and reunion in a monogamous mammal (Microtus ochrogaster). International Behavioral Neuroscience Society Meeting, abstract 23. Witt, E.D., 1994. Mechanisms of alcohol abuse and alcoholism in adolescents: a case for developing animal models. Behav. Neural Biol. 62, 168 – 177.
.