Immediate early gene expression associated with induction of brooding behavior in Japanese quail

Hormones and Behavior 46 (2004) 19 – 29 www.elsevier.com/locate/yhbeh

Immediate early gene expression associated with induction of brooding behavior in Japanese quail Michael G. Ruscio * and Elizabeth Adkins-Regan Department of Psychology, Cornell University, USA Received 14 October 2003; revised 2 February 2004; accepted 3 February 2004 Available online 22 April 2004

Abstract Certain species can be induced to foster infant or neonatal animals through the process of sensitization. We induced brooding behavior in adult Japanese quail through repeated exposure to foster chicks across five 20-min trials. Brooding behavior was characterized by a bird allowing chicks to approach and remain underneath its wings while assuming a distinctive stationary crouching posture, preening, and feather fluffing. Birds who did not show brooding behavior actively avoided chicks. Among the birds that brooded chicks, females brooded chicks for longer durations compared to males. Brooding females continued a regular daily egg laying pattern; males showed no significant changes in testosterone levels after exposure to chicks. In a second experiment, we measured expression of two immediate early gene (IEG) protein products, ZENK and Fos, to identify the brain regions activated or inhibited by brooding behavior in females. ZENK and Fos expression in brooding or sensitized females (SF) were compared with expression in nonmaternal females with chicks (NMF) and with females without chicks and with blocks as control objects (BL). There was a reduced density of ZENK-like immunoreactive (ZENK-lir) cells in the medial preoptic nucleus (POM) in NMF birds. In SF birds, the density of Fos-like immunoreactive (Fos-lir) cells was elevated in the bed nucleus stria terminalis, medial portion (BSTm), and ectostriatum (E). These experiments begin to define the neural circuitry underlying brooding behavior in Japanese quail, and establish a model for future studies of the neural mechanisms of avian parental behavior. D 2004 Elsevier Inc. All rights reserved. Keywords: Fos; ZENK; Coturnix japonica; Parental behavior; Avian; Sensitization; Maternal behavior; POM; BSTm; Ectostriatum

Introduction Parenting marks the onset of an entirely new repertoire of behaviors intended to ensure the survival of offspring. Among vertebrate species, the specific parental behaviors displayed and the factors that facilitate their onset vary, yet the functional purpose of parental behavior remains the same. Typically, parental behavior occurs after a series of integrated social interactions (mating, and in some species, pairbonding) and physiological changes (such as those associated with gestation or incubation) preceding and following birth or hatching (Clutton-Brock, 1991). However, in some species, the mere presentation of foster neonatal conspecifics elicits parental behavior through ‘‘sensitization’’. The process of sensitization has been repeatedly demonstrated, is well * Corresponding author. Brain Body Center, Department of Psychiatry (mc 912), University of Illinois at Chicago, 1601 West Taylor Street, Chicago, IL 60612. E-mail address: [email protected] (M.G. Ruscio). 0018-506X/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.yhbeh.2004.02.002

characterized in species of mammals and birds, and often results in the display of the full repertoire of parental behaviors (Lonstein and DeVries, 2000; Richard-Yris et al., 1983; Richard-Yris, 1993; Rosenblatt, 1967; Wang and Buntin, 1999). The present set of experiments establishes a paradigm for sensitization in adult female and male Japanese quail (Coturnix japonica) and examines the neural basis for this behavior in females, using immunocytochemistry to measure immediate early gene (IEG) expression. The following studies are the first to use IEG expression to identify the neural circuitry associated with parental behavior in a galliform species or in any bird other than the ring dove. Parental behavior (including nest building, incubation, posthatch care) in Japanese quail (and many other galliform birds) is typically performed by females, according to observations from semi-natural environments (Hess et al., 1976; Madge and McGowan, 2002; Mills et al., 1997; Nichols, 1991). Newly hatched chicks are quite precocial and can walk, feed, see, and hear upon hatching, but cannot effectively thermoregulate until about 12 days of age (Mills

20

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

et al., 1997). They require brooding by an adult bird to survive. Brooding consists of an adult bird allowing chicks to nestle underneath its slightly raised wings while assuming a distinct crouching posture (Hess et al., 1976). Several galliform species including chickens, turkeys, and bobwhite quail (Colinus virginianus) can be sensitized to foster chicks and will brood them (Opel and Proudman, 1988; Richard-Yris et al., 1983; Vleck and Dobrott, 1993), even castrated male chickens can be sensitized to brood (Goodale, 1916). Japanese quail males and females on short day lengths (nonreproductively active) subject to artificial selection for sociality also brood foster chicks (Richard-Yris, 1993). Although brooding behavior has been characterized in some avian species, little is known about its neural basis. Immediate early gene (IEG) expression is an appropriate approach to identify nuclei that show differential activity between brooding females and controls. ZENK and Fos are the protein products of the IEGs zenk and c-fos. They are known to indicate cellular activity and have been repeatedly shown to reflect behavioral differences in both birds and mammals (Balthazart et al., 1996; Hoffman et al., 1993; Sharp et al., 1996). Looking at both Fos and ZENK provides a more thorough description of the neural activity associated with brooding behavior. Not every area in the brain expresses every single IEG, and certain areas may more readily express one IEG than another. For example, during quail sexual behavior, Fos and ZENK are expressed in a variety of the same forebrain areas and throughout the telencephalon. However, some areas show a difference in expression of one IEG and not the other (Ball et al., 1997; Meddle et al., 1999; Tlemcßani et al., 2000). Although IEG techniques are descriptive by nature, we hypothesize that differential expression in association with brooding is more likely to occur in certain nuclei. Ample evidence supports the idea that among vertebrates, there is substantial evolutionary conservation of the neural circuitry associated with motivated behaviors, including sex and parental behavior (Butler and Hodos, 1996; Newman, 2002). Because there is no prior information regarding the neural basis of parental behavior in this species or in any other galliform, we focused (but did not limit) our investigation of IEG expression on nuclei most likely to be involved with maternal behavior based upon the following criteria: nuclei that were (1) involved in maternal behavior in other galliforms and the ring dove and/or (2) potentially homologous to nuclei involved with maternal behavior in rodents and/or (3) involved with sexual behavior in quail. Ultimately, identification of the neural basis of brooding behavior in this species will demonstrate how homologous neural mechanisms contribute to species-specific patterns of maternal behavior. Incubation in turkey hens is perturbed by lesions of the preoptic area and the ventromedial nucleus of the hypothalamus (VMH) (Youngren et al., 1989); however, the effect on brooding behavior has not been measured. The ring dove (Streptopelia risoria) has provided an excellent model of the neural basis of avian parental

behavior. In ring doves, some nuclei in the brain (including the medial preoptic area or POM, and the VMH) are hormonally sensitive and directly involved in maternal behavior (Buntin, 1996; Slawski and Buntin, 1995). In particular, the POM shows an increase in IEG expression during incubation (Sharp et al., 1996). Although much is known about the neural basis of parental behavior in the ring dove, it is a biparental species with altricial young and some features of its parental behavior are unique to its lineage (Columbiformes). Interestingly, there are some apparent similarities in the neural components that underlie the production of parental behavior in rats and ring doves (Buntin, 1996; Numan, 1994). In both species, the medial preoptic area is part of a series of interconnected forebrain nuclei that concentrate sex steroid receptors and are involved with motivated behaviors. Lesions of the medial preoptic area impair species-specific components of maternal behavior (Numan et al., 1998; Slawski and Buntin, 1995). Additionally, in rodents, both ZENK and Fos are expressed in association with maternal behavior in areas that include the medial preoptic area (MPOA) and the bed nucleus of the stria terminalis (BNST) (Numan, 1994; Numan and Numan, 1995). It is important to note that a sensitized virgin female rat will show maternal behavior and a pattern of c-fos activity in the brain that is similar to maternal postpartum females (Numan and Numan, 1994). This demonstrates that maternal behavior in virgin and postpartum females utilizes similar neural components despite their different hormonal conditions. In Japanese quail, both Fos and ZENK are expressed in nuclei associated with sexual behavior, such as the POM and the BSTm (Ball et al., 1997; Balthazart et al., 1996; D’Hondt et al., 1999). Both the POM and BSTm are very likely homologous to the rodent MPOA and BNST (reviewed in Adkins-Regan, 1996; Aste et al., 1998; Balthazart et al., 1996) and are hypothesized to be involved with parental/brooding behavior in this species. In the first experiment, we induced brooding behavior in lab-reared, singly housed, long day, reproductively active but naive female and male adult Japanese quail through repeated exposures to foster chicks. Based upon previous reports (Hess et al., 1976; Nichols, 1991), we hypothesized that if brooding behavior was shown, it was more likely to be shown by females. If females did show brooding behavior, we anticipated that it could interrupt egg laying. Exposure to foster chicks is known to cause a drop in sex steroid levels and cessation of egg laying in chickens (Richard-Yris et al., 1983, 1998a,b). In males, we measured circulating testosterone levels. Based upon results in bobwhite quail (a different family of quail) (Vleck and Dobrott, 1993) and several other avian species (reviewed in Marler et al., 2003), we predicted that testosterone would decrease in a male showing parental behavior. In the second experiment, we used immunocytochemistry to detect IEG protein products ZENK and Fos in neural tissue. We compared ZENK and Fos expression in sensi-

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

tized brooding females (SF) with expression in nonmaternal females (NMF) and with females in an identical environment without chicks and with control objects (blocks) (BL). We predicted that there would be one or more nuclei showing differences in levels of expression of ZENK, Fos, or both in SF females compared with NMF or BL females.

Materials and methods The following procedures were independently reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at Cornell University. Animals All adult birds used as test subjects in this experiment were lab-reared Japanese quail 4 –8 months of age, individually housed on long day lengths (18L:6D), and fed ad libitum. Domesticated female Japanese quail typically lay one egg every 24 h when kept on long (reproductive) day lengths. Any female that was not laying regularly (at least five eggs over 7 days) was not used in this study. All males that appeared reproductively active were crowing and had enlarged cloacal glands (a testosterone-dependent characteristic that is more pronounced in long day males). All chicks used as stimulus birds were hatched in house from artificially incubated eggs obtained from Truslow Farms, MD. The chicks were housed together in a heated brooder except when used as stimuli during testing. Experiment 1. Induction of brooding behavior in adult female and male Japanese quail Brooding behavior testing and behavioral measures. A total of 31 females and 21 males were randomly picked from the existing laboratory colony (of approximately 100 birds) and tested for the induction of brooding behavior. Testing consisted of five 20-min exposures to four 2- to 5day-old chicks in a 45.72  48.26 cm Plexiglask arena. The day before testing, each animal was given one 20-min habituation period in the arena. The testing room was maintained at 20– 22jC, and was illuminated only by a 75 W heat lamp directed away from the testing arena. There was one trial on the first day followed by two per day over the next 2 days. Each trial was preceded by a 5-min habituation period in the arena. The ObserverR (Noldus) program was used to collect all data. The following behaviors were measured in female and male subjects in each trial. Latency to brood four chicks: Time from the beginning of the trial to the onset of brooding all four chicks simultaneously. Brooding behavior consisted of a bird allowing chicks to approach and remain under its wings. Brooding all four chicks simultaneously provided the most conservative and definitive measure of brooding behavior. It was possible

21

for one or two chicks to ‘‘sneak’’ underneath a bird without the bird actually showing any type of brooding. For the bird to allow all four chicks to come under its wings, it would invariably have to remain stationary and not actively avoid the chicks. Duration of Brooding four chicks: The total time in each 20-min trial that a bird was brooding all four chicks. Preen: A preen was counted every time a bird moved its head to preen its back, wing, or breast plumage. Another preen was not counted until the bird stopped the previous preen, returned its head to an upright position, and then returned to preening. Although preening occurs in other behavioral contexts, preening in brooding tests appeared to be directed toward chicks already underneath a bird’s wings. Typically, preening would not occur until after tactile contact with the chicks; therefore, we did not distinguish allopreening from self-preening. Given the close proximity of the chicks with the bird, it was not possible to distinguish precisely which animal’s feathers were affected by each preen. Wing raise: The bird gently raises its wing slightly away from its body while remaining stationary. The number of wing raises was counted. There are prior reports describing quail vocalizations associated with brooding or parental behavior in this species (Richard-Yris, 1993). Throughout all of our testing, none of our birds made any audible vocalizations; therefore, we did not use this as a criterion for our testing. Blood sample collection for testosterone assay in males. Blood samples were taken from males 1 day before testing (pretesting) and 1 day after testing (post-testing). Blood was obtained through wing vein puncture, using a heparinized tube. Samples were centrifuged and the plasma was separated. Radioimmunoassays were performed at the Cornell University Veterinary Diagnostic lab, using Coata-Count Total Testosterone kit from Diagnostic Products Corporation. The sensitivity of the assay is reported at 4 ng/dl (0.04 ng/ml; 0.14 nmol/l) with less than 5% cross reactivity with dihydrotestosterone. Statistical analyses. Behavioral data were analyzed using Mann –Whitney U tests. Testosterone levels in brooding and nonbrooding males were compared using t tests, pooled variance. Paired t tests were used to compare pre- and posttesting testosterone levels in brooding and nonbrooding males. All were two-tailed tests. Experiment 2. ZENK and Fos expression in sensitized, nonmaternal, and control female quail Brooding behavior testing, behavioral measures, and group assignment Birds were divided into three groups with six to eight animals in each group. Experimental group, sensitized females (SF) (N = 6): These females were exposed to chicks as in experiment 1. A

22

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

bird was considered a sensitized female if she brooded all four chicks simultaneously, regardless of the latency or duration. All showed brooding behavior by the second trial and continued to display it through the fifth trial. Nonmaternal females (NMF) (N = 8): These females were exposed to chicks as in experiment 1. No female in this group showed any brooding behavior throughout the five trials. Chicks pursued the females, tried to make contact with them, and often made distress calls. However, the females evaded the chicks for the duration of the trial. Females with blocks as control objects (BL) (N = 8): Females were placed in the testing arena for five 20-min trials. On each trial, four square blue blocks (2.54 cm) were placed in the testing arena with the female. This group controlled for neural activity caused by being handled and placed in a test arena. The four blocks were chosen to control for effects of visual stimulation caused by exposure to four novel objects (as opposed to visual stimulation from chicks). Comparisons among these three groups allowed for a more precise interpretation of IEG activity. For example, in a given nucleus, if the SF group was different from the two control groups, it would suggest that the nucleus is most likely involved with brooding behavior, because no other conditions elicited similar levels of IEG expression. Alternatively, if both the SF and NMF groups were similar to each other, but different from the BL group, the IEG expression could be the result of sensory stimulation caused by chicks, and not specific to brooding behavior. If the NMF group was lower than either of the other two groups, it would indicate that there may be inhibition impeding brooding. Lastly, no differences across groups would indicate baseline activity or normal function independent of manipulations, that is, a control nucleus. ZENK and Fos immunocytochemistry and histology Sets of three females (one from each group) were tested consecutively. Immediately following each female’s fifth trial, she was isolated for 1 h and euthanized. Tissue from each set of three females was processed for immunocytochemistry in parallel. In two cases, only brains from the NMF and BL groups were processed for ZENK immunocytochemistry in parallel, due to unequal group sizes. All birds were perfused with 0.9% saline solution followed by a 4% paraformaldehyde solution. Brains were postfixed overnight in PBS with 30% sucrose, embedded in a 30% sucrose, 10% gelatin solution, and then sectioned on a freezing microtome. Brains were sliced in 40 Am coronal sections. Every third section was processed for ZENK immunocytochemistry. Sections were quenched in H2O2, then incubated in goat serum, then incubated at 4jC for 36 –40 h in a rabbit anti-ZENK (EGR-1) polyclonal primary antibody (Santa Cruz, 588, sc-110) at a 1/2000 dilution. (The ZENK protein is a member of the Krox family, and is an acronym for the gene’s other names: Zif-286, EGR-1, NGFI-A, and

Krox-24.) This was followed by incubation in a 1/1000 dilution of a goat anti-rabbit secondary solution (Jackson Labs), then incubation with an ABC Vectastain Elite KitR. Primary, secondary, and ABC solutions were made in PBS with 0.1% triton-X. A diaminobenzodiazepine (DAB) reaction was done to visualize the stained cells. PBS washes were used between all of the preceding steps with the exception of Tris buffer used before and following the DAB reaction. Another third of all the sections from the first four sets of brains processed (1 brain from each group; 12 brains total) were incubated using the Fos antibody. Because this antibody is not commercially available, the amount of tissue that could be processed was limited. Within these four brains from each group, we further selected a range of consecutive sections to process. Based upon preliminary results with the ZENK antibody, we chose sections that most closely overlapped with areas that had previously shown differences. Sections were stained from the most rostral end of the ectostriatum (E) to slightly past the most caudal end of the nucleus rotundus. The protocol for Fos immunocytochemistry was identical to the ZENK protocol with the exception of the primary antibody incubation. The primary antibody (gift from Luc Berghman and Els D’Hondt) was raised against an avian (chicken) Fos sequence and developed specifically for use in birds, which may make it more sensitive to differences in activation by more precise binding to the quail Fos antigen. It has been extensively validated and its specificity has been demonstrated in quail (D’Hondt et al., 1999; Tlemcßani et al., 2000). Sections were incubated at 4jC for 36– 40 h in a 1/3000 dilution. The remaining third of the sections from each bird were nissl stained. Quantification of ZENK-lir and Fos-lir cells A Spot RT color camera and a Nikon Eclipse E800 microscope were used for counting immunoreactive cells. Individual sections were visualized under a 10 objective, digitized using Spot RT v 3.0 software, and analyzed using NIH image. In all nuclei of interest, density counts were made on serial sections stained with the same antibody (120 Am apart). The border around each nucleus was drawn by the experimenters and cells within the border were counted manually. Cells were considered to be ZENK-like immunoreactive (ZENK-lir) or Fos-lir only if a brown/black nucleus with definable borders was visible. The cells with definable borders were typically more darkly stained, but we made no distinction based upon the intensity. Density was calculated as the number of cells/per unit area (Am2) on individual sections throughout a given nucleus. The sections were coded and the experimenter was blind to the condition of each bird. Selection of the nuclei to quantify was based upon areas that were observed to show ZENK or Fos expression and those that were hypothesized to be involved with brooding behavior. Neuroanatomical definition of these nuclei was

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

based primarily on the Kuenzel and Masson (1988) chick atlas. The exceptions to this were the definitions of the borders of the bed nucleus stria terminalis (BSTm, based upon Aste et al., 1998) and of the POM (based upon Aste et al., 1998 and Tlemcß ani et al., 2000). Previous research in quail demonstrates that important differences in IEG expression occur within different rostral – caudal levels of some nuclei including the POM (Tlemcß ani et al., 2000). To ensure that sections at the same rostral – caudal level were compared across groups, sections from each nucleus were identified and labeled for analysis by their position relative to the landmarks commissura anterior (CA) and tractus septomesencephalicus (TSM) (salient fiber tracts visible without any type of stain). For example, the first section in a given nucleus (moving rostral to caudal) in which the TSM appeared was labeled TSM. A section with the designation TSM + 1 would be the next section stained with the same antibody (or nissl stain), 120 Am caudal (i.e., labels were relative to each third of the brain stained with either ZENK, Fos, or nissl). Therefore, three different sections from different thirds of the brain (ZENK, Fos, or nissl) could have the designation TSM. Serial rostral to caudal sections were sampled from each nucleus as described below. POM: Cell counts were made on eight serial sections. The first count began just caudal to the emergence of the TSM, ran through to the CA and continued caudally (Figs. 1B –D). AM: AM was a control nucleus; therefore, our sample focused on sections where the AM, POM, and BSTm were visible through three serial sections (Fig. 1D). BSTm: Cell counts through the BSTm were made on five serial sections, beginning at the CA, and proceeded four sections caudally (Figs. 1D,E). Ectostriatum (E): Density counts through the E were made on four sections. Only sections stained for Fos were counted, as no apparent or discrete ZENK expression was observed in the E. Counts began before the emergence of the TSM and then moved caudally through three sections (Figs. 1A,B). Statistical analyses A two-way ANCOVA was used to analyze the data across groups and sections for ZENK stained sections. To control for differences in each run (each set of three brains processed in parallel) of immunocytochemistry, run was entered as a covariate. If a test was significant for a particular nucleus, Fig. 1. Schematic of the quail brain. Drawings represent rostral (A) to caudal (E) coronal sections. Definitions of nuclei were based upon Kuenzel and Masson (1988) and Aste et al. (1998). 3V, third ventricle; AM, nucleus anterior medialis hypothalami; BSTl, bed nucleus stria terminalis, pars lateralis; BSTm, bed nucleus stria terminalis, pars medialis; CA, commissura anterior; CPa, commissura pallii; E, ectostriatum; FPL, fasciculus prosencephali lateralis; OM, occipitomesencephalic tract; POM, nucleus preopticus medialis; TSM, tractus septomesencephalicus. Drawings approximate section designations for analysis: A, tsm; B, tsm + 2; C, tsm + 4; D, ca; E, ca + 4.

23

Fisher’s Least Significant Difference post hoc tests were performed. The same statistical analyses were used for Fos tissue; however, it was not necessary to use run as a covariate because there was less variability between runs.

24

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

Results Experiment 1 Seventeen of the 31 females (54.8%) and 9 of 21 males (48.2%) displayed brooding behavior. A brooding bird would allow all four chicks to approach and remain under its wings. While the chicks were underneath the wings of a brooding bird, it remained very still with the exception of preening its wings and gently lifting its feathers. It is important to note that the integrated display of these behaviors is unique to the expression of brooding behavior and is not seen in any other contexts. The display of brooding behavior typically marks a dramatic change in behavior as birds, who are initially very fearful of chicks, rapidly become sensitized to them and allow them to approach and be brooded. In tests where birds displayed brooding behavior, it was often shown toward the end of the first trial and always shown by the second trial. The behavior of the 14 females and 11 males that did not show brooding behavior was a marked contrast to the parental birds. They actively avoided the chicks and seemed fearful of them, rarely brooding even one chick. None of them ever brooded two or more chicks. Some of these nonbrooding birds were not tested for all five trials. As these birds would try to avoid the chicks, they would often step on some, potentially harming them. For the safety of the chicks, five of the birds were tested for only three trials and two birds for only four trials. Comparison between female and male brooding behavior There was no significant difference between males and females in their latency to respond to chicks (Fig. 2, top); however, the duration of brooding all four chicks was significantly different (U = 116, N1 = 17, N2 = 9, P < 0.05) (Fig. 2, bottom). This can be attributed to the mean brooding bout duration within each trial. Males would brood four chicks for a shorter duration, move (typically several times), and then return to brooding four chicks. In contrast, females would typically continue to brood all four chicks once they had started. The mean brooding bout duration across all five trials was significantly longer in females (X = 304 s, SE F 45.95) compared with males (X = 28.96 s, SE F 9.65) (U = 124, N1 = 14, N2 = 9, P < 0.01). (Brooding bout duration data were not available for the first three brooding females.) Regardless of their behavior toward chicks, all females continued to lay eggs through the duration of the testing and 2 days afterwards. Testosterone levels There was no significant change between pretesting and post-testing testosterone levels in males that showed brooding behavior. There was no significant change between pretesting and post-testing nonbrooding males either. Tes-

Fig. 2. Mean F SE latency (seconds) to brood (top), and mean F SE duration (seconds) brooding (bottom) of all four chicks across trials 1 – 5. Each trial was 1200 s. Females (N = 17) and males (N = 9). The sex difference in brooding duration was significant (P < 0.05). Only data from brooding birds are shown.

tosterone levels in brooding compared with nonbrooding males were not different pretesting or post-testing. Experiment 2 Behavior All SF birds brooded four chicks during their five trials. All females were brooding by the second trial (X = 803.30 s, SE F 106.84). By the fifth trial, all females were brooding chicks for more than half the total trial time (X = 998.26 s, SE F 46.05). IEG measurements reflect behavior during the fifth trial. ZENK All ZENK-lir cells had a dark nuclear stain. Generally, label appeared specific within the borders of the nuclei quantified below. In addition to these nuclei, there was limited and apparently less discrete staining throughout portions of the telencephalon, including the ectostriatum. POM: There was a significant effect of group [ F(2,103) = 5.868, P < 0.05]. Run, the covariate, was significant [ F(1,103) = 15.978, P < 0.01]. In section CA, the NMF group had a significantly lower density of ZENK-lir cells than either the SF ( P < 0.05) or the BL groups ( P < 0.05). In section CA +

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

25

1, the NMF group again had significantly lower density than SF ( P < 0.05) (Fig. 3, top, and Figs. 4A – C). BSTm and AM: Neither showed significant differences (Fig. 3, middle and bottom). Fos The overall pattern of activation in Fos stained sections was noticeably different from the ZENK stained sections. In general, fewer areas showed Fos expression when compared with ZENK. Fos activation was visible in the BSTm and E. Fos labeled cells were also seen in the nucleus rotundus and other thalamic nuclei but with no apparent group differences. There was little Fos staining in the POM. BSTm: There was a significant effect of group [ F(2, 44) = 12.24, P < 0.01]. In section CA + 4, the SF group had significantly higher levels of Fos expression than either NMF (P < 0.01) or BL (P < 0.01) groups (Fig. 5, top). E: There was a significant effect of group [ F(2,31) = 11.45, P < 0.001]. In section TSM + 2, the SF group had

Fig. 3. Mean F SE density of ZENK-lir cells across sections (rostral – caudal). Top: POM, P < 0.05; different numbers of asterisks indicate significant post hoc differences (Fisher’s LSD) between groups at that section (P < 0.05). In every run of immunocytochemistry, density was lowest in the NMF bird for CA section. Middle: BSTm, P = 0.14. Bottom: AM, P = 0.30. SF (sensitized female: N = 6), NMF (nonmaternal female; N = 8) and BL (female with blocks as control; N = 8). CA, commissura anterior; TSM, tractus septomesencephalicus.

Fig. 4. ZENK-lir expression in the POM (section CA). (A) SF bird. (B) NMF bird. (C) SF bird (higher magnification). (SF and BL groups were not significantly different). Scale bars = 100 Am.

26

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

Fig. 5. Mean F SE density of Fos-lir cells across sections (rostral – caudal). Top: BSTm, P < 0.01. Bottom: Ectostriatum (E), P < 0.001. Different numbers of asterisks indicate significant post hoc differences (Fisher’s LSD) between groups at that section (P < 0.01). N = 4 per group. SF, sensitized female; NMF, nonmaternal female; and BL, female with blocks as control. CA, commissura anterior; TSM, tractus septomesencephalicus.

significantly higher levels of Fos expression than either NMF (P < 0.01) or BL (P < 0.01) groups (Fig. 5, bottom).

Discussion Induction of brooding With a relatively short exposure to four foster chicks, both female and male Japanese quail show brooding behavior with near equal likelihood, although females brood for longer durations. Every female tested for brooding behavior was laying daily before testing and continued to lay during and following testing. Additionally, males did not show a significant change in plasma testosterone regardless of their behavioral response to chicks. In several avian and mammalian species, the onset of parental behavior coincides with changes in steroid hormones (Buntin, 1996; Richard-Yris et al., 1983, 1987; Storey et al., 2000; Vleck and Dobrott, 1993; Wynne-Edwards and Reburn, 2000). However, the current data along with previous research (reviewed in Eisner, 1960) suggest that this generalization does not necessarily apply to brooding behavior.

There are several possible reasons why female quail did not interrupt egg laying and males did not show a drop in testosterone levels with the onset of brooding behavior. One possibility is that large chronic changes in steroid hormone levels are not necessary for brooding behavior in Japanese quail. This certainly would not be first case where a type of parental behavior is uncoupled from large changes in adult steroid hormone levels (Eisner, 1960; Lea et al., 1996; Lonstein and DeVries, 2000, 2001; Rosenblatt, 1967). A more natural sequence of events that typically precedes the onset of brooding behavior (mating, nest building and incubation) may facilitate changes in steroid hormones. However, the current data demonstrate that Japanese quail are capable of showing brooding behavior in the absence of this sequence. With respect to males, we cannot entirely rule out the possibility that a change in testosterone may have occurred, but was not detected. Our sample size may have precluded detecting a change in testosterone levels. However, other studies with comparable sample sizes have demonstrated significant behaviorally induced (initial dominance encounters) changes in testosterone in this species (Ramenofsky, 1984). A final possibility is that the duration of exposure to chicks may have been too short to induce a hormonal change and additional exposure time would have induced a change in steroid hormones. Alternatively, there may have been a transitory change in testosterone levels immediately following exposure to the chicks that was not detected by our sample 3 days later. Such transitory changes in testosterone accompanying parental behavior occur in the biparental hamsters (Phodopus campbelli) (Reburn and WynneEdwards, 1999). However, testosterone samples taken from bobwhite quail 2 days after exposure to foster chicks show a decrease with brooding behavior (Vleck and Dobrott, 1993). Further, experimentation is necessary to determine the precise role of steroid hormones in Japanese quail brooding behavior. The percentage of Japanese quail responding to the stimulus in the current study (about 50%) is similar to other sensitization studies using chronic or longer exposure times in both birds and mammals (Lonstein and DeVries, 2000; Richard-Yris et al., 1983, 1987; Vleck and Dobrott, 1993; Wang and Buntin, 1999), although the source of this variation in quail is unknown at this point. Differences in IEG expression Significant differences across groups in either ZENK or Fos expression were seen in the POM, BSTm, and E. However, it is important to note ZENK and Fos did not always show the same pattern of differences within the same nucleus. It is not uncommon for different IEGs (including Fos and ZENK) to show different patterns of activation caused by identical behaviors (Ball et al., 1997; Kalinichev et al., 2000; Meddle et al., 1999; Numan et al.,

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

1998; Tlemcß ani et al., 2000). The use of multiple antibodies provides a more comprehensive picture of activity as some nuclei may more readily express a certain IEG or changes in IEG expression with behavior. The fact that our Fos antibody was raised against an avian Fos sequence (D’Hondt et al., 1999), whereas our ZENK anti-body was raised against a mammalian ZENK sequence, may also have contributed to these differences. Although further research is required to determine the precise role of each nucleus in brooding behavior, the patterns of IEG expression along with the known neuroendocrinology of this species and related comparative evidence allow inferences about function. It was not surprising that differences were found in IEG expression within the POM and BSTm. Both the POM and BSTm are parts of a series of forebrain nuclei that concentrate sex steroid receptors and are involved with other motivated behaviors in this species, particularly male sexual behavior (Adkins-Regan, 1996; Balthazart et al., 1996). In the ring dove, the POM is crucial for parental behavior (reviewed in Buntin, 1996; Sharp et al., 1996). Homologues of the POM and BSTm (MPOA and BNST) are wellestablished components of maternal behavior (and other social behaviors) in rodent species as demonstrated by lesion and IEG studies (Kalinichev et al., 2000; Newman, 2002; Numan, 1994). Within the BSTm, Fos expression was significantly higher in the SF group. Interestingly, the avian BSTm has a high concentration of vasotocinergic cell, fibers, and receptors (Aste et al., 1998). The role of this neuropeptide (and its mammalian homologue, vasopressin) in affiliative behaviors in both birds and mammals is well known (Goodson and Bass, 2001). It is possible that differences in vasotocin activity might be responsible for the variation in brooding responsiveness seen in our population of quail. Vasopressin receptor distribution is a source of variation in parental responsiveness in rodents (Francis et al., 2002; Hammock and Young, 2002), and vasotocin has been linked with other social behaviors in this species (Balthazart et al., 1996). Within the POM, SF and BL groups showed significantly higher levels of ZENK expression compared to the NMF group. Differences were confined to two discrete sections (CA, CA + 1). It is interesting that male sexual behavior also results in differences in IEG expression within the POM at or near the level of the CA (Ball et al., 1997; Meddle et al., 1999; Tlemcß ani et al., 2000). Homologous circuitry involved in both maternal and sexual behavior has been shown in the rat (reviewed in Newman, 2002). It could be that the same area within the POM in male and female quail influences sexual and maternal behavior, respectively. It is also interesting to note that the sexual dimorphism in this structure is most pronounced at the level of the CA (Thompson and Adkins-Regan, 1994). Whether males show a different pattern of activity in the POM when brooding than females remains to be determined.

27

The lack of difference between the SF and BL groups within the POM suggests that a baseline level of activity of IEG expression is necessary for brooding behavior to occur in response to appropriate stimuli, whereas in the NMF group, this activity is inhibited. IEGs have been previously used to characterize inhibition of maternal behavior in rodents (Sheehan et al., 2000). Exactly how the POM in NMF females is inhibited in the current study is not clear at this point, but it may be associated with a fear response. The absence of a parental response to foster conspecifics is often characterized as a fearful or avoidance response (Numan, 1994). Portions of the amygdaloid complex are involved in responses to fear in birds and mammals (Butler and Hodos, 1996; Newman, 2002). One possibility is that portions of the amygdaloid complex, which have an output to the POA via the BST, are involved with this inhibition. Within the caudal E, Fos expression in the SF group was significantly greater compared with both NMF and BL groups. It is likely that this difference was related to the visual stimulus of the chicks. The E is part of the tectofugal visual pathway in birds that is believed to correspond to the colliculothalmocortical pathway in mammals (Gunturkin, 1994; Husband and Shimuzu, 1999). The E receives visual projections via the retina-optic tectum-nucleus rotundus pathway, and lesions along this pathway (including lesions of the E) produce deficits in visual discrimination tasks (Butler and Hodos, 1996; Engelage and Bischof, 1993). The difference between the SF and NMF groups suggests that it is not the mere sight of the chicks that activates the E, but perhaps a different level of visual attention toward the chicks among brooding females. Efferents from the E (including projections to the neostriatum frontale, pars lateralis, area tempo-parieto-occipitalis and neostriatum intermedium, pars lateralis) suggest that it also integrates visual stimuli with other sensory information (Husband and Shimuzu, 1999). To effectively brood all four chicks, a female sees the position of the chicks, hears their distress calls, feels them underneath her wings, and preens her feathers to ensure that all the chicks are covered. This coordinated effort may require both activation within the E and modification of efferent information to other regions. Visual stimuli may play a particularly important part in parental behavior in birds. Rodents olfactory (Numan, 1994) as well as tactile stimulation associated with suckling (Lonstein and Stern, 1997) likely have a unique influence on the initiation and maintenance of maternal behavior. In lieu of these types of stimulation, visual stimuli may be the most prominent source of information leading to the initiation of parental behavior. Japanese quail as a model for the neural basis of avian parental behavior Do the patterns of IEG expression in sensitized females represent activation of the same neural mechanisms that

28

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29

would control brooding behavior under more natural conditions? It has been clearly established in birds (Buntin, 1996) and mammals (Bridges, 1996) that sex steroid hormones play some role in the induction and maintenance of maternal behavior. Compared to brooding females in the current study, maternal females under natural conditions would have been incubating eggs for over 2 weeks and would have different levels of circulating sex steroid hormones that would likely act upon the central nervous system including the POM and BSTm. However, maternally behaving female quail in any hormonal state would likely show similar patterns of IEG activation. In rats, postpartum females and sensitized virgin females show no significant difference in levels of Fos expression throughout the forebrain (Numan and Numan, 1994). This suggests that in rats, and perhaps quail as well, maternal behavior utilizes the same neural pathways in animals with different hormonal states. The present study begins to identify the neural circuitry underlying maternal behavior in quail showing differences in IEG expression in brooding females when compared with controls. Although these results are promising, further studies are needed to conclusively identify the precise role of the POM, BSTm, and E in quail brooding behavior as well as other nuclei that may be involved with this and other specific components of parental behavior.

Acknowledgments We are grateful to Steve Lamb and the Cornell Veterinary Diagnostic Laboratory for the testosterone assays. We thank Jennifer Sun, Jackie Maffucci, and Jenna Carroll for assistance with data collection and Robert Johnston for the use of his computer software and microscope equipment. We also thank Els D’Hondt and Luc Berghman for their generosity in providing the Fos antibody. We are grateful to Tim DeVoogd and Sarah Newman for providing valuable advice regarding several aspects of these studies. This research was supported by NSF Grant IBN 9514088.

References Adkins-Regan, E., 1996. Neuroanatomy of sexual behavior in the male Japanese quail from top to bottom. Poult. Avian Biol. Rev. 7, 193 – 204. Aste, N., Balthazart, J., Absil, P., Grossmann, R., Mulhbauer, E., VigliettiPanzica, C., Panzica, G.C., 1998. Anatomical and neurochemical definition of the nucleus of the stria terminalis in Japanese quail (Coturnix japonica). J. Comp. Neurol. 396, 141 – 157. Ball, G.F., Tlemcßani, O., Balthazart, J., 1997. Induction of the Zenk protein after sexual interactions in male Japanese quail. NeuroReport 8, 2965 – 2970. Balthazart, J., Tlemcßani, O., Ball, G.F., 1996. Do sex differences in the brain explain sex differences in the hormonal induction of reproductive behavior? What 25 years of research on the Japanese quail tells us. Horm. Behav. 30, 627 – 661. Bridges, R., 1996. Biochemical basis of parental behavior in the rat. In:

Rosenblatt, J.S., Snowdon, C.T. (Eds.), Parental Care: Evolution, Mechanisms and Adaptive Significance. Academic Press, New York, pp. 215 – 237. Buntin, J.D., 1996. Neural and hormonal control of parental behavior in birds. In: Rosenblatt, J.S., Snowdon, C.T. (Eds.), Parental Care: Evolution, Mechanisms and Adaptive Significance. Academic Press, New York, pp. 161 – 202. Butler, A., Hodos, W., 1996. Comparative Vertebrate Neuroanatomy. Wiley-Liss, New York. Clutton-Brock, T.H., 1991. The Evolution of Parental Care. Princeton Univ. Press, New Jersey. D’Hondt, E., Vermeiren, J., Peeters, K., Balthazart, J., Tlemcßani, O., Ball, G.F., Duffy, D.L., Vandesande, F., Berghman, L.R., 1999. Validation of a new antiserum directed towards the synthetic C-terminus of the FOS protein in avian species: immunological, physiological and behavioral evidence. J. Neurosci. Methods 91, 31 – 45. Engelage, J., Bischof, H., 1993. Organization of the tectofugal pathway in birds: a comparative review. In: Zeigler, H.P., Bischof, H. (Eds.), Vision, Brain and Behavior in Birds. MIT Press, Cambridge, MA, pp. 137 – 158. Eisner, E., 1960. The relationship of hormones to the reproductive behavior of birds, referring especially to parental behaviour: a review. Anim. Behav. 8, 155 – 179. Francis, D.D., Young, L.J., Meaney, M.J., Insel, T.R., 2002. Naturally occurring differences in maternal care are associated with the expression of oxytocin and vasopressin (V1a) receptors: gender differences. J. Neurosci. 14, 349 – 353. Goodale, H.D., 1916. Notes on the behaviour of capons when brooding chicks. J. Anim. Behav. 6, 319 – 324. Goodson, J.L., Bass, A.H., 2001. Social behavior functions and related anatomical characteristics of vasotocin/vasopressin systems in vertebrates. Brain Res., Brain Res. Rev. 35, 246 – 265. Gunturkin, O., 1994. Functional organization of the avian visual system. In: Andrew, R.J. (Ed.), Neural and Behavioural Plasticity: The Use of the Domestic Chick as a Model. Oxford Univ. Press, New York, pp. 92 – 108. Hammock, E.A.D., Young, L.J., 2002. Variation in the vasopressin V1a receptor promoter and expression: implications for inter and intraspecific variation in social behavior. Eur. J. Neurosci. 16, 399 – 402. Hess, E.H., Petrovich, S.B., Goodwin, E.B., 1976. Induction of parental behavior in Japanese quail (Coturnix japonica). J. Comp. Physiol. Psychol. 90, 244 – 251. Hoffman, G.E., Smith, M.S., Verbalis, J.G., 1993. c-Fos and related immediate early gene products as markers of activity in neuroendocrine systems. Front. Neuroendocrinol. 14, 173 – 213. Husband, S.A., Shimuzu, T., 1999. Efferent projections of the ectostriatum in the pigeon (Columbia livia). J. Comp. Neurol. 406, 329 – 345. Kalinichev, M., Rosenblatt, J.S., Nakabeppu, Y., Morrell, J.I., 2000. Induction of c-fos-like and fosB-like immunoreactivity reveals forebrain neuronal populations involved differentially in pup-mediated maternal behavior in juvenile and adult rats. J. Comp. Neurol. 416, 45 – 78. Kuenzel, W.J., Masson, M., 1988. Stereotaxic atlas of the brain of the chick (Gallus domesticus). Johns Hopkins Univ. Press, Baltimore. Lea, R.W., Richard-Yris, M.A., Sharp, P.J., 1996. The effect of ovariectomy on concentrations of plasma prolactin and LH on parental behavior in the domestic fowl. Gen. Comp. Endocrinol. 101, 115 – 121. Lonstein, J.S., DeVries, G.J., 2000. Sex differences in the parental behavior of rodents. Neurosci. Biobehav. Rev. 24, 669 – 689. Lonstein, J.S., DeVries, G.J., 2001. Social influences on parental and nonparental responses towards pups in virgin female prairie voles (Microtus ochrogaster). J. Comp. Psychol. 115, 53 – 61. Lonstein, J.S., Stern, J.M., 1997. Somatosensory contributions to c-fos activation within the caudal periaqueductal gray of lactating rats: effects of perioral, rooting, and suckling stimuli from pups. Horm. Behav. 32, 155 – 166. Madge, S., McGowan, P., 2002. Pheasants, Partridges, and Grouse. Princeton Univ. Press, Princeton.

M.G. Ruscio, E. Adkins-Regan / Hormones and Behavior 46 (2004) 19–29 Marler, C.A., Bester-Meredith, J.K., Trainor, B.C., 2003. Paternal behavior and aggression: endocrine mechanisms and nongenomic transmission of behavior. Adv. Study Behav. 32, 263 – 323. Meddle, S.L., Foidart, A., Wingfield, J.C., Ramenofsky, M., Balthazart, J., 1999. Effects of sexual interactions with a male on? Fos-like immunoreactivity in the female quail brain. J. Neuroendocrinol. 11, 771 – 784. Mills, A.D., Crawford, L.L., Domjan, M., Faure, J.M., 1997. The behavior of the Japanese or domestic quail Coturnix japonica. Neurosci. Biobehav. Rev. 21, 261 – 281. Newman, S.W., 2002. Pheromonal signals access the medial extended amygdala: one node in proposed social behavior network. In: Pfaff, D.W. (Ed.), Hormones, Brain and Behavior. Academic Press, Boston, pp. 17 – 32. Nichols, C.R., 1991. A comparision of the reproductive and behavioural differences in feral and domestic Japanese quail. Master of Science thesis, University of British Columbia. Numan, M., 1994. Maternal behavior. In: Knobil, E., Neill, J.D. (Eds.), The Physiology of Reproduction, second ed. Raven Press, New York, pp. 221 – 302. Numan, M., Numan, M.J., 1994. Expression of Fos-like immunoreactivity in the preoptic area of maternally behaving virgin and postpartum rats. Behav. Neurosci. 108, 379 – 394. Numan, M., Numan, M.J., 1995. Importance of pup-related sensory inputs and maternal performance for the expression of Fos-like immunoreactivity in the preoptic area and ventral bed nucleus of the stria terminalis of postpartum rats. Behav. Neurosci. 109, 135 – 149. Numan, M., Numan, M.J., Marzella, S.R., Palumbo, A., 1998. Expression of c-fos, fos B, and egr-1 in the medial preoptic area and bed nucleus of the stria terminalis during maternal behavior in rats. Brain Res. 792, 348 – 352. Opel, H., Proudman, J.A., 1988. Effects of poults on plasma concentrations of prolactin in turkey hens incubating without eggs or a nest. Br. Poult. Sci. 31, 791 – 800. Ramenofsky, M., 1984. Agonistic behaviour and endogenous plasma hormones in male Japanese quail. Anim. Behav. 32, 698 – 708. Reburn, C.J., Wynne-Edwards, K.E., 1999. Hormonal changes in males in a naturally biparental and a uniparental mammal. Horm. Behav. 35, 163 – 176. Richard-Yris, M.A., 1993. Comportement parental chez les gallinacaes: importance du facteur emotivite´ dans la vitesse d’ e´mergence des re´ponses parentales: apport du mode`le caille japonaise. In: Picard, M., Porter, R., Signoret, J.-P. (Eds.), Comportement et adaptation des animaux domestiques aux contraintes de l’e´levage: bases techniques du bien-eˆtre animal. INRA Publications, Paris, pp. 127 – 139. Richard-Yris, M.A., Garnier, D.H., Leboucher, G., 1983. Induction of ma-

29

ternal behavior and some hormonal and physiological correlates in the domestic hen. Horm. Behav. 17, 345 – 355. Richard-Yris, M.A., Leboucher, G., Chadwick, A., Garnier, D.H., 1987. Induction of maternal behavior in incubating and non-incubating hens: influence of hormones. Physiol. Behav. 40, 193 – 199. Richard-Yris, M.A., Guemene, D., Lea, R.W., Sharp, P.J., Bedecarrats, G., Foraste, M., Wauters, A.M., 1998a. Behaviour and hormone concentrations in nest deprived and renesting hens. Br. Poult. Sci. 39, 309 – 317. Richard-Yris, M.A., Sharp, P.J., Wauters, A.M., Guemene, D., Richard, J.P., Foraste, M., 1998b. Influence of stimuli from chicks on behavior and concentrations of plasma prolactin and luteinizing hormone in incubating hens. Horm. Behav. 33, 139 – 148. Rosenblatt, J.S., 1967. Nonhormonal basis of maternal behavior in the rat. Science 156, 1512 – 1517. Sharp, P.J., Li, Q., Georgiou, G., Lea, R.W., 1996. Expression of fos-like immunoreactivity in the hypothalamus of the ring dove (Streptopelia risoria) at the onset of incubation. J. Neuroendocrinol. 8, 291 – 298. Sheehan, T.P., Cirrito, J., Numan, M.J., Numan, M., 2000. Using c-Fos immunocytochemistry to identify forebrain regions that may inhibit maternal behavior in rats. Behav. Neurosci. 114, 337 – 352. Slawski, B.A., Buntin, J.D., 1995. Preoptic lesions disrupt prolactininduced parental feeding behavior in ring doves. Horm. Behav. 29, 248 – 266. Storey, A.E., Walsh, C.J., Quinton, R.L., Wynne-Edwards, K.E., 2000. Hormonal correlates of paternal responsiveness in new and expectant fathers. Evol. Hum. Behav. 21, 79 – 95. Thompson, R.R., Adkins-Regan, E., 1994. Photoperiod affects the morphology of a sexually dimorphic nucleus within the preoptic area of the male Japanese quail. Brain Res. 667, 201 – 208. Tlemcßani, O., Ball, G.F., D’Hondt, E., Vandesande, F., Sharp, P.J., Balthazart, J., 2000. Fos induction in the Japanese quail brain after expression of appetitive and consummatory aspects of male sexual behavior. Brain Res. Bull. 52, 249 – 262. Vleck, C.M., Dobrott, S.J., 1993. Testosterone, antiandrogen and alloparental behavior in bobwhite quail foster fathers. Horm. Behav. 27, 92 – 107. Wang, Q., Buntin, J.D., 1999. The roles of stimuli from young, previous breeding experience, and prolactin in regulating parental behavior in ring doves (Streptopelia risoria). Horm. Behav. 35, 241 – 245. Wynne-Edwards, K.E., Reburn, C.J., 2000. Behavioral endocrinology of mammalian fatherhood. Trends Ecol. Evol. 15, 464 – 467. Youngren, O.M., El Halawani, M.E., Phillips, R.E., Silsby, J.L., 1989. Effects of preoptic and hypothalamic lesions in female turkeys during a photoinduced reproductive cycle. Biol. Reprod. 41, 610 – 617.