An approximate line attractor in the hypothalamus encodes an aggressive state.

The hypothalamus regulates innate social behaviors, including mating and aggression. These behaviors can be evoked by optogenetic stimulation of specific neuronal subpopulations within MPOA and VMHvl, respectively. Here, we perform dynamical systems modeling of population neuronal activity in these nuclei during social behaviors. In VMHvl, unsupervised analysis identified a dominant dimension of neural activity with a large time constant (>50 s), generating an approximate line attractor in neural state space. Progression of the neural trajectory along this attractor was correlated with an escalation of agonistic behavior, suggesting that it may encode a scalable state of aggressiveness. Consistent with this, individual differences in the magnitude of the integration dimension time constant were strongly correlated with differences in aggressiveness. In contrast, approximate line attractors were not observed in MPOA during mating; instead, neurons with fast dynamics were tuned to specific actions. Thus, different hypothalamic nuclei employ distinct neural population codes to represent similar social behaviors.

The Neuropeptide Tac2 Controls a Distributed Brain State Induced by Chronic Social Isolation Stress.

Chronic social isolation causes severe psychological effects in humans, but their neural bases remain poorly understood. 2 weeks (but not 24 hr) of social isolation stress (SIS) caused multiple behavioral changes in mice and induced brain-wide upregulation of the neuropeptide tachykinin 2 (Tac2)/neurokinin B (NkB). Systemic administration of an Nk3R antagonist prevented virtually all of the behavioral effects of chronic SIS. Conversely, enhancing NkB expression and release phenocopied SIS in group-housed mice, promoting aggression and converting stimulus-locked defensive behaviors to persistent responses. Multiplexed analysis of Tac2/NkB function in multiple brain areas revealed dissociable, region-specific requirements for both the peptide and its receptor in different SIS-induced behavioral changes. Thus, Tac2 coordinates a pleiotropic brain state caused by SIS via a distributed mode of action. These data reveal the profound effects of prolonged social isolation on brain chemistry and function and suggest potential new therapeutic applications for Nk3R antagonists.

Transformations of neural representations in a social behaviour network.

Mating and aggression are innate social behaviours that are controlled by subcortical circuits in the extended amygdala and hypothalamus1-4. The bed nucleus of the stria terminalis (BNSTpr) is a node that receives input encoding sex-specific olfactory cues from the medial amygdala5,6, and which in turn projects to hypothalamic nuclei that control mating7-9 (medial preoptic area (MPOA)) and aggression9-14 (ventromedial hypothalamus, ventrolateral subdivision (VMHvl)), respectively15. Previous studies have demonstrated that male aromatase-positive BNSTpr neurons are required for mounting and attack, and may identify conspecific sex according to their overall level of activity16. However, neural representations in BNSTpr, their function and their transformations in the hypothalamus have not been characterized. Here we performed calcium imaging17,18 of male BNSTprEsr1 neurons during social behaviours. We identify distinct populations of female- versus male-tuned neurons in BNSTpr, with the former outnumbering the latter by around two to one, similar to the medial amygdala and MPOA but opposite to VMHvl, in which male-tuned neurons predominate6,9,19. Chemogenetic silencing of BNSTprEsr1 neurons while imaging MPOAEsr1 or VMHvlEsr1 neurons in behaving animals showed, unexpectedly, that the male-dominant sex-tuning bias in VMHvl was inverted to female-dominant whereas a switch from sniff- to mount-selective neurons during mating was attenuated in MPOA. Our data also indicate that BNSTprEsr1 neurons are not essential for conspecific sex identification. Rather, they control the transition from appetitive to consummatory phases of male social behaviours by shaping sex- and behaviour-specific neural representations in the hypothalamus.

Distinct hypothalamic control of same- and opposite-sex mounting behaviour in mice.

Animal behaviours that are superficially similar can express different intents in different contexts, but how this flexibility is achieved at the level of neural circuits is not understood. For example, males of many species can exhibit mounting behaviour towards same- or opposite-sex conspecifics1, but it is unclear whether the intent and neural encoding of these behaviours are similar or different. Here we show that female- and male-directed mounting in male laboratory mice are distinguishable by the presence or absence of ultrasonic vocalizations (USVs)2-4, respectively. These and additional behavioural data suggest that most male-directed mounting is aggressive, although in rare cases it can be sexual. We investigated whether USV+ and USV- mounting use the same or distinct hypothalamic neural substrates. Micro-endoscopic imaging of neurons positive for oestrogen receptor 1 (ESR1) in either the medial preoptic area (MPOA) or the ventromedial hypothalamus, ventrolateral subdivision (VMHvl) revealed distinct patterns of neuronal activity during USV+ and USV- mounting, and the type of mounting could be decoded from population activity in either region. Intersectional optogenetic stimulation of MPOA neurons that express ESR1 and vesicular GABA transporter (VGAT) (MPOAESR1∩VGAT neurons) robustly promoted USV+ mounting, and converted male-directed attack to mounting with USVs. By contrast, stimulation of VMHvl neurons that express ESR1 (VMHvlESR1 neurons) promoted USV- mounting, and inhibited the USVs evoked by female urine. Terminal stimulation experiments suggest that these complementary inhibitory effects are mediated by reciprocal projections between the MPOA and VMHvl. Together, these data identify a hypothalamic subpopulation that is genetically enriched for neurons that causally induce a male reproductive behavioural state, and indicate that reproductive and aggressive states are represented by distinct population codes distributed between MPOAESR1 and VMHvlESR1 neurons, respectively. Thus, similar behaviours that express different internal states are encoded by distinct hypothalamic neuronal populations.

Identification of the FSH-RH as the other gonadotropin-releasing hormone.

In vertebrates, folliculogenesis and ovulation are regulated by two distinct pituitary gonadotropins: follicle-stimulating hormone (FSH) and luteinizing hormone (LH). Currently, there is an intriguing consensus that a single hypothalamic neurohormone, gonadotropin-releasing hormone (GnRH), regulates the secretion of both FSH and LH, although the required timing and functions of FSH and LH are different. However, recent studies in many non-mammalian vertebrates indicated that GnRH is dispensable for FSH function. Here, by using medaka as a model teleost, we successfully identify cholecystokinin as the other gonadotropin regulator, FSH-releasing hormone (FSH-RH). Our histological and in vitro analyses demonstrate that hypothalamic cholecystokinin-expressing neurons directly affect FSH cells through the cholecystokinin receptor, Cck2rb, thereby increasing the expression and release of FSH. Remarkably, the knockout of this pathway minimizes FSH expression and results in a failure of folliculogenesis. Here, we propose the existence of the "dual GnRH model" in vertebrates that utilize both FSH-RH and LH-RH.

Gaining insights into the internal states of the rodent brain through vocal communications.

Animals display various behaviors during social interactions. Social behaviors have been proposed to be driven by the internal states of the animals that reflect their emotional or motivational states. However, the internal states that drive social behaviors are complex and are often difficult to interpret. Many animals species, such as rodents, use vocalizations for communication in various social contexts. This review provides an overview of the current understanding of mouse vocal communications, its underlying neural circuitry, and the potential to use vocal communications as a readout for the animal's internal states during social interactions.

Flexibility of neural circuits regulating mating behaviors in mice and flies.

Mating is essential for the reproduction of animal species. As mating behaviors are high-risk and energy-consuming processes, it is critical for animals to make adaptive mating decisions. This includes not only finding a suitable mate, but also adapting mating behaviors to the animal's needs and environmental conditions. Internal needs include physical states (e.g., hunger) and emotional states (e.g., fear), while external conditions include both social cues (e.g., the existence of predators or rivals) and non-social factors (e.g., food availability). With recent advances in behavioral neuroscience, we are now beginning to understand the neural basis of mating behaviors, particularly in genetic model organisms such as mice and flies. However, how internal and external factors are integrated by the nervous system to enable adaptive mating-related decision-making in a state- and context-dependent manner is less well understood. In this article, we review recent knowledge regarding the neural basis of flexible mating behaviors from studies of flies and mice. By contrasting the knowledge derived from these two evolutionarily distant model organisms, we discuss potential conserved and divergent neural mechanisms involved in the control of flexible mating behaviors in invertebrate and vertebrate brains.

The Mouse Action Recognition System (MARS) software pipeline for automated analysis of social behaviors in mice.

The study of naturalistic social behavior requires quantification of animals' interactions. This is generally done through manual annotation-a highly time-consuming and tedious process. Recent advances in computer vision enable tracking the pose (posture) of freely behaving animals. However, automatically and accurately classifying complex social behaviors remains technically challenging. We introduce the Mouse Action Recognition System (MARS), an automated pipeline for pose estimation and behavior quantification in pairs of freely interacting mice. We compare MARS's annotations to human annotations and find that MARS's pose estimation and behavior classification achieve human-level performance. We also release the pose and annotation datasets used to train MARS to serve as community benchmarks and resources. Finally, we introduce the Behavior Ensemble and Neural Trajectory Observatory (BENTO), a graphical user interface for analysis of multimodal neuroscience datasets. Together, MARS and BENTO provide an end-to-end pipeline for behavior data extraction and analysis in a package that is user-friendly and easily modifiable.

The Multi-Agent Behavior Dataset: Mouse Dyadic Social Interactions.

Multi-agent behavior modeling aims to understand the interactions that occur between agents. We present a multi-agent dataset from behavioral neuroscience, the Caltech Mouse Social Interactions (CalMS21) Dataset. Our dataset consists of trajectory data of social interactions, recorded from videos of freely behaving mice in a standard resident-intruder assay. To help accelerate behavioral studies, the CalMS21 dataset provides benchmarks to evaluate the performance of automated behavior classification methods in three settings: (1) for training on large behavioral datasets all annotated by a single annotator, (2) for style transfer to learn inter-annotator differences in behavior definitions, and (3) for learning of new behaviors of interest given limited training data. The dataset consists of 6 million frames of unlabeled tracked poses of interacting mice, as well as over 1 million frames with tracked poses and corresponding frame-level behavior annotations. The challenge of our dataset is to be able to classify behaviors accurately using both labeled and unlabeled tracking data, as well as being able to generalize to new settings.

Imaging neuropeptide release at synapses with a genetically engineered reporter.

Research on neuropeptide function has advanced rapidly, yet there is still no spatio-temporally resolved method to measure the release of neuropeptides in vivo. Here we introduce Neuropeptide Release Reporters (NPRRs): novel genetically-encoded sensors with high temporal resolution and genetic specificity. Using the Drosophila larval neuromuscular junction (NMJ) as a model, we provide evidence that NPRRs recapitulate the trafficking and packaging of native neuropeptides, and report stimulation-evoked neuropeptide release events as real-time changes in fluorescence intensity, with sub-second temporal resolution.

Morphological Analysis of the Axonal Projections of EGFP-Labeled Esr1-Expressing Neurons in Transgenic Female Medaka.

Some hypothalamic neurons expressing estrogen receptor α (Esr1) are thought to transmit a gonadal estrogen feedback signal to gonadotropin-releasing hormone 1 (GnRH1) neurons, which is the final common pathway for feedback regulation of reproductive functions. Moreover, estrogen-sensitive neurons are suggested to control sexual behaviors in coordination with reproduction. In mammals, hypothalamic estrogen-sensitive neurons release the peptide kisspeptin and regulate GnRH1 neurons. However, a growing body of evidence in nonmammalian species casts doubt on the regulation of GnRH1 neurons by kisspeptin neurons. As a step toward understanding how estrogen regulates neuronal circuits for reproduction and sex behavior in vertebrates in general, we generated a transgenic (Tg) medaka that expresses enhanced green fluorescent protein (EGFP) specifically in esr1-expressing neurons (esr1 neurons) and analyzed their axonal projections. We found that esr1 neurons in the preoptic area (POA) project to the gnrh1 neurons. We also demonstrated by transcriptome and histological analyses that these esr1 neurons are glutamatergic or γ-aminobutyric acidergic (GABAergic) but not kisspeptinergic. We therefore suggest that glutamatergic and GABAergic esr1 neurons in the POA regulate gnrh1 neurons. This hypothesis is consistent with previous studies in mice that found that glutamatergic and GABAergic transmission is critical for estrogen-dependent changes in GnRH1 neuron firing. Thus, we propose that this neuronal circuit may provide an evolutionarily conserved mechanism for regulation of reproduction. In addition, we showed that telencephalic esr1 neurons project to medulla, which may control sexual behavior. Moreover, we found that some POA-esr1 neurons coexpress progesterone receptors. These neurons may form the neuronal circuits that regulate reproduction and sex behavior in response to the serum estrogen/progesterone.

Evolutionally Conserved Function of Kisspeptin Neuronal System Is Nonreproductive Regulation as Revealed by Nonmammalian Study.

The kisspeptin neuronal system, which consists of a neuropeptide kisspeptin and its receptor Gpr54, is considered in mammals a key factor of reproductive regulation, the so-called hypothalamic-pituitary-gonadal (HPG) axis. However, in nonmammalian vertebrates, especially in teleosts, existence of kisspeptin regulation on the HPG axis is still controversial. In this study, we applied multidisciplinary techniques to a teleost fish, medaka, and examined possible kisspeptin regulation on the HPG axis. First, we generated knockout medaka for kisspeptin-related genes and found that they show normal fertility, gonadal maturation, and expression of gonadotropins. Moreover, the firing activity of GnRH1 neurons recorded by the patch clamp technique was not altered by kisspeptin application. Furthermore, in goldfish, in vivo kisspeptin administration did not show any positive effect on HPG axis regulation. However, as kisspeptin genes are completely conserved among vertebrates except birds, we surmised that kisspeptin should have some important nonreproductive functions in vertebrates. Therefore, to discover novel functions of kisspeptin, we generated a gpr54-1:enhanced green fluorescent protein (EGFP) transgenic medaka, whose gpr54-1-expressing cells are specifically labeled by EGFP. Analysis of neuronal projection of gpr54-1:EGFP-expressing neurons showed that these neurons in the ventrolateral preoptic area project to the pituitary and are probably involved in endocrine regulation other than gonadotropin release. Furthermore, combination of deep sequencing, histological, and electrophysiological analyses revealed various novel neural systems that are under control of kisspeptin neurons-that is, those expressing neuropeptide Yb, cholecystokinin, isotocin, vasotocin, and neuropeptide B. Thus, our new strategy to genetically label receptor-expressing neurons gives insights into various kisspeptin-dependent neuronal systems that may be conserved in vertebrates.

Whole brain-pituitary in vitro preparation of the transgenic medaka (Oryzias latipes) as a tool for analyzing the differential regulatory mechanisms of LH and FSH release.

Two types of gonadotropins, luteinizing hormone (LH) and follicle stimulating hormone (FSH), are important pituitary hormones for sexual maturation and reproduction, and both of them are centrally regulated by gonadotropin-releasing hormone (GnRH) from the hypothalamus. In mammals, these two gonadotropins are secreted from a single type of gonadotrope. The mechanisms of differential regulation by GnRH of the release of two types of gonadotropins with different secretory profiles are still unknown. In teleosts, however, LH and FSH are secreted from separate cellular populations, unlike in mammals. This feature makes them useful for studying the regulatory mechanisms of LH and FSH secretions independently. Here, we generated transgenic medaka lines that express Ca(2+) indicator protein, inverse-pericam, specifically in the LH or FSH cells. We performed cell-type-specific Ca(2+) imaging of LH and FSH cells, respectively, using the whole brain-pituitary preparations of these transgenic fish in which all neural circuits and GnRH neuronal projection to the pituitary are kept intact. LH and FSH cells showed different Ca(2+) responses to GnRH. The results suggest differential regulation mechanisms for LH and FSH release by GnRH. Moreover, we also succeeded in detecting the effect on LH cells of endogenous GnRH peptide, which was released by electrical stimulation of the axons of GnRH1 neurons. Thus, our newly developed experimental model system using the whole brain-pituitary in vitro preparation of the transgenic medaka is a powerful tool for analyzing the differential regulatory mechanisms of the release of LH and FSH by multisynaptic neural inputs to the pituitary.

Neurobiological study of fish brains gives insights into the nature of gonadotropin-releasing hormone 1-3 neurons.

Accumulating evidence suggests that up to three different molecular species of GnRH peptides encoded by different paralogs of gnrh genes are expressed by anatomically distinct groups of GnRH neurons in the brain of one vertebrate species. They are called gnrh1, gnrh2, and gnrh3. Recent evidence from molecular, anatomical, and physiological experiments strongly suggests that each GnRH system functions differently. Here, we review recent advancement in the functional studies of the three different GnRH neuron systems, mainly focusing on the electrophysiological analysis of the GnRH-green fluorescent protein (GFP) transgenic animals. The introduction of GFP-transgenic animals for the electrophysiological analysis of GnRH neurons greatly advanced our knowledge on their anatomy and electrophysiology, especially of gnrh1 neurons, which has long defied detailed electrophysiological analysis of single neurons because of their small size and scattered distribution. Based on the results of recent studies, we propose that different electrophysiological properties, especially the spontaneous patterns of electrical activities and their time-dependent changes, and the axonal projections characterize the different functions of GnRH1-3 neurons; GnRH1 neurons act as hypophysiotropic neuroendocrine regulators, and GnRH2 and GnRH3 neurons act as neuromodulators in wide areas of the brain.

Time-of-day-dependent changes in GnRH1 neuronal activities and gonadotropin mRNA expression in a daily spawning fish, medaka.

GnRH neurons in the preoptic area and hypothalamus control the secretion of GnRH and form the final common pathway for hypothalamic-pituitary-gonadal axis regulation in vertebrates. Temporal regulation of reproduction by coordinating endogenous physiological conditions and behaviors is important for successful reproduction. Here, we examined the temporal regulation of reproduction by measuring time-of-day-dependent changes in the electrical activity of GnRH1 neurons and in levels of expression of pituitary gonadotropin mRNA using a daily spawning teleost, medaka (Oryzias latipes). First, we performed on-cell patch-clamp recordings from GnRH1 neurons that directly project to the pituitary, using gnrh1-green fluorescent protein transgenic medaka. The spontaneous firing activity of GnRH1 neurons showed time-of-day-dependent changes: overall, the firing activity in the afternoon was higher than in the morning. Next, we examined the daily changes in the pituitary gonadotropin transcription level. The expression levels of lhb and fshb mRNA also showed changes related to time of day, peaking during the lights-off period. Finally, we analyzed effects of GnRH on the pituitary. We demonstrated that incubation of isolated pituitary with GnRH increases lhb mRNA transcription several hours after GnRH stimulation, unlike the well-known immediate LH releasing effect of GnRH. From these results, we propose a working hypothesis concerning the temporal regulation of the ovulatory cycle in the brain and pituitary of female medaka.

Regular pacemaker activity characterizes gonadotropin-releasing hormone 2 neurons recorded from green fluorescent protein-transgenic medaka.

GnRH2 is a molecule conserved from fish to humans, suggesting its important functions. However, recent studies have shown that GnRH2 neurons project widely in the brain but not to the pituitary, which suggests their functions other than stimulation of gonadotropin secretion. In contrast to the wealth of knowledge in GnRH1 and GnRH3 neuronal systems, the GnRH2 neuronal system remains to be studied, and there has been no single cell approach so far, partly because of the lack of GnRH2 system in rodents. Here, we generated GnRH2-green fluorescent protein (GFP) transgenic medaka for the first single cell electrophysiological recording from GnRH2 neurons in vertebrates. Whole-cell and on-cell patch clamp analyses revealed their regular pacemaker activities that are intrinsic to the GnRH2 neurons. Pacemaker activities of GnRH2 neurons were not peculiar to medaka because dwarf gourami GnRH2 neurons also showed similar pacemaker activities. By comparing with spontaneous action currents from GFP-expressing GnRH1 and GnRH3 neurons in the adult transgenic medaka, which were already in our hands, we have demonstrated that GnRH2 neurons show pacemaker activity similar to nonhypophysiotropic GnRH3 neurons but not to hypophysiotropic GnRH1 neurons. Thus, by taking advantage of medaka brain, which has all three GnRH neuronal systems with different axonal projection patterns and thus different functions, we have gained insights into the close relationship between the pattern of spontaneous electrical activity and the functions of the three. Moreover, the three types of GnRH-GFP transgenic medaka will provide useful models for studying multifunctional GnRH systems in future.