The role of the endocannabinoid system in female reproductive tissues

AEA acts as a partial agonist at CB1 and as a weak/partial agonist at CB2 [5]. Interestingly, AEA is reported to demonstrate promiscuous binding activity [11], as it can trigger various signaling pathways via a number of different receptors, both extracellularly at CB1and CB2, intracellularly at the transient receptor potential vanilloid-1 (TRPV1) channel, and in the nucleus via the peroxisome proliferator-activated receptors (PPARs) [7]. More recently, studies have suggested that AEA may also act at the level of the mitochondria via CB1 receptors situated on the mitochondrial outer membrane [14].

The biosynthesis of AEA occurs in two steps. First, AEA is released from phospholipid precursors in the plasma membrane to a phosphatidylethanolamine, leading to the formation of N-acylphosphatidylethanolamine (NAPE). In the second step, a type D phospholipase (NAPE-PLD) catalyzes the formation of AEA from its NAPE precursor [5]. AEA is then rapidly taken up by cells [8, 15]. Movement of AEA across the phospholipid bilayer is thought to occur by simple diffusion or endocytosis [8, 15] since AEA is uncharged and lipid soluble. Other studies strongly suggest the involvement of a putative endocannabinoid membrane transporter (EMT) [8, 16, 17] which allows AEA to be rapidly shuttled to its intracellular targets [8, 15]. Some of the intracellular targets for AEA identified to date include AEA intracellular binding proteins (AIBPs), namely albumin, heat shock protein 70 (Hsp70) and fatty acid binding protein-5 and -7 (FABP-5 and -7) [8]. It has been proposed that FABPs are principally involved in AEA trafficking and breakdown, as the use of a novel reversible FABP inhibitor, BMS309403, partially reduced AEA uptake [15]. Ultimately, AEA is broken down by intracellular hydrolases into ethanolamine and arachidonic acid [5, 7]. A key regulator of AEA activity is the serine hydrolase fatty acid amide hydrolase (FAAH) which is bound to intracellular membranes, particularly the endoplasmic reticulum (ER) and nuclear membrane [7].

Belonging to the monoacylglycerol (MAG) family of endocannabinoids, 2-AG acts equally at CB1 and CB2 as a full potent agonist, but it has not been shown to act at the TRPV1 receptor [5, 11]. Tissue levels of 2-AG are markedly higher than that of AEA within the same tissue [5]. Coupled with its full agonistic activity at cannabinoid receptors, it has been proposed as the primary endogenous agonist of both CB1 and CB2 [8].

The biosynthesis of 2-AG involves the combined action of two membrane-bound enzymes: phospholipase C (PLC) and diacylglycerol lipase (DAGL) [5]. Unlike AEA, only a few studies have investigated the mechanisms underlying the rapid cellular uptake of 2-AG, which is surprising given that it is more abundant than AEA [8]. While the mechanism(s) may be dependent upon the cell type [8, 16, 18], the most likely routes of entry for 2-AG may be via endocytosis, simple diffusion, the same EMT as AEA or other transporters [8, 16].

Once within the cytosol, 2-AG becomes a substrate for the chief catalytic enzyme monoacylglycerol lipase (MAGL), associated with the inner membrane, which degrades 2-AG to glycerol and arachidonic acid [8]. In addition to MAGL, two other 2-AG hydrolases have been identified: α,β-hydrolase-6 and -12 (ABHD-6 and -12), which are integral cell membrane proteins and thought to share the catalytic triad with MAGL [8]. In addition to the consensus regarding the putative degradative enzymes, these prototypical endocannabinoids may also be degraded via oxidation by cyclooxygenase (COX), lipoxygenase (LOX), or cytochrome P450 [5]. These and other biosynthetic and degradative pathways for AEA and 2-AG are discussed at length in Fezza et al [5].

THC was the first exogenous ligand of cannabinoid receptors to be discovered [13]. The physiological effects of THC are clinically concerning and need to be well delineated. Since this compound is highly lipophilic [5, 12, 19], it can be readily sequestered into adipose tissue, resulting in a rapid decrease in plasma concentrations [19], followed by a slow release into circulation over extended periods of time [12]. This tissue distribution permits longer-lasting stimulation of the cannabinoid receptors. This is unlike that of the locally released endocannabinoids AEA and 2-AG, which are rapidly inactivated by their transporters and hydrolases [12].

Endogenously, other fatty acid derived compounds, including N-arachidonoyldopamine (NADA), 2-arachidonoylglycerylether (noladin ether) and O-arachidonoyl ethanolamine (virodhamine) have also been identified as endocannabinoids, although information on their function and biological relevance is somewhat limited [4, 5, 10]. Interestingly, a novel group of ligands has been identified, referred to as retro-anandamides, which are characterized by a reversal in the positions of the carbonyl and the amido groups [20]. While these compounds demonstrate reduced affinity for CB1 and CB2 receptors as compared to AEA, they are resistant to FAAH catabolism resulting in increased stability relative to AEA [20]. These same authors, in an earlier report, identified the first metabolically stable AEA analogue, (R)-methanandamide, which exhibited significantly greater affinity for CB1 and resistance to FAAH degradation when compared to AEA [20]. Despite their presence, these and other ligands have received less scientific attention than AEA and 2-AG, perhaps due to the difficulty involved in isolating them from biological tissues [5].

CB1 and CB2 belong to a large superfamily of seven-transmembrane spanning GPCRs [4]. AEA or 2-AG ligand binding to either CB1 or CB2 leads to multiple signal transduction mechanisms, including the inhibition of adenylate cyclase [4], a common target for activated G proteins, and consequent decrease in intracellular cAMP levels, increased potassium influx, and/or inhibition of certain calcium channels, thus reducing calcium influx [21, 22]. The intracellular signals arising from these cascades subsequently leads to the regulation of growth, proliferation, and/or differentiation [6].

CB1 receptors are found mostly within the central nervous system [22]. Peripherally, CB1 receptors have been identified in the spleen, heart, adrenal gland, ovaries, endometrium, testes, among others [1, 22]. Furthermore CB1 receptors have also been localized intracellularly on the mitochondrial outer membrane [14]. Mitochondria regulate the energy demands of the cell, thus compromises in its function, from aberrant cannabinoid signaling [23, 24], will deregulate energy metabolism [25]. For example, THC induced mitochondrial dysfunction has been associated with pathologies such as stroke [26]. Mechanistically, mitochondrial CB1 receptors are thought to modulate complex I activity via a process involving soluble adenylyl cyclase [27]. Since mitochondrial function controls apoptosis, disruption of this role can impact the process of producing quality gametes, subsequently interfering with embryogenesis and lead to the production poor quality embryonic stem cells [28]. Furthermore, ovarian ageing is also associated with increased accumulation of mitochondrial DNA mutations which are likely to affect mitochondrial biogenesis and impact oocyte quality [29]. Additionally, placental oxidative stress is also linked to mitochondrial dysfunction [30] and may impact vascular remodeling in the placenta [31] which is important for tissue oxygenation and organ function. Dysregulation of placental vascular development can result in a number of adverse pregnancy outcomes such as intrauterine growth restriction and preeclampsia [32, 33]. CB1 receptors are also found in the hypothalamus, the central regulator of energy homeostasis, further implicating a role for the ECS in energy balance. Moreover, the preoptic area of the hypothalamus contain CB1 receptors, from which secretory neurons for gonadotropin-releasing hormone (GnRH) are located [34].

Like the CB1 receptor, CB2 is also a G protein coupled receptor demonstrating 44% sequence homology to CB1 [6, 22]. CB2 receptors are predominantly found peripherally within cells of the immune system, such as lymphocytes and macrophages [23]. El-Talatini et al confirmed the presence of CB2 receptors in the ovarian cortex, ovarian medulla, and ovarian follicles from human samples [1], which followed a similar staining pattern as CB1 in the same tissues [1]. However, Wang et al found that in murine oocytes, the action of endocannabinoids were mediated by CB1 receptor activation, not that of CB2. [35]. This interspecies difference highlights the importance of utilizing human reproductive tissues as a means to study the impact of the ECS on its function/dysregulation [1].

Endocannabinoids are thought to also target a number of other orphan receptors. These include the GPR55 (also known as CB3) and GPR119 receptors which have signaling mechanisms distinct from CB1/CB2 [3, 16, 36]. Other receptors targeted by endocannabinoids include TRPV1, cytosolic target for AEA; and nuclear PPAR. Pertwee et al provide a detailed review of these other receptors and their pharmacology [10].

Polycystic ovarian syndrome (PCOS) is a common endocrine pathology characterized by oligo-anovulation, hyperandrogenism, and the appearance of polycystic ovaries upon ultrasonography. This condition affects approximately 5–15% of women of reproductive age [17, 60]. Data from epidemiological studies demonstrate that PCOS often affects obese women, most of which present with hyperinsulinemia, which are risk factors associated with type II diabetes [17]. Obesity exaggerates insulin resistance leading to compensatory hyperinsulinemia and increased risk of type II diabetes [61]. These metabolic derangements have important implications in the pathogenesis of PCOS. Increased adiposity leads to an increase in serum leptin, which blocks LH secretion [61], leading to infertility and anovulation [17, 61]. Furthermore, hyperinsulinemia increases ovarian steroid hormone production by blocking LH leading to the cessation of follicular development [61].

Studies on the etiology and pathogenesis of PCOS typically have focussed on the association with metabolic syndrome [60], with few studies to date examining the impact of the ECS on PCOS. Indeed, the cardinal features of PCOS, such as insulin resistance and obesity, might be influenced by the ECS. It is well known that the ECS regulates energy balance by regulating appetite, food intake, and glucose metabolism [52]. Importantly, the presence of AEA has been shown to result in insulin hypersecretion and insulin resistance via activation of CB1 in pancreatic islet cells [53] (Fig. 2). In addition, evidence suggests that the ECS can effect ovarian function via modulation of pathways involved in energy balance and metabolic control [52] since obesity is associated with menstrual irregularities, chronic oligo-anovulation, and infertility [52]. Evidence using animal models of obesity, although limited, suggest a connection between increased adiposity and dysregulation of AEA and 2-AG [62].

A recent study by Cui et al aimed to establish a link between the ECS and PCOS using non-obese subjects [53]. Within the context of the proliferative and secretory phases of the menstrual cycle, the endometrium exhibited significantly reduced levels of FAAH in the subjects with PCOS when compared to non-PCOS infertile women who served as the control subjects. Given that FAAH is the catabolic enzyme for AEA, it is biologically plausible that AEA levels may be elevated in patients who suffer from PCOS [53]. Indeed, in obese and type II diabetic patients, serum levels of AEA and 2-AG are significantly higher than that seen in women of healthy weight [52]. Using real-time PCR, these authors demonstrated that CB1 transcripts from omental adipose tissue were lower in PCOS subjects vs. those without PCOS [52]. However, they did not observe significant differences in adipose tissue CB2 transcript expression between the two groups. Interestingly, Cui et al, report that endometrial CB1 expression levels were not different between the aforementioned groups, nor did it fluctuate during the menstrual cycle [53]. In a subsequent study by Cui et al [63], non-obese women with PCOS were treated with Diane-35 (a mix of antiandrogens and estrogens) and metformin (an oral hypoglycemic) which significantly reversed the increase in plasma AEA levels [63]. These findings support a role of the ECS in non-obese women with PCOS, suggesting that the ECS, mainly via increased serum AEA and reduced endometrial FAAH expression, potentiates the progression of PCOS [63] (Fig. 2). Intriguingly, Cui et al relied on the endometrial expression of FAAH, rather than that which is expressed within the ovary. Endometrial FAAH expression was thought to correlate with serum AEA expression; however, serum AEA does not necessarily correlate to that found within tissues. The proposed reasoning was that FAAH, as the main hydrolase for AEA, fluctuates during the menstrual cycle; however, a larger sample size is required to support their findings, along with a direct analysis of ovarian FAAH expression [63]. Studies in obese women with and without PCOS are warranted to ascertain the impact of the ECS on the pathogenesis of PCOS in this cohort. Taken together, the ECS may be intimately involved with the progression of PCOS, likely due to its key role in energy homeostasis, whose components may be useful clinically as biomarkers of PCOS.

A significant amount of research focus has been directed towards how cannabinoids, particularly the plant-derived and synthetic cannabinoids, effect the initiation and progression of cancer [64]. Synthetic cannabinoids (such as prescription Marinol®), have been widely used to address the negative symptoms associated with chemotherapy, namely neuropathic pain, loss of appetite, nausea and vomiting [3, 5, 23, 65‐67]. Beyond the reported palliative effects of phyto- and synthetic cannabinoids, these molecules are gaining recognition for their role in the pathogenesis of cancer [3]. Indeed, various in vivo and in vitro studies have demonstrated that the endocannabinoid system is able to initiate apoptosis and autophagy, induce cell cycle arrest, mount inflammatory responses against malignant cells, and block angiogenic and metastatic processes [3, 67]. In short, cannabinoids exert a variety of anticancer effects by disturbing the signaling pathways involved in malignant transformation and thus tumour progression [67].

Several mechanisms are thought to mediate the anticancer effects of cannabinoids. It has been suggested that stimulation of TRPV1 or PPARγ, along with inhibition of COX2 might mediate the apoptotic and anti-proliferative effects of AEA and synthetic cannabinoids. Furthermore, several signaling pathways governing cell survival, proliferation, and apoptosis, including p38, MAPK, cAMP, PI3K-PKB, and others are activated by ligand-receptor interactions at the classical cannabinoid receptors [67, 68].

Endocannabinoid effects in different tumour models are highly variable and are likely a result of the differential expression of cannabinoid receptors, along with variability between tissue types. Indeed, reports have suggested that differential expression of cannabinoid receptors between normal and cancerous cells may impact the pathogenesis of a malignant tumour [3]. The current view is that cancerous cells express greater levels of CB1 and/or CB2 as compared to normal cells [3, 69]. This finding has been demonstrated in hepatocellular carcinoma, breast cancer, lymphoma, and prostate cancer cells lines [3]. However, a paucity of data exists regarding the role of the ECS in the pathogenesis of ovarian cancer.

To the best of our knowledge, there is only one study that specifically looked at the ECS in the context of ovarian cancer. This recent study by Messalli et al, utilized histochemical analyses in human ovarian tumours and revealed a tumor grade-dependent expression of CB1 receptors. The benign ovarian tumors displayed weak-moderate expression of CB1, with borderline tumors showing a similar trend. In sharp contrast, invasive tumors showed moderate-strong CB1 expression [69]. Curiously, the authors only analyzed FAAH expression in the borderline phenotype, which paralleled the expression pattern of CB1 [69]. They postulated that while low cannabinoid levels may activate proliferative pathways (in noncancerous cells), a higher concentration results in anti-proliferative and apoptotic events in cancerous phenotypes. Importantly, and as previously stated, the CB1 or CB2 expression in cancer cells, including the increase reported by Messalli et al, does not necessarily correlate with the expression pattern from the healthy tissue of origin [68]. Indeed, it has been reported that cannabinoid receptors and their endogenous ligands are generally upregulated in the cancer phenotype when compared to its non-cancerous phenotype. Such elevated levels of cannabinoid receptor expression permit exogenous cannabinoids, administered at therapeutic doses, to impair tumour progression by inducing apoptosis (reviewed in [68, 70]). Furthermore, the expression of the various components of the ECS is not homogenous across all cancers and the mechanisms are complex, thus by pharmacological or genetic manipulations of the ECS, further investigations into the mechanistic connections between the ECS and cancer progression are warranted [67, 69].