We thus asked whether enhanced AEA levels in

We thus asked whether enhanced AEA levels in the KPT-335 reduce the level of age-related inflammatory changes and slow down brain ageing. To answer this question we compared the morphology of microglia and level of pro-inflammatory cytokines in the hippocampus of young and old wild-type and FAAH−/− mice.
Material and methods
Results
Discussion In this paper we demonstrate that young FAAH−/− mice have an increased hippocampal microglia density, an increased microglia cell size, and elevated IL-1β level when compared to age-matched wild type animals. This phenotype reminded us of the age-related changes observed in wild type mice. The data therefore indicate that the brain milieu may be shifted towards a more pro-inflammatory state in young animals lacking FAAH. Interestingly also, the age-related increase in GFAP expression was not affected by the FAAH deletion. This finding was highly unexpected, because previous publications showed an anti-inflammatory effect of the pharmacological blockade [12], [16] or genetic deletion [33] of FAAH. Moreover, genetic deletion of the CB1 receptor, the major target of AEA [3], [34], [35], led to early onset of age-related changes in the brain [36], [37] including microgliosis and astrocytosis [38]. Furthermore, it was reported that the cardiac ageing assessed as presence of cardiac dysfunction, inflammatory gene expression and apoptosis in the heart was slowed down in the FAAH−/− animals [39]. We therefore expected to see a rather reduced level of age-related neuroinflammation in animals lacking FAAH, which was clearly not the case. The results were also surprising, because pharmacological blockade of FAAH with URB 597 in old rats decreased the expression of microglia activation markers and pro-inflammatory cytokines in the hippocampus [40]. It is possible that compensatory changes due to the life-long deletion of FAAH contributed to the increased pro-inflammatory state in young mice. Indeed, higher elevation in AEA levels was found after acute treatment with FAAH inhibitor than after genetic deletion of FAAH [41]. On the other hand, the 15-fold elevation of AEA levels in the hippocampus of FAAH−/− mice did not lead to a change in CB1 receptor density or affinity [42]. Cyclooxygenase 2 is involved in the degradation of AEA in the brain [43] and this pathway may be prominent in FAAH−/− mice. The enhanced level of prostamides [43] in the mouse brain and possibly prostaglandins (measured in keratinocytes) [44] due to this alternative degradation pathway and the induction of arachidonic acid pathway due to the enhanced AEA levels [45] together can lead to an elevated chronic pro-inflammatory signalling in FAAH−/− animals. It is important to note that an anti-inflammatory effect of AEA on microglia was detected after strong activation with either LPS [13], [16] or with LPS/TNFα [10]. Also the neuroprotective effect on hippocampal slice culture was present after NMDA treatment [15]. In vivo, enhanced AEA levels proved to be protective in a viral model of multiple sclerosis [20] and in traumatic brain injury [12]. In normal ageing, however, a pro-inflammatory milieu develops slowly over an extended period of time and the neuroinflammation is weaker than under most pathological conditions. Thus, it is possible that AEA blocks the acute induction of inflammatory processes after a strong trigger, whereas it does not counteract slowly progressing pro-inflammatory changes. Nevertheless, the question why young healthy animals showed enhanced microglial activity still deserves further studies. An increase in the number and size of microglia was observed both in wild-type and FAAH−/− mice in ageing. Importantly, the number of microglia was higher in the FAAH−/− mice as compared with wild-type animals in the young age group. Furthermore, the size of microglia was larger in FAAH−/− mice than in wild-type in both age groups. Together these changes are indicative of a more pronounced pro-inflammatory activity state of microglia in young FAAH−/− mice. We wondered whether microglia with a large cell size form a distinct cell population or if there is rather a shift in the size distribution. Thus, we measured the area of individual Iba-1 positive cell bodies and plotted their distribution (Fig. S1). In each group we found a log-normal distribution with a single peak, thus strongly suggesting a continuum in the morphological states.