Neural transection between preoptic area and septum inhibits maternal behavior in female and male rats

Neurobiological mechanisms underlying oxytocin-mediated parental behavior in rodents

Parental behavior is essential for mammalian offspring to survive. Because of this significance, elucidating the neurobiological mechanisms that facilitate parental behavior has received strong interest. Decades of studies utilizing pharmacology and molecular biology have revealed that in addition to its facilitatory effects on parturition and lactation, oxytocin (OT) promotes the expression of parental behavior in rodents. Recent studies have also described the modulation of sensory processing by OT and the interaction of the OT system with other brain regions associated with parental behavior. However, the precise neurobiological mechanisms underlying the facilitation of caregiving behaviors by OT remain unclear. In this Review, I summarize the findings from rats and mice with a view toward integrating past and recent progress. I then review recent advances in the understanding of the molecular, cellular, and circuit mechanisms of OT-mediated parental behavior. Based on these observations, I propose a hypothetical model that would explain the mechanisms underlying OT-mediated parental behavior. Finally, I conclude by discussing some major remaining questions and propose potential future research directions.

Association between parental behaviors and structural plasticity in the brain of male rodents

In recent decades, human fathers across the globe have shown a substantial increase in their engagement in paternal caregiving behaviors. Despite the growing interest, the precise neurobiological mechanisms underlying caregiving behaviors in males remain unclear. Neurobiological studies conducted on rodents have advanced our understanding of the molecular, cellular, and circuit-level mechanisms. Typically, sexually naïve males exhibit aggression toward offspring, while fathers display parental behaviors. This drastic behavioral plasticity may be associated with changes in connections among specific regions or cell types. Recent studies have begun to describe this structural plasticity by comparing neural connections before and after fatherhood. In this Perspective, we summarize the findings from four well-studied rodent species, namely prairie voles, California mice, laboratory rats, and laboratory mice, with a view toward integrating past and current progress. We then review recent advances in the understanding of structural plasticity for parental behaviors. Finally, we discuss remaining questions that require further exploration to gain a deeper understanding of the neural mechanisms underlying paternal behaviors in males, including their possible implications for the human brain.

Comparison of medial preoptic, amygdala, and nucleus accumbens lesions on parental behavior in California mice (Peromyscus californicus)

We have previously shown that medial preoptic area (MPOA) lesions disrupt parental behavior in both male and female California mice (P. californicus). In the present study, we compare the effects of lesions in the MPOA, with those in the basolateral amygdala (BA) and nucleus accumbens (NA) on male and female parental behaviors in the biparental California mouse. A male or multiparous female from each male-female pair was given an electrolytic or sham lesion in the MPOA, BA, or NA and tested for parental responsiveness. Since female P. californicus show postpartum estrus, they were likely pregnant during parental testing. MPOA lesions produced deficits in both male and female parental behaviors, and BA lesions disrupted male, and to a lesser extent, female parental behavior. NA lesions produced mild effects on pup-retrieval in males and no effect on parental behavior in females. However, NA lesions incompletely destroyed the NA shell, the region most relevant for maternal behavior in rats, and should be investigated further. These results support a role for the MPOA and BA in both male and female parental behaviors.

Sex difference and postnatal change of maternal behavioral patterns in juvenile male and female rats

Juvenile rats are known to show certain elements of maternal behavior. In this experiment, to investigate sex difference and postnatal change of retrieving and pup-cleaning (licking) behaviors in juvenile rats, these behaviors were recorded using new observation method at 20, 30 and 45 days of age in female and male Wistar rats. At 20 days of age, maternal behavior was observed in a common plastic observation cage (test A) and then test B was performed. In the test B, observation was carried out using a cage with a wooden box that was open on one side, helping the juveniles to establish a nest. As the results of day 20, most rats in all groups showed licking behavior in both the test A and B. The incidence of retrieving behavior increased from the test A to the test B with the box in both sexes, especially in males (p<0.01). The box is thought to play a facilitative role in induction of retrieving. Moreover, the incidence in males was higher than that in females in the test B (p<0.001). At 30 and 45 days of age, only a test B with box was performed. The incidences of licking and retrieving behaviors at 30 days of age were decreased significantly compared to those at 20 days of age in both sexes(p<0.001). Further decrease from 30 days to 45 days was observed. These results suggest that in juvenile rat, incidence of retrieving behavior in males is higher than that in females but there is no sex difference in incidence of licking behavior. Potency to show these behaviors decreases acutely before puberty in rats.

Fast neurotransmission in the rat medial preoptic nucleus

The functional properties of neurotransmission in the medial preoptic nucleus (MPN) were studied in a brain slice preparation from young male rats. The aims were to evaluate the thin slice preparation for studying evoked synaptic responses in MPN neurons, to characterize the fast responses triggered by activation of presynaptic nerve fibers in the MPN, and to identify the involved receptor types. Presynaptic stimulation within the MPN evoked postsynaptic voltage and current responses that were blocked by 200 microM Cd2+ or by 2.0 microM tetrodotoxin and were attributed to action potential-evoked transmitter release. The relation to stimulus strength and comparison with spontaneous synaptic currents suggested that in many cases only one presynaptic nerve fiber was excited by the stimulus. Furthermore, the transmission was probabilistic in nature, with frequent failures. Thus, response probability, most likely reflecting transmitter release probability, could be evaluated in the thin slice preparation. Evoked excitatory postsynaptic currents recorded under voltage-clamp conditions were, due to kinetics, I-V relation, and pharmacological properties, attributed to AMPA/kainate receptors and NMDA receptors, whereas inhibitory currents were attributed to GABAA receptors. No responses that could be attributed to glycine or other types of primary transmitters were detected. Although serotonin (5-HT) did not appear to function as a primary transmitter, glutamate- as well as GABA-mediated transmission was suppressed by 500 microM 5-HT, with a clear reduction in response probability observed. 5-HT also reduced the frequency, but not the amplitude, of spontaneous postsynaptic currents and was therefore ascribed a presynaptic site of action.

Differential fos activation in virgin and lactating mice in response to an intruder

Lactating (L) mice display fierce aggression towards novel, male mice, while virgin (V) mice do not. This study compares patterns of brain activation in V and L mice in response to a novel intruder using immunohistochemical detection of Fos (Fos-IR). Animals were sampled 120 min after either a sham or real 10 min test with a male intruder. L mice were aggressive towards intruders, but V mice were not. In general, Fos-IR for both groups increased with exposure to an intruder, with L mice showing higher increases in Fos-IR than V mice. In only medial preoptic nucleus and ventral portion of bed nucleus of stria terminalis (BNST) was Fos-IR significantly increased in both groups with testing. In V mice, testing resulted in Fos-IR increases in an additional 10 regions examined that did not reach significance in L mice, including lateral septum, lateral and medial preoptic areas, and anterior hypothalamus. Fos-IR also increased with testing in nine regions unique to L mice, including the mitral and granular layers of accessory olfactory bulb, regions of the amygdala, dorsal BNST, and caudal portions of the hypothalamic attack area. These increases in Fos-IR with testing suggest alterations in the circuitry governing response to pheromonal cues and imply some commonalities between the circuitries governing maternal aggression and intermale aggression. These results support the hypothesis that pregnancy and lactation induce substantial changes in brain circuitry and function; changes that enable maternal defense of offspring by altering the neural response to an intruder male.

Medial preoptic lesions disrupt parental behavior in both male and female California mice (Peromyscus californicus)

California mice (Peromyscus californicus) are monogamous and naturally biparental, making them an ideal species in which to study the neural basis of paternal behavior. A male or female from each male-female pair was given an electrolytic or sham lesion in the medial preoptic area (MPOA), an area known to be critical for the expression of maternal behavior in rats, and retested for parental responsiveness. MPOA-lesioned males and females showed significantly longer latencies to show parental behavior and spent significantly less time near pups, sniffing pups, and licking pups than sham-lesioned mice. However, MPOA lesions did not reduce time spent hovering over pups. The results suggest that the neural mechanisms mediating paternal behavior are similar to those mediating maternal behavior in this species.

FOS and FOSB expression in the medial preoptic nucleus pars compacta of maternally active C57BL/6J and DBA/2J mice

C57BL/6J and DBA/2J inbred mice differ in aspects of maternal behavior and in the morphology of the medial preoptic nucleus (MPO), suggesting a possible association. DBA/2J mice have a compact subnucleus in the MPO, the MPOpc, that is sexually dimorphic and absent in C57BL/6J mice. To determine whether MPOpc cells are activated by maternal behavior, FOS and FOSB immunohistochemistry was performed on brain sections of C57BL/6J and DBA/2J mothers following the return of their pups after a separation of 2 days. In both light and dark phases of the daily cycle, stimulation of DBA/2J mothers evoked an increase in FOS- and FOSB-immunoreactivity in the MPOpc. Stimulated C57BL/6J mice, which lack the MPOpc, did not show an increase in cellular activity in the corresponding MPO region. Cells immediately lateral to the MPOpc were activated by pup stimulation, in both strains. These results suggest that MPOpc cells are active during maternal behavior, and that strain differences in maternal behavior are related to anatomical differences in the MPO.

Neural mediation of nursing and related maternal behaviors

Nursing is the behavioral concomitant of lactation and the most generalizable maternal behavior across mammals. In lactating rats nursing often occurs in the kyphotic (upright crouched) posture; like the neuroendocrine determinants of milk synthesis and release, kyphosis requires suckling by the young. The dam's active pronurturant behaviors, such as retrieval and licking of pups, requires perioral somatosensory stimulation, which is often a precursor of kyphosis as well, and is inhibited by suckling. The sequential nature of maternal behaviors and the dissociations in their somatosensory regulation are critical to understanding their neural mediation, as exemplified by our recent work in lactating rats. We found that the caudal lateral and ventrolateral midbrain periaqueductal gray (cPAGl,vl) is a sensorimotor integration site for the kyphotic nursing posture. Destruction of the cPAGl,vl, or increased activity of the inhibitory neurotransmitter GABA within it, severely reduced kyphosis, increased nursing in more atypical postures, and had little or no effect on pronurturance. Various forebrain sites are known to mediate retrieval and licking of pups. Inhibition of dopaminergic activity in the nucleus accumbens of dams via microinfusions of a mixed D1/D2 dopamine receptor antagonist, cis-flupenthixol (FLU), dose-dependently reduced these active behaviors, while increasing nursing duration. Retrieval was inhibited, however, only by infusions of FLU that included the nucleus accumbens shell, which is reciprocally connected with other sites implicated in retrieval of pups. Thus, maternal behavior is not a unitary process but rather a complex category consisting of sequential behavioral components that have their own sensory and neural determinants.

Exposure to a novel stimulus reduces anxiety level in adult and aging rats

Male Wistar rats aged 3, 15 and 24 months were isolated and housed individually for 72 h prior to being subjected to inanimate objects (two objects per rat, each 1.5 cm in diameter and 4 cm in length, made of surgical gauze). Following the exposure to the objects, rats were subsequently tested in an elevated plus-maze. The inanimate objects induced locomotor activity, chewing and transportation of the object. This changed to social interaction and play-like behavioral activity in rats of all ages that were kept in small groups with a latency of 1 to 2 min. One hour after the start of exposure, the animals were tested in the elevated plus-maze to measure anxiety behavior. It was found that all age groups spent significantly more time in the open arm of the elevated plus-maze than the corresponding controls. Latencies to first entry into open arms were also significantly lowered. The number of entries to the open or to the dark arm, however, did not show statistical difference, indicating that the novel object-induced activity failed to exert influence on gross motor activity in the elevated plus-maze. In conclusion, the stimulation elicited by the exposure to novel stimulus (inanimate objects) resulted in a significant reduction of anxiety level both in adult and in aging rats.

Strain and sex differences in the morphology of the medial preoptic nucleus of mice

The medial preoptic nucleus (MPO), which is involved in sexual and maternal behaviors, contains neuronal clusters that have been described as being sexually dimorphic in size and neuropeptide content in a variety of species. A subnucleus in DBA/2J (D2) inbred mice, called the pars compacta of the MPO (MPOpc), is absent in C57BL/6J (B6) inbred mice (Robinson et al. [1985] J. Neurogenet. 2:381-388). We report here on experiments that further characterize strain and sex differences in medial preoptic morphology in D2 and B6 inbred mice. A prominent MPOpc, located within the caudal part of the MPO and dorsal to the suprachiasmatic nucleus, was present in both male and female D2 animals but was absent from B6 animals. MPOpc neurons were darkly stained for Nissl substance and larger than neurons in the surrounding MPO. In D2 brains, galanin-immunoreactive (-ir), oxytocin-ir, vasopressin-ir, and NADPH diaphorase-positive neurons were concentrated within the MPOpc. Fewer similar neurons in the comparable region of the MPO of B6 mice suggests that the absence of the MPOpc is due to absence of these neurons rather than a less compact organization. In D2 animals, the density of galanin-ir neurons in the MPOpc was sexually dimorphic, with higher numbers of galanin-ir neurons in females. Strain differences in galanin-ir, oxytocin-ir, vasopressin-ir, and NADPH diaphorase staining appeared to be limited to the MPOpc. Cholecystokinin-immunoreactive neurons, which have been reported to be numerous in the sexually dimorphic central subdivision of the MPO of rats, were sparse in the MPO of D2 and B6 mice. Confirmation of the MPOpc as an accessory magnocellular neurosecretory nucleus was obtained by finding labeling of MPOpc neurons after injection of DiI into the posterior pituitary.

Maternal behavior stimulates c-fos activity within estrogen receptor alpha-containing neurons in lactating rats

Estradiol and other hormones are thought to be critical for the onset, but not maintenance, of maternal behavior in rats. Maternal behavior is instead maintained postpartum by tactile stimulation that dams receive during interactions with pups, and many neural sites implicated in the control of maternal behavior show elevated c-fos activity in response to this stimulation. Many of these sites also contain neurons that express the alpha subtype of the estrogen receptor (ERalpha). Because of possible interactions between tactile stimulation from pups, c-fos, and ERalpha in the lactating rat brain, we determined if populations of cells that show increased c-fos activity after maternal behavior in lactating rats also contain ERalpha. Dams were separated from their pups for 48 h beginning on day 5 postpartum. On day 7 postpartum, experimental dams were reunited with pups and mother-litter interactions were observed for 60 min. Control dams received no pup stimulation. Subjects were sacrificed 60 min later and brain sections were double immunolabeled for the Fos and ERalpha proteins. As expected, the number of ERalpha-immunoreactive (ERalpha-ir) neurons did not differ between the two groups in the eight areas analyzed (lateral region of the lateral septum, posterodorsal medial amygdala, dorsal and ventral medial preoptic area, dorsal and ventral bed nucleus of the stria terminalis, lateral habenula, and ventrolateral caudal periaqueductal gray). Consistent with previous reports, maternal dams had 2- to 7-fold more Fos-immunoreactive (Fos-ir) neurons in these sites compared with nonstimulated controls. Maternal dams had significantly more Fos-ir neurons that also contained ERalpha-ir in all sites, with the greatest increases in the ventral medial preoptic area, lateral habenula, and ventral bed nucleus of the stria terminalis. Between approximately 25 and 45% of the Fos-ir cells in the sites examined also expressed ERalpha. Thus, a substantial number of neurons that are genomically activated during maternal behavior contain ERalpha, raising the possibility that the postpartum display of maternal behavior can be influenced by ERalpha activity.

Maternal behavior in female C57BL/6J and DBA/2J inbred mice

Inbred strains of mice exhibit different patterns of maternal behavior, providing material for studies of genetic influences on the expression of maternal behavior. Beginning 1 day after birth, maternal behavior was recorded daily for 14 days in the first and second litters of C57BL/6J (B6) and DBA/2J (D2) mothers. D2 mice had higher pup survival than B6 mice, and pup survival was higher in both strains in second litters than in first litters. D2 mothers spent more time engaged in maternal behavior, especially resting with, crouching over, and nursing pups than B6 mothers with first litters, but not with second litters. Not all measures of maternal behavior were correlated with pup survival; with both litters, B6 mothers retrieved pups faster than D2 mothers.

Forebrain expression of c-fos due to active maternal behaviour in lactating rats

To reveal brain sites simultaneously active during the expression of maternal behaviour in lactating rats, we used immunocytochemical visualization of the nuclear protein product Fos of the immediate-early gene c-fos as a marker of neuronal activity. After a 48 h separation from their litter, day 7 postpartum dams received a 1 h period of physical interaction with pups either capable or incapable of suckling, inaccessible pups in a wire-mesh box, an empty box, or no stimulation. Physical interaction with pups elicited high levels of pronurturant maternal behaviour (retrieval, licking, mouthing), and suckling elicited nursing behaviour as well. Exposure to the box, with or without pups, elicited high levels of investigatory sniffing, self-grooming, and general activity. Distal stimulation from pups did not differentially activate Fos in any of 20 sites, including olfactory-processing structures such as the piriform cortex and medial amygdala. Physical interaction with pups, with or without suckling, elicited higher levels of Fos-immunoreactive nuclei than that of other conditions in numerous sites, including many previously implicated in maternal behaviour (medial preoptic nucleus, nucleus accumbens, lateral septum, lateral habenula, and the bed nucleus of the stria terminalis). Similar group patterns of Fos expression also occurred in sites not previously implicated in maternal behaviour (somatosensory cortex and paraventricular thalamic nucleus). Interaction with nonsuckling pups elicited the highest levels of Fos in the cortical amygdala, whereas suckling did not activate higher Fos than nonsuckling interaction in any site included in this report, including hypothalamic nuclei involved in lactation (paraventricular, supraoptic, and arcuate). There was little or no Fos in cingulate cortex, olfactory tubercle, medial septum, medial habenula, or ventromedial hypothalamus. These data suggest that trigeminal stimuli received by lactating rats during the performance of pronurturant maternal behaviour promote cellular activity resulting in neuronal expression of c-fos in many forebrain sites including the medial preoptic nucleus, several sites connected with it that are part of the mesotelencephalic dopamine system, and in the somatosensory cortex. In contrast, in these forebrain sites suckling does not elicit greater levels of Fos than that seen in nonsuckled rats and distal stimuli from pups are ineffective in increasing Fos levels compared with non-stimulated controls.

Cuts between the septum and preoptic area increase ultrasound production, lordosis, and body weight in female hamsters

Studies of the mechanisms for female-typical mating behavior have focused on the ventromedial hypothalamus, and on the decrements in lordosis caused by lesions of this structure. However, opposed changes of comparable size are produced, at least in rats, by horizontal cuts extending forward from the anterior commissure (anterior roof deafferentation, or ARD). This suggests the existence of a lordosis-inhibiting system of forebrain structures that may include the lateral septum and preoptic area. To test the generality of this system, ovariectomized hamsters in hormone-induced estrus were observed for levels of ultrasound production and lordosis during tests with male conspecifics. In addition, subjects were observed for lordosis responses to light manual stimulation. Upon the completion of these tests, subjects received control treatments or ARD prior to a second round of behavioral observations. These postoperative tests revealed clear ARD-stimulated increases in ultrasound production and body weight. In contrast, the facilitation of lordosis was more subtle, appearing in tests with manual stimulation, but not in response to males. These results, then, demonstrate some consistency across species in the effects of ARD. At the same time, however, they suggest species differences in the magnitude of these effects, in turn, suggesting species differences in the dependence of receptivity on forebrain lordosis-inhibiting mechanisms.

The efferent connections of the lateral septal nucleus in the guinea pig: projections to the diencephalon and brainstem

The anterograde Phaseolus vulgaris-leucoagglutinin (PHA-L) tracing technique was used to determine the distribution of efferent fibers originating in the lateral septal nucleus of the guinea pig. For complementary detection of the chemical identity of the target neurons, double-labeling immunocytochemistry was performed with antibodies to PHA-L and to vasopressin, oxytocin, vasoactive intestinal polypeptide, serotonin or dopamine beta-hydroxylase, respectively. The hypothalamus received the majority of the PHA-L-stained septofugal fibers. Here, a specific topography was observed. (1) The medial and lateral preoptic area, (2) the anterior, lateral, dorsal, posterior hypothalamic and retrochiasmatic area, (3) the supraoptic, paraventricular, suprachiasmatic, dorsomedial, caudal ventromedial and arcuate nuclei, and (4) the tuberomammillary, medial and lateral supramammillary, dorsal and ventral premammillary nuclei always contained PHA-L-labeled fibers. The rostral portion of the ventromedial nucleus and the medial and lateral mammillary nucleus only occasionally showed weak terminal labeling. In other diencephalic areas, termination of PHA-L-labeled fibers was observed in the epithalamus and the nuclei of the midline region of the thalamus. In the mesencephalon, terminal varicosities occurred in the ventral tegmental area, interfascicular and interpeduncular nucleus, and periaqueductal gray. In addition, the dorsal and medial raphe nuclei of the metencephalon, together with the locus coeruleus and the dorsal tegmental nucleus, received lateral septal efferents.