Opposite Effects of Methanandamide on Lipopolysaccharide-Induced Prostaglandin E2 and F2α Synthesis in Uterine Explants from Pregnant Mice

Prostaglandins (PG) are effective abortifacients and are important mediators of lipopolisaccharide (LPS)-induced embryonic resorption (ER). Besides, anandamide (AEA) has been described as one of the major endocannabinoids present in the uterus suggesting that it might play a role in reproduction. It has been reported that high levels of AEA are associated with pregnancy failure and that LPS increases AEA production. Also, it has been observed that AEA modulates PG production in different tissues. In this sense, we studied whether LPS-induced PG production is modulated by AEA and we also assessed the effect of this endocannabinoid on PG metabolism in an in vitro model. Uterine explants from BALB/c implantation sites were cultured in the presence of LPS plus cannabinoid receptor (CB) specific antagonists and PG production was assessed. Then, we studied the effect of exogenous AEA on different steps of PG metabolic pathway. We showed that AEA is involved in LPS-induced PG biosynthesis. Also, we observed that AEA exerts opposite effects on PGE2 and PGF2α biosynthesis, by inhibiting PGE2 production and increasing PGF2α levels. We suggest that AEA could be involved in the mechanisms implicated in LPS-induced ER. A better understanding of how AEA could be affecting ER could help developing specific interventions to prevent this pathology.


Introduction
Intrauterine infection plays a major role in the pathogenesis of early pregnancy loss. It has been reported that Gram-negative organisms induce preterm labor and embryonic loss by triggering the release of various proinflammatory molecules, such as cytokines, growth factors and prostaglandins [1,2,3]. Prostaglandins are important paracrine regulators of uterine function in normal and pathological pregnancies and are used clinically to induce abortion and stimulate parturition [4]. Our previous results [1] showed that in vivo administration of lipopolysaccharide (LPS), a component of Gram-negative bacteria, increased prostaglandin E 2 (PGE 2 ) and prostaglandin F 2a (PGF 2a ) production in the uterus of early pregnant mice. We have also observed that, in mice challenge with LPS, the in vivo administration of cyclooxygenase (COX) inhibitors prevented LPS-induced embryonic resorption (ER).
Anandamide (arachidonoylethanolamide, AEA) belongs to a group of endogenous lipids termed ''endocannabinoids'' [5] and is an agonist of type-1 (CB1) and type-2 (CB2) cannabinoid receptors. It has been described as one of the major endocannabinoids present in the uterus and this suggests that it might play a role in reproduction [6]. It has been reported that low levels of AEA are beneficial for implantation and trophoblast outgrowth while increased AEA concentrations are associated with retarded embryo development, fetal loss and pregnancy failure [7]. Previous results indicate that LPS increases AEA levels in human peripheral lymphocytes [8] and in murine macrophages [9]. Our previous results suggest that LPS could be increasing AEA levels in uterine explants by inhibiting its degradation and also by enhancing NAPE-PLD expression, one of its synthesizing enzymes [10]. In addition, recent research has revealed that AEA regulates prostaglandin production in human gestational tissues, cerebral microvascular endothelium and in rat pheochromocytoma PC12 cells [11,12,13]. Although several lines of evidence indicate that both, cannabinoids and AEA, stimulate arachidonic acid (AA) release in several tissues [14,15,16], the relationship between endocannabinoids and PG metabolism is not fully understood. In the present study we investigated whether LPS-induced PG production is modulated by AEA and we also determined the effect of this endocannabinoid on PG biosynthesis and catabolism in uterine explants from pregnant mice.

Materials and Methods
Animals BALB/c 8-to 12-week-old virgin female mice were paired with 8-to 12-week-old BALB/c males and the day of appearance of a coital plug was taken as day 0 of pregnancy. Animals were housed in cages under controlled conditions of light (14 h light, 10 h dark) and temperature (23-25uC) and received murine chow and water ad libitum. Mice were sacrificed by cervical dislocation on day 7 of pregnancy and in each implantation site the uterus and the decidua were separated. Uterine explants (composed of myometrial cells and uterine adventitia) were weighed and culture as described below.

Ethics Statement
The experimental procedures reported here were approved by the Animal Care Committee of the Center for Pharmacological and Botanical Studies of the National Research Council (CEFYBO -CONICET) and were carried out in accordance with the Guide for Care and Use of Laboratory Animals (NIH).
Quantitative Polymerase Chain Reaction (qPCR) Analysis cDNA was synthesized from total RNA as described by Vercelli et al. [10]. Real-time quantitative PCR was performed with a Corbett-Rotor Gene system (Qiagen, Argentina) using EVA green (Biotium Inc., CA, USA) as the detection agent. Oligonucleotide primers for m-PGES-1 and b-actin were synthesized as described by Kubota et al. [18] and by Wang et al. [19], respectively. Specificity of the PCR reaction was controlled by the generation of melting curves. The relative gene expression levels were calculated using the comparative Ct (DDCt) method [20]. Data was normalized to b-actin and mPGES-1 mRNA levels under control conditions (no m-AEA) were set to 1 (dotted line). Experiments were independently run three times. In each experiment, cDNA samples were performed in triplicate.

Western Blot Analysis
Tissues were homogenized and centrifuged as described by Vercelli et al. [21]. Samples were separated by electrophoresis and transferred to a nitrocellulose membrane. Blots were incubated overnight with the primary antibodies and 30 minutes with anti-b-actin, washed with buffer (10 mM Tris, 100 mM NaCl and 0.1% Tween 20, pH 7.5) followed by 1 h incubation with horse radish peroxidase-conjugated anti-rabbit secondary antibody, and developed using the enhanced chemiluminescence western blot system. Photographs of the membranes were taken using a digital camera and analysed using the Image J software package.

PGE 2 and PGF 2a Determination
The amount of PGE 2 and PGF 2a were assayed in culture supernatants by specific radioimmunoassay (RIA) as reported by Farina et al. [22]. Briefly, specific antisera for both PGF 2a

Determination of PGE Synthase (PGES) Activity
PGES activity in uterine explants was measured by assessment of conversion of PGH 2 to PGE 2 as reported by Murakami et al. [23] with minor modifications. Briefly, tissues were homogenized in a 10 mM Tris-HCl buffer (pH = 8) and after centrifugation the supernatants were used as the enzyme source. One hundred micrograms of protein were incubated with 2 mg of PGH 2 for 1 min at 24uC in 0.1 ml of 1 M Tris-HCl, pH = 8, containing 2 mM GSH. After terminating the reaction by the addition of 20 ul of HCl 6N, PGE 2 content in the supernatants was quantified by radioimmunoassay. Results were expressed as pg PGE2/mg protein/h.

COX-2 Enzyme Activity Determination
COX-2 activity was determined by measuring the disappearance of the radiolabelled substrate [ 14 C]-arachidonic acid using a modified method previously reported [24]. Briefly, 100 mg of uterine explants were homogenized in 1 ml cold fresh buffer (50 mM Tris-HCl and 1 mM EDTA, pH = 8), centrifuged at 20000 xg for 60 min at 4uC and supernatant was used for subsequent determination of enzyme activity. One hundred micrograms of protein were added to each incubation tube in  [25]. The mixture was then incubated at 37uC for 30 min. Termination was achieved by addition of chloroform:methanol (1:1, v/v). The unreacted [ 14 C]-AA was resolved in the organic layer of a solvent system of ethyl acetate:hexane:acetic acid:distilled water (100:50:20:100 v/v) mixture. The distribution of radioactivity on the plate was counted in a scintillation counter by scraping off the corresponding spots detected in the plate. The area of each radioactive peak corresponding to AA was calculated and expressed as a percentage of the total radioactivity of the plates. For each sample, COX-1 activity was determined by calculating the rate of loss of [ 14 C]-AA incubated with selective COX-2 inhibitor. Conversely, the COX-2 activity of each corresponding sample was determined by calculating the rate of loss of [ 14 C]-AA incubated without selective COX-2 inhibitor, and subtracting from this value that of COX-1. The values of COX-2 and COX-1 were corrected by subtracting the [ 14 C]-AA disappearance values due to other enzymatic activities and non-enzymatic reactions. Enzyme activity is reported as nmol of disappeared [ 14 C]-AA/ mg protein/h. The optimal reaction conditions were previously determined (data not shown).

Data Analysis and Statistical Procedures
Statistical analysis was performed using the Graph Pad Prism Software (San Diego, CA, USA). Comparisons between values of different groups were performed using one-way ANOVA. Significance was determined using Tukey's multiple comparison tests for unequal replicates or Student t Test. All values presented in this study represent means 6 SEM. Differences between means were considered significant when p was 0.05 or less.

LPS-induced PGE 2 and PGF 2a Production is Mediated by Endocannabinoids
As mentioned before, previous results showed that in vivo administration of LPS increased PGE 2 and PGF 2a production in the uterus of early pregnant mice [1]. Figure 1 (A and B) shows that uterine explants incubated for 24 h in the presence of LPS (1 mg/ml) induced PGE 2 and PGF 2a production in vitro. To evaluate whether endogenous cannabinoids could modify LPSinduced PG production, uterine explants were incubated for 24 h in the presence of LPS (1 mg/ml), LPS plus AM251 (CB1 receptor antagonist) or LPS plus SR144528 (CB2 receptor antagonist) and PGE 2 and PGF 2a levels were assessed by RIA. LPS-induced PGE 2 synthesis was observed to be significantly increased when uterine explants were incubated with LPS and AM251 (10 nM) ( Figure 1A). Nevertheless, incubation of the endotoxin with the CB2 receptor antagonist (10 nM) had no effect on PGE 2 increased levels due to LPS in the same tissue ( Figure 1A). Higher concentrations of the CB2 receptor antagonist were also used with the same results (data not shown). On the other hand, LPSinduced PGF 2a synthesis was partially but significantly decreased when uterine explants were incubated both with LPS and AM251 (10 nM) or with LPS and SR144528 (100 nM) ( Figure 1B). PG levels remained unchanged when tissues were incubated either with AM251 or SR144528 alone. These findings suggest that endocannabinoids could exert opposite effects on PGE 2 and PGF 2a biosynthesis, by inhibiting PGE 2 production and increasing PGF 2a levels.
Among endocannabinoids AEA has been reported to regulate PG production in several tissues [12,13]. Mitchell et al. have shown that AEA significantly reduced PGE 2 production in human decidual tissue [11]; on the other hand, Someya et al. showed that AEA stimulates PGF 2a formation in PC12 cells [13]. Regarding LPS-induced PG production is mediated by endocannabinoids and considering that AEA could be involved in this process, we next analyzed if exogenous AEA could modify PG biosynthesis. First, uterine explants were incubated with methanandamide (m-AEA, a stable synthetic AEA analog) and PGE 2 and PGF 2a levels were determined by RIA. We observed that explants incubated with m-AEA 100 nM showed a significantly increase of PGF 2a levels ( Figure 2B), while PGE 2 levels were significantly reduced (Figure 2A). We also used an inhibitor of AEA degradation (URB597, 1 uM) with the same results (Figure 2A and B). To test whether AEA mediates LPS-induced PG production, uterus was incubated in the presence of LPS (1 mg/ml) plus m-AEA (100 nM) or LPS (1 mg/ml) plus URB597 (1 uM) for 24 h and PG levels were assessed by RIA. Figure 2A shows that PGE 2 production induced by LPS was partially but significantly decreased when tissues were incubated either with LPS plus m-AEA or with LPS plus URB597. On the other hand, PGF 2a production was significantly increased when tissues were incubated with LPS plus m-AEA ( figure 2B).

Effect of LPS on CB1 and CB2 mRNA and Protein Levels
The data showed in figure 1 suggest a possible involvement of CB1 and CB2 on PG production upon stimulation with LPS. In order to determine the effect of LPS on CB1 and CB2 expression in uterine explants, tissues were incubated for 24 h in the presence of LPS (1 mg/ml) and mRNA and protein levels were assessed. We observed the presence of both CB receptors, CB1 and CB2, in pregnant murine uterus. Furthermore, our results show that the endotoxin caused an increase in CB1 expression ( Figure 3A and B) while CB2 levels remained unchanged ( Figure 3C and D).

LPS-induced PG Synthesis and Cyclooxygenase (COX) Expression is Mediated by Endocannabinoids
Since PG synthesis first involves the release of arachidonic acid which is converted to prostaglandin H 2 (PGH 2 ) by cyclooxygenase 1 (COX-1) or cyclooxygenase 2 (COX-2) enzymes, uterine explants were incubated for 24 h in the presence of LPS (1 ug/ ml) alone, LPS plus AM251 (10 nM) or LPS plus SR144528 (100 nM) and COX-1 and COX-2 mRNA and protein levels were assessed. RT-PCR and western-blot analysis indicated that LPS significantly increased the expression of COX-2, whereas COX-1 mRNA and protein levels remained unchanged ( Figures 4A and  B). On the other hand, LPS-induced COX-2 expression (both mRNA and protein) was observed to be partially but significantly decreased when uterine explants were incubated either with LPS and AM251 or with LPS and SR144528 ( Figure 4A and B). COX  levels remained unchanged when tissues were incubated with AM251 or SR144528 alone (data not shown).
Regarding these results, we next evaluated COX-2 activity by measuring the disappearance of the radiolabelled substrate [ 14 C]arachidonic acid. Figure 5 shows a 3-fold increase in COX-2 activity when tissues were incubated with LPS alone. Moreover, LPS-induced COX-2 activity was partially but significantly decreased when uterine explants were incubated either with LPS and AM251 or with LPS and SR144528. These findings suggest that endocannabinoids mediate LPS-induced COX-2 activity.
As stated before, several lines of evidence indicate that both, cannabinoids and AEA, stimulate arachidonic acid (AA) release in several tissues [14,15,16]. Thus, we next analyzed if exogenous AEA could modify COX-1 and COX-2 mRNA and protein levels. RT-PCR and western-blot analysis indicated that m-AEA significantly increased the expression of COX-2, whereas COX-1 mRNA and protein levels remained unchanged (Figures 6A and  B). We next evaluated COX-2 activity. Figure 7 shows a 2-fold increase in COX-2 activity when tissues were incubated either with m-AEA (100 nM) or URB597 (1 uM) alone. These results suggest that AEA could be mediating, at least partially, LPSinduced COX-2 activity.

AEA: a Possible Role in PGE 2 Metabolism
Our data show that AEA-induced PGF 2a production correlates with the increased expression and activity of COX-2. Nevertheless, decreased production of PGE 2 induced by AEA is not explained by an augmentation of COX-2 expression and activity. In this sense, PGE 2 levels could fluctuate due to inhibition of its synthesis, an increased catabolism or both. PGE 2 synthesis involves the activity of PGE synthase (PGES) which catalyzes conversion of COX-derived PGH 2 to PGE 2 . So far, three different isoforms have been identified with the capacity to synthesize PGE 2 [26]. These include membrane-associated (or microsomal) PGE 2 synthase-1 and -2 (mPGES-1 and mPGES-2, respectively), and  cytosolic PGES (cPGES). Regarding this evidence, we first investigate whether endogenous cannabinoids could modify the expression of these enzymes and thus, we evaluated the effect of LPS plus CB receptor antagonists on uterine mRNA levels of PGES isoforms. Our results show that only mPGES-1 mRNA levels were significantly increased when uterine explants were incubated in the presence of LPS ( Figure 8B). Moreover, LPSinduced mPGES-1 expression was observed to be significantly increased when uterine explants were incubated with LPS and AM251 (10 nM) or with LPS and the CB2 receptor antagonist (100 nM) ( Figure 8B). We next evaluated if endocannabinoids could modulate mPGES-1 protein levels. Western blot analysis showed that LPS-induced mPGES-1 expression was significantly increased when tissues were incubated with LPS plus CB receptor antagonists ( Figure 9).
As we mentioned before, AEA reduces PGE 2 production in human decidual tissue [11]. Considering that LPS-induced mPGES-1 expression is mediated by endocannabinoids and regarding AEA could be involved in this process, we next analyzed if exogenous AEA could modify the expression of mPGES-1. Our results show that both mPGES-1 mRNA and protein levels were significantly reduced when uterine explants were incubated in the presence of m-AEA (100 nM) ( Figure 10A and B, respectively).
Regarding the very small decline in mPGES-1 mRNA (although significant) and the substantial decrease in mPGES-1 protein we decided to make a more quantitative approach by performing a real time PCR at early times. We assessed mPGES-1 mRNA levels when uterine explants were incubated either for 3 h or 6 h in the presence of m-AEA (100 nM). We observed that mPGES-1 mRNA levels remained unchanged after 3 h incubation in the presence of m-AEA. Nevertheless, uterine explants incubated for 6 h in the presence of the synthetic analog reduced mPGES-1 mRNA levels up to nearly 25% ( Figure 11). We next investigate whether exogenous AEA could modify the production of PGE 2 by modulating PGES activity. Thus, uterine explants were incubated in the presence of m-AEA (100 nM) alone or m-AEA plus indomethacin (INDO 1 mM [27], a non-selective COX inhibitor) and conversion of PGH 2 to PGE 2 was assessed by RIA. Incubation with INDO prevents COX-derived PGE 2 de novo synthesis from AA release. We observed that m-AEA significantly decreased PGES activity ( Figure 12). These findings suggest that AEA could be downregulating uterine PGE 2 levels by inhibiting its synthesis.
Another enzyme which may influence the level and biological activity of PGs is 15-hydroxyprostaglandin dehydrogenase . It is responsible for converting PGs into biologically inactive 15-keto derivative [28]. We first studied the effect of AEA on 15-PGDH mRNA and protein levels in uterine explants incubated with m-AEA 100 nM for 24 h. RT-PCR and westernblot analysis indicated that m-AEA did not modify 15-PGDH expression (data not shown). Next, we studied whether AEA could modify the production of PGE 2 metabolite (PGEM, biologically inactive). In this sense, PGEM quantification was assessed in culture supernatants by Cayman's PGE metabolite EIA kit as described in the Materials and methods section. At the same time, PGE 2 levels were assessed in the same sample by RIA. Our results  show that m-AEA (100 nM) significantly decreased PGE 2 levels corroborating the results showed before, and at the same time, significantly increased PGEM production in the same samples ( Figure 13).
In several tissues, including deciduas and fetal membranes, PGE 2 -9-ketoreductase (9-KPR) converts PGE 2 into PGF 2a thus regulating PGE 2 /PGF 2a ratio in these tissues [29,30]. Therefore, we studied the effect of AEA on 9-KPR activity in uterine explants incubated with m-AEA 100 nM for 24 h. We found that although we detected 9-KPR activity in the uterus of pregnant mice m-AEA did not modify it in this tissue (data not shown).
These findings suggest a possible role of AEA on PGE 2 metabolism and, in this sense, AEA could be downregulating uterine PGE 2 levels not only by inhibiting its synthesis but also by increasing its degradation.

Discussion
Embryonic resorption induced by sepsis is viewed as a serious obstetric problem but it may also be seen from an evolutionary perspective as a pathophysiological mechanism to protect the maternal reproductive potential by expelling infected material. Prostaglandins are potent stimulators of myometrial contractility [31] and are used clinically to induce abortion and stimulate parturition [4]. Particularly, both PGE 2 and PGF 2a stimulate myometrial contractility and are produced during labor by gestational tissues [32,33]. Our previous results showed that in vivo administration of LPS increased PGE 2 and PGF 2a production in the uterus of early pregnant mice. Also, we have observed that, in mice challenge with LPS, the in vivo administration of COX inhibitors prevented LPS-induced ER [1]. Collectively, these data suggest that although PGs are important paracrine regulators of uterine function in normal pregnancies, in an inflammatory setting such as sepsis, PGs have deleterious effects.
On the other hand, several lines of evidence show that LPS increases AEA levels [8,9] and also that AEA modulates prostaglandin production in different tissues [11,12,13].
Based on the above evidence, we hypothesized that LPSinduced PG synthesis is mediated by endocannabinoids. Our results showed that endocannabinoids are involved in LPSinduced PG biosynthesis and also had different effects on PG biosynthetic pathways. While LPS-induced PGF 2a production was partially abrogated through both CB1 and CB2 receptors, LPSinduced PGE 2 production was significantly increased when uterine explants were incubated with LPS plus a selective CB1 inhibitor (AM251) compared to explants incubated with LPS alone. On the other hand, CB2 receptor appeared not to be involved in LPSinduced PGE 2 production. Considering that other groups' findings show that, among endocannabinoids, AEA modulates PG biosynthetic pathways differently [11,12,13] we next evaluated whether AEA mediates LPS-induced PG production. Thus, we assessed uterine PGE 2 and PGF 2a production in the presence of LPS plus a stable synthetic analog of AEA called m-AEA, or LPS plus an inhibitor of AEA degradation, URB597. Our findings showed that PGE 2 production induced by LPS was partially but significantly decreased when tissues were incubated either with LPS plus m-AEA or with LPS plus URB597. On the other hand, PGF 2a production was significantly increased when tissues were incubated with LPS plus m-AEA. Together, these results suggest that AEA may be involved in LPS-induced PG production and that it modulates PG biosynthetic pathways differently for each PG. This is in accordance with the findings from other groups where it has been reported that AEA stimulates PGF 2a formation in PC12 cells [13] and also exerts a significant reduction in PGE 2 production in human decidual tissue [11]. We have also shown that endocannabinoids mediate LPS-induced COX-2 activity, mRNA and protein expression and AEA could be involved in this process. Although we showed that mAEA had opposite effects on PGE 2 and PGF 2a production, our previous results demonstrate that LPS induced the increase of PGs both in vivo [1] and in vitro. Also, LPS-induced COX-2 activity and expression was partially but significantly decreased when tissues were incubated with LPS and CB receptor antagonists. Taking this into consideration, it seems that AEA partially mediates LPS-induced PG production and therefore, we cannot rule out that other mechanisms (and/or other endocannabinoids) may be involved in this process.
On the other hand, we detected CB1 expression in the uterus of pregnant mice and also we reported for the first time the presence of CB2 in this tissue, contrary to other groups' findings that suggest the absence of CB2 both in the oviduct and in the uterus of mice [34,35]. Moreover, we have shown that LPS modulated CB1 mRNA and protein expression although CB2 levels remained unchanged. This is in accordance with the work of Matias et al. [36] and Do et al. [37] where they reported that LPS modulates CB1 and CB2 expression in dendritic cells.
Taking into consideration that AEA-induced PGF 2a production correlates with an increased expression and activity of COX-2 but decreased production of PGE 2 induced by AEA is not in accordance with the regulation of COX-2 levels, we investigated whether endogenous cannabinoids could modify the expression of the three different isoforms with capacity to synthesize PGE 2 identified so far, m-PGES-1, mPGES-2 and cPGES. Our results showed that only mPGES-1 expression was induced in the presence of LPS. This is in accordance with previous reports where a preferential coupling of COX-2 with mPGES-1 is shown in infectious response [38] and their expression is induced by proinflammatory mediators [23]. Moreover, LPS-induced mPGES-1 expression was observed to be significantly increased when uterine explants were incubated with LPS and CB receptor antagonists suggesting the participation of endocannabinoids in this process.
We also investigated the effect of exogenous AEA on different steps of PGE 2 metabolic pathway. Our results showed that AEA downregulated PGE 2 levels most likely due to inhibition of its synthesis and upregulation of its degradation. We have observed that AEA significantly reduced mPGES-1 mRNA and protein levels, which probably contribute to the decrease of PGE 2 synthesis. These results are in accordance with Navarrete et al. [39] who reported that N-arachidonoyldopamine (NADA, another member of the endocannabinoid family) is a potent inhibitor of PGE 2 production through a mechanism that involves reduction in the synthesis of mPGES-1 protein in LPS-activated microglia. Taking this evidence into consideration we cannot rule out that other endogenous CB1 and CB2 agonists (such as NADA and 2-AG, 2-arachidonoyl-glycerol) could also be involved in the action of LPS. The fact that AEA reduces PGE 2 levels could be considered as an anti-inflammatory effect of this endocannabinoid.
On the other hand, although several groups have reported that uterine 15-PGDH activity is regulated by progesterone during the entire pregnancy [40,41] this is the first time that an endocannabinoid modulation of PGE 2 degradation is demonstrated.
The above reported data and our previous results [1] show that LPS induces PGE 2 and PGF 2a production. Nevertheless, it is possible that PGE 2 /PGF 2a ratio more than the concentration reached by each of these PGs, may play a critical role in many biological functions, such as contractility. In fact, the altered ratio of PGE 2 /PGF 2a was described as one of the factors involved in the retention of fetal membranes in cattle [42,43,44]. In our case, where we studied PGE 2 and PGF 2a participation in LPS-induced embryonic resorption [1] there always seemed to be a predominance of PGF 2a over PGE 2 . The difference in the amount of PGF 2a and PGE 2 that reaches the uterus may play a role in the control of myometrial contractility. PGF 2a mediates myometrial contractions associated with the onset of labor [32]. Although it is less clearly established, PGE 2 is thought to exert the opposite action and favor the establishment of pregnancy. Moreover, Slater et al. [45] have reported that PGE 2 has anti-inflammatory and relaxatory effects on human myometrial smooth muscle.
Other groups have shown that LPS increases AEA production [8,9] and in spite of the fact that we did not measure direct changes in endocannabinoid concentrations in culture media after LPS challenge, the use of CB specific antagonists and the use of m-AEA allow us to hypothesized that AEA could be involved in the mechanisms implicated in LPS-induced embryonic resorption. We propose that AEA participates in LPS-induced COX-2 expression and activity, increasing PGF 2a levels and downregulating PGE 2 production modulating PG's ratio in this tissue. AEA inhibitory effect on PGE 2 production may be compensated by LPS stimulus (which may involve other endogenous CB receptor agonists and/ or other mechanisms) and the final outcome is likely to be a PGE 2 / PGF 2a ratio that would favor myometrial contractions which in turn contribute to fetal expulsion in an inflammatory setting.
In conclusion, this study suggests that AEA acts as a procontractility agent modulating the production of PGs induced by LPS. This and our previous results showing that AEA participates in the increase of nitric oxide synthesis induced by LPS [10] demonstrate that this endocannabinoid could contribute to the mechanisms associated with pathological reproductive events such as septic abortion.