Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 21 July 2019

Oxytocin and Plasticity of Social Behavior

Abstract and Keywords

Oxytocin plays well-known roles in modulating social behavior in mammals. Oxytocin function depends on the brain circuitry it modulates, which is determined by the cell-type specific expression of the oxytocin receptor and the network integration of those cells. This review describes emerging evidence for the neural network mechanisms of oxytocin in behavioral plasticity in adults and development. A role for oxytocin in modulating excitatory/inhibitory balance and improving signal:noise processing is an emerging mechanism of function. This emerging literature calls for developmental studies of oxytocin modulation of signal:noise processing in socially naïve circuits. In this review, current oxytocin research findings are placed within the coordinate system of the Uncanny Valley Hypothesis as a model to better understand the role of oxytocin in experience-dependent development and adulthood, to translate research results among diverse mammalian species, and to generate testable predictions for future research.

Keywords: oxytocin, oxytocin receptor, social behavior, Uncanny Valley, social expertise, affiliation, signal:noise, excitatory inhibitory balance, experience-dependent development


Oxytocin (OXT) has pleiotropic functions throughout the organism, depending on the tissue-specific expression of oxytocin receptors (OXTR). A significant body of literature, covered in numerous recent and comprehensive reviews (Caldwell & Albers, 2016; Ma, Shamay-Tsoory, Han, & Zink, 2016; Marlin & Froemke, 2016; Miller & Caldwell, 2015; Numan & Young, 2015; Sannino, Chini, & Grinevich, 2016; Shamay-Tsoory & Abu-Akel, 2016), has established a prominent role for oxytocin in the brain basis of maternal care and adult socio-sexual behavior. Oxytocin is most well-known for its roles in adults in the detection/discrimination of familiar individuals, and the modulation of reward that permits the formation of social preferences. All adult social behavior relies heavily on social detection, perception, reward, and action. The brain capacity for species-typical adult social behavior requires experience-dependent developmental input. An emerging literature suggests a role for oxytocin in activity-dependent development. This review focuses on the central thesis that oxytocin enhances signal:noise processing across multiple nodes of social brain circuits, and that these findings from adults would be very informative from a developmental perspective. The available evidence raises the possibility that oxytocin in experience-dependent sensory development may have underexplored explanatory power for the roles of oxytocin in adult species-typical social behavior. This developmental role for oxytocin is conceptualized in the coordinate system of a well-characterized reductionist model, the Uncanny Valley Hypothesis. This framework accommodates prior research on oxytocin as well as generates testable predictions for future research.

Causal experimental designs by a variety of techniques to block or enhance oxytocin signaling have played a critical role in elucidating the neural mechanisms of action of oxytocin on social behavior. From the earliest studies delivering oxytocin into the brain of estrogen-primed virgin female rats (Pedersen, Ascher, Monroe, & Prange, 1982; Pedersen & Prange, 1979), to recent cell-type selective manipulations in knock-out and transgenic mice (Dolen, Darvishzadeh, Huang, & Malenka, 2013), the consistent theme is that oxytocin modulates complex behaviors and behavioral plasticity necessary for diverse social environments. Emerging evidence indicates that oxytocin plays roles in multiple nodes of social brain circuits, including modulation of social reward and signal:noise processing in primary sensory systems. How these processes play out during development as the socially naïve infant is just beginning to internalize the social world is a promising area for more research (Hammock, 2015). Here I review the recent strong causal evidence linking oxytocin to multiple nodes of social behavior circuits. The focus here is primarily on oxytocin in animal model research, as this research permits site specific manipulations needed to infer direct causation in social behavior’s many component parts.

Oxytocin Regulates Social Recognition Behavior at Multiple Nodes of Social Recognition Circuitry

Oxytocin has a well-established role in modulating social recognition behavior. Social recognition behavior is sometimes used interchangeably with “social memory,” although caution is warranted, as “social memory” is one potential interpretation of the social recognition behavior. Rodents tend to spend less and less time investigating a social stimulus with repeated presentations (habituation), and this habituated investigation time is specific to the familiar individual and not social contact in general, because a “dishabituation” can be elicited with the introduction of a novel social stimulus: investigation time returns to the high levels observed in the first trial once a novel social stimulus is introduced. It is also possible to present a familiar and novel animal at the same time and record where the individual spends its time investigating. In this two-choice assay, more time investigating a novel stimulus animal compared to a familiar animal is consistent with social recognition of the familiar individual. Whether tested serially in the habituation/dishabituation social recognition assay, or tested concurrently in the preference test, social recognition behavior is modulated by numerous factors, including whether or not the animal is tested after a period of social isolation or social housing (Kogan, Frankland, & Silva, 2000). Social housing results in improved social recognition as the time delay between exposure to a “familiar” animal and the test condition can be prolonged without deleterious effects on the social recognition behavior. In contrast, animals that are single-housed appear to be less likely to show a preference for the stranger over the familiar animal, which is interpreted as an impairment in social recognition. Using the habituation/dishabituation assay in male oxytocin wild type (OXT WT) and oxytocin knock-out (OXT KO) mice investigating ovariectomized female mice and intact female and male mice, Ferguson and colleagues (Ferguson, Aldag, Insel, & Young, 2001; Ferguson et al., 2000) demonstrated a robust role for oxytocin in modulating social recognition behavior: OXT KO mice failed to habituate their investigation of the same mouse over repeated trials. The absence of social recognition behavior was not explained by general deficits in memory, olfactory, or habituation processes. This deficit was rescued by intra-amygdala infusions of oxytocin.

Rats also display habituation of investigation behavior with repeated presentations of the same stimulus animal or social odor. As with mice, presentation of a novel animal after habituation reinstates high levels of social investigation, leading to the interpretation of the habituation/dishabituation behavior as a social memory process. Further, when tested concurrently, rats also spend more time investigating a novel rat compared to a familiar rat. Is it possible that oxytocin could modulate primary sensory systems before those signals could even make it to the amygdala to impact social recognition? Where in the brain is this sensory discrimination taking place? Emerging evidence suggests that oxytocin can play a top-down role in refining the incoming sensory information necessary to detect and discriminate social signals. Oettl et al. (Oettl et al., 2016) recently demonstrated a role for oxytocin in enhancing the signal:noise processing of olfactory social stimuli at the level of the olfactory bulb. The anterior olfactory nucleus is rich with oxytocin receptors in rats. The authors demonstrated that oxytocin receptor-expressing cells in the anterior olfactory nucleus were responsive to oxytocin and the selective oxytocin receptor agonist, [4-threonine, 7-glycine]oxytocin, (TGOT) (Busnelli, Bulgheroni, Manning, Kleinau, & Chini, 2013; Lowbridge, Manning, Haldar, & Sawyer, 1977). Oxytocin and TGOT enhanced the frequency of spontaneous excitatory postsynaptic currents (EPSCs). Using optogenetics, they were able to enhance endogenous oxytocin release onto the anterior olfactory nucleus, with the same effect on enhanced sEPSC frequency. An oxytocin receptor antagonist, OTA, blocked these effects. These glutamatergic oxytocin receptor expressing cells of the anterior olfactory nucleus project to the main olfactory bulb where they make contact with perisomatic region of granule cells, which are a local inhibitory interneuron. Oettl et al. measured the response of the granule cells after TGOT was applied to or oxytocin was released optogenetically into the anterior olfactory nucleus. They observed an enhancement of EPSC frequency in the granule cells. This effect was blocked when the connection between the anterior olfactory nucleus and the olfactory bulb was severed, demonstrating that the potentiating effect of oxytocin on distant granule cells was through excitatory drive from the anterior olfactory nucleus. Granule cells make inhibitory contact with the mitral cells, which are the main output cells of the olfactory bulb via the lateral olfactory tract. Oxytocin applied to the anterior olfactory nucleus increased the frequency of inhibitory currents of the mitral cells. When oxytocin enhances the activity of the AON, which enhances the activity of the granule cells, the net effect is to quiet the noise coming from the mitral cells to higher-order areas. Therefore, oxytocin modulation of this circuit is to enhance the signal:noise processing. In sum, oxytocin activity in the anterior olfactory nucleus modulated primary sensory processing of granule cell neurons in the olfactory bulb, representing a top-down process for enhancing social signal detection.

Oxytocin in the anterior olfactory nucleus led to a clear enhancement of signal:noise processing, but did it lead to detectable changes in social discrimination? Using optogenetics to enhance endogenous release of oxytocin in adult rats, Oettl et al. (2016) observed that compared to controls, oxytocin-enhanced rats spent more time investigating the anogenital area of a novel juvenile during a first sampling phase. After 2 hours, which is beyond the usual social recognition behavior capacity of rats, those rats that had stimulated oxytocin during the sample phase also showed stronger social recognition behavior when tested (more time spent investigating a novel juvenile than investigating the juvenile with whom they engaged two hours prior). Because oxytocin release was enhanced during the sampling phase and this directly influenced the sampling behavior that occurred, this experiment set the stage to proceed to investigate a potential role for oxytocin in enhancing sensory cues during the initial sampling and encoding of social recognition. This enhanced oxytocin signaling was not effective for increasing investigation of an object or subsequent object recognition behavior. A future test would be needed to see whether it would facilitate object recognition if the objects had been painted with different social or non-social odors. Importantly, because endogenous oxytocin could be acting in several locations throughout the brain via several receptor types to facilitate social recognition behavior, further experiments were needed to dissect the circuit by which endogenous oxytocin can enhance social exploration and social recognition behavior. To test the specific role for oxytocin receptor in the anterior olfactory nucleus, the experimenters introduced a rAAV-Cre virus into the anterior olfactory nucleus of floxed Oxtr mice. After using this viral approach to delete oxytocin receptor from the anterior olfactory nucleus, the animals were tested for their social recognition behavior. While they had heightened anogenital investigation of a novel mouse during the sampling phase compared to wild type controls, they failed to show social recognition behavior after 30 minutes. This demonstrates that oxytocin receptor in the anterior olfactory nucleus is necessary for social recognition behavior. This deficit was specific for social recognition behavior, as the mice showed no disruption in non-social odor recognition or discrimination. Social contexts result in enhanced release of oxytocin (Landgraf & Neumann, 2004; Neumann, 2007, 2008), so in the absence of social context, endogenous oxytocin activity is predicted to be low. In these non-social control experiments, exogenous oxytocin was not delivered, nor were endogenous oxytocin levels enhanced in the anterior olfactory nucleus. Therefore, the lack of effect for non-social odor recognition or discrimination could be a result of a lack of oxytocin release in those contexts. Choe et al. (2015) recently provided evidence that increasing oxytocin signaling in the absence of social context does not necessarily enhance recognition for non-social stimuli: oxytocin’s effects were specific to social stimuli but independent of social valence.

Male mice prefer sexually receptive females over non-receptive females. Choe et al. (2015) determined that if a novel non-social odor is presented with a sexually receptive female, male mice will later spend more time with the paired odor than with another novel odor. In contrast, males do not show a conditioned odor preference for odors paired with non-receptive females in diestrus. This difference in preference learning is thought to be mediated by the high salience of the receptive female group. The preference learning can be blocked by a non-peptidergic oxytocin receptor antagonist (L-368,899) given during the conditioning phase. Additionally, if the endogenous levels of oxytocin release are enhanced via optogenetics during conditioning, then males will form a preference for odors paired with unreceptive diestrus females. Stimulating oxytocin release in a nonsocial context did not enhance conditioned place preferences, indicating that the enhanced oxytocin release specifically facilitated social learning.

Where is oxytocin exerting this influence? Choe et al. (2015) demonstrated by in situ hybridization that Oxtr mRNA was present in the piriform cortex, and by immunohistochemistry determined that there were oxytocin positive nerve terminals that innervated the piriform cortex. In a later set of experiments, instead of using an odor as a conditioning stimulus, Choe et al. used direct optogenetic stimulation of the piriform cortex. During conditioning, optogenetic stimulation of the piriform cortex served as the conditioning stimulus paired with a receptive female. During the testing phase, instead of looking for place preference for an odorant, the animals were measured for a place preference for optogenetic activation of the piriform cortex. Animals conditioned in this way preferred the arena contingent with optogenetic stimulation of the piriform cortex as a conditioning stimulus. To test the specific contribution of oxytocin receptors in the piriform cortex on social odor preference learning, Choe et al. selectively eliminated oxytocin receptors in the piriform cortex of floxed Oxtr mice with a bilateral injection of a Cre recombinase virus co-injected with the channelrhodopsin virus that infected numerous cells of the piriform cortex. The animals with a selective deletion of oxytocin receptors from the piriform cortex did not show preference behavior for the optogenetic-paired chamber. Thus, these data strongly implicate oxytocin receptors in the piriform cortex in social odor preference learning, which is a type of appetitive learning. Choe et al. then asked whether oxytocin could facilitate learning to aversive social stimuli. To test this role for oxytocin, male mice were introduced into the home cage of a large aggressive CD-1 male mouse. After establishing aggression, a novel odor was paired with the aggressor. If this results in aversive conditioning, then the subjects should avoid this odor in a subsequent behavioral test. Odor pairings with an aggressive CD-1 did elicit avoidance behaviors to that odor. Subjects injected with an oxytocin receptor antagonist (L-368-899) or with a conditional deletion of oxytocin receptors in the piriform cortex failed to make an association between the odor and the aggressive encounter with the CD-1 mouse, as demonstrated by the duration of investigation of the CD-1 mouse. As with appetitive conditioning, instead of using an odor as the conditioning stimulus, optogenetic stimulation of the piriform cortex was used. Pairing of piriform cortex stimulation with interaction with an aggressive CD-1 mouse resulted in avoidance of a chamber with that optogenetic stimulation. This association was blocked with an oxytocin receptor antagonist (L-368-899) and also blocked in mice without oxytocin receptors in the piriform cortex. In sum, oxytocin receptors in the piriform cortex of mice are required for the ability to associate novel odors with social stimuli. This was true for appetitive stimuli such as a receptive female and aversive stimuli such as an aggressive male. Further, while effective for both appetitive and aversive social stimuli (i.e. valence independent), this role for oxytocin appeared to be selective for social stimuli and ineffective for non-social stimuli.

Signal:Noise Modulation as a General Mechanism of Oxytocin Function

As described previously, oxytocin modulates very early olfactory sensory processing with a significant role in rodent social behavior. There are numerous oxytocin receptors throughout the main and accessory olfactory bulbs in rodents to contribute to this chemosensory signal processing. Clearly, there are oxytocin receptors outside of these olfactory areas in rodents, as well as other species. For example, receptor mapping studies in primates have identified numerous oxytocin receptors in primary visual sensory areas (Freeman, Inoue, Smith, Goodman, & Young, 2014; Freeman, Smith, Goodman, & Bales, 2016; Freeman, et al., 2014). As primates rely heavily on visual input for social stimulus detection, this raises the possibility that oxytocin might modulate visual processing to enhance social visual signals. This is an important future research direction. Even in rodents, there is ample evidence that oxytocin modulates signal:noise processing in other brain areas, including primary sensory areas and other downstream networks.

The amygdala is an exemplary brain area where oxytocin appears to change the status of network activity. The behavioral effects of oxytocin in the medial amygdala on social recognition behavior were described earlier. But there are additional amygdala nuclei impacted by oxytocin. Huber, Veinante, & Stoop (2005) demonstrated that oxytocin enhanced evoked spikes in the oxytocin receptor expressing portion of the central amygdala (CeL/C) of rats. In contrast, oxytocin inhibited evoked spikes in the vasopressin 1a receptor expressing portion (CeM) of the central amygdala, through inhibitory connections from the CeL/C to the CeM. In the CeM, it looks as though oxytocin increases spike probability in oxytocin receptor expressing cells. This indicated that on a network level, oxytocin might change the threshold for neural responses, and in particular for an oxytocin-mediated reduction of amygdala output to brainstem fear response mechanisms.

Owen et al. (2013) demonstrated that oxytocin enhances signal:noise processing in the hippocampus, and that oxytocin does this by enhancing the activity of inhibitory interneurons. In the hippocampus, oxytocin lowers baseline activity and enhances signals. At baseline, pyramidal cells of CA1 from prepared slices fire at seemingly random intervals and also when the axons of Schaffer collaterals are stimulated. However, in the context of a selective oxytocin receptor agonist, TGOT, the pyramidal cells are silent, except when the Schaffer collaterals are stimulated, where they respond robustly. Thus, oxytocin reduced background activity and increased stimulated activity of hippocampal CA1 pyramidal cells. To achieve this network effect, TGOT increased the activity of fast-spiking interneurons in CA1 and enhanced the frequency of spontaneous inhibitory postsynaptic currents (IPSCs) at CA1 pyramidal cells, consistent with the reduced activity of CA1 cells under TGOT in the absence of stimulation. In contrast, TGOT enhanced evoked activity in pyramidal cells, and this effect depended on GABA A receptors. This role of oxytocin in the hippocampus—as a node in a network—could further enhance the signal:noise processing of social inputs. This example demonstrates how a diffusely delivered neuroendocrine modulator associated with social contexts may alter stimulus-evoked processing with spatial and temporal precision.

So far, several converging examples point to a role for oxytocin modulation of signal:noise processing. Electrophysiological and simulation studies indicate that excitatory/inhibitory balance controls network stability and the sensitivity to external inputs, i.e. “signals” (e.g., Economo & White, 2012). Excitatory/inhibitory balance is the scaled relationship between excitatory and inhibitory neuronal activity. It is most often studied in the context of network excitation by pyramidal cells and inhibition by fast spiking interneurons. An imbalance in these two processes is thought to underlie both autism and schizophrenia (Gao & Penzes, 2015). Causal evidence exists for disruptions to excitatory/inhibitory balance in the neocortex on social behavior (Yizhar et al., 2011). Therefore, a disruption in neocortical excitatory/inhibitory balance in the neocortex may result in poor social behavior outcomes by reducing cortical sensitivity to signals, in effect a disturbance in signal:noise processing. Given that oxytocin may modulate signal:noise processing in other circuits and other sensory modalities, is there evidence that oxytocin-enhanced signal:noise processing in the neocortex could actually impact social behavior? An illustrative example of the power of oxytocin enhanced signal:noise processing in the neocortex has recently emerged. As originally described in estrogen-primed rats, oxytocin can promote the onset of maternal behavior, which includes the retrieval of isolated pups to the nest (Pedersen et al., 1982; Pedersen & Prange, 1979). Marlin et al. (2015) demonstrated that oxytocin promotes retrieval behavior in inexperienced, pup-naïve virgin mice by enhancing the signal:noise processing in the auditory cortex. This enhanced signal detection of pup vocalizations permits the female to better detect distal pup cues, which is a necessary first step in motivated maternal retrieval behavior. To establish that oxytocin has these effects in auditory cortex, Marlin et al. (2015) demonstrated that oxytocin receptors were present in auditory cortex. By developing a selective oxytocin receptor antibody, they observed that oxytocin receptors were present on 30–40% of parvalbumin and somatostatin positive inhibitory interneurons, although these accounted for fewer than 20% of all oxytocin receptor positive neurons. In particular, oxytocin receptors were most prevalent in the left auditory cortex; while there was oxytocin innervation from the paraventricular nucleus to the neocortex, it did not appear to be lateralized. Temporary inactivations with local injections of the GABA agonist muscimol of the left, but not the right, auditory cortex impaired pup retrieval behavior in experienced mothers. This demonstrated that the left auditory cortex is required for maternal retrieval behavior in mice. To determine whether there is a role for oxytocin in the left cortex for retrieval in naïve virgins, oxytocin itself was infused or animals were optogenetically stimulated to release their own oxytocin into the left auditory cortex. These oxytocin-enhanced mice were quicker to retrieve pups than animals treated with saline. Finally, an oxytocin receptor antagonist (either the peptide receptor antagonist d(CH2)5[Tyr(Me)2,Thr4,Orn8,des-Gly-NH29]-vasotocin, or the non-peptide antagonist L-368,899) delivered to the left auditory cortex in experienced mothers did not significantly reduce retrieval behavior. This suggests that maternal mice have already solidified auditory mechanisms that no longer depend on oxytocin receptor activation. An important test would be to apply an oxytocin receptor antagonist to the left auditory cortex of virgin mice to see whether that reduces their onset of retrieval.

Electrophysiological recordings in the left auditory cortex while listening to pup vocalizations demonstrated a reliable neural activity in females with experience retrieving pups compared to naïve virgins. Importantly, in experienced females, excitation and inhibition were correlated in response to pup calls. When oxytocin was given to inexperienced virgin females along with pup calls, over a period of time the left auditory cortex transformed from a state of unreliable responding to pup calls to a state of strongly responding to pup calls. The immediate effect of the addition of oxytocin was a reduction in inhibition. Longer term changes over several hours led to correlated excitation and inhibition. Specifically, from trial to trial, they measured EPSC and IPSC responses in cortical neurons to pup calls in these initially pup-naïve females. In these inexperienced females, the initial trial to trial variability in EPSC or IPSCs was very high. However, the addition of oxytocin along with exposure to pup calls led to a rapid induction of consistent EPSC responses to pup calls. The IPSCs became more consistent in their response to pup calls after the first 45 minutes. Thus, oxytocin facilitated the encoding of experience. While there are some dissimilarities in the approach used, these observations in the auditory cortex are similar to the effects of oxytocin on evoked activity in the hippocampus where oxytocin enhanced the signal:noise ratio. Unanswered questions are whether an oxytocin receptor antagonist can block this auditory transition from pup call naïve to call responsive, although it is clear that once the transition from pup-naïve to pup-experienced has happened, oxytocin receptor is no longer necessary for reliable responses to pup calls. Further, it is still unclear what the electrophysiological effects of oxytocin alone or in combination with input are on oxytocin receptor expressing cells specifically. Ongoing approaches such as this will help clarify the circuit based mechanisms of the role of oxytocin in the plasticity of behavior.

Oxytocin Modulates Social Reward

The role of oxytocin receptor on neural network properties critically depends on the cell-type expression of the oxytocin receptor, not to mention the subcellular localization (e.g., distal dendrites vs. soma). If oxytocin receptor is present on local circuit interneurons and oxytocin increases the firing of local circuit interneurons, the net effect could be circuit inhibition. However, if receptors are present on glutamatergic projection neurons, and the effect of oxytocin is still enhanced spike probability, then the neural network outcome will be drastically different. Even this description is oversimplified, in that the presumed effects on the “circuit” depend on which cell or cells in the circuit are focal or downstream of oxytocin’s effects. In an elegant example of careful manipulations of which cells are manipulated and measured, Dolen et al. (2013) assessed neural mechanisms of social reward in the nucleus accumbens of adult male mice.

To assess social reward, social housing and social isolation were paired with novel types of bedding. Mice were then tested for their conditioned bedding preferences. Socially housed mice were removed from their home cage and placed into a test box with two compartments each with different kinds of novel bedding material. Mice were free to explore these two compartments. Then, mice were placed in a social environment with multiple mice and one of the new kinds of bedding. After this “social conditioning” to the new bedding, mice were placed in a cage by themselves with the other new bedding (“isolate conditioning”). Finally, mice were returned to the two chambered test box to determine if any preferences had developed for one bedding type over the other. Mice conditioned and tested in this way showed a place preference for the bedding type that had been paired with other mice—they did not prefer the bedding associated with isolation. This conditioned bedding preference was prevented by either systemic (intraperitoneal) injections of oxytocin receptor antagonist (L-368,899) or direct injections of this oxytocin receptor antagonist into the nucleus accumbens during conditioning. These injections did not affect locomotor behavior in the test. Further, cocaine conditioned place preference was not affected by the oxytocin receptor antagonist, demonstrating a selectivity for oxytocin on social conditioning.

Next, Dolen and colleagues determined if the endogenous source of oxytocin from the hypothalamus made direct connections with the nucleus accumbens. A retrograde tract-tracing virus injected into the nucleus accumbens revealed oxytocinergic projections from the paraventricular nucleus to the nucleus accumbens, but no apparent connections from the supraoptic nucleus to the nucleus accumbens. Double label immunohistochemistry for markers of various cell types in an Oxtr-venus reporter mouse indicated that parvalbumin positive and GFAP positive cells co-labeled venus reporter (oxytocin receptor) expressing cells in the nucleus accumbens. In addition to cells within the nucleus accumbens that express the Oxtr-venus reporter, a retrograde tract-tracing virus injected into the nucleus accumbens revealed numerous brain areas with oxytocin receptor expressing projections to the nucleus accumbens.

These anatomical findings lead to the following question: are the effects of oxytocin receptor in the nucleus accumbens due to local oxytocin receptor expressing cells or oxytocin receptor expressing cells from other brain areas that project to the nucleus accumbens? To differentially manipulate locally produced versus distally produced oxytocin receptors, two Cre recombinase viruses were employed in the floxed oxytocin receptor mouse line. An AAV-Cre-GFP virus eliminated oxytocin receptor expression locally. After eliminating locally produced oxytocin receptor, the social conditioning behavior was still intact, suggesting that oxytocin receptors produced by cells within the nucleus accumbens are not required for social conditioning of place preference. In contrast, a RBV-Cre-GFP virus injected into the nucleus accumbens deleted oxytocin receptors through a retrograde mechanism, thus eliminating the pool of oxytocin receptors in the nucleus accumbens that are on the terminals of neurons that project to the nucleus accumbens but have their cell bodies in other brain areas. These mice, with deletions of “presynaptic” oxytocin receptors in the nucleus accumbens, failed to form socially conditioned preference for bedding.

Which of these oxytocin receptor expressing projections to the nucleus accumbens are critical for the social conditioned place preference behavior? Through a series of anatomical investigations, the dorsal raphe was narrowed down as a suitable candidate for oxytocin dependent modulation of the accumbens: oxytocin receptor expressing serotonergic dorsal raphe cells projected to the nucleus accumbens. When the AAV-Cre-GFP virus was injected into the dorsal raphe to specifically eliminate oxytocin receptor expression in the raphe, this was sufficient to prevent socially conditioned place preference. Instead of enhancing spike probability of evoked spikes, oxytocin in the nucleus accumbens induced long term depression (LTD), or a decrease in EPSCs. Oxytocin did not alter IPSCs in medium spiny neurons of the nucleus accumbens. Oxytocin-induced LTD was apparent in both dopamine D1 and D2 expressing cells of the nucleus accumbens. Critically, the LTD was more pronounced in animals that had been isolated. The combined electrophysiological and anatomical experiments in this study support a network gating mechanism for paraventricular nucleus (PVN) oxytocin neuron terminals in the nucleus accumbens: oxytocin from PVN terminals binds to oxytocin receptors on the terminals of raphe inputs to the nucleus, this causes enhanced serotonin release within the accumbens which bind to serotonin 1b receptors on the presynaptic glutamatergic inputs to the nucleus accumbens. These serotonin receptors inhibit glutamatergic vesicle release onto medium spiny neurons in the nucleus accumbens, which represents the focal mechanism of oxytocin induced LTD in the nucleus accumbens. Indeed, direct blockade of serotonin 1b receptors in the nucleus accumbens also induced LTD even in the presence of oxytocin and blocked socially conditioned place preference behavior. While the effect of oxytocin at the medium spiny neuron is LTD, this is made possible by the social context-dependent enhanced release of oxytocin from the PVN into the nucleus accumbens, detected by the oxytocin receptors on serotonergic terminal afferents in the accumbens. This example illustrates the value of precise circuit manipulations and identifying the cells that are directly affected by oxytocin and also the downstream network effects that permit neural mechanisms of social behavior plasticity.

These data open up a great deal of possibilities in designing future experiments. An important question remains—how does an animal know that an environmental stimulus is “social” or worthy of oxytocin-release? Seminal behavioral experiments in rodents demonstrate that developmental experience shapes the definition of social preference stimuli (e.g., Denenberg, Hudgens, & Zarrow, 1964). Presumably, it is developmental experience that would likewise determine what environmental stimuli cause oxytocin release.

Oxytocin in Experience-Dependent Development

As already described, significant evidence from the past several years indicates that oxytocin influences the signal:noise processing of oxytocin receptor expressing circuits and their interconnected networks. The emerging argument is that by modulating network properties, oxytocin can influence the detection, perception, salience, and valence of oxytocin eliciting stimuli, which by definition, should be social stimuli. These processes are distributed throughout the brain in various nodes of the social brain network. A fundamental question that remains to be addressed is how do socially naïve newborns know to engage the oxytocin system in social contexts? Further, how does oxytocin modulate signal:noise processing in development? Finally, are the known developmentally transient “infant patterns” of oxytocin receptor an important mechanism of sensitive period plasticity for social brain development?

We have hypothesized (Hammock & Levitt, 2013; Hammock, 2015; Vaidyanathan & Hammock, 2016) that the transient expression of oxytocin receptor in the neocortex of developing mammals permits a special plasticity for social cues. For example, the plasticity that was observed by Marlin et al. (2015) did not require associative learning mechanisms. The females were passively exposed to pup calls, and their detection was enhanced by oxytocin in the neocortex. Consider the emerging literature on perceptual narrowing (Grossmann, Missana, Friederici, & Ghazanfar, 2012; Lewkowicz & Ghazanfar, 2012; Steckenfinger & Ghazanfar, 2009; Takahashi et al., 2015) for language and for face processing. In developing human infants, there are periods of multipotentiality for stimulus detection. However, with exposure, the kinds of stimuli that can be perceived become narrow. Certain phonemic distinctions become almost impossible for non-native listeners to detect. It is not so difficult to imagine that a heightened developmental plasticity for these processes might be achieved by enhanced signal:noise processing by oxytocin. A transient period of heightened receptor expression in the neocortex (Hammock & Levitt, 2013) might make this a putative sensitive period mechanism. While non-associative (passive exposure) may be sufficient, associative processes would most certainly further facilitate learning.

There is evidence that oxytocin may enhance experience-dependent multisensory neocortical plasticity during this period of heightened neocortical oxytocin receptor expression. Zheng et al. (2014) demonstrated that sensory deprivation in the first two post-natal weeks of mice results in decreased oxytocin production in the hypothalamus, specifically the PVN and not the SON measured at 2 weeks of age. Beginning on the first post-natal day, mice were either whisker deprived, reared in constant darkness, or had typical sensory experience. At 2 weeks of age, the activity of pyramidal neurons of layers II/III of neocortex was assessed. Animals with early sensory deprivation had reduced baseline spontaneous EPSCs in the area of cortex representing the primary sensory modality as well as other primary sensory areas. Deprived animals also had reduced evoked activity, including for modalities other than the deprived modality; surprisingly, even whisker deprivation resulted in reduced light-evoked firing rates and signal:noise ratios in layers II/III of primary visual cortex. By screening the telencephalon and diencephalon for potential transcriptomic changes that might be responsible for deprivation-induced changes in neocortical activity, reduced oxytocin coding mRNA levels were observed in both dark rearing and whisker deprivation conditions. Following up on the transcriptomic analysis, immunohistochemistry and electrophysiology revealed less oxytocin peptide production as well as decreased electrophysiological activity in oxytocin neurons of the PVN. This indicates that by some unknown mechanism, unimodal sensory deprivation leads to reduced oxytocin production in the neonate. In contrast, sensory enrichment enhanced oxytocin production in developing mice. Oxytocin injection in vivo was sufficient to rescue the effects of sensory deprivation on excitatory synaptic transmission measured in pre-weaning mice. Oxytocin enhancement of mEPSC frequency was blocked by the selective oxytocin receptor antagonist (d(CH2)51,Tyr(Me)2,Thr4,Orn8,des-Gly-NH29)-vasotocin or atosiban. These results are important because they suggest a broader sensory system role for oxytocin that may permit oxytocin to bias sensory system development toward social inputs. What is critically needed at this point is an assessment of the acute behavioral impact of oxytocin modulation of neocortical excitability at these pre-weaning ages and a determination of the long term impacts of oxytocin-induced neocortical multi-sensory plasticity. Is it the case that oxytocin is a protective factor or a plasticity factor? In other words, in a sensory poor environment such as whisker deprivation, does oxytocin alone serve to protect the brain in the long term, or alternatively, if oxytocin is a plasticity factor, would it enshrine the impoverished environment in the neural substrates of sensory processing? Additionally, rats and mice differ in their transient developmental profile of oxytocin receptors. While mice have robust transient oxytocin receptor expression in all primary sensory cortices (Hammock & Levitt, 2013), rats have a transient oxytocin receptor peak in the cingulate cortex (Tribollet et al., 1989), with little appreciable oxytocin receptor in primary sensory areas. Do rats also show deprivation induced reductions in oxytocin production and does this selectively impact electrophysiological properties of the cingulate cortex and not primary sensory cortices? As the results in mice have appreciable clinical relevance for the combination of broad neocortical pathology in autism spectrum disorders and numerous associations with oxytocin system and ASD, these species differences in infant patterns of oxytocin receptors may be informative. According to the human transcriptome atlas (Kang et al., 2011), developmental oxytocin receptor expression in human neocortex may be more mouse-like than rat-like.

Reductionist Models of Oxytocin in Social Behavioral Plasticity

As research on oxytocin function in the brain has grown rapidly over the past several decades, so have the attempts to frame the role of oxytocin into simplified theoretical models that summarize the outcomes of numerous studies in humans and other animals. Shamay-Tsoory and Abu-Akel (2016) recently reviewed the many reductionist theories the role of oxytocin in social behavior (Figure 1A–D). Each subsequent reductionist model has been refined by newly available data. The newest of these reductionist models, the salience hypothesis of oxytocin proposed by Shamay-Tsoory and Abu-Akel, is particularly powerful because it describes a role for oxytocin in increasing the salience of social stimuli. In its current formulation, it specifically relies on the modulation of social salience through oxytocin interaction with dopaminergic circuits (Shamay-Tsoory & Abu-Akel, 2016). While this model incorporates a prominent and evidence-based role for oxytocin modulation of dopaminergic circuits to regulate salience, the salience model overlooks a potential role for oxytocin modulation of signal:noise ratios in several other nodes of social brain circuits including primary sensory processing. Oxytocin modulation of sensory inputs may be an important first step in increasing the salience of social inputs which then can facilitate motivated behavior.

In addition to neglecting oxytocin’s emerging role in primary sensory processing, none of the reductionist models proposed thus far include a role for oxytocin in the development of the social brain. None capture the fundamental dilemma in social behavior in the newborn— what is “social” to a socially naïve infant? We know that perinatal manipulations of oxytocin affect social behavior in adulthood (Carter, 2003; Cushing & Kramer, 2005; Hammock, 2015), although the mechanisms of action are relatively unknown. A combination of heightened sensitivity due to enhanced receptor expression profiles in development (Hammock & Levitt, 2013; Hammock, 2015; Vaidyanathan & Hammock, 2016) as well as a potential for increased signal:noise processing as observed in mature brains might be a robust mechanism for the effects of oxytocin on developmental plasticity. To account for the contribution of developmental oxytocin across multiple nodes of the social brain, here, I propose both organizational (developmental) and activational (adulthood) roles for oxytocin in the coordinate system of the Uncanny Valley Hypothesis. This model of oxytocin incorporates prior research in adults and developing mammals and proposes a key role for developmental oxytocin in setting up a brain to respond to social cues (Figure 1E). As will be explained in more detail, this model incorporates sensory expertise and affiliation on two orthogonal axes. I propose that during development, oxytocin helps define both of these axes, and during adulthood, oxytocin modulates the position along these axes.

Oxytocin and Plasticity of Social BehaviorClick to view larger

Figure 1. Schematic depiction of reductionist models of the role of oxytocin in social behavior. These models are presented roughly in the order that they have emerged from the growing evidence in the literature. Thus, the supporting evidence for each subsequent theoretical formulation is built upon the prior available evidence and refined with newly available evidence. (A) In the prosocial model, oxytocin enhances affiliative social behavior, such as partner preference behavior and maternal care. (B) In the fear reduction model, oxytocin promotes affiliative social behavior by reducing fear states. Because oxytocin can achieve both prosocial behavior and fear reduction, and these two systems antagonize each other, the two models could be combined to achieve the same result of oxytocin enhancing affiliation. (C) The simple model of oxytocin enhancing social behavior directly through enhanced social reward or indirectly through reduction in fear was called in to question by the data supporting the In-group/Out-group model wherein the prosocial effects of oxytocin seem to depend on social context. Oxytocin enhanced prosocial behaviors in the context of the in-group, while antagonizing prosocial behaviors with out-group members. (D) In the salience model, oxytocin is hypothesized to increase the salience of and attention to social stimuli by enhancing dopaminergic signaling, regardless of the valence of the stimulus. Because salience is independent of valence as depicted on a vertical axis in A–C, salience is depicted as orthogonal to the affiliation axis. (E) The coordinate system of Mori’s Uncanny Valley Hypothesis combines the features of all of the reductionist models to provide a robust model (H1) that will generate testable predictions about the role of oxytocin in species-typical social behavior. Specifically, this framework posits that developmental oxytocin enhances the development of the x and y axes. Along the x-axis, developmental oxytocin increases signal:noise processing of species typical social sensory stimuli to shape social expertise, thus creating a “social brain” with high social stimulus salience. Along the y-axis, developmental oxytocin would determine the capacity for affiliation—an individual’s achievable peak along the y-axis. In adulthood, oxytocin continues to modulate the position of the x- and y-axes of the Uncanny Valley coordinate system. See text.

The Uncanny Valley Hypothesis was first proposed by Mori in 1970 (“Bukimi No Tani” literally translated as “valley of eeriness” (Mori, 1970, 2012)) to describe the eerie feelings adults experience when engaging with humanoid robots and prostheses. The original formulation of the Uncanny Valley was mapped out on a two-dimensional coordinate system with human likeness on the x-axis (from not at all humanoid at the origin to fully human) and shinwakan, literally translated as “affinity” on the y-axis. The highest values on the y-axis could roughly be equated with strong preference or attachment while negative scores on the y-axis would be associated with aversion, fear, or perhaps disgust. Mori’s Uncanny Valley landscape is depicted on this coordinate system. At relatively low levels of human likeness, affinity is near zero or neutral levels. However, as an object, such as a robot, has more human-like features (e.g., WALL·E from the 2008 Disney/Pixar Animation Studios film of the same name), the preference/affinity value on the y-axis rises. As human likeness increases along the x-axis, so does affinity on the y-axis, until the humanoid features are very close to human likeness, but disturbingly “off” (e.g., the human animations in The Polar Express (2004, Warner Brothers Pictures) fell into this Uncanny Valley for many adult viewers). Thus, Mori’s Uncanny Valley is the non-linear dip in affinity for visual stimuli that are close to human, but not human enough. This model of aversion to humanoid features that do not quite meet the criteria for full human likeness, has been used to organize theoretical constructs of in-group love/ out-group derision, xenophobia, racism, and norm violations. Specifically, in contrast to an in-group member who would rank to the far right on the x-axis and highly on the y-axis, an out-group member, while still highly salient, would fall into the Uncanny Valley as “sub-human” and a potential threat rather than as a potential collaborator. I propose that the coordinate system of Mori’s Uncanny Valley Hypothesis is a convenient framework for accommodating prior research on the roles of oxytocin in adulthood. Further, I propose that oxytocin may be an important modulator of the developmental specification of the capacity for affinity (affiliation) on the y-axis and salience due to social expertise comprising the x-axis.

The y-axis looks most like the combination of the prosocial (Figure 1A) and fear reduction models (Figure 1B) of oxytocin. The scale of the y-axis would be specified during development. In rats, there is good evidence that affiliative social orienting mechanisms develop before fear-based aversion learning. Specifically, in the first postnatal week, infant rats learn to approach stimuli paired with the mom, and even stimuli paired with aversive stimuli such as tail shocks (Sullivan, 2003). Therefore, the capacity for affiliation as depicted by the positive value of the y-axis starts to develop before the negative valence of this axis. The model predicts that oxytocin during this very early developmental period would strengthen this process and perhaps enhance the capacity for affiliation in the future. In rats, around the second week of life, infants begin to show conditioned fear in the absence of stress buffering provided by the mother. In the Uncanny Valley model, this would be represented by the first appearance of negative values along the y-axis. The first appearance of conditioned fear in neonatal rats involves stress hormone activity (corticosterone) in the amygdala, which is buffered by maternal presence during a transitional sensitive period (Moriceau, Wilson, Levine, & Sullivan, 2006). Oxytocin antagonizes the hypothalamic pituitary adrenal release of corticosterone (Kramer, Cushing, & Carter, 2003; Smith & Wang, 2014; Uvnas-Moberg, 1998), so again, oxytocin would increase affiliation through reducing stress and fear related mechanisms. There is good evidence from work in neonatal rats, monkeys, and humans that newborn mammals can use parental presence as a buffer to stressors (Gunnar et al., 2015; Levine, 2001). After this developmental sensitive period, maternal presence is significantly less effective at buffering against fear learning.

As observed in adults, oxytocin is hypothesized to enhance signal:noise processing in primary sensory systems during developmental social encounters. This would facilitate the experience-dependent foundation of the x-axis of the Uncanny Valley coordinate system. This prediction suggests that oxytocin activity in concert with social experience during development will refine the neural representation of a social sensory template or “norm” for conspecifics against which all social sensory stimuli would be compared. This social template would represent the “in-group,” which we know is defined by developmental social experience, even in rodents (Denenberg et al., 1964). This template comprises the sensory system capacity to detect norm violations among social stimuli. A norm-violation would trigger “out-group” aversion responses. In adulthood, an expected norm violation might be a human of a different race (if the rearing environment was racially homogeneous), or different mannerisms, language, dialect, and so forth. A prediction of the x-axis role for oxytocin would be that oxytocin guides the development of this axis in ethologically relevant contexts. Oxytocin levels are presumed to be highest during sensory stimulation normally provided by a caregiver. If oxytocin enhances the signal:noise processing of these sensory stimuli, then such socially relevant stimuli are more likely to contribute to the neural representation of the experienced social environment, leading to expertise on those stimuli—in effect, “social expertise.” This framework predicts that if high levels of oxytocin are added during development as a clinical treatment but in the absence of ethologically typical social environment, then oxytocin might contribute to the development of expertise for other non-social stimuli in the environment. In an extreme case, consider a clinically delivered enhanced oxytocin signal in a low stimulus neglectful environment. The developing brain in this case, might become a stimulus detection expert for non-social object cues such as shadows, doors, building sounds, etc. As theorized for face processing, this could lead to an expected norm for these objects and enhanced affiliation for stimuli that closely match the developmentally specified sensory template. This would lead to enhanced norm violation detection mechanisms for objects that do not match the template in adulthood. Thus, individuals with high oxytocin levels in low social contexts might develop a particular sensitivity to and pre-occupation with non-social objects due to developmentally acquired expertise enhanced by oxytocin. In sum, the proposed role for oxytocin in the x-axis of the Uncanny Valley model is to enhance signal:noise processing during the developmental acquisition of species-typical socially relevant sensory inputs. This would promote the development of the neural representation of typical social stimuli. Such expected stimuli could include own-race faces (if reared in a racially homogenous environment), language phonemes, the feel of another’s skin, and body movements.

After experience-dependent developmental sensitive periods for the acquisition of affiliative capacity (scale of the y-axis) and social expertise (specification of the x-axis), oxytocin could continue to shape the Uncanny Valley landscape by changing the stimulus position along the x and y-axes. This role is proposed both during development and in adulthood. Because in the adult, the y-axis cannot distinguish between oxytocin enhancing prosocial behavior directly or indirectly through reducing fear, the y-axis represents the first two reductionist theories (prosocial and fear reduction). At a neural level, moving along this axis would require considering the net effect of affiliative networks (ventral forebrain reward circuits) and fear circuitry (amygdala). Because mature patterns of activity permit the frontal cortex to down regulate the amygdala to provide a mechanism of top-down suppression of fear, it might be possible to dissociate fear processes from affiliative processes in adults by looking at the change in the shape of the Uncanny Valley over time. The prediction is that initial responses to norm-violating “uncanny stimuli” might yield an Uncanny Valley response (i.e. aversion to out-group), but these responses could be actively suppressed with a brief delay by top-down frontal cortical suppression of the amygdala. Importantly, the Uncanny Valley landscape is not necessarily fixed, but rather it is hypothesized to be further influenced by oxytocin in ventral forebrain reward circuits, in fear circuits, and even top-down cognitive control of the position along the y-axis. For example, a person with high levels of cognitive control, perhaps obtained through mindfulness practices, may be able to eliminate the Uncanny Valley aversion response to out-group members. This model predicts that children reared in welcoming multicultural environments might not have to rely on top-down control to minimize Uncanny Valley responses as adults, as their developmentally defined in-group and social norm definitions might be more inclusive.

The Social Salience model and the coordinate system of the Uncanny Valley have many of the same predictions, such as the potential for exacerbation of social threat detection in contexts with high oxytocin levels and high-threat stimuli. Similarly, both models predict that chronic oxytocin therapy must be given concurrently with appropriate social behavior therapy. These predictions are testable (falsifiable) with currently available research tools. The main refinement offered by the coordinate system of the Uncanny Valley model for oxytocin over other reductionist models is that it gives a primary role to oxytocin in the developmental emergence of social behavior through enhancing the capacity for affiliation on the y-axis and enhanced signal:noise processing in primary sensory systems for the development of social expertise on the x-axis, which creates the foundation for oxytocin’s role in adulthood. Therefore, one benefit of the model is the ability to make predictions about atypical oxytocin exposures during development. After typical development with typical levels of oxytocin, for adults the model would predict that low oxytocin is associated with low levels of social salience. However, if oxytocin is low during development, the model predicts that the capacity for affiliation in the future is blunted and the developing x-axis will not reach as far in terms of “expertise.” This may blunt Uncanny effects, even in the presence of high levels of experimenter-delivered oxytocin. This is a falsifiable prediction that could be tested in rodent models. This developmental role for oxytocin in shaping the expertise and affiliation landscape necessary for the emergence of the Uncanny Valley will no doubt require refinement as new evidence emerges.


Recent rapid progress in understanding the network integration of oxytocin has led to the hypothesis that oxytocin enhances signal:noise ratios. As described in detail earlier, this is evident across sensory modalities (e.g., olfactory, auditory, visual, somatosensory) as well as in brain regions downstream of primary sensory processing such as the hippocampus and the nucleus accumbens. The expected behavioral effect of enhanced signal:noise processing is consistent with observed improvement in social recognition behavior and capacity for social reward, for example. More broadly, this neural circuit level understanding informs human models of social behavior and cognition, and is highly consistent with a “social salience” role for oxytocin, but includes a role for oxytocin during development to specify a coordinate system in which oxytocin can function in adulthood. Thus, oxytocin’s net effect may be to enhance stimulus detection in social contexts and further amplify downstream signals to better integrate social cues with internal physiological needs. Oxytocin modulation of excitatory/inhibitory balance ties together previously distinct literatures on autism pathology, namely disrupted cortical balance and neuropeptides (Hammock, 2015; Hammock & Young, 2006; Hammock & Levitt, 2006; Marlin & Froemke, 2016).

The Uncanny Valley model of oxytocin elaborates on this Stimulus Salience model by placing a special emphasis on the developmental refinement of the social brain. The Uncanny Valley model of oxytocin predicts that oxytocin facilitates the sensory-dependent development of social expertise on the x-axis and capacity for affiliation on the y-axis of Mori’s Uncanny Valley coordinate system. After developmental sensitive periods, oxytocin can continue to modulate the individual’s position along the x- and y-axes of the Uncanny Valley coordinate system.

Future Directions

Despite rapid recent progress, numerous important questions remain. As alluded to earlier, it will be valuable to gain a better understanding of how socially naïve infant brains are able to engage their oxytocin system in appropriate social contexts. Each species thus far investigated has an “infant pattern” of oxytocin receptor availability wherein some brain regions have a peak of expression in the neonatal period. The brain area(s) with this infant pattern seems to vary across species. The infant pattern may represent an underappreciated sensitive period for the experience-dependent development of species-typical social expertise. Additionally, defining the features of putative oxytocin-mediated sensitive periods for experience-dependent developmental plasticity has important implications for translational models of the mechanisms of benefit of social support during early infancy. The Uncanny Valley model of oxytocin provides an accommodating and accessible framework to probe the role of oxytocin the development of social behavior, with strong translational predictive value. Understanding the neural processes of sensitive period plasticity has been an ongoing goal for translational mental health research to yield better intervention tools for neglected and abused children and children with other environmental or genetic hurdles in development.


Busnelli, M., Bulgheroni, E., Manning, M., Kleinau, G., & Chini, B. (2013). Selective and potent agonists and antagonists for investigating the role of mouse oxytocin receptors. Journal of Pharmacology and Experimental Therapeutics, 346(2), 318–327. doi:10.1124/jpet.113.202994Find this resource:

Caldwell, H. K., & Albers, H. E. (2016). Oxytocin, vasopressin, and the motivational forces that drive social behaviors. Current Topics in Behavioral Neuroscience, 27, 51–103. doi:10.1007/7854_2015_390Find this resource:

Carter, C. S. (2003). Developmental consequences of oxytocin. Physiology and Behavior, 79(3), 383–397.Find this resource:

Choe, H. K., Reed, M. D., Benavidez, N., Montgomery, D., Soares, N., Yim, Y. S., & Choi, G. B. (2015). Oxytocin mediates entrainment of sensory stimuli to social cues of opposing valence. Neuron, 87(1), 152–163. doi:10.1016/j.neuron.2015.06.022Find this resource:

Cushing, B. S., & Kramer, K. M. (2005). Mechanisms underlying epigenetic effects of early social experience: the role of neuropeptides and steroids. Neuroscience and Biobehavior Reviews, 29(7), 1089–1105. doi:S0149-7634(05)00067-9 [pii] 10.1016/j.neubiorev.2005.04.001Find this resource:

Denenberg, V. H., Hudgens, G. A., & Zarrow, M. X. (1964). Mice reared with rats: Modification of behavior by early experience with another species. Science, 143(3604), 380–381.Find this resource:

Dolen, G., Darvishzadeh, A., Huang, K. W., & Malenka, R. C. (2013). Social reward requires coordinated activity of nucleus accumbens oxytocin and serotonin. Nature, 501(7466), 179–184. doi:10.1038/nature12518Find this resource:

Economo, M. N., & White, J. A. (2012). Membrane properties and the balance between excitation and inhibition control gamma-frequency oscillations arising from feedback inhibition. PLoS Computational Biology, 8(1), e1002354. doi:10.1371/journal.pcbi.1002354Find this resource:

Ferguson, J. N., Aldag, J. M., Insel, T. R., & Young, L. J. (2001). Oxytocin in the medial amygdala is essential for social recognition in the mouse. Journal of Neuroscience, 21(20), 8278–8285.Find this resource:

Ferguson, J. N., Young, L. J., Hearn, E. F., Matzuk, M. M., Insel, T. R., & Winslow, J. T. (2000). Social amnesia in mice lacking the oxytocin gene. Nature Genetics, 25(3), 284–288.Find this resource:

Freeman, S. M., Inoue, K., Smith, A. L., Goodman, M. M., & Young, L. J. (2014). The neuroanatomical distribution of oxytocin receptor binding and mRNA in the male rhesus macaque (Macaca mulatta). Psychoneuroendocrinology, 45, 128–141. doi:10.1016/j.psyneuen.2014.03.023Find this resource:

Freeman, S. M., Smith, A. L., Goodman, M. M., & Bales, K. L. (2016). Selective localization of oxytocin receptors and vasopressin 1a receptors in the human brainstem. Social Neuroscience, 1–11. doi:10.1080/17470919.2016.1156570Find this resource:

Freeman, S. M., Walum, H., Inoue, K., Smith, A. L., Goodman, M. M., Bales, K. L., & Young, L. J. (2014). Neuroanatomical distribution of oxytocin and vasopressin 1a receptors in the socially monogamous coppery titi monkey (Callicebus cupreus). Neuroscience, 273, 12–23. doi:10.1016/j.neuroscience.2014.04.055Find this resource:

Gao, R., & Penzes, P. (2015). Common mechanisms of excitatory and inhibitory imbalance in schizophrenia and autism spectrum disorders. Currents in Molecular Medicine, 15(2), 146–167.Find this resource:

Grossmann, T., Missana, M., Friederici, A. D., & Ghazanfar, A. A. (2012). Neural correlates of perceptual narrowing in cross-species face-voice matching. Developmental Science, 15(6), 830–839. doi:10.1111/j.1467-7687.2012.01179.xFind this resource:

Gunnar, M. R., Hostinar, C. E., Sanchez, M. M., Tottenham, N., & Sullivan, R. M. (2015). Parental buffering of fear and stress neurobiology: Reviewing parallels across rodent, monkey, and human models. Social Neuroscience, 10(5), 474–478. doi:10.1080/17470919.2015.1070198Find this resource:

Hammock, E., & Levitt, P. (2013). Oxytocin receptor ligand binding in embryonic tissue and postnatal brain development of the C57BL/6J mouse. Frontiers in Behavioral Neuroscience, 7. doi:10.3389/fnbeh.2013.00195Find this resource:

Hammock, E. A. (2015). Developmental perspectives on oxytocin and vasopressin. Neuropsychopharmacology, 40(1), 24–42. doi:10.1038/npp.2014.120Find this resource:

Hammock, E. A., & Young, L. J. (2006). Oxytocin, vasopressin and pair bonding: implications for autism. Philosophical Transactions of the Royal Society of London, B, Biological Sciences, 361(1476), 2187–2198. doi:Q301VXX105162721 [pii] 10.1098/rstb.2006.1939Find this resource:

Hammock, E. A. D., & Levitt, P. (2006). The discipline of neurobehavioral development: The emerging interface of processes that build circuits and skills. Human Development, 49(5), 294–309. doi: 10.1159/000095581Find this resource:

Huber, D., Veinante, P., & Stoop, R. (2005). Vasopressin and oxytocin excite distinct neuronal populations in the central amygdala. Science, 308(5719), 245–248.Find this resource:

Kang, H. J., Kawasawa, Y. I., Cheng, F., Zhu, Y., Xu, X., Li, M., … Sestan, N. (2011). Spatio-temporal transcriptome of the human brain. Nature, 478(7370), 483–489. doi:nature10523 [pii] 10.1038/nature10523Find this resource:

Kogan, J. H., Frankland, P. W., & Silva, A. J. (2000). Long-term memory underlying hippocampus-dependent social recognition in mice. Hippocampus, 10(1), 47–56.Find this resource:

Kramer, K. M., Cushing, B. S., & Carter, C. S. (2003). Developmental effects of oxytocin on stress response: single versus repeated exposure. Physiology and Behavior, 79(4–5), 775–782.Find this resource:

Landgraf, R., & Neumann, I. D. (2004). Vasopressin and oxytocin release within the brain: A dynamic concept of multiple and variable modes of neuropeptide communication. Frontiers in Neuroendocrinology, 25(3–4), 150–176. doi:10.1016/j.yfrne.2004.05.001Find this resource:

Levine, S. (2001). Primary social relationships influence the development of the hypothalamic—pituitary—adrenal axis in the rat. Physiology and Behavior, 73(3), 255–260.Find this resource:

Lewkowicz, D. J., & Ghazanfar, A. A. (2012). The development of the uncanny valley in infants. Developmental Psychobiology, 54(2), 124–132. doi:10.1002/dev.20583Find this resource:

Lowbridge, J., Manning, M., Haldar, J., & Sawyer, W. H. (1977). Synthesis and some pharmacological properties of [4-threonine, 7-glycine]oxytocin, [1-(L-2-hydroxy-3-mercaptopropanoic acid), 4-threonine, 7-glycine]oxytocin (hydroxy[Thr4, Gly7]oxytocin), and [7-Glycine]oxytocin, peptides with high oxytocic-antidiuretic selectivity. Journal of Medicinal Chemistry, 20(1), 120–123.Find this resource:

Ma, Y., Shamay-Tsoory, S., Han, S., & Zink, C. F. (2016). Oxytocin and social adaptation: Insights from neuroimaging studies of healthy and clinical populations. Trends in Cognitive Sciences, 20(2), 133–145. doi:10.1016/j.tics.2015.10.009Find this resource:

Marlin, B. J., & Froemke, R. C. (2016). Oxytocin modulation of neural circuits for social behavior. Developmental Neurobiology. doi:10.1002/dneu.22452Find this resource:

Marlin, B. J., Mitre, M., D’Amour J. A., Chao, M. V., & Froemke, R. C. (2015). Oxytocin enables maternal behaviour by balancing cortical inhibition. Nature, 520(7548), 499–504. doi:10.1038/nature14402Find this resource:

Miller, T. V., & Caldwell, H. K. (2015). Oxytocin during development: Possible organizational effects on behavior. Frontiers in Endocrinology (Lausanne), 6, 76. doi:10.3389/fendo.2015.00076Find this resource:

Mori, M. (1970). The uncanny valley. Energy, 7, 33–35.Find this resource:

Mori, M. (2012). The uncanny valley. IEEE Robotics & Automation Magazine, 19, 98–100. doi:10.1109/Mra.2012.2192811Find this resource:

Moriceau, S., Wilson, D. A., Levine, S., & Sullivan, R. M. (2006). Dual circuitry for odor-shock conditioning during infancy: corticosterone switches between fear and attraction via amygdala. Journal of Neuroscience, 26(25), 6737–6748.Find this resource:

Neumann, I. D. (2007). Stimuli and consequences of dendritic release of oxytocin within the brain. Biochemistry Society Transactions, 35(5), 1252–1257. doi:BST0351252 [pii] 10.1042/BST0351252Find this resource:

Neumann, I. D. (2008). Brain oxytocin: A key regulator of emotional and social behaviours in both females and males. Journal of Neuroendocrinology, 20(6), 858–865. doi:JNE1726 [pii] 10.1111/j.1365-2826.2008.01726.xFind this resource:

Numan, M., & Young, L. J. (2015). Neural mechanisms of mother-infant bonding and pair bonding: Similarities, differences, and broader implications. Hormones and Behavior. doi:10.1016/j.yhbeh.2015.05.015Find this resource:

Oettl, L. L., Ravi, N., Schneider, M., Scheller, M. F., Schneider, P., Mitre, M., … Kelsch, W. (2016). Oxytocin enhances social recognition by modulating cortical control of early olfactory processing. Neuron, 90(3), 609–621. doi:10.1016/j.neuron.2016.03.033Find this resource:

Owen, S. F., Tuncdemir, S. N., Bader, P. L., Tirko, N. N., Fishell, G., & Tsien, R. W. (2013). Oxytocin enhances hippocampal spike transmission by modulating fast-spiking interneurons. Nature, 500(7463), 458–462. doi:10.1038/nature12330Find this resource:

Pedersen, C. A., Ascher, J. A., Monroe, Y. L., & Prange, A. J., Jr. (1982). Oxytocin induces maternal behavior in virgin female rats. Science, 216(4546), 648–650.Find this resource:

Pedersen, C. A., & Prange, A. J., Jr. (1979). Induction of maternal behavior in virgin rats after intracerebroventricular administration of oxytocin. Proceedings of the National Academy of Sciences U S A, 76(12), 6661–6665.Find this resource:

Sannino, S., Chini, B., & Grinevich, V. (2016). Lifespan oxytocin signaling: Maturation, flexibility and stability in newborn, adolescent, and aged brain. Developmental Neurobiology. doi:10.1002/dneu.22449Find this resource:

Shamay-Tsoory, S. G., & Abu-Akel, A. (2016). The social salience hypothesis of oxytocin. Biological Psychiatry, 79(3), 194–202. doi:10.1016/j.biopsych.2015.07.020Find this resource:

Smith, A. S., & Wang, Z. (2014). Hypothalamic oxytocin mediates social buffering of the stress response. Biological Psychiatry, 76(4), 281–288. doi:10.1016/j.biopsych.2013.09.017Find this resource:

Steckenfinger, S. A., & Ghazanfar, A. A. (2009). Monkey visual behavior falls into the uncanny valley. Proceedings of the National Academy of Sciences USA, 106(43), 18362–18366. doi:10.1073/pnas.0910063106Find this resource:

Sullivan, R. M. (2003). Developing a sense of safety: The neurobiology of neonatal attachment. Annals of the New York Academy of Sciences, 1008, 122–131.Find this resource:

Takahashi, D. Y., Fenley, A. R., Teramoto, Y., Narayanan, D. Z., Borjon, J. I., Holmes, P., & Ghazanfar, A. A. (2015). Language development: The developmental dynamics of marmoset monkey vocal production. Science, 349(6249), 734–738. doi:10.1126/science.aab1058Find this resource:

Tribollet, E., Charpak, S., Schmidt, A., Dubois-Dauphin, M., & Dreifuss, J. J. (1989). Appearance and transient expression of oxytocin receptors in fetal, infant, and peripubertal rat brain studied by autoradiography and electrophysiology. Journal of Neuroscience, 9(5), 1764–1773.Find this resource:

Uvnas-Moberg, K. (1998). Antistress pattern induced by oxytocin. News in Physiological Sciences, 13, 22–25.Find this resource:

Vaidyanathan, R., & Hammock, E. A. (2016). Oxytocin receptor dynamics in the brain across development and species. Developmental Neurobiology. doi:10.1002/dneu.22403Find this resource:

Yizhar, O., Fenno, L. E., Prigge, M., Schneider, F., Davidson, T. J., O’Shea, D. J., … Deisseroth, K. (2011). Neocortical excitation/inhibition balance in information processing and social dysfunction. Nature, 477(7363), 171–178. doi:nature10360 [pii] 10.1038/nature10360Find this resource:

Zheng, J. J., Li, S. J., Zhang, X. D., Miao, W. Y., Zhang, D., Yao, H., & Yu, X. (2014). Oxytocin mediates early experience-dependent cross-modal plasticity in the sensory cortices. Nature Neuroscience, 17(3), 391–399. doi:10.1038/nn.3634Find this resource: