Intrahypothalamic effects of oxytocin on PVN CRH neurons in response to acute stress

Abstract

Much of the centrally available oxytocin (OT) is synthesized in magnocellular neurons located in the paraventricular nucleus of the hypothalamus. This same area is home to parvocellular corticotropin-releasing hormone (CRH) synthesizing neurons that regulate activation of the hypothalamic-pituitary-adrenal (HPA) axis. A large body of data indicates that complex interactions between these systems inextricably link central OT signaling with the neuroendocrine response to stress. This review focuses on a small but diverse set of cellular and synaptic mechanisms that have been proposed to underlie intrahypothalamic OT/CRF interactions during the response to acute stress.

Introduction

In a biological context, stress is broadly defined as a real or perceived challenge to homeostasis of an organism. Acute stressors can be physical and/or psychological. In both cases, the response to acute stress is mediated in large part by coordinated neuroendocrine and autonomic responses, both of which depend heavily on the activity of distinct populations of neurons in the paraventricular nucleus of the hypothalamus (PVN). Neurons of the PVN are broadly classified based on location and morphology as either magnocellular or parvocellular, and each of these groups is further divided based on neurochemical phenotype, projection target, and function [1]. For example, parvocellular neurons located in the PVN that synthesize corticotropin releasing hormone (CRH) and project to the median eminence play a clear and well-established role in mediating the neuroendocrine response to stress. More specifically, stress-induced activation of these neurons promotes CRH-mediated release of adrenocorticotropic hormone (ACTH) from the anterior pituitary, which in turn acts in the inner adrenal cortex (the zona fasciculata) to promote synthesis and release of glucocorticoids into the systemic circulation. As a result, central mechanisms that regulate the activity of PVN CRH neurons directly regulate, moment to moment, the level of activation of the HPA axis, and thus the level of circulating glucocorticoids [2,3].

A lot of what is known about central regulation of the neuroendocrine response to stress is dependent on activation of central glucocorticoid receptors (GRs), which exist both within and outside of the hypothalamus, and which can produce both genomic and non-genomic effects that ultimately regulate the activity of PVN CRH neurons over a wide range of time scales [4]. Collectively, GR-dependent mechanisms are intimately involved in setting the basal level of HPA activation, and in modulating the response to both chronic and acute stressors [for additional review see 5–8]. That said, substantial and compelling evidence indicates that GR-independent crosstalk between anatomically and neurochemically distinct types of hypothalamic neurons is likely to also play a prominent role in modulating the neuroendocrine response to stress, potentially on a somewhat faster time scale. Oxytocin, a neuropeptide which is synthesized in a subset of both parvocellular and magnocellular PVN neurons [9], is a particularly interesting signaling molecule in this regard. This review will briefly cover evidence indicating that PVN OT neurons can exert powerful regulatory effects on stress-induced activation of the HPA axis and will then discuss efforts to date to reveal the detailed cellular and/or synaptic mechanisms involved.

Evidence that OT/CRF interactions in the hypothalamus modulate the neuroendocrine response to stress

Before discussing the role of PVN OT neurons in stress responding, it is important to note that both parvocellular and magnocellular PVN OT neurons are involved in homeostatic processes that are not directly related to the HPA axis. For example, many parvocellular PVN OT neurons project to the brainstem, spinal cord, and SON, where they likely contribute to the regulation of autonomic output, breathing, and nociception [9–14]. By contrast, most magnocellular PVN OT neurons project through the median eminence to the posterior pituitary, where they can release oxytocin into the neurohypophysis and thus directly modulate circulating levels of oxytocin in the periphery [15–17]. This type of OT release contributes to the induction of uterine contractions and milk letdown in females [18], and is involved in body fluid homeostasis [19]. The lack of strong axonal arborizations within the PVN, from either parvocellular or magnocellular OT neurons [9,20,21], might make them seem unlikely candidates for direct modulation of HPA activity. However, magnocellular OT neurons also have a clear ability to release oxytocin, in both a calcium- and activity-dependent manner, from dense core vesicles located in their dendrites [22–24]. It is this dendritic release of oxytocin into the PVN that seems most likely to modulate the activity of PVN CRH neurons during or immediately after an acute stressor.

The possibility that intrahypothalamic OT signaling might play a key, and relatively rapid, role in modulating the neuroendocrine response to stress is suggested by the observation that OT releasing dendrites of magnocellular OT neurons are found in close proximity to parvocellular PVN CRH neurons, a subset of which are OTR positive [25,26]. It is further suggested by the observation that homeostatic challenges that activate PVN OT neurons, such as acute hypernatremia [27,28], are associated with subsequent reductions in stress-induced activation of the HPA axis [25,29–31]. Similarly, other conditions associated with natural increases in central oxytocin concentration, such as lactation, are also associated with markers of decreased HPA function [32,33]. Further, it is notable that a wide variety of both physical and psychological stressors that cause an acute increase in HPA activity also promote central release of OT [34–40]. Functionally, this could occur if both PVN OT and PVN CRH neurons receive descending excitatory input from the same stress-activated limbic circuits, and yet it may also be an indication of intrahypothalamic excitatory signaling from PVN CRH to PVN OT neurons [e.g. see 26,41,42].

Additional, and arguably more direct, evidence for meaningful OT/CRF interactions during stress responding comes from studies that evaluate stress responsiveness after directly activating, blocking, or genetically removing central OT receptors. For example, intranasal or intracerebroventricular administration of oxytocin blunts the HPA axis response to acute stress and decreases genetic markers of activity in CRH neurons [43–46]. Conversely, intracerebral administration of an OTR antagonist increases both basal and stress-induced activity of the HPA axis [47–49]. Finally, data from OT knockout mice suggests that OT regulates stress-induced expression of CRH mRNA in the PVN [50].

Collectively, the data described above provide strong evidence that OT/CRF interactions are likely to play an important role in regulating the neuroendocrine response to stress, further highlight the PVN as one probable site for this interaction, and finally underscore that in most contexts the effect of central OT release on HPA activity is inhibitory [for additional review see 51–56]. Intriguingly, despite the strength and amount of data supporting these conclusions, it has been difficult to conclusively and effectively reveal the detailed cellular and/or synaptic mechanisms involved. Thus, this mini-review will focus largely on two of the most well studied mechanisms through which activation of PVN OT neurons may act locally in the hypothalamus to provide rapid feedback inhibition to PVN CRH neurons during an acute stressor.

Proposed mechanisms for intrahypothalamic OT/CRH interactions as they relate to stress responding

Local OT release creates a tonic OTR-mediated inhibitory tone by activating OTRs on PVN CRH neurons

The first hypothesis, stated simply, is that OT released from PVN OT neurons creates a local paracrine signal, which activates OTRs expressed by PVN CRH neurons, and produces an inhibitory current, which ultimately reduces the probability that the cells will fire action potentials in response to an excitatory input. Initial evidence for this hypothesis came from studies in rats demonstrating that acute hypernatremia both activates hypothalamic OT neurons, as measured by changes in circulating levels of oxytocin [28], and decreases activation of the HPA axis, as measured by stress-induced changes in plasma concentrations of ACTH and CORT [29]. Subsequent experiments, in both rats and mice, revealed that acute hypernatremia is also associated with increased activation of PVN OT neurons and decreased activation of PVN CRH neurons, as reported by immunohistochemistry for c-Fos [29,30]. Further, in vitro electrophysiological experiments in Type II/neurosecretory PVN parvocellular neurons in rats, identified by electrophysiological properties [57], and in genetically identified PVN CRH neurons in mice, revealed a clear OTR-mediated inhibitory tone that was blocked by bath application of a selective OTR antagonist and that was present only in animals that had been exposed to acute hypernatremia [25,31]. Notably, this tonic OTR-mediated inhibitory current was not associated with significant changes in spontaneous glutamatergic or GABAergic transmission, was still apparent in the presence of antagonists for glutamatergic and GABAergic receptors, was blocked by inhibiting GPCR dependent signaling only in the patched neuron, was absent in mice where OTRs were conditionally deleted from CRH neurons, and was associated with a clear reduction in neuronal gain [25]. Collectively, these data make a strong case in favor of the stated hypothesis. That said, it is important to note that the amount of OTR mediated inhibitory tone produced in PVN CRH neurons by acute hypernatremia was observed to vary substantially from cell to cell, and that both our group and others have been unable to produce a comparable inhibitory current in PVN CRH neurons using acute bath application of an OTR agonist in tissue slices extracted from naïve animals [25,58]. Notably, however, we were able to produce a similar OTR dependent inhibitory current in PVN CRH neurons by incubating naïve slices in 200 nM oxytocin for a minimum of 1 hour [25]. These additional observations highlight that it may be valuable for future research to attempt to better identify specific mechanisms that regulate expression of functional OTRs by PVN CRH neurons, and to carefully consider both concentration and time of exposure when working with exogenous agonists. Further, more broadly, it seems plausible that acute hypernatremia in combination with acute psychological stress may produce more robust activation of PVN OT neurons, and/or more effective local OT release, than either stimulus alone.

Local OT release promotes GABAergic inhibition of PVN CRH neurons.

The PVN itself contains few GABAergic neurons, and yet the ability of GABAergic signaling to robustly modulate the activity of the HPA axis is well documented [59–61]. Indeed, the PVN is surrounded by nuclei that contain a high density of GABAergic neurons, many of which project into parvocellular and/or magnocellular regions [62–64]. Several lines of evidence suggest local oxytocin signaling may be able to modulate GABAergic inputs from the peri-PVN region to the PVN in a way that down regulates activation of the HPA axis. For example, it has been reported that OT acting in the PVN blunts stress-induced activation of PVN CRH neurons, and associated anxiety-like behaviors, via a mechanism that involves increased GABA release in the PVN, and that is blocked by concurrent administration of a GABA_A receptor antagonist [65]. While this study did not directly implicate a specific population of OTR sensitive GABAergic neurons in a peri-PVN region, it seems likely that such neurons could be involved. For example, GABAergic neurons that synthesize melanin-concentrating hormone (MCH) are located in the lateral hypothalamus, are depolarized by OT, release GABA in the PVN, and have been specifically implicated in OT-mediated induction of maternal behaviors [66,67]. Further, OTR-mediated excitation of GABAergic neurons is likely to occur in other stress/anxiety related circuits [68–70]. These types of findings clearly suggest that paracrine and/or synaptic activation of OTRs on a subset of GABAergic neurons in the peri-PVN region is likely to increase GABA release in the PVN in a way that modulates the response to acute stressors. That said, there is still much to learn about the specific population(s) of OTR positive local GABAergic neurons that project to the PVN, about the extent to which they are activated by paracrine vs. synaptic release of OT, and about how the GABA they release ultimately impacts excitability of PVN CRH neurons. In the end, the answers, particularly to the last question, are likely to be complicated. Consider that recent work has elegantly demonstrated that both presynaptic and extrasynaptic GABAergic receptors in the PVN can contribute meaningfully to GABAergic regulation of PVN CRH neurons, and further indicated that tonic GABAergic currents observed in these neurons are enhanced by acute exposure to corticosteroids [71]. Given that neither activation of OT neurons produced by hypernatremia [25], nor acute bath application of an OTR agonist [58], causes a significant increase in sIPSC frequency observed in PVN neurons, these data highlight that OTR-mediated modulation of tonic GABAergic signaling in PVN CRH neurons may warrant further careful evaluation. While perhaps not a mechanism that works on the same time scale being considered here, it is also worth noting that OTRs in other systems have been implicated for their ability to modulate the functional impact of GABAergic signaling by modulating chloride transporters that set the reversal potential for current flow through ionotropic GABA receptors [72], and regulation of potassium / chloride co-transporters in PVN CRH neurons has been demonstrated to regulate stress-induced activation of the HPA axis [73,74].

Conclusions

In conclusion it is important to highlight that the two main mechanisms covered here are not mutually exclusive and thus could each contribute meaningfully to OT-mediated regulation of the HPA axis during the response to an acute stressor, although perhaps to different extents in different contexts. It’s also important to note that while the focus here has been on intrahypothalamic OT/CRH interactions that can occur rapidly during response to an acute stressor, any such mechanisms revealed will ultimately need to be viewed in the context of both genomic and non-genomic (but largely slower / longer term) GR-mediated feedback, and in the context of additional di- and poly-synaptic circuits where extrahypothalamic effects of both OT and CRH clearly also contribute to modulation of the stress response. Notably, the story will likely also become more complicated by consideration of off-target (OTR independent) effects of OT, and/or by off-target activation of OTRs (by other endogenous agonists). That said, continuing to develop a better understanding of the specific cellular and synaptic mechanisms through which centrally released OT acts within the hypothalamus to regulate the activity of PVN CRH neurons is an important goal, and one that holds real potential to help meaningfully advance therapeutic strategies for stress related disorders.