Our data evidence that prophylactic sub-chronic PEA administration distinctly effects peripheral immune system and hippocampus, at both transcriptional and lipid molecular level-the exception is COX-2 pathway, which is downregulated in both tissues. Moreover, the downregulation of the COX-2 mRNA is time-point specific (3.5 h post PEA-injection) for both regions. This finding supports the hypothesis of PEA’s action across brain-periphery axis to decrease pro-inflammatory environment via transcriptional modulation of COX-2 pathway. Hence, splenic and hippocampal transcriptional downregulation of COX-2 pathway upon prophylactic PEA administration could balance the hippocampal over-expression of COX-2 upon brain insult (Lerner et al., 2018) and thus prevent development of neuroinflammation (Post et al., 2018). Similarly, PEA’s downregulation of COX-2 pathway in brain and periphery could contribute to building up resistance to bacterial infections and sepsis development (Heide et al., 2018). The transcriptional regulation of splenic and hippocampal cPLA₂ is divergent (increase in spleen and decrease in HC) and, interestingly, with different consequences on the lipidome in the two regions: in spleen with the exception of PC 38:6, all other phospholipids were downregulated, while, in HC no change of the membrane PLs was detected, except the LPA 20:4 (Figure 9). Even though an upregulation of cPLA₂ and breakdown of membrane PLs is a hallmark of many diseases, in such cases it is accompanied by the increased production of pro-inflammatory lipids, lipid peroxidation and free fatty acids (Farooqui et al., 2004). This is not the case here in spleen (Figure 2, 5, 9), where it is evident that the transcriptional upregulation of cPLA₂ is accompanied by a downregulation of COX-2 and 15-LOX with a concurrent upregulation of the 5-LOX mRNA (Figure 7, 9). The increase of resolvin D1 (Figure 6), which is reportedly an anti-inflammatory lipid, along with concurrent decrease of AA and downstream COX and LOX- lipid derivatives (11β-PGF₂α, 12(S)- and 15(S)-HETE, respectively) (Figure 2C, 5) evidence a shift towards provision of a pro-resolving lipid environment in the spleen via transcriptional activation of the cPLA₂, 5-LOX, with 5-LOX rendering conversion of omega-3 fatty acids in pro-resolving lipid environment (Figure 9). In line with this mechanism, the omega-6 fatty acids are not used to increase the AA pool (Figure 9) for inflammatory signals, but more likely to remodel/resynthesize splenic membrane PLs, indicated by an increase of PC 38:6 (3.5 h post-injection). The plasma omega 3- and omega-6 fatty acids, as well as PC 38:6 have the same alteration pattern as spleen, which could explain the PEA’s role in positively influencing lipid metabolism in obesity and associated inflammatory consequences (Hoareau et al., 2009).

The downregulation of the splenic pro-inflammatory signaling in favor of pro-resolution signaling is also supported by the transcriptional downregulation of the autotaxin, a major modulator of acute and chronic inflammation (Knowlden and Georas, 2014; Valdés-Rives and González-Arenas, 2017), and by the subsequent decrease of its extracellularly produced LPA lipids (LPA 20:4 and C16:0, respectively) (Figure 2C, 7B). LPA is one of the main signaling lipids in spleen that is multi-functional in modulating immune response (Saba, 2004). Recently, autotaxin-LPA pathway has been implicated as a main player in aberrant immune responses and development of inflammatory responses. The PEA’s effect on this pathway is therefore supporting its role as a positive immunomodulatory and warrants further investigation. PEA prophylactic administration modulates (increases) the level of S1P which is a vital lipid signal in the splenic immune function. The increase of S1P likely occurs through the breakdown of SM 36:1 (Figure 9) and conversion through SPH to S1P. S1P, via binding to its S1P1R receptors, is the bioactive lipid regulator of both innate and adaptive immune system and serves a plethora of biological functions in the splenic immune system. From our data it is impossible to reveal the role of PEA-derived S1P increase. However, considering the reported immunomodulatory properties of PEA it is pertinent to conclude that PEA-derived S1P increase must be part of a beneficial immune system boosting, which certainly needs further investigation. A time-specific (50 min) increase of IL1-β was also detected in spleen. Even though this molecule is mainly described as a pro-inflammatory cytokine that activates the cPLA₂ leading to pro-inflammatory cascades, it is obvious that this was not the case in spleen. In contrast, IL1-β and cPLA₂ activation is accompanied by a shift in signaling towards decreasing inflammatory cascade and boosting pro-resolving lipid environment (Figure 6, 9). This is also supported by the fact that IL1-β increase is terminated after 50 min. It is therefore, more likely that (short) activation of IL1-β boosts in this case immune defense and functions as an immunoadjuvant upon PEA prophylaxis (Dinarello, 1998). In contrast, in HC mRNA levels of both, IL1-β and cPLA₂ are downregulated along with those of MPAK11 and COX-2. These indicate a distinct regulation by PEA of neuroinflammatory pathway in HC than inflammatory pathway in spleen. The concerted decrease of these molecules in hippocampus point towards a MAPK11-induced downregulation of cytokine production and cytosolic cPLA2 with a downstream regulation of COX-2 as well (Wang et al., 2008). Nevertheless, this broad downregulation of the hippocampal pro-inflammatory tone explains the reported anti neuro-inflammatory properties of the PEA in various brain injuries, recently demonstrated for multiple sclerosis as well (Orefice et al., 2016). LPC 20:4 has reportedly been implicated in triggering inflammation via activation of COX-2 (Brkić et al., 2012). Because in hippocampus, the mRNA of the COX-2 is down-regulated, but LPC 20:4 up-regulated, modulation of hippocampal COX-2 is independent of LPC 20:4. It follows then that MAPK- induced cPLA₂/ IL1-β /COX-2 pathway is the more likely venue for PEA modulation of hippocampal inflammation. What is the source and function of hippocampal upregulation of LPC 20:4 upon PEA administration is not apparent from our data and remains to be determined. Similarly, increase and function of 5(S)-, 12(S)-, and 20- HETE in HC remains an open question to be investigated, especially, since none of the investigated LOX- mRNAs are altered, and provision of more data on CYP450 pathway is required. Indeed, in our study neither the translation of genes nor enzymatic activities were investigated which would answer and guide the interpretation of such lipid changes, e.g., hippocampal increase of LPC 20:4 and 5(S)-, 12(S)- and 20- HETE and/or plasma increase of PE 38:4, PC 38:6, LPC 20:4, SM 34:1, and SM 36:1.

Of general note is that the biochemical processes leading to the significant changes of the targeted lipidome upon PEA administration are time-wise diverse, as evident from the time-resolution lipid analysis. A broader lipidome change occurs at 3.5 h than at 50 min post-administration (Figure 2–4). We, therefore consider that generally, investigating a temporal dynamic of lipidome is advantageous to exhaustively reveal lipids involved in a particular biological context such as here described.

A particular feature of prophylactic PEA effect in the HC is the transcriptional downregulation of Bdnf gene and upregulation of Crh gene. Bdnf-mRNA downregulation by PEA could counterbalance aberrant increase elicited by brain insults such as the case in epilepsy (Lerner et al., 2018), hence downregulating the underlying increased neuronal activity rendering decreased seizure intensity (Post et al., 2018). More recently, a hippocampal neuronal subpopulation was shown to express Crh which modulates hippocampal excitability and maintain adaptive network excitability. Inhibition of Crh-expressing neurons increased locomotor activity while selective ablation of Crh-neurons in HC led to increased seizure susceptibility (Hooper et al., 2018). In view of this line of evidence, the transcriptional upregulation of hippocampal Crh gene upon PEA prophylactic administration is really interesting, and opens new venue of research to explain anticonvulsant effects of PEA administration. Collectively, hippocampal decreased neuroinflammatory pathways MAPK11/IL1-β/cPLA₂/COX-2 and decreased hippocampal excitability as indicated by decreased Bdnf mRNA along with increased Crh mRNA expression could contribute to the neuroprotective, anti-neuroinflammatory, and anticonvulsant properties described for PEA in epilepsy and other brain insults. A transcriptional activation of PPARα gene by PEA at the time points investigated here in control mice was not detected, which certainly does not exclude its activation.