- Email: [email protected]
Male pheromones initiate prolactin-induced neurogenesis and advance maternal behavior in female mice
Available online at www.sciencedirect.com
Hormones and Behavior 53 (2008) 509 – 517 www.elsevier.com/locate/yhbeh
Male pheromones initiate prolactin-induced neurogenesis and advance maternal behavior in female mice Caroline M. Larsen ⁎, Ilona C. Kokay, David R. Grattan Centre for Neuroendocrinology and Department of Anatomy and Structural Biology, Box 913, University of Otago, Dunedin 9001, New Zealand Received 11 September 2007; revised 23 November 2007; accepted 27 November 2007 Available online 15 December 2007
Abstract Prolactin is required for rapid onset of maternal behavior after parturition, inducing adaptive changes in the maternal brain including enhanced neurogenesis in the subventricular zone during pregnancy. The resultant increase in olfactory interneurons may be required for altered processing of olfactory cues during the establishment of maternal behavior. Pheromones act through olfactory pathways to exert powerful effects on behavior in rodents and also affect prolactin secretion. Hence, this study aimed to investigate the effect of male pheromones on neurogenesis and maternal behavior in female mice. Virgin female mice were housed individually or in split-cages where they had pheromonal but not physical contact with a male. Maternal behavior was assessed in a foster pup retrieval paradigm. Some mice were injected with bromodeoxyuridine, and the labeled cells visualized using immunohistochemistry. The data show that exposure to male pheromones, for a duration equivalent to a murine pregnancy, advanced maternal behavior in both virgin and postpartum female mice. The pheromone action was dependent on prolactin and ovarian steroids, and was associated with increased cell proliferation in the subventricular zone and subsequent increases in new neurons in the olfactory bulb. Moreover, the effect of pheromones on both cell proliferation and maternal behavior could be induced solely through administration of exogenous prolactin to mimic the pheromone-induced changes in prolactin secretion. The data suggest that male pheromones induce a prolactin-mediated increase in neurogenesis in female mice, resulting in advanced maternal behavior. © 2007 Elsevier Inc. All rights reserved. Keywords: Maternal behavior; Neurogenesis; Mice; Prolactin
Introduction Multiple adaptations occur in the maternal brain to prepare the mother for the demands of pregnancy and lactation. Prolactin, acting in the brain, has been implicated in the onset of some maternal behaviors in rodents (Bridges et al., 1990, 1997; Bridges and Ronsheim, 1990). Steroid-primed virgin rats exhibit maternal behavior earlier when given central infusions of prolactin, whereas maternal behavior is delayed when bromocriptine is used to block endogenous prolactin secretion (Bridges and Ronsheim, 1990). Moreover, prolactin receptor-deficient mice have markedly impaired maternal behavior (Lucas et al., 1998), clearly demonstrating a requirement for prolactin action. Based on prolactin receptor localization in the medial preoptic area (MPOA) (Bakowska and Morrell, 1997; Pi and Grattan, ⁎ Corresponding author. E-mail address: [email protected] (C.M. Larsen). 0018-506X/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.yhbeh.2007.11.020
1998), and localized microinjection studies (Bridges and Ronsheim, 1990), the MPOA is seen as a probable site of prolactin action in the brain in inducing maternal behavior. Prolactin also mediates a pregnancy-induced increase in neurogenesis in the subventricular zone (SVZ) of the adult forebrain that results in the incorporation of new neurons into olfactory pathways of the maternal brain (Shingo et al., 2003). It is postulated that this change results in altered olfactory perception to facilitate maternal behavior. Pheromones act through the nose to influence behavior in most mammalian species (Meredith, 2001). Pheromones suppress prolactin secretion under certain circumstances (Li et al., 1989, 1990), and thus may influence maternal behavior. Pheromones are processed in the olfactory (Yoon et al., 2005) and vomeronasal systems (Keverne and de la Riva, 1982), with tertiary projections via the amygdala to the MPOA (Jia et al., 1997; Li et al., 1989, 1990). As pheromones affect serum prolactin levels, and indirectly affect areas of the brain implicated in maternal behavior
510
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
(Numan and Numan, 1994), we hypothesized that female mice exposed to male mouse pheromones would show reduced levels of prolactin, resulting in suppressed maternal behavior. Previous research has demonstrated that elevated prolactin during pregnancy, or 6 days of prolactin infusion (icv or subcutaneous), increases the number of new neurons that are made in the SVZ in the brain (Shingo et al., 2003). Therefore, we further hypothesized that sustained contact with male pheromones might alter neurogenesis in the SVZ and thus influence maternal behavior. The aim of this study was to test these hypotheses by housing adult virgin female or pregnant female mice either alone, or in divided cages in the presence of a male such that the mouse could have pheromonal but not physical contact (split-cages). To evaluate the effect of pheromones, we examined maternal behavior in both virgin female and postpartum mice in a pup retrieval paradigm. To determine whether pheromones affected neurogenesis, we examined the uptake of bromodeoxyuridine (BrdU) in the SVZ after varying periods of time exposed to male mouse pheromones, as well as chronic changes in BrdU-labeled neurons in the olfactory bulb. Some of these data have been reported previously, in abstract form (Larsen et al., 2006). Materials and methods Subjects Virgin maternally naive 6–9 week old C57BL/6J female mice were either housed in 32 × 16 × 18 cm individual cages or in 28 × 54 × 18 cm split cages. The split cages had a 54 × 18 cm metal divider, which bisected the cage longitudinally, with a series of 0.1 cm holes allowing pheromonal but not physical contact with another mouse. In addition, maternally naive virgin mice were mated with a stud male to examine if pheromone exposure throughout pregnancy affected maternal behavior expression to foster pups. These mice were considered day 1 pregnant when sperm was present in a vaginal smear. The mice were housed under a 12 h light–dark cycle, with ad libitum access to food and water. The University of Otago Animal Ethics Committee approved all experiments.
Maternal behavior testing Female naive mice or day 2 postpartum mice had three foster pups placed at the opposite end of the cage from their nest. In each test session, mice were continuously observed and any behavior recorded i.e. exploration, grooming and sniffing of the pups, nest building, pup retrieval, or crouching over the pups. Full maternal behavior was defined as being when the mouse had gathered all 3 pups to her nest and was crouching over them. As there were no significant differences in any of the parameters within a group, and each measured variable reflected the overall trend within a group, only the data for full maternal behavior has been included. In the first experiment, maternal behavior was compared between individually-caged mice (n = 6) and mice housed in a splitcage with a male mouse (n = 6). Split-caged mice were housed in the split cage for 21 days prior to testing, and remained in the split-cage for the duration of the experiment. Testing was initially repeated daily for 6 days (data not shown), but was limited to 4 days after the first experiment as after this point there were no notable time differences. Testing was performed between 9:30–11 am. After completion of testing on day 4, animals were immediately killed by cervical dislocation, trunk blood collected and centrifuged, and the plasma frozen at − 20 °C until required. To determine whether any effects were specifically caused by pheromones, the experiments were repeated by housing female mice in individual cages for 21 days with 0.05 g of male urine soaked bedding placed daily in their cage for 21 days (n = 6, four male mice were group housed to provide urine soaked sawdust). A control group (n = 6) had their cage lid lifted briefly daily and their
bedding disturbed. Both groups were then tested daily for maternal behavior for 4 days. To evaluate the role of ovarian hormones in any pheromone induced actions, female mice were bilaterally ovariectomized under halothane anesthesia 1 week prior to the experiment, and then housed either alone for 21 days (n = 6) or in a split cage with a male for 21 days (n = 6). Maternal behavior was then tested daily for 4 days, and compared with ovary-intact females either housed alone for 21 days (n = 6) or in split-cages for 21 days (n = 6). Finally, to determine whether pregnant animals might also be affected by pheromones, female mice were mated overnight with a stud male and deemed to be day 1 pregnant when sperm was visualized in a cervical smear. The male was then removed, and females were exposed to male urine soaked bedding for the duration of the pregnancy and subsequently tested for maternal behavior to foster pups on day 2 postpartum (n = 6).
Time course of effect of pheromones on prolactin secretion and on maternal behavior Virgin female maternally naive mice were either housed alone, or in a split cage with a male for 1, 2, 4, 7, 10, or 21 days (all groups n = 6). All groups were then tested for maternal behavior, as described above. Additional groups were sacrificed after 1, 3.5 and 6 h of pheromonal contact for collection of trunk blood for hormone measurements. Serum prolactin from the various groups was measured by double antibody radioimmunoassay (RIA) using NIH reagents. Mouse prolactin (NIDDK-mouse PRL, batch #AFP6476C) was iodinated by the chloramine-T method and used at a concentration of 15,000 cpm/tube. Primary antiserum (NIDDK rabbit anti-mouse PRL-RIA-9, batch # AFP-131078) was added at 1:70,000. Data are expressed in terms of a mouse prolactin standard curve generated using the same prolactin preparation as that used for iodination. Sensitivity was 2 ng/ml. All samples were run in a single assay, with an intra assay % C.V. 2.5%.
Effect of pheromone-induced changes in serum prolactin levels on maternal behavior To evaluate the role of prolactin in pheromone effects on maternal behavior, female mice housed in individual cages were injected s.c. with a slow release preparation of prolactin (50 μg of purified ovine prolactin in a 25% polyvinylpyrrolidone (PVP): 25% H2O: 50% NaHCO3 buffer) or an identical vehicle every 12 h for 72 h (equivalent to 24–72 h of pheromone exposure). To block the pheromone-induced rise in endogenous prolactin, split-caged housed female mice were injected s.c. with bromocriptine (10 mg/kg; 300 μg bromocriptine mesylate (Sigma) dissolved in a minimum amount of ethanol and made up to 0.1 ml with sesame oil), or an identical vehicle, every 12 h, for the first 24–72 h of pheromonal exposure. All groups (n = 6) were tested for maternal behavior 18 days after manipulation of prolactin levels, to simulate the period of time required for the advancement of maternal behavior following the hyperprolactinemia seen from 24–72 h of pheromone exposure.
Effect of male pheromones on neurogenesis in adult female mice Split-caged female mice after 3, 7 or 14 days of pheromonal contact, or individually-caged mice (all groups n = 6) were injected with a series of 6 i.p. BrdU (Sigma) injections 2 h apart (12 mg/100 g of body weight, dissolved in phosphate buffer). The final dose was given 2 h before perfusion. A further group was injected on day 7 of pheromonal contact and killed 14 days later. BrdU-injected mice were anaesthetized with pentobarbitone and perfused with 4% paraformaldehyde. Brains were postfixed for 24 h and then cryoprotected in sucrose prior to being frozen at − 80 °C until tissue processing. Coronal sections (20 µm) of the olfactory bulb and subventricular zone were cut in a cryostat and BrdU-labeled cells detected using immunohistochemistry (mouse monoclonal anti-BrdU, 1:200, Dako, Carpinteria, CA). Fluorescence immunohistochemistry was performed to establish the cell type of the Brdu labeled cells. The primary antibodies used were rat monoclonal anti-BrdU (1:100 AbD, Serotec, OBT0030 S), mouse anti-neuN, (1:50, Chemicon MAB377), and guinea pig anti-doublecortin (1:100, Chemicon, AB5910). Secondary antibodies used were AlexaFluor tagged Alexa 488 (green) and Alexa 568 (red). The fluorescent stained sections were counterstained with 4′6-diamidino-2-phenylindole (DAPI,
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517 Molecular probes, D-1306). For quantification, subventricular zone labeled cells were observed under 200× magnification, and positively labeled cells counted in a systematic manner after a random start, such that all labeled nuclei were counted in 1-in-6 serial sections 120 µm apart for a total of 14 sections throughout the subventricular zone (approximately Bregma + 1.94 mm– 3.9 mm). For the olfactory bulb, 1-in-3 serial sections throughout the entire olfactory bulb were analyzed and all labeled cells counted. Data are presented as the total counts collected, and are not corrected for the total number of sections. To investigate the role of endogenous prolactin in pheromone-induced neurogenesis, groups of split-caged and individually-housed mice were injected s.c. with bromocriptine for the first 24–72 h of pheromonal exposure (as described above). Mice were then injected with BrdU on day 7 of pheromonal contact and processed for immunohistochemistry, as above.
Statistical Analysis The data are expressed as mean ± SEM. Maternal behavior data were analyzed using a repeated measures ANOVA, with p b 0.05 used as the level of significance. Serum prolactin levels (ng/ml) and total counts of Brdu-labeled cells were compared by one-way ANOVA and Fishers PLSD post hoc test.
Results
511
mice, they showed a significantly reduced time to retrieve all pups on day 2 compared to day 1 (p b 0.05). From the third day of maternal behavior testing there were no further significant differences between the two groups. Therefore, the prolonged presence of a male, or brief bouts of physical contact with a pup over several days, caused enhanced maternal behavior. Male pheromones advance maternal behavior To determine whether the observed effect was specifically mediated by pheromones, we examined whether exposure to male mouse urine was sufficient to observe the enhanced maternal behavior. Pheromones are bound to major urinary proteins, and hence placing urine-soaked bedding in with the female will expose her to pheromones without the confounding effects of a companion. Individually-housed females exposed to male urine for 21 days were significantly faster to express full maternal behavior on day 1 of testing, compared to controls (Fig. 1B, p b 0.05). Thus, the advancement of maternal behavior was specifically mediated by male mouse pheromones.
The presence of a male mouse advances maternal behavior Individually-housed virgin female mice spontaneously expressed full maternal behavior to foster mouse pups on day 1 of maternal behavior testing (Fig. 1A). While there is some variability between different mouse strains, the spontaneous expression of full maternal behavior on the first day of testing is consistent with other accounts in the literature, which show that females of the inbred C57BL/6J mouse strain exhibit spontaneous pup retrieval (Brown et al., 1999). On day 2, they showed a significant reduction in time to express full maternal behavior (p b 0.05). While there was a trend for further decreases in time until the fourth day, this difference was not significant. By contrast with the individually-housed mice, virgin females splitcaged with a male were significantly faster to express full maternal behavior on both day 1 and day 2 of maternal behavior testing (Fig. 1A, p b 0.05). As seen with individually-housed
Ovarian steroid hormones are essential for the male mouse advancement of maternal behavior To evaluate the role of ovarian hormones in mouse maternal behavior, virgin female mice were ovariectomized (ovx), housed in individual or split cages for 21 days, and then tested daily for 4 days for maternal behavior. As seen previously, split-caged intact mice were significantly faster on day 1 and 2 to express full maternal behavior compared to individually-caged intact mice (Fig. 1C, p b 0.05). Ovx, individually-caged mice showed no differences in the spontaneous expression of maternal behavior compared to individually-caged controls. Maternal behavior in the ovx, split-caged animals, however, was not significantly different from that in individually-housed mice. Hence, removal of the ovarian steroid hormones completely prevented the pheromone-induced advancement of maternal behavior.
Fig. 1. Data presented as mean ± SEM, all groups n = 6. (A) Female mice split-caged with a male mouse expressed full maternal behavior more rapidly than controls (⁎p b 0.05). Both individually-caged and split-caged mice were faster (+p b 0.05) to express maternal behavior on day 2 of testing. (B) Individually-caged females exposed to male urine for 21 days (pheromone exposed) expressed full maternal behavior more rapidly than mice housed alone (⁎p b 0.05). (C) Intact split-caged females expressed full maternal behavior more rapidly than controls (⁎p b 0.05). In contrast, ovariectomized split-caged females were not different from intact mice housed alone.
512
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
The male needs to be present for longer than 14 days for the pheromonal advancement of maternal behavior to occur To establish the time course of the male pheromone-induced advancement of maternal behavior, we performed maternal behavior testing after varying periods of pheromone exposure. In contrast to the effect of long-term pheromone exposure, females housed with a male in a split cage for 48 h showed significantly impaired maternal behavior (p b 0.05, Fig. 2). At all other measured time points up to 14 days of pheromone exposure, there were no significant differences in maternal behavior compared to individually-housed controls. However, after 21 days, the split-caged females were significantly faster (p b 0.05) to express full maternal behavior. Hence, pheromone-induced advancement of maternal behavior took between 14–21 days to manifest. Pheromone exposure leads to a sustained increase in serum prolactin levels Male mouse pheromones have been reported to acutely suppress prolactin secretion (Li et al., 1989, 1990; Ryan and Schwartz, 1980). Therefore, we examined serum prolactin levels in the split-caged mice. Although there was an initial trend for reduced prolactin, this was not significant (p = 0.09), and by 24 h of exposure serum prolactin levels were significantly increased compared to controls (p b 0.05). Serum prolactin levels remained elevated until 72 h of contact (Fig. 3A). After this time, prolactin levels returned to normal, and there were no further significant differences until day 21 when the levels were significantly reduced (p b 0.05), compared to controls (Fig. 3A). Sustained hyperprolactinemia advances maternal behavior 2 weeks later To evaluate the role of the initial pheromone-induced rise in prolactin, individually-housed females were given injections of prolactin every 12 h for 72 h to mimic the pheromone-induced changes. When subsequently tested for maternal behavior 18 days later, they were significantly faster to express full maternal behavior, compared to controls given injections of
Fig. 3. Data presented as mean ± SEM, all groups n = 6. (A) Split-caged females showed increased serum prolactin levels from 24–72 h of pheromone exposure (⁎p b 0.05). Levels were decreased after 21 days (⁎p b 0.05). (B) Individuallyhoused mice injected with slow release prolactin from 24–72 h of pheromonal contact, were faster to express maternal behavior 18 days later (⁎p b 0.05), whereas, split-caged females injected with bromocriptine from 24–72 h of pheromonal contact showed no differences in maternal behavior expression 18 days later, compared to vehicle-treated controls.
vehicle alone (Fig. 3B). Conversely, split-caged mice given 12 hourly injections of bromocriptine from day 1 to day 3 of pheromonal contact with a male, to block the pheromone-induced increase in prolactin, showed no differences in expression of maternal behavior compared to controls when behavior was tested 18 days later (Fig. 3B). Hence, the pheromone-induced advance of maternal behavior was dependent on the rise in prolactin that followed exposure to male pheromones. Pheromones initiate a prolactin-dependent increase in mitogenesis of neural precursor cells in the subventricular zone
Fig. 2. Data presented as mean ± SEM, all groups n = 6. Female mice split-caged with a male for 2 days had impaired maternal behavior (+p b 0.05). In contrast, split-caged mice with a male for 21 days, expressed advanced maternal behavior (⁎p b 0.05). All other measured time points were not different from individuallyhoused controls.
Split-caged females showed a significant, 227% increase in BrdU-labeled cells in the SVZ after 7 days of pheromonal contact, compared to individually-housed controls (Fig. 4A, p b 0.05). This change reduced to control levels by 14 days of contact. Ovariectomy caused a significant decrease in BrdUlabeled cells in the SVZ in both split-caged and individuallyhoused mice compared to intact controls (Fig. 4B, p b 0.05), and completely prevented the pheromone-induced increase in cell proliferation in this region. The data suggest that a time and ovarian steroid-dependent action of male pheromones stimulates
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
513
Fig. 4. Data presented as mean ± SEM, all groups n = 6. (A) Female mice split-caged with a male for 7 days showed an increase in BrdU-labeled cells in the SVZ (⁎p b 0.05), which reverted to baseline levels by day 14, compared to controls. (B) BrdU-labeled cells were increased in the SVZ after 7 days (⁎p b 0.05), in intact split-caged females. Ovariectomized (OVX) mice were less than intact (+p b 0.05), and showed no response to pheromone exposure. Representative images of the groups in Figs. 4 A and B are shown below. Inset shows high power view of BrdU-labeled cells. (Lat. V = lateral ventricle, CC = corpus callosum, SVZ = indicates region of the subventricular zone, adjacent to the lateral wall of the lateral ventricle). (C) Split-caged mice injected with vehicle from 24–72 h of pheromonal contact, showed an increase in neurogenesis after 7 days of contact with the male (⁎p b 0.05), compared to individually-housed controls injected with vehicle. In contrast, split-caged females injected with bromocriptine from 24–72 h of pheromonal contact, showed no differences in neurogenesis after 7 days of contact with a male, compared to controls.
cell proliferation in the SVZ. To establish if the pheromoneinduced increase in mitogenesis in the SVZ was mediated by the period of hyperprolactinemia from 24–72 h of pheromonal contact, prolactin secretion was inhibited by bromocriptine over this 48-hour period (as described above). Split-caged females injected with vehicle from 24–72 h of pheromone contact showed a significant increase in BrdU-labeled cells in the SVZ
on day 7 of pheromone contact compared to individually-housed controls (Fig. 4C). In contrast, split-caged females injected with bromocriptine from 24–72 h of pheromone contact showed no differences in number of BrdU-labeled cells after 7 days of contact with a male, compared to vehicle-injected controls (Fig. 4C). Thus, suppression of the pheromone-induced period of hyperprolactinemia, earlier found to prevent the pheromone-
514
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
induced enhancement of maternal behavior, also abolished the pheromonal-induced increase in neurogenesis. Pheromones increase the amount of new neurons in the olfactory bulb Split-caged females injected with BrdU on day 7 of pheromonal contact also showed a marked 68% increase in labeled cells in the olfactory bulb 14 days later compared to individually-housed controls (Fig. 5). Again, the pheromoneinduced increase in olfactory neurons was dependent on the presence of ovarian hormones (Figs. 5). Ovx split-caged and individually-caged mice both showed a significant decrease in BrdU-labeled cells in the olfactory bulb compared to intact controls. Cells generated in the SVZ can become new neurons or glial cells. Our preliminary analysis suggests that a significant proportion of the BrdU-labeled cells, both in the SVZ and in the OB could be labeled with neuron specific markers. In split-caged mice, 96% of BrdU immunoreactive cells in the SVZ co-expressed doublecortin, and 88% of BrdU immunoreactive cells in the OB co-expressed Neu N (Fig. 5). Male pheromones advance maternal behavior to foster pups in day 2 postpartum females To determine whether the pheromone effect seen in virgin animals could also influence mice during a normal pregnancy,
Fig. 6. Data presented as mean ± SEM, n = 6. Virgin female mice exposed to male urine for 21 days (pheromone exposed) expressed full maternal behavior more rapidly than virgin controls (+p b 0.05). Postpartum mice expressed maternal behavior to foster pups more rapidly than pheromone and nonpheromone exposed virgin mice (⁎p b 0.05). Individually-caged day 1 pregnant females exposed to male urine for the duration of the pregnancy (pheromone exposed) expressed full maternal behavior more rapidly on day 2 postpartum than non-pheromone exposed day 2 postpartum mice (+p b 0.05).
we housed pregnant mice in the presence of male pheromones. As expected, individually-housed day 2 postpartum female mice spontaneously expressed maternal behavior to foster pups on day 1 of maternal behavior testing (Fig. 6) Similar to the pheromonal advancement of maternal behavior seen in virgin mice, however, the day 2 postpartum mice exposed to male pheromones for the duration of the pregnancy were significantly faster to express full maternal behavior to foster pups than
Fig. 5. Data presented as mean ± SEM, n = 6. BrdU-labeled cells were increased in the olfactory bulb 14 days after injection (⁎p b 0.05), in intact split-caged females, in the glomerular and periglomerular layers of the olfactory bulb. Ovariectomized split-caged mice again showed no response to pheromone exposure and were less than intact mice (+p b 0.05). Photomicrographs A–F show representative examples of the immunohistochemistry from intact animals (A, B) and ovariectomized animals (C, D), with controls animals in the upper panels (A, C) and pheromone exposed in the lower panels (B, D). Examples of merged images of OB cells immunoreactive for BrdU (green), NeuN (red), counterstained with DAPI unclear stain (blue) are also shown (E, F). Counting of dual labeled cells showed an 88% increase in the number of new neurons in the granule and periglomerular layers of the olfactory bulb of split-caged housed mice.
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
individually housed post partum females. Thus, the pheromone effect was not restricted to virgin females, and could influence behavioral adaptation during a normal pregnancy. Discussion Virgin female or day 2 postpartum mice exposed to male mouse pheromones for 21 days were significantly faster to express full maternal behavior to foster pups than mice that had been housed individually. One possible explanation for the enhanced maternal behavior was that it was being triggered by the presence of a companion or an enriched environment. The advancement of maternal behavior induced by split-cage housing with a male, however, was completely replicated by exposure to male urine alone, suggesting it was specifically mediated by pheromones. Although ovariectomy did not alter normal expression of maternal behavior, it completely blocked the pheromoneinduced advancement of maternal behavior. The pheromone effect could also be replicated by treatment with exogenous prolactin approximately 2 weeks prior to maternal behavior testing, to mimic the endogenous changes in prolactin that were induced by pheromone exposure. Moreover, the pheromone effect could be completely prevented by blocking pheromoneinduced prolactin secretion. These findings suggest that an extended period of exposure to male mouse pheromones exerted a prolactin-dependent action to prime female mice to more rapidly express full maternal behavior. The advancement in maternal behavior was dependent on ovarian steroids and correlated with a prolactin- and time-dependent increase in cell proliferation in the SVZ and a subsequent increase in new neurons in the olfactory bulb. Importantly, our original observation of pheromoneinduced neurogenesis (Larsen et al., 2006) has recently been repeated by others (Mak et al., 2007). The data suggest that the pheromone-induced incorporation of new neurons into olfactory pathways, mediated by increased prolactin secretion, may facilitate expression of maternal behavior. Both odorants and pheromones affect reproductive behaviors. Blockade of olfactory receptors leaves many reproductive behaviors intact (Powers and Winans, 1973), however, whereas blockade of vomeronasal receptors severely affects many reproductive behaviours (Meredith, 1986; Powers and Winans, 1975). Mice lacking Trpc2, an ion channel only expressed in the vomeronasal organ (Liman et al., 1999), have deficits in gender discrimination and maternal behavior (Kimchi et al., 2007). Pheromones have been shown to act through both olfactory and vomeronasal pathways (Keverne and de la Riva, 1982; Yoon et al., 2005), but the behaviors seen in Trpc2 deficient mice infer that the vomeronasal organ mediates gender-specific reproductive behavior in mice. Exposure to male urine-soaked bedding selectively initiates Fos activity in the accessory bulb and the medial amygdala (Kang et al., 2006), consistent with the hypothesis that a vomeronasal pathway is being activated. We cannot eliminate the possibility, however, that olfactory cues might be contributing to the observed response. Given the critical role of prolactin in the onset of maternal behavior, our initial hypothesis, that the females exposed to male pheromones would show impaired maternal behavior, was based
515
on previous studies reporting that male mouse pheromones lower serum prolactin levels in female mice. Pheromones carried by male mouse major urinary proteins activate the tuberoinfundibular dopamine (TIDA) neurons in the arcuate nucleus of the hypothalamus, which in turn leads to suppression of prolactin secretion (Li et al., 1989, 1990). This mechanism may underpin the “Bruce effect”, whereby early pregnancy in mice is terminated if a foreign male mouse is encountered (Keverne, 1998), through suppression of prolactin secretion and subsequent loss of luteotrophic support for the corpus luteum. In the present model, however, serum prolactin levels were only briefly and not significantly lower in the split-caged mice after one hour of male mouse contact. It has been demonstrated that a male mouse of a different strain from the female is required to produce the “Bruce effect” in inbred mouse colonies (Parkes and Bruce, 1961). As the males in the split cages were from the same inbred colony as the females, this possibly explains why the drop in serum prolactin was only brief and not significant. In contrast to our expectation, we observed a pheromoneinduced period of sustained hyperprolactinemia. It is well established that prolactin treatment advances maternal behavior in rats (Bridges et al., 1997), and in mice (Voci and Carlson, 1973). During pregnancy the placenta produces placental lactogens (Colosi et al., 1988), which circulate in serum, and bind to prolactin receptors with equal affinity as prolactin (Lee and Voogt, 1999), ensuring a continuously high level of prolactin receptor stimulation. Thus, both pregnancy and pheromone exposure are characterized by a sustained period of hyperprolactinemia. Manipulation of prolactin levels, either by administering prolactin to individually-housed animals or blocking prolactin in pheromone-exposed animals, demonstrated that this rise in serum prolactin levels was essential for the enhanced maternal behavior. At all other time points tested prior to the significant advancement of maternal behavior at 21 days, the split-cage mice had no significant differences in the time taken to express maternal behavior compared to individually housed controls. Hence, although the pheromone-induced increase in prolactin occurred almost immediately, pheromonal advancement of maternal behavior took between 14 and 21 days to manifest. Interestingly, 21 days is the same duration as in a rodent pregnancy, suggesting that pheromonal contact associated with the mating interaction might facilitate behavioral adaptations required for subsequent maternal behavior. The prolonged time required to observe the enhancement of maternal behavior after pheromone exposure led us to hypothesize that the effect might involve generation of new neurons in the olfactory bulb. It has previously been shown that prolactin stimulates an increase in neurogenesis in the SVZ during early pregnancy (Shingo et al., 2003), and these new neurons take approximately 14–21 days to be integrated into olfactory circuits (Lois and Alvarez-Buylla, 1994) Thus, changes in neurogenesis would be expected to take 2–3 weeks before they could exert an effect on animal behavior. A prolactin-dependent increase in interneurons incorporated into the olfactory bulb is correlated with the enhanced maternal behavior seen both after pregnancy (Shingo et al., 2003) and after sustained pheromonal exposure (present data). The amount of new olfactory bulb granule cell
516
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
layer interneurons is important for the ability to discriminate between different odors (Desmaisons et al., 1999; Gheusi et al., 2000; Laurent et al., 1996; Shepherd and Greer, 1998). The pheromone-induced increase in inhibitory interneurons in the olfactory bulb may therefore potentially alter the processing of odors. Thus, we propose that sustained exposure to pheromones stimulates a prolactin-induced increase in neurogenesis in the SVZ, which alters maternal behavior some 2–3 weeks later. Steroid hormones have been shown to be essential for prolactin induction of maternal behavior in rats (Bridges et al., 1997). Estrogen may influence maternal behavior directly, through estrogen receptors in the MPOA (Rosenblatt et al., 1994), or indirectly by influencing PRL secretion (Lieberman et al., 1981; Pilotte et al., 1984) or prolactin receptor expression (Muccioli et al., 1991; Mustafa et al., 1995; Pi et al., 2003). Ovariectomy had no effect on maternal behavior expression in individually-housed mice, suggesting that estrogen is not crucial for either the initiation or continuation of spontaneous maternal behavior in the mouse. However, ovariectomy prevented both the increase in neurogenesis and the advancement in maternal behavior induced by male mouse pheromones. It seems likely that the presence of estrogen may be permissive to allow normal prolactin action on this function, by maintaining expression of prolactin receptors in the brain (Anderson et al., in press). As there is often a large amount of redundancy in behaviors crucial for a species survival, it is possible that while estrogen is not essential for murine maternal behavior, that the estrogendependent pheromonal enhancement of maternal behavior is an additional biological safety measure, to ensure that dependent offspring are rapidly nurtured soon after birth. By contrast with the pheromone-induced enhancement of maternal behavior after 21 days exposure, after 48 h of pheromonal contact there was a striking impairment in maternal behavior that returned to baseline levels by 72 h of male contact. This acute inhibitory effect could correlate with the Bruce effect, where the presence of an unknown male mouse will terminate pregnancy for up to 2 days after conception (Bruce, 1959). When the female mouse mates, she forms an imprinted memory of the odor of any males present, which leads to suppression of estrous (Bellringer et al., 1980; Rosser and Keverne, 1985). When an unfamiliar male is present within 48 h of copulation, pregnancy is blocked and estrous is induced (Parkes and Bruce, 1961). In addition, new dominant males kill pups they have not sired (Huck et al., 1982). It is possible then, that to optimize survival of the offspring, the mother will not spend energy on either continuing with a pregnancy (Bruce, 1959), or on maternal behavior (present data), if an unfamiliar dominant male is present. Maternal behavior is essential for survival of mammalian offspring. Normally, during the 21 days of a rodent pregnancy, a profound range of adaptations occur in the maternal brain that lead to rapid maternal behavior expression after birth. The pheromonal influences described above may represent a previously unrecognized contribution to the signals underlying these adaptations. As the pheromonal enhancement of maternal behavior and increase in neurogenesis were both mediated by prolactin, our present results suggest that pheromones are acting through a common pathway to enhance neurogenesis and maternal be-
havior. We propose that the male pheromone-induced hyperprolactinemia initiates increased cell proliferation in the SVZ with a subsequent increase in olfactory bulb interneurons. It seems likely that the change in olfactory function, acting through tertiary projections to the MPOA, enhances maternal behavior. Interestingly, although the changes are dependent on prolactin and ovarian steroids in the mother, the changes can occur without the influence of the hormones from the placenta and the fetus. Placental and fetal hormones normally provide multiple signals during pregnancy to ensure establishment of maternal behavior at term. While pheromones have previously been reported to exert powerful actions on the reproductive system, this study shows for the first time that male pheromones potentially complement the prolactin-mediated establishment of maternal behavior. References Anderson G.M., Kieser D.S., Grattan D.R., in press. Prolactin suppression of pulsatile LH secretion requires an estradiol-dependent upregulation of prolactin receptor mRNA. Endocrinology. Bakowska, J.C., Morrell, J.I., 1997. Atlas of the neurons that express mRNA for the long form of the prolactin receptor in the forebrain of the female rat. J. Comp. Neurol. 386 (2), 161–177. Bellringer, J.F., Pratt, H.P., Keverne, E.B., 1980. Involvement of the vomeronasal organ and prolactin in pheromonal induction of delayed implantation in mice. J. Reprod. Fertil. 59 (1), 223–228. Bridges, R.S., Numan, M., Ronsheim, P.M., Mann, P.E., Lupini, C.E., 1990. Central prolactin infusions stimulate maternal behavior in steroid-treated, nulliparous female rats. Proc. Natl. Acad. Sci. U. S. A. 87 (20), 8003–8007. Bridges, R.S., Robertson, M.C., Shiu, R.P., Sturgis, J.D., Henriquez, B.M., Mann, P.E., 1997. Central lactogenic regulation of maternal behavior in rats: steroid dependence, hormone specificity, and behavioral potencies of rat prolactin and rat placental lactogen I. Endocrinology 138 (2), 756–763. Bridges, R.S., Ronsheim, P.M., 1990. Prolactin (PRL) regulation of maternal behavior in rats: bromocriptine treatment delays and PRL promotes the rapid onset of behavior. Endocrinology 126 (2), 837–848. Brown, R.E., Mathieson, W.B., Stapleton, J., Neumann, P.E., 1999. Maternal behavior in female C57BL/6J and DBA/2J inbred mice. Physiol. Behav. 67 (4), 599–605. Bruce, H.M., 1959. An exteroceptive block to pregnancy in the mouse. Nature 184, 105. Colosi, P., Ogren, L., Southard, J.N., Thordarson, G., Linzer, D.I., Talamantes, F., 1988. Biological, immunological, and binding properties of recombinant mouse placental lactogen-I. Endocrinology 123 (6), 2662–2667. Desmaisons, D., Vincent, J.D., Lledo, P.M., 1999. Control of action potential timing by intrinsic subthreshold oscillations in olfactory bulb output neurons. J. Neurosci. 19 (24), 10727–10737. Gheusi, G., Cremer, H., McLean, H., Chazal, G., Vincent, J.D., Lledo, P.M., 2000. Importance of newly generated neurons in the adult olfactory bulb for odor discrimination. Proc. Natl. Acad. Sci. U. S. A. 97 (4), 1823–1828. Huck, U.W., Carter, C.S., Banks, E.M., 1982. Natural or hormone-induced sexual and social behaviors in the female brown lemming (Lemmus trimucronatus). Horm. Behav. 16 (2), 199–207. Jia, C., Goldman, G., Halpern, M., 1997. Development of vomeronasal receptor neuron subclasses and establishment of topographic projections to the accessory olfactory bulb. Brain Res. Dev. Brain Res. 102 (2), 209–216. Kang, N., Janes, A., Baum, M.J., Cherry, J.A., 2006. Sex difference in Fos induced by male urine in medial amygdala-projecting accessory olfactory bulb mitral cells of mice. Neurosci. Lett 398 (1–2), 59–62. Keverne, E.B., 1998. Vomeronasal/accessory olfactory system and pheromonal recognition. Chem. Senses 23 (4), 491–494. Keverne, E.B., de la Riva, C., 1982. Pheromones in mice: reciprocal interaction between the nose and brain. Nature 296 (5853), 148–150. Kimchi, T., Xu, J., Dulac, C., 2007. A functional circuit underlying male sexual behaviour in the female mouse brain. Nature 448 (7157), 1009–1014.
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517 Larsen, C., Kokay, I., Grattan, D., 2006. The presence of a male mouse initiates mitogenesis in the subventricular zone in virgin C57B6 mice. Front. Neuroendocrinol. 27 (1), 80–81. Laurent, G., Wehr, M., Davidowitz, H., 1996. Temporal representations of odors in an olfactory network. J. Neurosci. 16 (12), 3837–3847. Lee, Y., Voogt, J.L., 1999. Feedback effects of placental lactogens on prolactin levels and Fos-related antigen immunoreactivity of tuberoinfundibular dopaminergic neurons in the arcuate nucleus during pregnancy in the rat. Endocrinology 140 (5), 2159–2166. Li, C.S., Kaba, H., Saito, H., Seto, K., 1989. Excitatory influence of the accessory olfactory bulb on tuberoinfundibular arcuate neurons of female mice and its modulation by oestrogen. Neuroscience 29 (1), 201–208. Li, C.S., Kaba, H., Saito, H., Seto, K., 1990. Neural mechanisms underlying the action of primer pheromones in mice. Neuroscience 36 (3), 773–778. Lieberman, M.E., Maurer, R.A., Claude, P., Wiklund, J., Wertz, N., Gorski, J., 1981. Regulation of pituitary growth and prolactin gene expression by estrogen. Adv. Exp. Med. Biol. 138, 151–163. Liman, E.R., Corey, D.P., Dulac, C., 1999. TRP2: a candidate transduction channel for mammalian pheromone sensory signaling. Proc. Natl. Acad. Sci. U. S. A. 96 (10), 5791–5796. Lois, C., Alvarez-Buylla, A., 1994. Long-distance neuronal migration in the adult mammalian brain. Science 264 (5162), 1145–1148. Lucas, B.K., Ormandy, C.J., Binart, N., Bridges, R.S., Kelly, P.A., 1998. Null mutation of the prolactin receptor gene produces a defect in maternal behavior. Endocrinology 139 (10), 4102–4107. Mak, G.K., Enwere, E.K., Gregg, C., Pakarainen, T., Poutanen, M., Huhtaniemi, I., Weiss, S., 2007. Male pheromone-stimulated neurogenesis in the adult female brain: possible role in mating behavior. Nat. Neurosci 10 (8), 1003–1011. Meredith, M., 1986. Vomeronasal organ removal before sexual experience impairs male hamster mating behavior. Physiol. Behav. 36 (4), 737–743. Meredith, M., 2001. Human vomeronasal organ function: a critical review of best and worst cases. Chem. Senses 26 (4), 433–445. Muccioli, G., Ghe, C., Di Carlo, R., 1991. Distribution and characterization of prolactin binding sites in the male and female rat brain: effects of hypophysectomy and ovariectomy. Neuroendocrinology 53 (1), 47–53. Mustafa, A., Bogdanovic, N., Nyberg, F., Suliman, I., Islam, A., Roos, P., Winblad, B., Adem, A., 1995. Effects of long-term ovariectomy and ovarian steroids on somatogenic binding sites in rat brain and liver. Neurosci. Lett 194 (3), 193–196.
517
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 (2), 379–394. Parkes, A.S., Bruce, H.M., 1961. Olfactory stimuli in mammalian reproduction. Science 134, 1049–1054. Pi, X., Zhang, B., Li, J., Voogt, J.L., 2003. Promoter usage and estrogen regulation of prolactin receptor gene in the brain of the female rat. Neuroendocrinology 77 (3), 187–197. Pi, X.J., Grattan, D.R., 1998. Distribution of prolactin receptor immunoreactivity in the brain of estrogen-treated, ovariectomized rats. J. Comp. Neurol. 394 (4), 462–474. Pilotte, N.S., Burt, D.R., Barraclough, C.A., 1984. Ovarian steroids modulate the release of dopamine into hypophysial portal blood and the density of anterior pituitary [3H]spiperone-binding sites in ovariectomized rats. Endocrinology 114 (6), 2306–2311. Powers, J.B., Winans, S.S., 1973. Sexual behavior in peripherally anosmic male hamsters. Physiol. Behav. 10 (2), 361–368. Powers, J.B., Winans, S.S., 1975. Vomeronasal organ: critical role in mediating sexual behavior of the male hamster. Science 187 (4180), 961–963. Rosenblatt, J.S., Wagner, C.K., Morrell, J.I., 1994. Hormonal priming and triggering of maternal behavior in the rat with special reference to the relations between estrogen receptor binding and ER mRNA in specific brain regions. Psychoneuroendocrinology 19 (5–7), 543–552. Rosser, A.E., Keverne, E.B., 1985. The importance of central noradrenergic neurones in the formation of an olfactory memory in the prevention of pregnancy block. Neuroscience 15 (4), 1141–1147. Ryan, K.D., Schwartz, N.B., 1980. Changes in serum hormone levels associated with male-induced ovulation in group-housed adult female mice. Endocrinology 106 (3), 959–966. Shepherd, G.M., Greer, C.A., 1998. In: Shepherd, G.M. (Ed.), The synaptic organisation of the brain. Oxford University Press, New York, pp. 159–203. Shingo, T., Gregg, C., Enwere, E., Fujikawa, H., Hassam, R., Geary, C., Cross, J.C., Weiss, S., 2003. Pregnancy-stimulated neurogenesis in the adult female forebrain mediated by prolactin. Science 299 (5603), 117–120. Voci, V.E., Carlson, N.R., 1973. Enhancement of maternal behavior and nest building following systemic and diencephalic administration of prolactin and progesterone in the mouse. J. Comp. Physiol. Psychol. 83 (3), 388–393. Yoon, H., Enquist, L.W., Dulac, C., 2005. Olfactory inputs to hypothalamic neurons controlling reproduction and fertility. Cell 123 (4), 669–682.
Hormones and Behavior 53 (2008) 509 – 517 www.elsevier.com/locate/yhbeh
Male pheromones initiate prolactin-induced neurogenesis and advance maternal behavior in female mice Caroline M. Larsen ⁎, Ilona C. Kokay, David R. Grattan Centre for Neuroendocrinology and Department of Anatomy and Structural Biology, Box 913, University of Otago, Dunedin 9001, New Zealand Received 11 September 2007; revised 23 November 2007; accepted 27 November 2007 Available online 15 December 2007
Abstract Prolactin is required for rapid onset of maternal behavior after parturition, inducing adaptive changes in the maternal brain including enhanced neurogenesis in the subventricular zone during pregnancy. The resultant increase in olfactory interneurons may be required for altered processing of olfactory cues during the establishment of maternal behavior. Pheromones act through olfactory pathways to exert powerful effects on behavior in rodents and also affect prolactin secretion. Hence, this study aimed to investigate the effect of male pheromones on neurogenesis and maternal behavior in female mice. Virgin female mice were housed individually or in split-cages where they had pheromonal but not physical contact with a male. Maternal behavior was assessed in a foster pup retrieval paradigm. Some mice were injected with bromodeoxyuridine, and the labeled cells visualized using immunohistochemistry. The data show that exposure to male pheromones, for a duration equivalent to a murine pregnancy, advanced maternal behavior in both virgin and postpartum female mice. The pheromone action was dependent on prolactin and ovarian steroids, and was associated with increased cell proliferation in the subventricular zone and subsequent increases in new neurons in the olfactory bulb. Moreover, the effect of pheromones on both cell proliferation and maternal behavior could be induced solely through administration of exogenous prolactin to mimic the pheromone-induced changes in prolactin secretion. The data suggest that male pheromones induce a prolactin-mediated increase in neurogenesis in female mice, resulting in advanced maternal behavior. © 2007 Elsevier Inc. All rights reserved. Keywords: Maternal behavior; Neurogenesis; Mice; Prolactin
Introduction Multiple adaptations occur in the maternal brain to prepare the mother for the demands of pregnancy and lactation. Prolactin, acting in the brain, has been implicated in the onset of some maternal behaviors in rodents (Bridges et al., 1990, 1997; Bridges and Ronsheim, 1990). Steroid-primed virgin rats exhibit maternal behavior earlier when given central infusions of prolactin, whereas maternal behavior is delayed when bromocriptine is used to block endogenous prolactin secretion (Bridges and Ronsheim, 1990). Moreover, prolactin receptor-deficient mice have markedly impaired maternal behavior (Lucas et al., 1998), clearly demonstrating a requirement for prolactin action. Based on prolactin receptor localization in the medial preoptic area (MPOA) (Bakowska and Morrell, 1997; Pi and Grattan, ⁎ Corresponding author. E-mail address: [email protected] (C.M. Larsen). 0018-506X/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.yhbeh.2007.11.020
1998), and localized microinjection studies (Bridges and Ronsheim, 1990), the MPOA is seen as a probable site of prolactin action in the brain in inducing maternal behavior. Prolactin also mediates a pregnancy-induced increase in neurogenesis in the subventricular zone (SVZ) of the adult forebrain that results in the incorporation of new neurons into olfactory pathways of the maternal brain (Shingo et al., 2003). It is postulated that this change results in altered olfactory perception to facilitate maternal behavior. Pheromones act through the nose to influence behavior in most mammalian species (Meredith, 2001). Pheromones suppress prolactin secretion under certain circumstances (Li et al., 1989, 1990), and thus may influence maternal behavior. Pheromones are processed in the olfactory (Yoon et al., 2005) and vomeronasal systems (Keverne and de la Riva, 1982), with tertiary projections via the amygdala to the MPOA (Jia et al., 1997; Li et al., 1989, 1990). As pheromones affect serum prolactin levels, and indirectly affect areas of the brain implicated in maternal behavior
510
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
(Numan and Numan, 1994), we hypothesized that female mice exposed to male mouse pheromones would show reduced levels of prolactin, resulting in suppressed maternal behavior. Previous research has demonstrated that elevated prolactin during pregnancy, or 6 days of prolactin infusion (icv or subcutaneous), increases the number of new neurons that are made in the SVZ in the brain (Shingo et al., 2003). Therefore, we further hypothesized that sustained contact with male pheromones might alter neurogenesis in the SVZ and thus influence maternal behavior. The aim of this study was to test these hypotheses by housing adult virgin female or pregnant female mice either alone, or in divided cages in the presence of a male such that the mouse could have pheromonal but not physical contact (split-cages). To evaluate the effect of pheromones, we examined maternal behavior in both virgin female and postpartum mice in a pup retrieval paradigm. To determine whether pheromones affected neurogenesis, we examined the uptake of bromodeoxyuridine (BrdU) in the SVZ after varying periods of time exposed to male mouse pheromones, as well as chronic changes in BrdU-labeled neurons in the olfactory bulb. Some of these data have been reported previously, in abstract form (Larsen et al., 2006). Materials and methods Subjects Virgin maternally naive 6–9 week old C57BL/6J female mice were either housed in 32 × 16 × 18 cm individual cages or in 28 × 54 × 18 cm split cages. The split cages had a 54 × 18 cm metal divider, which bisected the cage longitudinally, with a series of 0.1 cm holes allowing pheromonal but not physical contact with another mouse. In addition, maternally naive virgin mice were mated with a stud male to examine if pheromone exposure throughout pregnancy affected maternal behavior expression to foster pups. These mice were considered day 1 pregnant when sperm was present in a vaginal smear. The mice were housed under a 12 h light–dark cycle, with ad libitum access to food and water. The University of Otago Animal Ethics Committee approved all experiments.
Maternal behavior testing Female naive mice or day 2 postpartum mice had three foster pups placed at the opposite end of the cage from their nest. In each test session, mice were continuously observed and any behavior recorded i.e. exploration, grooming and sniffing of the pups, nest building, pup retrieval, or crouching over the pups. Full maternal behavior was defined as being when the mouse had gathered all 3 pups to her nest and was crouching over them. As there were no significant differences in any of the parameters within a group, and each measured variable reflected the overall trend within a group, only the data for full maternal behavior has been included. In the first experiment, maternal behavior was compared between individually-caged mice (n = 6) and mice housed in a splitcage with a male mouse (n = 6). Split-caged mice were housed in the split cage for 21 days prior to testing, and remained in the split-cage for the duration of the experiment. Testing was initially repeated daily for 6 days (data not shown), but was limited to 4 days after the first experiment as after this point there were no notable time differences. Testing was performed between 9:30–11 am. After completion of testing on day 4, animals were immediately killed by cervical dislocation, trunk blood collected and centrifuged, and the plasma frozen at − 20 °C until required. To determine whether any effects were specifically caused by pheromones, the experiments were repeated by housing female mice in individual cages for 21 days with 0.05 g of male urine soaked bedding placed daily in their cage for 21 days (n = 6, four male mice were group housed to provide urine soaked sawdust). A control group (n = 6) had their cage lid lifted briefly daily and their
bedding disturbed. Both groups were then tested daily for maternal behavior for 4 days. To evaluate the role of ovarian hormones in any pheromone induced actions, female mice were bilaterally ovariectomized under halothane anesthesia 1 week prior to the experiment, and then housed either alone for 21 days (n = 6) or in a split cage with a male for 21 days (n = 6). Maternal behavior was then tested daily for 4 days, and compared with ovary-intact females either housed alone for 21 days (n = 6) or in split-cages for 21 days (n = 6). Finally, to determine whether pregnant animals might also be affected by pheromones, female mice were mated overnight with a stud male and deemed to be day 1 pregnant when sperm was visualized in a cervical smear. The male was then removed, and females were exposed to male urine soaked bedding for the duration of the pregnancy and subsequently tested for maternal behavior to foster pups on day 2 postpartum (n = 6).
Time course of effect of pheromones on prolactin secretion and on maternal behavior Virgin female maternally naive mice were either housed alone, or in a split cage with a male for 1, 2, 4, 7, 10, or 21 days (all groups n = 6). All groups were then tested for maternal behavior, as described above. Additional groups were sacrificed after 1, 3.5 and 6 h of pheromonal contact for collection of trunk blood for hormone measurements. Serum prolactin from the various groups was measured by double antibody radioimmunoassay (RIA) using NIH reagents. Mouse prolactin (NIDDK-mouse PRL, batch #AFP6476C) was iodinated by the chloramine-T method and used at a concentration of 15,000 cpm/tube. Primary antiserum (NIDDK rabbit anti-mouse PRL-RIA-9, batch # AFP-131078) was added at 1:70,000. Data are expressed in terms of a mouse prolactin standard curve generated using the same prolactin preparation as that used for iodination. Sensitivity was 2 ng/ml. All samples were run in a single assay, with an intra assay % C.V. 2.5%.
Effect of pheromone-induced changes in serum prolactin levels on maternal behavior To evaluate the role of prolactin in pheromone effects on maternal behavior, female mice housed in individual cages were injected s.c. with a slow release preparation of prolactin (50 μg of purified ovine prolactin in a 25% polyvinylpyrrolidone (PVP): 25% H2O: 50% NaHCO3 buffer) or an identical vehicle every 12 h for 72 h (equivalent to 24–72 h of pheromone exposure). To block the pheromone-induced rise in endogenous prolactin, split-caged housed female mice were injected s.c. with bromocriptine (10 mg/kg; 300 μg bromocriptine mesylate (Sigma) dissolved in a minimum amount of ethanol and made up to 0.1 ml with sesame oil), or an identical vehicle, every 12 h, for the first 24–72 h of pheromonal exposure. All groups (n = 6) were tested for maternal behavior 18 days after manipulation of prolactin levels, to simulate the period of time required for the advancement of maternal behavior following the hyperprolactinemia seen from 24–72 h of pheromone exposure.
Effect of male pheromones on neurogenesis in adult female mice Split-caged female mice after 3, 7 or 14 days of pheromonal contact, or individually-caged mice (all groups n = 6) were injected with a series of 6 i.p. BrdU (Sigma) injections 2 h apart (12 mg/100 g of body weight, dissolved in phosphate buffer). The final dose was given 2 h before perfusion. A further group was injected on day 7 of pheromonal contact and killed 14 days later. BrdU-injected mice were anaesthetized with pentobarbitone and perfused with 4% paraformaldehyde. Brains were postfixed for 24 h and then cryoprotected in sucrose prior to being frozen at − 80 °C until tissue processing. Coronal sections (20 µm) of the olfactory bulb and subventricular zone were cut in a cryostat and BrdU-labeled cells detected using immunohistochemistry (mouse monoclonal anti-BrdU, 1:200, Dako, Carpinteria, CA). Fluorescence immunohistochemistry was performed to establish the cell type of the Brdu labeled cells. The primary antibodies used were rat monoclonal anti-BrdU (1:100 AbD, Serotec, OBT0030 S), mouse anti-neuN, (1:50, Chemicon MAB377), and guinea pig anti-doublecortin (1:100, Chemicon, AB5910). Secondary antibodies used were AlexaFluor tagged Alexa 488 (green) and Alexa 568 (red). The fluorescent stained sections were counterstained with 4′6-diamidino-2-phenylindole (DAPI,
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517 Molecular probes, D-1306). For quantification, subventricular zone labeled cells were observed under 200× magnification, and positively labeled cells counted in a systematic manner after a random start, such that all labeled nuclei were counted in 1-in-6 serial sections 120 µm apart for a total of 14 sections throughout the subventricular zone (approximately Bregma + 1.94 mm– 3.9 mm). For the olfactory bulb, 1-in-3 serial sections throughout the entire olfactory bulb were analyzed and all labeled cells counted. Data are presented as the total counts collected, and are not corrected for the total number of sections. To investigate the role of endogenous prolactin in pheromone-induced neurogenesis, groups of split-caged and individually-housed mice were injected s.c. with bromocriptine for the first 24–72 h of pheromonal exposure (as described above). Mice were then injected with BrdU on day 7 of pheromonal contact and processed for immunohistochemistry, as above.
Statistical Analysis The data are expressed as mean ± SEM. Maternal behavior data were analyzed using a repeated measures ANOVA, with p b 0.05 used as the level of significance. Serum prolactin levels (ng/ml) and total counts of Brdu-labeled cells were compared by one-way ANOVA and Fishers PLSD post hoc test.
Results
511
mice, they showed a significantly reduced time to retrieve all pups on day 2 compared to day 1 (p b 0.05). From the third day of maternal behavior testing there were no further significant differences between the two groups. Therefore, the prolonged presence of a male, or brief bouts of physical contact with a pup over several days, caused enhanced maternal behavior. Male pheromones advance maternal behavior To determine whether the observed effect was specifically mediated by pheromones, we examined whether exposure to male mouse urine was sufficient to observe the enhanced maternal behavior. Pheromones are bound to major urinary proteins, and hence placing urine-soaked bedding in with the female will expose her to pheromones without the confounding effects of a companion. Individually-housed females exposed to male urine for 21 days were significantly faster to express full maternal behavior on day 1 of testing, compared to controls (Fig. 1B, p b 0.05). Thus, the advancement of maternal behavior was specifically mediated by male mouse pheromones.
The presence of a male mouse advances maternal behavior Individually-housed virgin female mice spontaneously expressed full maternal behavior to foster mouse pups on day 1 of maternal behavior testing (Fig. 1A). While there is some variability between different mouse strains, the spontaneous expression of full maternal behavior on the first day of testing is consistent with other accounts in the literature, which show that females of the inbred C57BL/6J mouse strain exhibit spontaneous pup retrieval (Brown et al., 1999). On day 2, they showed a significant reduction in time to express full maternal behavior (p b 0.05). While there was a trend for further decreases in time until the fourth day, this difference was not significant. By contrast with the individually-housed mice, virgin females splitcaged with a male were significantly faster to express full maternal behavior on both day 1 and day 2 of maternal behavior testing (Fig. 1A, p b 0.05). As seen with individually-housed
Ovarian steroid hormones are essential for the male mouse advancement of maternal behavior To evaluate the role of ovarian hormones in mouse maternal behavior, virgin female mice were ovariectomized (ovx), housed in individual or split cages for 21 days, and then tested daily for 4 days for maternal behavior. As seen previously, split-caged intact mice were significantly faster on day 1 and 2 to express full maternal behavior compared to individually-caged intact mice (Fig. 1C, p b 0.05). Ovx, individually-caged mice showed no differences in the spontaneous expression of maternal behavior compared to individually-caged controls. Maternal behavior in the ovx, split-caged animals, however, was not significantly different from that in individually-housed mice. Hence, removal of the ovarian steroid hormones completely prevented the pheromone-induced advancement of maternal behavior.
Fig. 1. Data presented as mean ± SEM, all groups n = 6. (A) Female mice split-caged with a male mouse expressed full maternal behavior more rapidly than controls (⁎p b 0.05). Both individually-caged and split-caged mice were faster (+p b 0.05) to express maternal behavior on day 2 of testing. (B) Individually-caged females exposed to male urine for 21 days (pheromone exposed) expressed full maternal behavior more rapidly than mice housed alone (⁎p b 0.05). (C) Intact split-caged females expressed full maternal behavior more rapidly than controls (⁎p b 0.05). In contrast, ovariectomized split-caged females were not different from intact mice housed alone.
512
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
The male needs to be present for longer than 14 days for the pheromonal advancement of maternal behavior to occur To establish the time course of the male pheromone-induced advancement of maternal behavior, we performed maternal behavior testing after varying periods of pheromone exposure. In contrast to the effect of long-term pheromone exposure, females housed with a male in a split cage for 48 h showed significantly impaired maternal behavior (p b 0.05, Fig. 2). At all other measured time points up to 14 days of pheromone exposure, there were no significant differences in maternal behavior compared to individually-housed controls. However, after 21 days, the split-caged females were significantly faster (p b 0.05) to express full maternal behavior. Hence, pheromone-induced advancement of maternal behavior took between 14–21 days to manifest. Pheromone exposure leads to a sustained increase in serum prolactin levels Male mouse pheromones have been reported to acutely suppress prolactin secretion (Li et al., 1989, 1990; Ryan and Schwartz, 1980). Therefore, we examined serum prolactin levels in the split-caged mice. Although there was an initial trend for reduced prolactin, this was not significant (p = 0.09), and by 24 h of exposure serum prolactin levels were significantly increased compared to controls (p b 0.05). Serum prolactin levels remained elevated until 72 h of contact (Fig. 3A). After this time, prolactin levels returned to normal, and there were no further significant differences until day 21 when the levels were significantly reduced (p b 0.05), compared to controls (Fig. 3A). Sustained hyperprolactinemia advances maternal behavior 2 weeks later To evaluate the role of the initial pheromone-induced rise in prolactin, individually-housed females were given injections of prolactin every 12 h for 72 h to mimic the pheromone-induced changes. When subsequently tested for maternal behavior 18 days later, they were significantly faster to express full maternal behavior, compared to controls given injections of
Fig. 3. Data presented as mean ± SEM, all groups n = 6. (A) Split-caged females showed increased serum prolactin levels from 24–72 h of pheromone exposure (⁎p b 0.05). Levels were decreased after 21 days (⁎p b 0.05). (B) Individuallyhoused mice injected with slow release prolactin from 24–72 h of pheromonal contact, were faster to express maternal behavior 18 days later (⁎p b 0.05), whereas, split-caged females injected with bromocriptine from 24–72 h of pheromonal contact showed no differences in maternal behavior expression 18 days later, compared to vehicle-treated controls.
vehicle alone (Fig. 3B). Conversely, split-caged mice given 12 hourly injections of bromocriptine from day 1 to day 3 of pheromonal contact with a male, to block the pheromone-induced increase in prolactin, showed no differences in expression of maternal behavior compared to controls when behavior was tested 18 days later (Fig. 3B). Hence, the pheromone-induced advance of maternal behavior was dependent on the rise in prolactin that followed exposure to male pheromones. Pheromones initiate a prolactin-dependent increase in mitogenesis of neural precursor cells in the subventricular zone
Fig. 2. Data presented as mean ± SEM, all groups n = 6. Female mice split-caged with a male for 2 days had impaired maternal behavior (+p b 0.05). In contrast, split-caged mice with a male for 21 days, expressed advanced maternal behavior (⁎p b 0.05). All other measured time points were not different from individuallyhoused controls.
Split-caged females showed a significant, 227% increase in BrdU-labeled cells in the SVZ after 7 days of pheromonal contact, compared to individually-housed controls (Fig. 4A, p b 0.05). This change reduced to control levels by 14 days of contact. Ovariectomy caused a significant decrease in BrdUlabeled cells in the SVZ in both split-caged and individuallyhoused mice compared to intact controls (Fig. 4B, p b 0.05), and completely prevented the pheromone-induced increase in cell proliferation in this region. The data suggest that a time and ovarian steroid-dependent action of male pheromones stimulates
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
513
Fig. 4. Data presented as mean ± SEM, all groups n = 6. (A) Female mice split-caged with a male for 7 days showed an increase in BrdU-labeled cells in the SVZ (⁎p b 0.05), which reverted to baseline levels by day 14, compared to controls. (B) BrdU-labeled cells were increased in the SVZ after 7 days (⁎p b 0.05), in intact split-caged females. Ovariectomized (OVX) mice were less than intact (+p b 0.05), and showed no response to pheromone exposure. Representative images of the groups in Figs. 4 A and B are shown below. Inset shows high power view of BrdU-labeled cells. (Lat. V = lateral ventricle, CC = corpus callosum, SVZ = indicates region of the subventricular zone, adjacent to the lateral wall of the lateral ventricle). (C) Split-caged mice injected with vehicle from 24–72 h of pheromonal contact, showed an increase in neurogenesis after 7 days of contact with the male (⁎p b 0.05), compared to individually-housed controls injected with vehicle. In contrast, split-caged females injected with bromocriptine from 24–72 h of pheromonal contact, showed no differences in neurogenesis after 7 days of contact with a male, compared to controls.
cell proliferation in the SVZ. To establish if the pheromoneinduced increase in mitogenesis in the SVZ was mediated by the period of hyperprolactinemia from 24–72 h of pheromonal contact, prolactin secretion was inhibited by bromocriptine over this 48-hour period (as described above). Split-caged females injected with vehicle from 24–72 h of pheromone contact showed a significant increase in BrdU-labeled cells in the SVZ
on day 7 of pheromone contact compared to individually-housed controls (Fig. 4C). In contrast, split-caged females injected with bromocriptine from 24–72 h of pheromone contact showed no differences in number of BrdU-labeled cells after 7 days of contact with a male, compared to vehicle-injected controls (Fig. 4C). Thus, suppression of the pheromone-induced period of hyperprolactinemia, earlier found to prevent the pheromone-
514
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
induced enhancement of maternal behavior, also abolished the pheromonal-induced increase in neurogenesis. Pheromones increase the amount of new neurons in the olfactory bulb Split-caged females injected with BrdU on day 7 of pheromonal contact also showed a marked 68% increase in labeled cells in the olfactory bulb 14 days later compared to individually-housed controls (Fig. 5). Again, the pheromoneinduced increase in olfactory neurons was dependent on the presence of ovarian hormones (Figs. 5). Ovx split-caged and individually-caged mice both showed a significant decrease in BrdU-labeled cells in the olfactory bulb compared to intact controls. Cells generated in the SVZ can become new neurons or glial cells. Our preliminary analysis suggests that a significant proportion of the BrdU-labeled cells, both in the SVZ and in the OB could be labeled with neuron specific markers. In split-caged mice, 96% of BrdU immunoreactive cells in the SVZ co-expressed doublecortin, and 88% of BrdU immunoreactive cells in the OB co-expressed Neu N (Fig. 5). Male pheromones advance maternal behavior to foster pups in day 2 postpartum females To determine whether the pheromone effect seen in virgin animals could also influence mice during a normal pregnancy,
Fig. 6. Data presented as mean ± SEM, n = 6. Virgin female mice exposed to male urine for 21 days (pheromone exposed) expressed full maternal behavior more rapidly than virgin controls (+p b 0.05). Postpartum mice expressed maternal behavior to foster pups more rapidly than pheromone and nonpheromone exposed virgin mice (⁎p b 0.05). Individually-caged day 1 pregnant females exposed to male urine for the duration of the pregnancy (pheromone exposed) expressed full maternal behavior more rapidly on day 2 postpartum than non-pheromone exposed day 2 postpartum mice (+p b 0.05).
we housed pregnant mice in the presence of male pheromones. As expected, individually-housed day 2 postpartum female mice spontaneously expressed maternal behavior to foster pups on day 1 of maternal behavior testing (Fig. 6) Similar to the pheromonal advancement of maternal behavior seen in virgin mice, however, the day 2 postpartum mice exposed to male pheromones for the duration of the pregnancy were significantly faster to express full maternal behavior to foster pups than
Fig. 5. Data presented as mean ± SEM, n = 6. BrdU-labeled cells were increased in the olfactory bulb 14 days after injection (⁎p b 0.05), in intact split-caged females, in the glomerular and periglomerular layers of the olfactory bulb. Ovariectomized split-caged mice again showed no response to pheromone exposure and were less than intact mice (+p b 0.05). Photomicrographs A–F show representative examples of the immunohistochemistry from intact animals (A, B) and ovariectomized animals (C, D), with controls animals in the upper panels (A, C) and pheromone exposed in the lower panels (B, D). Examples of merged images of OB cells immunoreactive for BrdU (green), NeuN (red), counterstained with DAPI unclear stain (blue) are also shown (E, F). Counting of dual labeled cells showed an 88% increase in the number of new neurons in the granule and periglomerular layers of the olfactory bulb of split-caged housed mice.
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
individually housed post partum females. Thus, the pheromone effect was not restricted to virgin females, and could influence behavioral adaptation during a normal pregnancy. Discussion Virgin female or day 2 postpartum mice exposed to male mouse pheromones for 21 days were significantly faster to express full maternal behavior to foster pups than mice that had been housed individually. One possible explanation for the enhanced maternal behavior was that it was being triggered by the presence of a companion or an enriched environment. The advancement of maternal behavior induced by split-cage housing with a male, however, was completely replicated by exposure to male urine alone, suggesting it was specifically mediated by pheromones. Although ovariectomy did not alter normal expression of maternal behavior, it completely blocked the pheromoneinduced advancement of maternal behavior. The pheromone effect could also be replicated by treatment with exogenous prolactin approximately 2 weeks prior to maternal behavior testing, to mimic the endogenous changes in prolactin that were induced by pheromone exposure. Moreover, the pheromone effect could be completely prevented by blocking pheromoneinduced prolactin secretion. These findings suggest that an extended period of exposure to male mouse pheromones exerted a prolactin-dependent action to prime female mice to more rapidly express full maternal behavior. The advancement in maternal behavior was dependent on ovarian steroids and correlated with a prolactin- and time-dependent increase in cell proliferation in the SVZ and a subsequent increase in new neurons in the olfactory bulb. Importantly, our original observation of pheromoneinduced neurogenesis (Larsen et al., 2006) has recently been repeated by others (Mak et al., 2007). The data suggest that the pheromone-induced incorporation of new neurons into olfactory pathways, mediated by increased prolactin secretion, may facilitate expression of maternal behavior. Both odorants and pheromones affect reproductive behaviors. Blockade of olfactory receptors leaves many reproductive behaviors intact (Powers and Winans, 1973), however, whereas blockade of vomeronasal receptors severely affects many reproductive behaviours (Meredith, 1986; Powers and Winans, 1975). Mice lacking Trpc2, an ion channel only expressed in the vomeronasal organ (Liman et al., 1999), have deficits in gender discrimination and maternal behavior (Kimchi et al., 2007). Pheromones have been shown to act through both olfactory and vomeronasal pathways (Keverne and de la Riva, 1982; Yoon et al., 2005), but the behaviors seen in Trpc2 deficient mice infer that the vomeronasal organ mediates gender-specific reproductive behavior in mice. Exposure to male urine-soaked bedding selectively initiates Fos activity in the accessory bulb and the medial amygdala (Kang et al., 2006), consistent with the hypothesis that a vomeronasal pathway is being activated. We cannot eliminate the possibility, however, that olfactory cues might be contributing to the observed response. Given the critical role of prolactin in the onset of maternal behavior, our initial hypothesis, that the females exposed to male pheromones would show impaired maternal behavior, was based
515
on previous studies reporting that male mouse pheromones lower serum prolactin levels in female mice. Pheromones carried by male mouse major urinary proteins activate the tuberoinfundibular dopamine (TIDA) neurons in the arcuate nucleus of the hypothalamus, which in turn leads to suppression of prolactin secretion (Li et al., 1989, 1990). This mechanism may underpin the “Bruce effect”, whereby early pregnancy in mice is terminated if a foreign male mouse is encountered (Keverne, 1998), through suppression of prolactin secretion and subsequent loss of luteotrophic support for the corpus luteum. In the present model, however, serum prolactin levels were only briefly and not significantly lower in the split-caged mice after one hour of male mouse contact. It has been demonstrated that a male mouse of a different strain from the female is required to produce the “Bruce effect” in inbred mouse colonies (Parkes and Bruce, 1961). As the males in the split cages were from the same inbred colony as the females, this possibly explains why the drop in serum prolactin was only brief and not significant. In contrast to our expectation, we observed a pheromoneinduced period of sustained hyperprolactinemia. It is well established that prolactin treatment advances maternal behavior in rats (Bridges et al., 1997), and in mice (Voci and Carlson, 1973). During pregnancy the placenta produces placental lactogens (Colosi et al., 1988), which circulate in serum, and bind to prolactin receptors with equal affinity as prolactin (Lee and Voogt, 1999), ensuring a continuously high level of prolactin receptor stimulation. Thus, both pregnancy and pheromone exposure are characterized by a sustained period of hyperprolactinemia. Manipulation of prolactin levels, either by administering prolactin to individually-housed animals or blocking prolactin in pheromone-exposed animals, demonstrated that this rise in serum prolactin levels was essential for the enhanced maternal behavior. At all other time points tested prior to the significant advancement of maternal behavior at 21 days, the split-cage mice had no significant differences in the time taken to express maternal behavior compared to individually housed controls. Hence, although the pheromone-induced increase in prolactin occurred almost immediately, pheromonal advancement of maternal behavior took between 14 and 21 days to manifest. Interestingly, 21 days is the same duration as in a rodent pregnancy, suggesting that pheromonal contact associated with the mating interaction might facilitate behavioral adaptations required for subsequent maternal behavior. The prolonged time required to observe the enhancement of maternal behavior after pheromone exposure led us to hypothesize that the effect might involve generation of new neurons in the olfactory bulb. It has previously been shown that prolactin stimulates an increase in neurogenesis in the SVZ during early pregnancy (Shingo et al., 2003), and these new neurons take approximately 14–21 days to be integrated into olfactory circuits (Lois and Alvarez-Buylla, 1994) Thus, changes in neurogenesis would be expected to take 2–3 weeks before they could exert an effect on animal behavior. A prolactin-dependent increase in interneurons incorporated into the olfactory bulb is correlated with the enhanced maternal behavior seen both after pregnancy (Shingo et al., 2003) and after sustained pheromonal exposure (present data). The amount of new olfactory bulb granule cell
516
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517
layer interneurons is important for the ability to discriminate between different odors (Desmaisons et al., 1999; Gheusi et al., 2000; Laurent et al., 1996; Shepherd and Greer, 1998). The pheromone-induced increase in inhibitory interneurons in the olfactory bulb may therefore potentially alter the processing of odors. Thus, we propose that sustained exposure to pheromones stimulates a prolactin-induced increase in neurogenesis in the SVZ, which alters maternal behavior some 2–3 weeks later. Steroid hormones have been shown to be essential for prolactin induction of maternal behavior in rats (Bridges et al., 1997). Estrogen may influence maternal behavior directly, through estrogen receptors in the MPOA (Rosenblatt et al., 1994), or indirectly by influencing PRL secretion (Lieberman et al., 1981; Pilotte et al., 1984) or prolactin receptor expression (Muccioli et al., 1991; Mustafa et al., 1995; Pi et al., 2003). Ovariectomy had no effect on maternal behavior expression in individually-housed mice, suggesting that estrogen is not crucial for either the initiation or continuation of spontaneous maternal behavior in the mouse. However, ovariectomy prevented both the increase in neurogenesis and the advancement in maternal behavior induced by male mouse pheromones. It seems likely that the presence of estrogen may be permissive to allow normal prolactin action on this function, by maintaining expression of prolactin receptors in the brain (Anderson et al., in press). As there is often a large amount of redundancy in behaviors crucial for a species survival, it is possible that while estrogen is not essential for murine maternal behavior, that the estrogendependent pheromonal enhancement of maternal behavior is an additional biological safety measure, to ensure that dependent offspring are rapidly nurtured soon after birth. By contrast with the pheromone-induced enhancement of maternal behavior after 21 days exposure, after 48 h of pheromonal contact there was a striking impairment in maternal behavior that returned to baseline levels by 72 h of male contact. This acute inhibitory effect could correlate with the Bruce effect, where the presence of an unknown male mouse will terminate pregnancy for up to 2 days after conception (Bruce, 1959). When the female mouse mates, she forms an imprinted memory of the odor of any males present, which leads to suppression of estrous (Bellringer et al., 1980; Rosser and Keverne, 1985). When an unfamiliar male is present within 48 h of copulation, pregnancy is blocked and estrous is induced (Parkes and Bruce, 1961). In addition, new dominant males kill pups they have not sired (Huck et al., 1982). It is possible then, that to optimize survival of the offspring, the mother will not spend energy on either continuing with a pregnancy (Bruce, 1959), or on maternal behavior (present data), if an unfamiliar dominant male is present. Maternal behavior is essential for survival of mammalian offspring. Normally, during the 21 days of a rodent pregnancy, a profound range of adaptations occur in the maternal brain that lead to rapid maternal behavior expression after birth. The pheromonal influences described above may represent a previously unrecognized contribution to the signals underlying these adaptations. As the pheromonal enhancement of maternal behavior and increase in neurogenesis were both mediated by prolactin, our present results suggest that pheromones are acting through a common pathway to enhance neurogenesis and maternal be-
havior. We propose that the male pheromone-induced hyperprolactinemia initiates increased cell proliferation in the SVZ with a subsequent increase in olfactory bulb interneurons. It seems likely that the change in olfactory function, acting through tertiary projections to the MPOA, enhances maternal behavior. Interestingly, although the changes are dependent on prolactin and ovarian steroids in the mother, the changes can occur without the influence of the hormones from the placenta and the fetus. Placental and fetal hormones normally provide multiple signals during pregnancy to ensure establishment of maternal behavior at term. While pheromones have previously been reported to exert powerful actions on the reproductive system, this study shows for the first time that male pheromones potentially complement the prolactin-mediated establishment of maternal behavior. References Anderson G.M., Kieser D.S., Grattan D.R., in press. Prolactin suppression of pulsatile LH secretion requires an estradiol-dependent upregulation of prolactin receptor mRNA. Endocrinology. Bakowska, J.C., Morrell, J.I., 1997. Atlas of the neurons that express mRNA for the long form of the prolactin receptor in the forebrain of the female rat. J. Comp. Neurol. 386 (2), 161–177. Bellringer, J.F., Pratt, H.P., Keverne, E.B., 1980. Involvement of the vomeronasal organ and prolactin in pheromonal induction of delayed implantation in mice. J. Reprod. Fertil. 59 (1), 223–228. Bridges, R.S., Numan, M., Ronsheim, P.M., Mann, P.E., Lupini, C.E., 1990. Central prolactin infusions stimulate maternal behavior in steroid-treated, nulliparous female rats. Proc. Natl. Acad. Sci. U. S. A. 87 (20), 8003–8007. Bridges, R.S., Robertson, M.C., Shiu, R.P., Sturgis, J.D., Henriquez, B.M., Mann, P.E., 1997. Central lactogenic regulation of maternal behavior in rats: steroid dependence, hormone specificity, and behavioral potencies of rat prolactin and rat placental lactogen I. Endocrinology 138 (2), 756–763. Bridges, R.S., Ronsheim, P.M., 1990. Prolactin (PRL) regulation of maternal behavior in rats: bromocriptine treatment delays and PRL promotes the rapid onset of behavior. Endocrinology 126 (2), 837–848. Brown, R.E., Mathieson, W.B., Stapleton, J., Neumann, P.E., 1999. Maternal behavior in female C57BL/6J and DBA/2J inbred mice. Physiol. Behav. 67 (4), 599–605. Bruce, H.M., 1959. An exteroceptive block to pregnancy in the mouse. Nature 184, 105. Colosi, P., Ogren, L., Southard, J.N., Thordarson, G., Linzer, D.I., Talamantes, F., 1988. Biological, immunological, and binding properties of recombinant mouse placental lactogen-I. Endocrinology 123 (6), 2662–2667. Desmaisons, D., Vincent, J.D., Lledo, P.M., 1999. Control of action potential timing by intrinsic subthreshold oscillations in olfactory bulb output neurons. J. Neurosci. 19 (24), 10727–10737. Gheusi, G., Cremer, H., McLean, H., Chazal, G., Vincent, J.D., Lledo, P.M., 2000. Importance of newly generated neurons in the adult olfactory bulb for odor discrimination. Proc. Natl. Acad. Sci. U. S. A. 97 (4), 1823–1828. Huck, U.W., Carter, C.S., Banks, E.M., 1982. Natural or hormone-induced sexual and social behaviors in the female brown lemming (Lemmus trimucronatus). Horm. Behav. 16 (2), 199–207. Jia, C., Goldman, G., Halpern, M., 1997. Development of vomeronasal receptor neuron subclasses and establishment of topographic projections to the accessory olfactory bulb. Brain Res. Dev. Brain Res. 102 (2), 209–216. Kang, N., Janes, A., Baum, M.J., Cherry, J.A., 2006. Sex difference in Fos induced by male urine in medial amygdala-projecting accessory olfactory bulb mitral cells of mice. Neurosci. Lett 398 (1–2), 59–62. Keverne, E.B., 1998. Vomeronasal/accessory olfactory system and pheromonal recognition. Chem. Senses 23 (4), 491–494. Keverne, E.B., de la Riva, C., 1982. Pheromones in mice: reciprocal interaction between the nose and brain. Nature 296 (5853), 148–150. Kimchi, T., Xu, J., Dulac, C., 2007. A functional circuit underlying male sexual behaviour in the female mouse brain. Nature 448 (7157), 1009–1014.
C.M. Larsen et al. / Hormones and Behavior 53 (2008) 509–517 Larsen, C., Kokay, I., Grattan, D., 2006. The presence of a male mouse initiates mitogenesis in the subventricular zone in virgin C57B6 mice. Front. Neuroendocrinol. 27 (1), 80–81. Laurent, G., Wehr, M., Davidowitz, H., 1996. Temporal representations of odors in an olfactory network. J. Neurosci. 16 (12), 3837–3847. Lee, Y., Voogt, J.L., 1999. Feedback effects of placental lactogens on prolactin levels and Fos-related antigen immunoreactivity of tuberoinfundibular dopaminergic neurons in the arcuate nucleus during pregnancy in the rat. Endocrinology 140 (5), 2159–2166. Li, C.S., Kaba, H., Saito, H., Seto, K., 1989. Excitatory influence of the accessory olfactory bulb on tuberoinfundibular arcuate neurons of female mice and its modulation by oestrogen. Neuroscience 29 (1), 201–208. Li, C.S., Kaba, H., Saito, H., Seto, K., 1990. Neural mechanisms underlying the action of primer pheromones in mice. Neuroscience 36 (3), 773–778. Lieberman, M.E., Maurer, R.A., Claude, P., Wiklund, J., Wertz, N., Gorski, J., 1981. Regulation of pituitary growth and prolactin gene expression by estrogen. Adv. Exp. Med. Biol. 138, 151–163. Liman, E.R., Corey, D.P., Dulac, C., 1999. TRP2: a candidate transduction channel for mammalian pheromone sensory signaling. Proc. Natl. Acad. Sci. U. S. A. 96 (10), 5791–5796. Lois, C., Alvarez-Buylla, A., 1994. Long-distance neuronal migration in the adult mammalian brain. Science 264 (5162), 1145–1148. Lucas, B.K., Ormandy, C.J., Binart, N., Bridges, R.S., Kelly, P.A., 1998. Null mutation of the prolactin receptor gene produces a defect in maternal behavior. Endocrinology 139 (10), 4102–4107. Mak, G.K., Enwere, E.K., Gregg, C., Pakarainen, T., Poutanen, M., Huhtaniemi, I., Weiss, S., 2007. Male pheromone-stimulated neurogenesis in the adult female brain: possible role in mating behavior. Nat. Neurosci 10 (8), 1003–1011. Meredith, M., 1986. Vomeronasal organ removal before sexual experience impairs male hamster mating behavior. Physiol. Behav. 36 (4), 737–743. Meredith, M., 2001. Human vomeronasal organ function: a critical review of best and worst cases. Chem. Senses 26 (4), 433–445. Muccioli, G., Ghe, C., Di Carlo, R., 1991. Distribution and characterization of prolactin binding sites in the male and female rat brain: effects of hypophysectomy and ovariectomy. Neuroendocrinology 53 (1), 47–53. Mustafa, A., Bogdanovic, N., Nyberg, F., Suliman, I., Islam, A., Roos, P., Winblad, B., Adem, A., 1995. Effects of long-term ovariectomy and ovarian steroids on somatogenic binding sites in rat brain and liver. Neurosci. Lett 194 (3), 193–196.
517
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 (2), 379–394. Parkes, A.S., Bruce, H.M., 1961. Olfactory stimuli in mammalian reproduction. Science 134, 1049–1054. Pi, X., Zhang, B., Li, J., Voogt, J.L., 2003. Promoter usage and estrogen regulation of prolactin receptor gene in the brain of the female rat. Neuroendocrinology 77 (3), 187–197. Pi, X.J., Grattan, D.R., 1998. Distribution of prolactin receptor immunoreactivity in the brain of estrogen-treated, ovariectomized rats. J. Comp. Neurol. 394 (4), 462–474. Pilotte, N.S., Burt, D.R., Barraclough, C.A., 1984. Ovarian steroids modulate the release of dopamine into hypophysial portal blood and the density of anterior pituitary [3H]spiperone-binding sites in ovariectomized rats. Endocrinology 114 (6), 2306–2311. Powers, J.B., Winans, S.S., 1973. Sexual behavior in peripherally anosmic male hamsters. Physiol. Behav. 10 (2), 361–368. Powers, J.B., Winans, S.S., 1975. Vomeronasal organ: critical role in mediating sexual behavior of the male hamster. Science 187 (4180), 961–963. Rosenblatt, J.S., Wagner, C.K., Morrell, J.I., 1994. Hormonal priming and triggering of maternal behavior in the rat with special reference to the relations between estrogen receptor binding and ER mRNA in specific brain regions. Psychoneuroendocrinology 19 (5–7), 543–552. Rosser, A.E., Keverne, E.B., 1985. The importance of central noradrenergic neurones in the formation of an olfactory memory in the prevention of pregnancy block. Neuroscience 15 (4), 1141–1147. Ryan, K.D., Schwartz, N.B., 1980. Changes in serum hormone levels associated with male-induced ovulation in group-housed adult female mice. Endocrinology 106 (3), 959–966. Shepherd, G.M., Greer, C.A., 1998. In: Shepherd, G.M. (Ed.), The synaptic organisation of the brain. Oxford University Press, New York, pp. 159–203. Shingo, T., Gregg, C., Enwere, E., Fujikawa, H., Hassam, R., Geary, C., Cross, J.C., Weiss, S., 2003. Pregnancy-stimulated neurogenesis in the adult female forebrain mediated by prolactin. Science 299 (5603), 117–120. Voci, V.E., Carlson, N.R., 1973. Enhancement of maternal behavior and nest building following systemic and diencephalic administration of prolactin and progesterone in the mouse. J. Comp. Physiol. Psychol. 83 (3), 388–393. Yoon, H., Enquist, L.W., Dulac, C., 2005. Olfactory inputs to hypothalamic neurons controlling reproduction and fertility. Cell 123 (4), 669–682.