Our results demonstrate, for the first time, that OEA is distributed in the brain as an intact molecule ( Figure S1 of Supplementary Material ) within a few minutes after its acute i.p. injection. Indeed, the time-course measurements performed in the present study suggest a rapid OEA permeation of specific brain areas, partaking in the circuits controlling eating behavior and metabolism and partially known for being involved in the eating-inhibitory effect of systemically administered OEA. Such areas include the AP and the NST in the brainstem, the hypothalamic Arc, the ME and the HIPPO.

Interestingly, OEA concentrations reached the highest increase at the level of the AP, where they were up to 3-fold higher in OEA- than in VEH-treated rats (305.3 ± 52.87 vs 100 ± 20.30 respectively) as compared to the other brain areas analyzed (Arc/ME, NST, and HIPPO), where OEA levels were only about 1.5-fold higher than in controls (Arc/Me: 157.1 ± 6.6 vs 100 ± 9.2; NST: 166.6 ± 5.6 vs 100 ± 2.7; HIPPO: 166.5 ± 11.3 vs 100 ± 3.6). These results suggest that the AP might be an important part of the central circuits involved in mediating OEA’s pharmacological effects, thus corroborating our previous findings of the necessary role of this area in mediating OEA effects on eating (17). Moreover, the observation that OEA reached high levels also in the NST starting at the same time as in the AP (5 min) suggests that, once permeated into the AP, OEA rapidly diffuses into the anatomically adjacent NST. Moreover, in the present study we also considered the Arc/ME and HIPPO because of their high density of PPAR-α (48, 49), the main receptor responsible for the eating-inhibitory effect of OEA (35). Here, we demonstrate for the first time that intact OEA physically reaches these brain areas already at 2.5 min after its i.p. administration. In our previous studies in freely eating rats OEA did not induce c-fos mRNA at the level of the Arc, nor did it change the expression of proopiomelanocortin, which is highly expressed in the Arc and generates peptide signals that control PVN and supraoptic nucleus activities (24). In contrast, Umehara et al. (50) observed that peripheral OEA increased the expression of c-Fos in the Arc of fasted wild type mice, without producing any effect in the Arc of histidine decarboxylase knock-out mice, suggesting a necessary role of histaminergic neurons in mediating c-Fos induction in the Arc by OEA (50). Presumably, the different species (mouse vs rat) used, and, even more importantly the different homeostatic states of the animals (fasted vs fed) are responsible for the divergent observations.

The increased OEA concentration observed at the level of the HIPPO at the first time point considered (2.5 min) is in line with several studies suggesting that the HIPPO represents the primary brain region receiving ligands from both the blood and the cerebrospinal fluid (51), by virtue of its anatomical position, lying alongside the choroid plexus with its rich blood supply and immediately adjacent to the cerebral ventricles (43, 51). Importantly, PPAR-α are expressed in the HIPPO (52), thus suggesting that this region might represent another direct target site for circulating OEA. The HIPPO is primarily recognized as the brain area implicated in learning and memory processes (52), and a variety of studies demonstrate an effect of exogenous OEA on cognitive features. In fact, it has been demonstrated that OEA controls short-term spatial memory and hippocampal neurogenesis (53). Likewise, Yang et al. (33) demonstrated that systemic OEA improves spatial cognitive deficits in a rat model of acute cerebral ischemic injury through enhancing neurogenesis in the HIPPO (33). Finally, we recently demonstrated that OEA exerts antidepressant-like effects in mice and that this is associated with an increased expression of hippocampal phospho- cAMP response element-binding protein, a feature common to other PPAR-α agonists with antidepressant-like effects (54, 55). Interestingly, it has been recently demonstrated that eating behavior might be influenced by hippocampal-dependent memory functions, including meal-related memories and conditional learned associations between food and post-ingestive effects [for review see (56)]. We therefore hypothesize that by binding to PPAR-α in the HIPPO, OEA might modulate memory processes both related to food and cognition.

Although previous studies suggested that OEA is ineffective in controlling eating after its intracerebroventricular injection (4), the present results indicate that the eating-inhibitory effect of OEA is paralleled by an OEA distribution in the brain that is consistent with the assumption of its direct effects in one or more brain areas. These controversial findings may be explained by the high expression of fatty acid amide hydrolase in the brain ventricular epithelium (57), which might limit the bioavailability of OEA in brain tissue after intracerebroventricular administration and, hence, explain its ineffectiveness via this route of administration. Future studies should address this possibility.

Although brain OEA levels were already increased at 2.5 min after its i.p. administration, the maximum increase was observed at the 15-min time-point in most of the areas considered. This observation fits the behavioral data showing the greatest inhibitory effect of OEA on eating at 15-30 min after administration. The food intake reducing effect of i.p. injected OEA might therefore reflect its concentration in the brain. This interpretation is consistent with our previous observations demonstrating the necessary role of key brain sites in mediating OEA’s behavioral actions (13, 14, 16–18, 58) and indicates that such areas are recruited by OEA directly, rather than being activated as second- or third-order circuits downstream of peripheral nerves.

A previous study reported OEA brain levels in rats after its acute oral administration (59) showing no increase as compared to vehicle-treated animals. However, this discrepancy is not surprising for several reasons. First, compared to the present investigation, in that study orally administered OEA presumably underwent substantial first-pass hepatic extraction, which most likely limited the bioavailability of the molecule significantly. Second, the analyses were performed on the total brain, without any brain area separation, which most likely missed possible increases of OEA levels in specific brain areas. Third, rats were sacrificed at later time points, which is presumably unsuited to catch possible differences in OEA brain content following an acute systemic administration, due to the rapid permeation of the molecule into the brain and its subsequent hydrolysis. These latter two issues, namely the analyses in the whole brain and late time-points might explain also the negative results on brain distribution of OEA reported in a recent study (75) based on the intravenous administration of deuterated-OEA (¹³C-OEA) to rats at the dose of 1 mg/kg.

In parallel to investigating OEA brain distribution, in the present study we also measured OEA plasma levels at the same time points. Our results demonstrate that OEA plasma levels significantly increase as early as 2.5 min (30-fold higher than control, 3,333 ± 531 vs 100 ± 5) after its i.p. injection, with a maximum increase registered at 15-30 min (60-fold higher than controls respectively 6,218 ± 424 vs 100 ± 5 and 6,394 ± 344 vs 100 ± 8). The combination of the OEA distribution data in brain and plasma indicates that, once administered, OEA acts as a humoral mediator rapidly reaching brain structures including the Arc/ME, AP and HIPPO, which allow not only its access to the brain, but also likely represent brain target sites by virtue of the presence of PPAR-α, the major receptor for OEA. Beyond the HIPPO characteristics already described above, it is important to mention that both AP and Arc/ME are sensory circumventricular midline structures located around the fourth and third ventricles, respectively, lacking a blood brain barrier, with neurons of both areas being more exposed to blood-borne agents (60). This is in line with previous studies suggesting that the AP and Arc/ME represent easy access sites for circulating endogenous and exogenous eating-relevant hormones including amylin, glucagon-like peptide-1 and ghrelin (61, 62). Interestingly, very recent evidence from the research team led by Dr D’Agostino (63), demonstrates that systemic administration of OEA (10 mg/kg) to food-deprived female mice produces a rapid and stable inhibition of Agouti-related protein positive neurons of the Arc, which are strongly activated during a negative energy balance to stimulate eating. In their (63) OEA’s inhibitory effect on these neurons was maximal at 5 min after i.p. administration, thus mimicking the time course of OEA brain distribution observed in our experiment.

The results obtained in Experiment 1, i.e., that the OEA-induced activation of brainstem and hypothalamic nuclei did not depend on intact subdiaphragmatic vagal afferents, is consistent with our previous behavioral results (15), and further supports the view that OEA may act as a blood-borne messenger. Interestingly, also the effects of OEA on the brainstem noradrenergic and the PVN oxytocinergic systems were not affected by SDA, as OEA still maintained its capability to increase the expression of DBH in the brainstem nuclei and the activation of OXY neurons in the PVN in SDA rats. The only nucleus analyzed that seems to be under the control of afferent fibers is the vTMN because SDA prevented the activation of this nucleus by OEA, also indicating that OEA does affect vagal afferent signaling.

Interestingly, a communication between the TMN and the NST has been described; in particular, histaminergic neurons of the TMN extend their fibers to the NTS (64, 65). However, there is no direct evidence suggesting that such input affects eating. The slight attenuation of the OEA-induced c-Fos expression in the SolC and DMV of SDA rats might be, at least in part, the consequence of the lacking activation of vTMN in the same animals, although future studies are needed to address this assumption. However, the significant increase observed anyway at both sites might be associated with the activity of still intact efferent vagal fibers projecting from the brainstem to the gut, already known to be involved in modulating other satiety/appetite-signals, such as CCK (66). This result might partially explain the loss of OEA’s eating-inhibitory effect observed in rats (35) after total subdiaphragmatic vagotomy and the failure of the SDA procedure to abolish this behavioral effect of OEA.

Perhaps the inhibition of eating exerted by OEA results from a combination of both, humoral and vagal pathways, such that the short-term effect on eating is due to the direct activation of brain nuclei shown here, whereas the sustained effects on satiety (as suggested by the lack of compensatory overfeeding) may be due to the subsequent activation of vagal efferents, which might relay an OEA signal from the brain to the gut. Future studies might address this hypothesis.

Finally, as both the biosynthetic and catabolic pathways of OEA are shared by other NAEs (2), we also investigated whether the OEA treatment might affect the levels of other NAEs, and 2-AG in both plasma and brain. Interestingly, while AEA was the only endogenous compound affected by OEA in the brain (decreased level in the HIPPO), in the plasma all the NAEs analyzed were increased except for 2-AG that was unaffected. Because AEA is well known to stimulate appetite by activating CB1 receptors (67), its decrease in the HIPPO might contribute to the eating-inhibitory effect of OEA. However, future studies are necessary to examine this possibility. Concerning the increased plasma level of NAEs, we might refer to the theory of the ‘entourage’ effect (68, 69). According to this concept, peripherally administered OEA might cause an accumulation of endogenous NAEs by preventing their degradation, due to the saturation of the hydrolyzing pathways, which are shared among the different NAEs. Although OEA has been found to be the most active NAE in reducing energy intake, evidence suggests that also other NAEs, such as PEA, might participate in this process (67, 70, 71). Therefore, we cannot exclude that circulating NAEs might contribute to OEA’s pharmacological effects both related or not to eating and metabolism.

However, these aspects were beyond the aim of the present work, and future studies are needed to address these points.