On the synaptic rearrangement in the hypothalamus and the periaqueductal gray in an animal model of obesity

Obesity is a worldwide problem, affecting peoples‘ health and burdening healthcare systems. It is a complex interaction of endocrine and neural mechanisms underlying food intake, in particular, one of its pathological forms leading to obesity. Interestingly, patients as well as animal models suffering from abnormal feeding behavior display altered nociception. A renown mouse model for obesity is the leptin-deficient ob/ob mouse, our laboratory has demonstrated that the mainly excitatory innervation of orexin-expressing neurons (OX-N) in the lateral hypothalamus (LH) of wt mice is rearranged in favor of the inhibitory inputs in LH of ob/ob mice. Furthermore, the vesicle release from the inhibitory inputs is suppressed by endocannabinoids (eCBs) activating cannabinoid receptor type 1 (CB1) activation. The eCBs, most likely 2-AG, originate from the postsynaptic terminals, where they are being synthesized and “released“ as a reaction to depolarization (a mechanism called: depolarization-induced suppression of inhibition or DSI). Here, I demonstrated that the functional excitatory innervation of OX-N did not differ between ob/ob and wt mice. Furthermore, the activation of presynaptic CB1 receptors suppressed the vesicle release from excitatory inputs, in both wt and ob/ob mice, to the same extent. The imbalance of functional excitatory and inhibitory inputs in ob/ob mice, putatively, leaves OX-N with a hyperpolarized membrane potential and a reduced firing activity. Activation of CB1 receptors, mainly located on inhibitory inputs, eventually activate OX-N by disinhibiting them. Orexinergic neurons possess vast projections throughout the brain, i.a. to the mesoaccumbal dopamine system and the hypothalamus-pituitary-adrenal (HPA) axis. The activation of these two circuits, putatively, results in the increased food intake seen in the ob/ob mouse. Another target area of OX-N is the periaqueductal gray (PAG), playing a key role in nociception via the descending antinociceptive pathways. It has been reported that patients as well as animal models suffering from abnormal feeding behavior also display altered nociception. Furthermore, the administration of orexin A (OX-A) has been demonstrated to suppress inhibitory postsynaptic currents in CB1 receptor activation-mediated way, eventually, resulting in a depolarization of the membrane potential of PAG neurons and, finally, in an increase of firing activity. Intriguingly, these in vitro observations translate to behavior, OX-A administration into PAG elevated the pain threshold in rats during the tail-flick test. ob/ob mice displayed an elevated level of OX-A in PAG, similar to the situation after OX-A administration, hence, we hypothesized that the PAG neurons projecting to the rostroventral medulla were more depolarized and have a higher firing activity in ob/ob mice compared to wt. Indeed, blocking the orexin 1 receptor hyperpolarized the membrane potential and reduced the firing activity of ob/ob PAG neurons, but not wt PAG neurons. Furthermore, ob/ob PAG neurons were more likely to initiate an action potential than wt PAG neurons. Thus, suggesting that the activation of PAG neurons by activated orexinergic inputs results in an elevated pain threshold by further activating the descending antinociceptive pathways. In conclusion, the switch of innervation onto OX-N in favor of inhibitory inputs, caused by the absence of leptin in the ob/ob mouse, activates OX-N by inhibiting their mainly inhibitory inputs by eCB-mediated CB1 receptor activation, resulting in the suppression of vesicle release (DSI). These disinhibited OX-N activate target areas throughout the brain, such as the mesoaccumbal dopamine system and HPA axis, thereby modulating feeding behavior. Furthermore, orexinergic projections to PAG depolarizes the membrane potential and increases the firing activity of PAG neurons projecting to the rostroventral medulla, thereby raising the pain threshold.