Title: Cell Type-Specific Hormonal Signaling Configures Hypothalamic Circuits for Parenting
Title: Cell Type-Specific Hormonal Signaling Configures Hypothalamic Circuits for Parenting
Abstract: Parenting behavior emerges from hormonally sensitive circuits, but how distinct circuit components are affected by, and contribute to, sex and state-dependent changes in infant caregiving remains unclear. Using cell type-specific approaches, we characterized two circuit nodes that are differentially configured by hormones to modulate infant-evoked behavior. An excitatory neuronal population in the anteroventral periventricular nucleus is only active in lactating mothers, increases virgin female caregiving when artificially stimulated and confers oxytocin sensitivity in mothers through a prolactin-STAT5b pathway. These neurons function upstream of another preoptic area population involved in male and female parenting, thus boosting caregiving by mothers. By contrast, androgen signaling in the latter preoptic population reshapes their intrinsic properties to promote pup-directed aggression, revealing cell type-specific tuning of social behavior circuits.
Main Text: Mammals invest considerable resources to nurture offspring and undergo extensive reorganization of behavioral and physiological priorities to maximize the survival and well-being of caregiver and infants. While parenting is essential for survival across generations, the lack of immediate benefits for the caregiver suggests that this behavior is, at least in part, driven by evolutionarily shaped, genetically pre-programmed neural circuits. Depending on the species, parental care is provided by one or both parents, sometimes the larger social group, and lasts for a few days, months, or years. Hormonal and environmental factors have been identified as key modulators of infant-directed behaviors, underscoring the complexity of the neural control of parenting according to the animal internal state within a given environmental context. How systemic factors such as hormones lead to highly specific and stereotyped changes in behavior depending on sex or mating status is poorly understood. Recent efforts have aimed to uncover the specific cell types and circuits that underlie parental care, but whether and how specific neuronal populations within a circuit are modulated to achieve the observed state and context-dependent infant-evoked behavior has not been addressed.
Laboratory mice display a range of parental care towards infants according to their sex and physiological state. Lactating mothers display highly robust and reliable parental behavior, whereas fathers and virgin females show weaker and more variable caregiving and virgin males are typically aggressive or neglectful towards pups. What factors and mechanisms determine differences in pup-directed behavior? Decades of work on the regulation of parental care identified hormones as key determinants of adult-infant interactions. Parabiosis experiments in which virgin females were exposed to the blood plasma of lactating mothers were sufficient to induce the retrieval of infants to the virgin's nest. Accordingly, a large number of hormones and neuropeptides including prolactin, estrogen, and oxytocin have been shown to play a key role in inducing maternal-like behaviors in otherwise non-parental animals. Conversely, in males, removal of circulating androgens through castration decreases aggression towards infants. Although these circulating hormones affect all tissues, key brain areas involved in the control of parental care express high levels of hormone receptors suggesting specific hormonal modulation of neural populations and circuits.
The medial preoptic area of the hypothalamus is essential for parental care across vertebrate species. MPOA lesions result in impaired parental care in birds and mammals,
whereas increased MPOA neural activity in fish, frogs, and mammals is associated with parental display.
Additional parenting-related populations reside in the anteroventral periventricular nucleus of the preoptic area. Tyrosine hydroxylase-expressing AvPe neurons were shown to influence maternal, but not paternal, behavior, in part through their monosynaptic projections onto oxytocin expressing neurons in the paraventricular nucleus of the hypothalamus and thus modulating state-dependent hormonal release to stimulate parenting behavior. The AvPe also contains an excitatory subset of the broad bombesin-like receptor three-expressing population in the preoptic area, distinguished by the co-expression of bombesin-like receptor three and vesicular glutamate transporter two. These AvPe bombesin-like receptor three positive vesicular glutamate transporter two positive neurons express high levels of prolactin receptor, and are Fos-positive in lactating mothers, but not virgin females or mated-males engaged in parenting, thus representing a prominent cell type displaying sex- and state-specific activity during parenting.
Multiple cell types have also been implicated in infant-directed aggression. The rhomboid nucleus of the bed nuclei of the stria terminalis shows increased activity in infanticidal animals, with estrogen receptor positive neurons driving attack toward infants. Likewise, neurons located in the perifornical area that express the neuropeptide urocortin three and vesicular glutamate transporter two have been implicated in infant-directed aggression typically observed in virgin males.
The precise identification of specific cell types and associated neural circuits underlying the control of infant-evoked behavior offers a unique opportunity to uncover how these populations undergo sex- and state-dependent activity regulation. Internal states and external environments can affect behavior through extrinsic, network-driven changes in neural activity as well as cell-intrinsic modulation of the corresponding circuits and cell types. Here we hypothesize that cell-intrinsic changes in transcriptional and biophysical properties of key regulatory hubs of parenting behavior are instrumental in driving observed differences in neural activity and behavior. Using single-cell genomics and cell type-specific genetic manipulations, we uncovered sex- and physiological state-dependent remodeling of cell types controlling parental care. We identified unique targets of hormonal signaling that drive changes in cell-intrinsic properties and determine the nature of adult-infant interactions. Overall, these findings demonstrate how internal states influence neural activity in a cell type-specific manner to configure neural circuits and modulate behavior.
A maternal-specific neural circuit
A maternal-specific neural circuit
To uncover cell-intrinsic mechanisms underlying the sex- and state-dependent modulation of parenting, we focused on three cell types previously associated with adult-infant interactions: AvPeBrs3+Vglut2+ neurons activated during mother-specific parenting, MPOAGal+CalcR+ neurons driving parenting across all parental animals, and PeFA Ucn3+Vglut2+ neurons involved in male and female infant-directed aggression. Among parental mice, mothers display the most robust infant-evoked responses compared to virgin females and fathers. To gain insights into the unique, cell type-specific, molecular features of parental regulation in lactating mothers we sought to better characterize AvPeBrs3+Vglut2+ neurons, an excitatory population restricted to the AvPe that was previously found to be exclusively activated in lactating mothers but not in parental virgin females nor fathers based on immediate-early gene expression analysis. AvPeBrs3+Vglut2+ neurons are part of a broader population of Brs3+ neurons distributed throughout the preoptic area that comprises cell types such as MPOAGal+CalcR+, AvPeBrs3+Vglut2+, BNSTCrh+/Vgft, and BNST/MPAAro+/Cplx3+, all of which are implicated in diverse social behaviors such as mating and aggression and active during parenting in animals of various physiological states, highlighting the need for more precise cell type-specific analysis.
To independently confirm the activity pattern of AvPeBrs3+Vglut2+ neurons and further assess their real-time dynamics during behavior, we used an intersectional viral strategy to direct conditional expression of the calcium reporter GCaMP6m in a Cre- and Flp- dependent manner in Brs3-Cre, Vglut2-Flp mutant mice. Fiber photometry imaging was performed in freely interacting mice. Because AvPeBrs3+Vglut2+ neurons lie along the ventricle and are difficult to access, a mirror-tipped fiber was used to achieve maximal optical coverage. AvPeBrs3+Vglut2+ neurons displayed a significant increase in activity upon the introduction of pups into the cage of mothers, but not fathers, virgin males nor virgin females, supporting results obtained with immediate-early genes. Activity peaked during the first minute of infant interaction and subsequently returned to baseline, revealing minimal activity in pup-related behaviors after the initial interaction. No increase in activity was seen when ad libitum fed animals were eating or investigating a size-matched food item. Similarly, cumulative activity throughout the duration of the assay was elevated only in mothers during pup presentation, with no changes in other groups during food exploration.
The observed pattern of mother-specific pup-induced activity suggests a specific role of AvPeBrs3+Vglut2+ neurons in the parenting of mothers. Because parental care by mothers is significantly more robust than in any other parental state, we hypothesized that this excitatory population may enhance the initiation of parental care by boosting the activity of downstream parental circuits common to all parental states. To test this hypothesis, we selectively expressed the excitatory chemogenetic receptor hM3Dq in AvPeBrs3+Vglut2+ neurons of virgin female mice. In vivo chemogenetic activation of AvPeBrs3+Vglut2+ neurons by injecting virgin females with deschloroclozapine significantly increased the time spent crouching and decreased the latency to pup retrieval and nest building compared to saline-injected controls. No effect was observed on time spent grooming the pups. Interestingly, although AvPeBrs3+Vglut2+ neuronal activity was only detected at the initial phase of parenting, artificial activation of this population strongly increases crouching, a behavior that typically occurs after activity has returned to baseline, suggesting a long-lasting effect of a AvPeBrs3+Vglut2+ neuronal activation. Altogether, activation of AvPeBrs3+Vglut2+ neurons in virgin females increases several aspects of parental behaviors. By contrast, chemogenetic silencing of AvPeBrs3+Vglut2+ neurons in mothers through conditional hM4Di expression and deschloroclozapine injection did not notably affect parenting related behaviors, suggesting redundant mechanisms by which parenting behavior in mothers is enhanced.