UNIVERSITÀ DEGLI STUDI DI PISA
DIPARTIMENTO DI FARMACIA
Corso di Laurea Magistrale in Chimica e Tecnologia
Farmaceutiche
Graduation Thesis
Quantification of endocannabinoids, arachidonic acid and
prostaglandins in tissues from mice treated with the substrate
specific COX-2 inhibitor 4’-O-methylhonokiol
Supervisors:
Dott. Andrea Chicca
Prof.ssa Clementina Manera
Candidate:
Marco Borsò
Contents
1 Introduction 6
1.1 Overview of the endocannabinoid system . . . 6
1.2 Endocannabinoids . . . 7
1.2.1 How they work . . . 7
1.2.2 AEA and 2AG . . . 8
1.2.3 Other endocannabinoids and endocannabinoid-like compounds . . . 11
1.2.4 Receptors and ligands . . . 12
1.2.4.1 CB1 and CB2 . . . 12
1.2.4.2 Other receptors . . . 14
1.3 Cyclooxygenase-2 (COX-2) . . . 14
1.4 4'-O-Methylhonokiol . . . 17
2 Aim of the thesis 20 3 Materials and Methods 21 3.1 Animals treatment . . . 21
3.2 LC-MS/MS . . . 22
3.2.1 Chemicals and reagents . . . 22
3.2.2 Samples extraction . . . 22
3.2.3 LC-MS/MS conditions . . . 22
3.2.4 LC-MS/MS validation . . . 28
3.3 CB2 receptor binding assay . . . 30
4 Results and discussion 31 4.1 LC-MS/MS results . . . 31
4.1.1 Kidney . . . 31
4.1.2 Brain . . . 34
4.1.3 Liver . . . 37
4.1.4 Lung . . . 39
4.1.5 Spleen . . . 42
4.2 CB2 receptor binding assay . . . 45
List of Figures
1.1 Chemical structure of Δ9-THC . . . 6
1.2 Components of the endocannabinoid system (Mark Rau) . . . 7
1.3 Chemical structure of AEA . . . 8
1.4 Chemical structure of 2AG . . . 8
1.5 Biosynthetic and metabolic pathways of AEA and 2AG . . . 10
1.6 Structure of a FAAH inhibitor (URB-597) and a MAGL inhibitor (JZL-184) . . . 11
1.7 Chemical structures of Virodhamine and NE . . . 11
1.8 Chemical structure of NADA . . . 12
1.9 Chemical structures of endocannabinoid-like compounds . . . 12
1.10 Pathways associated with CB receptors activation . . . 12
1.11 CB1 and CB2 structures . . . 13
1.12 Chemical structures of CP-55,940 and WIN-55,212-2 . . . 14
1.13 Prostaglandins cascade . . . 15
1.14 COX-2 activity on AEA and 2AG . . . 16
1.15 Mechanism of a COX-2 SSI . . . 17
1.16 Chemical structures of Magnolol, Honokiol and MH . . . 17
1.17 Magnolia grandi ora . . . 18
3.1 SWISS CD1 mouse . . . 21
3.2 Gradient curve of the positive mode. . . 23
3.3 Gradient curve of the negative mode. . . 23
3.4 Mass spectrum of PGE2 upon negative ionization. . . 24
3.5 2AG and 2AG-d5 . . . 25
3.6 AEA and AEA-d4 . . . 25
3.7 MH and MH-d6 . . . 25
3.8 Corticosterone and Corticosterone-d4 . . . 25
3.9 LEA and LEA-d4 . . . 26
3.10 OEA and OEA-d4 . . . 26
3.11 PEA and PEA-d5 . . . 26
3.12 Progesterone and Progesterone-d9 . . . 26
3.13 2AGE and 2AG-d5 . . . 27
3.14 AA and AA-d8 . . . 27
3.15 PGE2 and PGE2-d4 . . . 27
3.16 PGD2 and PGE2-d4 . . . 27
4.1 Kidney: MH . . . 32
4.2 Kidney: AEA (a), 2AG (b), NE (c) . . . 32
4.3 Kidney: AA (a), PGE2 (b) . . . 32
4.4 Kidney: PEA (a), OEA (b), LEA (c) . . . 33
4.5 Kidney: Corticosterone (a), Progesterone (b) . . . 33
4.6 Brain: MH . . . 34
4.7 Brain: AEA(a), 2AG(b) . . . 35
4.8 AEA and 2AG adapted from Chicca et al Journal of Neuroin ammation . . . 35
4.9 Brain: AA(a), PGE2(b), PGD2(c) . . . 36
4.10 Brain: PEA(a), OEA(b), LEA(c) . . . 36
4.11 Brain: Corticosterone(a), Progesterone(b) . . . 36
4.12 Liver: MH . . . 37
4.13 Liver: AEA(a), 2AG(b), NE(c) . . . 37
4.14 Liver: AA(a), PGD2(b) . . . 38
4.15 Liver: PEA(a), OEA(b), LEA(c) . . . 38
4.16 Liver: Corticosterone(a), Progesterone(b) . . . 38
4.17 Lung: MH . . . 39
4.18 Lung: AEA(a), 2AG(b) . . . 40
4.19 Lung: AA(a), PGE2(b), PGD2(c) . . . 40
4.20 Lung: PEA(a), OEA(b), LEA(c) . . . 41
4.21 Lung: Corticosterone(a), Progesterone(b) . . . 41
4.22 Spleen: MH . . . 42
4.23 Spleen: AEA(a), 2AG(b) . . . 43
4.25 Spleen: PEA(a), OEA(b), LEA(c) . . . 44
4.26 Spleen: Corticosterone . . . 44
Chapter 1
Introduction
1.1 Overview of the endocannabinoid system
The endocannabinoid system (ECs) remained unknown till the rst part of 1990th, when the rst cannabi-noid receptor, cannabicannabi-noid receptor 1 (CB1), was discovered in the brain [1]. The cannabicannabi-noid receptor 2 (CB2), instead, was found in immune cells in 1993 [2].The presence of these two type of G-protein coupled receptors (GPCR) could be only explained with the presence of endogenous eCBs. In fact, immediately after the discovery of the receptors, a series of endogenous ligands for the cannabinoid re-ceptors have been identi ed and called endocannabinoids (eCBs). Among them, the two most present in biological tissues and most studied, are N-arachidonoylethanolamine, also known as andandamide (AEA) and 2-arachidonylglycerol (2-AG) [3]. ECBs are lipid mediators that mime the action of Δ9-tetrahydrocannabinol (Δ9-THC), an exogenous psychoactive compound present in Cannabis plants (Fig-ure 1.1).
Figure 1.1: Chemical structure of Δ9-THC
From that moment chemistry, metabolism, biochemistry and pharmacology of ECs have been strongly
investigated (Figure 1.2). It is implicated in the regulation of several physiological processes like immune functions, neuroprotection, pain regulation, cancer and cardiovascular diseases. It is clear that, a dysreg-ulation of ECs could lead to pathological conditions so, the interest in trying to nd new pharmacological strategies in order to reactivate the normal activity of this system, is still growing [4].
Figure 1.2: Components of the endocannabinoid system (Mark Rau)
1.2 Endocannabinoids
1.2.1 How they work
Endocannabinoids are synthetized on demand in a Ca2+-dependent manner in response to physiological
stimuli. This means that the biosynthesis is immediately followed by their release. Neuron depolarization or activation of Gq-coupled receptors, lead to an increase of [Ca2+] which is the main cause of the release
of eCBs from phospholipidic precursors [5]. ECBs act as retrograde messengers [6], meaning that they are released from the postsynaptic cell and activate receptors present on the presynaptic terminal. This activation of presynaptic receptors can produce an alteration on the production of neurotransmitters, a process known as transient and long lasting reduction [7]. The presynaptic eects that eCBs have on the release of neurotransmitters, are called in a dierent way depending on the involvement of γ-Aminobutyric Acid (GABA) or Glutamate transmissions. The depolarization-induced suppression of inhibition (DSI) is activated from a voltage dependent in ux of Ca2+ on the postsynaptic terminal. The eect on the
presynaptic terminal is the inhibition of GABA release, meaning a reduction of the inhibitory signal. The depolarization-induced suppression of excitation (DSE), acts with the same mechanism of DSI, but opposite in sign because of the glutamatergic target, and the nal eect is a reduction of excitatory signal [8]. The retrograde mechanism of eCBs is also important for the long-term synaptic plasticity, divided in long-term depression (LTD) and long-term potentation (LTP). ECBs induce LTD in the striatum and in the nucleus, whereas they prevent LTP in the hippocampus [7].
1.2.2 AEA and 2AG
The two most abundant eCBs in tissues are AEA and 2AG. AEA has been discovered and then isolated in 1992 [9]. Chemically, it is the ethanolamide of arachidonic acid (AA) (Figure 1.3), it belongs to N-acetylethanolamines (NAEs) family and it acts as a partial agonist on both CB1 and CB2 [10].
Figure 1.3: Chemical structure of AEA
AEA is biosynthetized following dierent pathways from phospholipidic precursors. The main one starts with an acyl chain transfer, catalyzed by a Ca2+-dependent N-acyltransferase (NAT), from the sn-1
position of a glycerophospholipid to the amino group of the hydroxyethyl moiety of phophatidylethanolamine (PE), to form N-acylphophatidylethanolamine (NAPE). Then, the phospholipase D (NAPE-PLD) cat-alyzes the hydrolysis of NAPE to form AEA and phosphatidic acid. In the second one, AEA formation is due to the deacylation of NAPE made by α/β-hydrolase 4 (ABHD4) to form glyceropohpho-anandamide and the following hydrolysis made by the phosphodiesterase glicerophophodiester-phosphodiesterase-1 (GDE1). Lastly, AEA can be synthetized from the hydrolysis of NAPE catalyzed by phospholipase C to form phosphoanandamide. The nal step is a phosphate group removing made by the phosphatase PTPN22 [11]. 2AG has been identi ed and isolated from the canine gut in 1995 [12]. It is the glycerol ester of AA (Figure 1.4) and acts as a full agonist on both cannabinoid receptors CB1 and CB2 [13].
The main biosynthetic pathway starts with the hydrolysis of membrane phospholipids catalyzed by Phospholipase C (PLC), to form 1,2-diacylglycerol (DAG). DAG is then converted to 2AG by diacylglyc-erol lipase (DAGL) [14]. Once the eCBs are released in the extracellular part, they activate the receptors and then, are removed by intracellular uptake and degraded [15]. Considering the lipophilic structure of most of the endocannabinoid molecules, it could be concluded that they easily pass through the cell mem-brane. Despite their structures, the tra cking of eCBs through the membrane is regulated by a putative endocannabinoid membrane transporter (EMT). EMT is responsible for the bidirectional movement of both 2AG and AEA, even if AEA showed higher selectivity towards the transporter. The interest in EMT activity on 2AG and AEA is still growing and several inhibitors of this transporter have been synthetized. These inhibitors could have a potential role and represent good alternatives in regulating eCBs levels [16]. Two hydrolases are implicated in the main inactivation pathway, both for AEA and 2AG. Regarding AEA, the main hydrolytic enzyme is the fatty acid amide hydrolase (FAAH). This is an integral membrane serine hydrolase protein, characterized by a Ser-Ser-Lys catalytic triade, which has the highest density in the brain and in the liver. It hydrolyzes the amidic bond to form AA and ethanolamide. AEA is the main target of FAAH respect to the other NAEs. The other enzyme responsible of the inactivation of AEA is N-acylethanolamine-hydrolyzing acid amidase (NAAA) which is mainly present in the immune cells [17]. The monoacyl-glycerol lipase (MAGL) is the main degrading enzyme for 2AG in the brain, responsible of 85% of total brain membrane 2AG hydrolase activity. Results of its hydrolytic action on the esteric bond are AA and glycerol. Other two minor enzymes present in the brain, responsible of the metabolism of 2AG, are α/β-Hydrolase-6 (ABHD-6) and α/β-Hydrolase-12 (ABHD-12). The results of their hydrolysis are the same as the ones of MAGL activity, AA and glycerol (Figure 1.5) [18].
Figure 1.5: Biosynthetic and metabolic pathways of AEA and 2AG [17]
The inhibition of FAAH and MAGL attracted recent studies as a possible mechanism to regulate and modulate eCBs levels. Several inhibitors have been synthetized and the scaold used as template for successive modi cation and improvement. The inhibition of FAAH activity lead to an increase of AEA levels. URB-597 (Figure 1.6) is a selective carbamate-based inhibitor of FAAH, without activity on other cannabinoid targets. Its activity is not accompanied by CB1-like behavioral side eects, known as tetrad: analgesia, hypomotility, hypothermia and catalepsy [19]. JZL-184 (Figure 1.6) have been described as highly e cacious and selective inhibitor of MAGL activity on 2AG. JZL-184 did not aect the levels of AEA, other NAEs or monoacylglycerols. Inhibitory activity of MAGL led to an increase of 2AG concentration in the brain but a series of cannabimimetic side eects have been shown [20].
(a) URB-597 (b) JZL-184
Figure 1.6: Structure of a FAAH inhibitor (URB-597) and a MAGL inhibitor (JZL-184)
1.2.3 Other endocannabinoids and endocannabinoid-like compounds
Except the well-known AEA and 2AG, the panorama of endocannabinoids includes other compounds. Virodhamine is the ethanolamine ester of AA (Figure 1.7), it acts as a full agonist on CB2 and as a partial agonist or antagonist on CB1 receptors [21]. Despite some current debates, the 2-arachidonyl-glyceryl ether (2AGE), also known as noladin-ether (NE) (Figure 1.7), is another compound classi ed as endocannabinoid and it binds to CB1 and weakly to CB2 receptors [22]. N-arachidonoyl-dopamine (NADA) (Figure 1.8) was discovered in the brain in 2002 and acts as an agonist on CB1 receptor and on the vanilloid receptor TRPV1 [23]. There are other compounds called endocannabinoid-like which exert cannabimetic eects without activating cannabinoid receptors and are able to prevent the enzy-matic metabolism of endocannabinoids [24]. Among these compounds, palmitoylethanolamide (PEA), oleoylethanolamide (OEA) and linoleoylethanolamide (LEA), which belong to the NAEs family, are the most studied. (Figure 1.9). PEA has an activity on both TRPV1 and PPARγ receptors that mediate its anti-in ammatory and analgesic eects [25]. Both PEA and OEA activate GPR55 receptors [26]. OEA and LEA are known for their eects on appetite [27] [28].
(a) Virodhamine (b) NE
Figure 1.8: Chemical structure of NADA
(a) PEA (b) OEA (c) LEA
Figure 1.9: Chemical structures of endocannabinoid-like compounds
1.2.4 Receptors and ligands
1.2.4.1 CB1 and CB2
CB1 and CB2 receptors are part of the GPCR superfamily, mainly coupled with Gi/o proteins. Once activated by a ligand, they both inhibit adenylyl cyclase (AC) and activate mitogen-activated protein kinase (MAPK). CB1 receptor is also coupled with Gs protein (Figure 1.10) [10].
Figure 1.10: Pathways associated with CB receptors activation [29]
There are homologies between CB1 and CB2 receptors: 44% for the total structure and 68% regarding the transmembrane domain (Figure 1.11) [2]. Since they are GPCR they show the typical structure of seven transmembrane domains.
Figure 1.11: CB1 and CB2 structures
CB1 receptors are most abundant in the Central Nervous System (CNS) and in the brain, especially in hippocampus, cerebellum and striatum [30]. The activation of CB1 has eects on cognition, memory, control of motor functions and analgesia [10]. CB1 receptors are more abundant in GABAergic interneu-rons than in glutamatergic principal neuinterneu-rons [31]. There are evidence of the presence of CB1 also in some peripheral tissues like gastrointestinal tract, liver, kidney, lungs [32]. CB2 receptors are mainly present in cells and organs of the immune system, such as tonsils, B and T cells, monocytes, macrophages and they are highly expressed in spleen tissue [33]. CB2 are also found in the brain and when activated, they are able to modulate the migration of immune cells and the release of cytokines in and outside the brain [32]. Since their discovery, the activity of agonist, antagonist and inverse agonist for these two type of recep-tors have been investigated. Most of the pharmacological eects are due to the activity on CB1 receprecep-tors which are mainly present in the CNS. Ligands for this type of receptor have been studied because of their potential role to treat pain, cardiovascular and gastrointestinal diseases, obesity. Unfortunately, these pharmacological bene ts are combined with undesirable psychotropic eects that limit the progression of CB1 ligands. Anyway, they now have an important role for in vitro and in vivo studies as comparison. Nowadays, the studies are focused on selective CB2 receptor ligands, especially for pain treatment and neurological diseases. The activation of CB2 receptors avoid the central side eects observed for CB1 [34]. CP-55,940 and WIN-55,512-2 are the two major non-selective endocannabinoid receptor agonists which
are used as control in laboratory experiments (Figure 1.12).
(a) CP-55,940 (b) WIN-55,212-2
Figure 1.12: Chemical structures of CP-55,940 and WIN-55,212-2
1.2.4.2 Other receptors
The presence of the G-protein coupled receptor GPR55 in brain, vascular endothelium and immune system has been showed and its activity is regulated by both natural and synthetic cannabinoid compounds. Virodhamine activates GPR55, CB1 and CB2 receptors with a higher selectivity compared with AEA. This activation is important for a cannabinoid vascular-tone controlling. Eects of PEA on immune system are mainly due to modulation of CB2 receptors but, it acts also as a potent agonist on GPR55 and this could have a role in immune cell migration and activation. It has been showed that AEA activated GPR55 with the same potency than CB1 and CB2. 2AG, PEA and Virodhamine, instead, are more potent at activating GPR55 than CB1 and CB2 [26]. Transient receptor potential vanilloid-type-1 (TRPV1) is a non-selective cation channel implicated in pain transmission, which can be found in both peripheral and central terminals of sensory neurons. This receptor can easily be activated during in ammation and the expression of TRPV1 is increased during neuropathic pain. The main exogenous ligand of TRPV1 is the capsaicin, present in chili peppers, which has an analgesic activity. Among the eCBs, AEA acts as a partial agonist on TRPV1 and together with CB1 activation, it decreases the neuropathic pain. Also NADA, 2AG and PEA are ligands that can activate these type of receptors [35].
1.3 Cyclooxygenase-2 (COX-2)
Cyclooxygenase (COX) is the enzyme responsible for the rst step in the metabolism of AA to gener-ate prostaglandins (PGs). PGH2s which are formed from COX-2 activity on AA, are metabolized by downstream synthases to form prostaglandin-E2 (PGE2), prostaglandin-D2 (PGD2), prostaglandin-F2α, prostacyclin (PGI2) and thromboxane-A2 (TxA2) (Figure 1.13). PGs are local mediators of in ammation and modulators of physiological functions, like maintenance of gastric mucosal integrity, modulation of
renal vascularization and renine release. The inhibition of this pathway, is the main contributor of the pharmacological eects of non-steroidal anti-in ammatory drugs. The most common side eects of this type of drugs are gastrointestinal damages and ulcer, caused by the inhibition of the formation of PGs in intestine and stomach. [36]
Figure 1.13: Prostaglandins cascade [37]
Two isoforms of COX exist and they are called COX-1 and COX-2. COX-1 was rst puri ed from ram seminal vesicles and was cloned in 1988 [38]. Subsequently, a second in ammatory-mediated COX isoform was identi ed and called COX-2 [39] [40]. COX-2 is a homodimer composed by two dierent monomers, but it behaves as functional heterodimer with an allosteric site and a catalytic site. The active site of COX-2 is located inside the catalytic domain, but recent studies discovered that the two monomers of this enzyme are functionally interdependent. COX-2 is also responsable for the metabolism of the two major eCBs, 2AG and AEA. It oxygenates and inactivates 2AG to form prostaglandin-glycerol esters (PG-GEs) and AEA to produce prostaglandin-ethanolamides (PG-EAs or prostamides) [41]. (Figure 1.14)
Figure 1.14: COX-2 activity on AEA and 2AG [42]
PG-GEs and prostamides act on dierent receptors as compared to classic prostaglandin, which have not been identi ed, yet [43]. They could have an action as precursor of common PGs or act as signal mediators [44]. These compounds are known to induce pro-in ammatory and hyperalgesic eects and to induce excitoxic damages, eects that could lead to an imbalance between neuroprotective eects of 2-AG and potential neurotoxic eects of its COX-2 metabolites [45]. Recent studies showed that the two monomers of COX-2 are functionally connected and the binding of a substrate or inhibitor at one of the two active sites, modi es the properties of the other one. Lower concentrations of inhibitors are required to inhibit the allosteric site (responsible of the eCBs oxygenation) than the ones required for the catalytic site (responsible of the AA oxygenation). Inhibitor binds to the rst subunit of the homodimer and induces a conformational change in the second subunit, responsible of the eCBs oxygenation. The result of this process is the inhibition of the eCBs oxygenation. In order to inhibit also the AA oxygenation, another molecule of inhibitor has to bind on the other subunit (Figure 1.15). These interesting ndings pushed studies to focus on the selective inhibition of COX-2, trying to inhibit eCBs oxygenation without
aecting AA and PGs. A compound with this type of action is called Substrate Speci c Inhibitor (SSI) of COX-2-mediated endocannabinoid oxygenation. It has been shown that (R)-Profens like Flurbiprofen, Naproxen and Ibuprofen act this type of COX-2 inhibition [46].
Figure 1.15: Mechanism of a COX-2 SSI [46]
1.4 4'-O-Methylhonokiol
4-O'-Methylhonokiol (MH), a biphenyl compound, (structure Figure 1.16) is the main component of the Magnolia grandi ora oil seeds (Figure 1.17) The seeds of this plant, have been known in the Mexican medicine for their antispasmolyitc and anti-in ammatory eects [47]. Magnolol and Honokiol (structure Figure 1.16) , known for their anti-in ammatory, anti-allergic and anti-bacteria activities, are secondary components of this oil. These bioactive compounds are mainly present in a dierent species of this plant, the Magnolia o cinalis [48].
Figure 1.16: Chemical structures of Magnolol, Honokiol and MH [47]
(a) Tree (b) Seeds
Figure 1.17: Magnolia grandi ora
MH is a selective ligand for CB2 receptor. It acts as a partial agonist (they can behave as agonists or antagonists depending on the conditions) in the GTPγS binding assay compared to the full agonist 2AG, which induces the maximal response. The GTPγS binding assay measures the level of G protein activation following agonist occupation of GPCR. MH is also known to act as full agonist in the CB2 receptor-mediated inhibition of forskolin-induced cAMP production. In this assay it reproduces the same eects of the full agonist 2AG in the cAMP pathways [49]. Finally, MH acts as a full agonist at the level of [Ca2+] intracellular. Dierent derivatives of MH have been tested to explore which parts of the
scaold were important for the CB2 binding. In this study, MH appeared to have the most active scaold compare to the others derivatives. Hydrogenation, epoxidation or hydroxylation of the side chains of MH did not change the binding a nity and, in some cases, it was abolished. Modi cation on the -OH and -OCH3, such as hydrophobic or bulky groups, led to a decrease in the binding activity [47]. A similar study has been done studying the activity of dierent derivatives on COX-1 and COX-2. Also in this case MH showed the best pro le, with a small selectivity towards COX-2. [50]. MH is studied for its anti-in ammatory and neuroprotective eects in Alzheimer Disease (AD) murine models. AD is the most common neurodegenerative disease which is characterized by cerebral deposition of the amyloid β peptide. This peptide is a critical factor in AD and leads to a series of neuronal damages due to the increase of cellular death. Anti-oxidative and anti-amyloidogenesis eects of MH ameliorates LPS-induced
memory impairment and inhibits neuroin ammation [51]. MH also suppresses the oxidative damages of β-amyloid induced memory impairment murine models [52]. It also shows anti-osteoclasteogenic eects [47]and induces anxiolytic eects through the activation of GABAergic transmission. GABA is the main inhibitory neurotransmitter and its system is the most common target of the anxiolytic drugs. MH has an agonistic eect on GABAA receptors which are heteromeric Cl− channels. Activation of these channels
leads to an in ux of Cl− into the cell that causes hyperpolarization and nally inhibition of cellular
activity [53]. In rat MH showed high systemic clearance and an extensive cytochrome P450 (CYP450) hepatic metabolism, where it is transformed in Honokiol. MH has a low oral bioavailability and this could be due to its extensive rst-pass hepatic metabolism. Importantly for its therapeutic application, is the ability to pass the Blood-Brain-Barrier (BBB) [54].
Chapter 2
Aim of the thesis
The selective inhibition of COX-2-mediated 2AG and AEA oxygenation without aecting AA and PGs levels could be an interesting alternative therapeutic opportunity to treat in ammatory processes. Com-pounds that acts in this manner are called substrate speci c inhibitors (SSIs) of COX-2-mediated en-docannabinoid oxygenation. In this way, the oxygenation of 2AG and AEA is inhibited as well as the formation of PG-GEs and prostamides, while of AA and PG levels should not be aected [46]. 4'-O-Methylhonokiol (MH) is a biphenyl compound extracted from oil seeds of Magnolia grandi ora. It is known for its several properties, such as in ammatory, anxiolytic, osteoclasteogenic, anti-oxidative and neuroprotective [47] [51] [52] [53]. MH acts as a CB2 receptor agonist and showed to be a substrate speci c inhibitor of COX-2-mediated endocannabinoid oxygenation [49]. There is evidence that COX-2 is expressed in critical location of mammalian kidney [55]. The SSI activity of MH on COX-2 in the kidney could play an important therapeutic role, since PGs have a key physiological role in this organ in terms of vascularization and protection.
The aim of my thesis was to quantify endocannabinoids and prostaglandins in mice that were injected with MH. In order to study the eects of this molecule in vivo, Swiss CD1 mice were treated with three dierent doses of MH or control, injected intraperitoneally (i.p.). After 2 hours the animals were sacri ced and organs collected. The quanti cation of endocannabinoids and prostaglandins in kidney, brain, liver, lung and spleen was done using a LC-MS/MS method currently used in the laboratory of Prof. Gertsch.
Chapter 3
Materials and Methods
3.1 Animals treatment
24 females, 8 weeks old SWISS CD1 mice (Figure 3.1) were used for this experiment. All animals were kept in a temperature-controlled environment with a 12 h light-dark cycle and were allowed free access to food and water at all times and were cared for in accordance with National Institute of Health (NIH) guidelines. Animals were injected intraperitoneally (i.p.) with dimethyl sulfoxide (DMSO) as vehicle and with 3mg/kg, 10mg/kg and 20mg/kg of MH. They were divided in four dierent groups, each one with a dierent treatment. After 2 hour mice were anesthetized and sacri ced. Brain, spleen, liver, kidneys and lungs were collected and stored at -80°C until use.
Figure 3.1: SWISS CD1 mouse
3.2 LC-MS/MS
3.2.1 Chemicals and reagents
All chemicals and reagents used for LC-MS/MS were of the purest analytical grade. AEA, AEA-d4, 2-AG, 2-AG-d5, PEA, PEA-d5, OEA, OEA-d4, LEA, LEA-d4, arachidonic acid (AA), AA-d8, PGD2, PGE2, Corticosterone and Progesterone were purchased from Cayman Chemical Europe, Tallinn, Estonia. MH was kindly provided by Prof. Schühly, University of Gratz ([47]). HPLC-grade methanol (CH3OH), HPLC-grade acetonitrile, ammonium acetate and formic acid were obtained from Sigma-Aldrich, Stein-heim, Germany. HPLC-grade ethyl acetate and hexane were obtained from Acros Organics, New Jersey, USA. Deionized water (18.2 MΩ Ö cm) was obtained from an ELGA Purelab Ultra Genetic system (VWS UK Ltd, ELGA LabWater, UK).
3.2.2 Samples extraction
For the extraction, tissues (brain, spleen, liver, kidneys and lungs) were weighted (~40mg) when still frozen and transferred to 2 mL Microcentrifuge vials (XXTu Microvials, BioSpec, Oklahoma, USA) containing 3 Chrome-Steel Beats (2.3 mm dia, BioSpec, Oklahoma, USA) and 0.1 M Formic Acid. Samples were homogenized using a 24 positions Mini bead beater (BioSpec Products, Inc, Bartlesville, OK, USA), then rapidly transferred to glass tubes containing 1.5 mL of ethyl acetate/hexane (9:1) 0.1% formic acid and 5µL of the Internal Standard mixture (IS), strongly vortexed for 30 seconds and sonicated (Sonicator Merck-Gruppe) in cold bath for 5 mins. Samples were centrifuged at 3000 rpm for 10 min at 4°C and kept for 1h at -20°C to freeze the aqueous phase. The upper organic phase was recovered in plastic tubes and, dried in a speed vacuum. The extracts were reconstituted in 50 µL of ACN:EtOH 8:2 and 10 µL were injected in the LC-MS/MS system.
3.2.3 LC-MS/MS conditions
A hybrid triple quadrupole 4000 QTRAP mass spectrometer (AB Sciex Concord, Ontario, Canada) was used with a Shimadzu UFLC (Shimadzu Corporation, Kyoto, Japan) with cooled auto-sampler. Sam-ple temperature was maintained at 4°C in the auto-samSam-pler prior to analysis. The LC columns were Reprosil-PUR C18 column (3 μm particle size; 2 Ö 50 mm, Dr. A. Maisch, High Performance LC-GMBH, Ammerbuch, Germany) maintained at 40° C. The system was operated in positive and negative mode with 2 dierent elution mobile phases and gradient method. In positive mode the elution mobile phases were: water, 2 mM ammonium acetate, 0.1% formic acid (solvent A) and methanol, 2 mM ammonium
acetate (solvent B). The gradient started at 15% B, increasing linearly to 70% at 3.5 min, then to 99% at 8 min, maintained until 12 min, with subsequent re-equilibration at 15% for further 2.5 min. The ow rate was 0.35mL/min. (Fig 3.2). In negative mode the gradient started at 5% B, increasing linearly to 40% at 3 min, then to 65% at 9 min and linearly again to 95% at 10min; this was maintained until 14 min, with subsequent re-equilibration at 5% for further 3 min. The ow rate was 0.3mL/min (Fig 3.3). In this mode the organic mobile phase (B) was substituted with acetonitrile 0.1% formic acid to minimize crosstalk between analytes with the same mass ([M-H]-, m/z 351 PGE2-D2) (Fig 3.4).
Figure 3.2: Gradient curve of the positive mode.
Figure 3.4: Mass spectrum of PGE2 upon negative ionization.
The dominant peak is m/z 351.2 which corresponds to the [PGE2-H]- ion. The spectrum shows the fragmentation pattern of m/z 351.2. The loss of H2O lead to two of the fragments observed: m/z 333.1 (-H2O) and 315 (-2H2O). PGD2, being isomer of PGE2, exhibits identical fragmentation patterns (iden-tical precursor and product ions) and requires the development of high resolution liquid chromatography method for their separation
Peaks were integrated, and the Analyst software version 1.5 (AB Sciex Concord, Ontario, Canada) was used for quanti cation. Identi cation of compounds in samples was con rmed by comparison of precursor and product ion m/z values and LC retention times with standards. The following Multiple Reaction Monitoring (MRM) transitions were monitored for quanti cation of the analytes: Positive mode: 2AG m/z 379287, 379203 (IS: 2-AG-d5 384287); AEA m/z 348133, 34862 (IS: AEA-d4 35266); MH m/z 281298, 281240, 281225 (IS: MH-d6 287246); Corticosterone m/z 347135, 347121, 34797 (IS: Corticosterone-d4 367121); LEA m/z 324109, 32462 (IS: LEA-d4 32866); OEA m/z 326309, 32662 (IS: OEA-d4 33066), PEA m/z 300283, 30062 (IS: PEA-d5 30562), Progesterone m/z 315109, 31597 (IS: PROG-d9 324100); NE m/z 365133, 365121 (IS: 2-AG-d5 384287). Negative mode: AA m/z 303259, 30359 (IS: AA-d8 31159), PGE2 m/z 351271, 351315 (IS: PGE2-d4 355319); PGD2 m/z 351189.1 (IS: PGE2-d4 355319). Typical extracted ion cromatograms are showed in the following gures.
Positive mode:
(a) 2AG (b) 2AG-d5
Figure 3.5: 2AG and 2AG-d5
2AG is known to isomerize into 1AG. The method was developed to minimize isomerization but if present the sum of the response of the two isomers was applied as suggested in several papers [56].
(a) AEA (b) AEA-d4
Figure 3.6: AEA and AEA-d4
(a) MH (b) MH-d6
Figure 3.7: MH and MH-d6
(a) Corticosterone (b) Corticosterone-d4
(a) LEA (b) LEA-d4
Figure 3.9: LEA and LEA-d4
(a) OEA (b) OEA-d4
Figure 3.10: OEA and OEA-d4
(a) PEA (b) PEA-d5
Figure 3.11: PEA and PEA-d5
(a) Progesterone (b) Progesterone-d9
(a) NE (b) 2AG-d5
Figure 3.13: 2AGE and 2AG-d5
Negative mode:
(a) AA (b) AA-d8
Figure 3.14: AA and AA-d8
(a) PGE2 (b) PGE2-d4
Figure 3.15: PGE2 and PGE2-d4
(a) PGD2 (b) PGE2-d4
Figure 3.16: PGD2 and PGE2-d4
PGE2-D2 are isomers and this is the best separation we could obtain using a C-18 reverse phase column. PGE2-d4 was used also as internal standard for PGD2.
Triplicate 11 calibration points were prepared in 0.1 % formic acid plus 1% bovine serum albumine (BSA), and the appropriate matrix. The concentration of calibrators moved from 20 ng/mL to 50000 ng/mL for AA, 0.2 ng/mL to 500 ng/mL for LEA, PEA, OEA, PGD2, PGE2, Corticosterone, from 0.08 ng/mL to 200 ng/mL for AEA and Progesterone, 2 ng/mL to 5000 ng/mL for 2-AG and 8 ng/mL to 20000 ng/mL for MH. The calibrators and the amount of ISs were speci cally design to measure the amount of the analytes in the biological matrix to be studied. To ensure that were constant background concentration of all endogenous analytes, bulk tissues were homogenized in 0.1% formic acid and used for validation and calibration data set (the background was subtracted to spiked concentration levels). The ratio (peak area of each analyte/peak area of internal standard), was calculated and used to ensure linearity of the method. The slope, intercept, and regression coe cient of those calibration lines were determined. Analytes amount were normalized to the tissue weight and plotted in GraphPad Prism 5 r. The statistical signi cance among groups was determined using a one-way ANOVA test followed by Tukey's test. Statistical dierences between control and treated groups were considered as signi cant if *P≤0.05 (**P≤0.01 and ***P≤0.001).
3.2.4 LC-MS/MS validation
In order to determine intraday (n=6) and interday (n=6 over a period of 8 separated days) precision and accuracy, analytes were spiked into liver tissue at 3 concentrations representing the entire range of concentration (LQC, low quality control, MQC medium quality control, HQC high quality control) and homogenized with extraction solvent. Precision was calculated from the relative standard deviation of the replicates (RSD), also known as coe cient of variation (CV), and accuracy was calculated by comparison of measured levels of spiked analytes with expected concentrations. A CV values of 20% was considered to be acceptable for accuracy and precision at the lower quality control. Recovery for the only non-endogenous analyte (MH) was calculated by comparison of the integrated area under the curve of the analyte at low, medium, and high concentration with 100% standards (neat solution). For all the other analytes, due to the endogenous nature of the eCBs, recovery was calculated by comparison of deuterated analogue (IS). Evaluation of ion suppression eects (matrix factor) was determined for non-endogenous analytes (MH) as the ratio between the peak area of analyte spike into matrix after extraction and the peak area of the analyte in neat solution. Values of 0.96 for LQC, 0.82 for MQC, 0.82 for HQC were obtained. For the analytes already present in blank matrix itself, evaluation was performed with the deuterated analogue. Matrix factor values were all closed to 1 (no matrix eects) except for LEA, PEA, OEA which values varied between 0.6-0.7 (ion suppression) indicating a signi cant matrix ionization eects in positive
mode eecting the ethanolamides. For the purpose of the study ethanolamides are minor analytes and the overall quality of the method, especially for endocannabinoids, resulted to be sensitive, precise and accurate. Examples of a set of validation data for an endogenous and non-endogenous analytes are showed in the next tables (Table 3.1 and Table 3.2 ).
INTERDAY INTRADAY LQC MQC HQC LQC MQC HQC Nominal concentration (ng/mL) 32 256 3200 32 256 3200 30.93 229.75 3176.28 34.83 239.98 2497.47 25.43 246.80 2768.59 34.50 253.16 3470.31 34.83 273.19 3382.83 30.95 257.13 2797.51 34.50 277.98 3207.69 34.29 228.42 2722.79 32.98 254.51 2993.86 35.78 227.75 2423.54 33.31 294.54 3292.70 31.97 234.82 2469.82 Mean 32.00 262.79 3136.99 33.72 240.21 2730.24 RSD (+/-) 3.50 23.51 222.46 1.85 12.47 392.01 C.V. (%) (Precision) 10.94 8.94 7.09 5.49 5.19 14.36 % Nominal conc. (Accuracy) 99.99 102.65 98.03 105.37 93.83 85.32
Matrix factor 96.20 81.71 81.20 Recovery 101.79 93.23 107.46
Table 3.1: Quality control data for 2AG
INTERDAY INTRADAY Nominal concentration (ng/mL) LQC MQC HQC LQC MQC HQC 32 256 3200 32 256 3200 9.10 59.44 828.23 10.53 57.30 739.50 8.23 65.45 739.87 8.78 69.72 764.86 8.64 70.16 886.43 6.71 73.12 811.64 8.02 66.75 827.75 10.98 65.27 894.20 9.87 60.23 835.22 12.17 54.18 750.38 7.73 64.45 835.22 10.25 64.13 766.11 Mean 8.60 64.41 825.45 9.90 63.95 787.78 RSD (+/-) 0.79 4.05 47.43 1.91 7.20 57.65 C.V. (%) (Precision) 9.19 6.28 5.75 19.29 11.25 7.32 % Nominal conc. (Accuracy) 107.45 100.65 103.18 123.80 99.93 98.47
Matrix factor 89.98 88.34 77.64 Recovery 101.09 94.33 92.28
3.3 CB2 receptor binding assay
For the CB2 binding experiment, 20 µg of Chinese Hamster Ovary (CHO) membranes overexpressing CB2 receptors were used. CHO-cell stably transfected with the hCB2 receptor are routinely cultured in the Gertsch's lab with the standard cell culture procedure (RPMI 1640 medium Sigma-Aldrich, Steinheim, Germany) supplemented with 10% fetal bovine serum (Invitrogen), 400μg/ml Geneticin (G418) (Invivo-gen), 1% penicillin-streptomycin (Sigma-Aldrich, Steinheim, Germany) at 37°C and 5% CO2. Membrane preparation is also a common procedure used in various assays in the lab [Chicca et al. 2016 in press]. The assay buer was made by 50mM Tris-HCl, 2.5mM ethylenediaminetetraacetic acid (EDTA), 5mM MgCl2 and 0.5 % BSA and was prepared at pH 7,5. Membranes were diluted in the assay buer and incubated for 1.5 hours at room temperature with 0.5 nM [3H]-CP55,940 (PerkinElmer Life Sciences, Waltham,
MA, USA) in the presence or absence of dierent concentration of MH using DMSO as control and (R)-WIN55,212-2 (Cayman Chemicals, Ann Arboe, MI, USA), to determine non-speci c binding. After the incubation time, membranes were ltered through a GF/B lter which was previously presoaked for 1.5 hours at room temperature with 0.1% of polyethyleneimine (PEI). Then the GF/B lter was washed twelve times with 150µL of the assay buer and dried. Finally, the lter was incubated for 20 minutes at room temperature with 45µL of the scintillation liquid MicroScint 20 (PerkinElmer Life Sciences, Waltaham, MA, USA) and radioactive signal (cpm) was read using a Top-counter Tri-carb 2100 TR (PerkinElmer Life Sciences, Waltaham, MA, USA).The concentration-dependent binding curve was measured for MH and also for WIN55,212-2 as a control. Results were analyzed using GraphPad Prism 5 ® and expressed as bound of [3H]-CP55,940 (% vehicle). Ki values were calculated using the Cheng-Pruso equation: Ki
Chapter 4
Results and discussion
4.1 LC-MS/MS results
4.1.1 Kidney
MH (Figure 4.1) at the doses of 10mg/kg and 20mg/kg induced a signi cant increase of AEA and 2AG levels in kidney as compared to vehicle control (Figure 4.2a and 4.2b). Intriguingly, the minor endo-cannabinoid noladin ether which was not detected in vehicle-treated animals and at the non-eective dose of 3mg/kg of MH, became detectable upon treatment with MH at the doses of 10mg/kg and 20mg/kg (Figure 4.2c). The increase of AEA, 2AG and NE could be due to the eect of MH on COX-2-mediated endocannabinoids oxygenation. On the contrary, AA and prostaglandin E2 (PGE2) levels were not af-fected by MH (Figure 4.3a e 4.3b). Prostaglandin D2 (PGD2) was below limit of detection. If MH had exerted non-selective inhibition of COX-2 activity (meaning that also the catalytic subunit responsable for the metabolism of AA was inhibited), a signi cant decrease of PGE2 concentration and a consequent increase of AA level should have been observed. Unfortunately, PG-GEs and PG-EAs were below the limit of detection.
Altogether this data strongly indicated that MH acts as a COX-2 SSI in vivo.
We also measured the levels of the other N-acetylethanolamines (NAEs) and hormones. The levels of PEA and OEA did not change while LEA showed an increase similarly to AEA (Figure 4.4). These data are in line with the lack of FAAH inhibition for MH [49]. Beyond FAAH hydrolysis, LEA can be metabolized by lypooxygenases and MH was shown to inhibit lypooxygenase-5 [50]. This could explain the speci c increase of LEA and not the other NAEs. Corticosterone and Progesterone levels were not aected by MH treatment (Figure 4.5).
Figure 4.1: Kidney: MH
(a) (b) (c)
Figure 4.2: Kidney: AEA (a), 2AG (b), NE (c)
(a) (b)
(a) (b) (c)
Figure 4.4: Kidney: PEA (a), OEA (b), LEA (c)
(a) (b)
4.1.2 Brain
The ability of MH to pass the blood brain barrier (BBB) in normal conditions was con rmed in this experiment (Figure 4.6a) in addition to the data showing that upon endotoxemia (LPS-induced), MH can reach bioactive concentrations in the brain ([49]). The concentration of MH was in the same range as a previous experiment made by Chicca et al. (Figure 4.6b). 2AG levels were not aected by MH treatment, while a decrease of AEA was observed (Figure 4.7). In the results shown by Chicca et al , a similar behavior of AEA was measured but the levels of 2AG increased (Figure 4.8). This dierence could be explained with the LPS challenge which induce a systemic in ammation. In this condition, COX-2-mediated oxygenation seems to be more relevant for 2AG metabolism, thus leading to a more pronounced pharmacological eect induced by MH. NE was not detectable in the brain, in accordance with literature. A signi cance decrease of AA levels could be seen with 20mg/kg of MH whereas PGE2 and PGD2 levels were not aected by the treatment (Figure 4.9).
In the brain PEA and OEA showed a similar decrease as AEA whereas LEA was only marginally aected (Figure 4.10). No signi cant changes could be found in Corticosterone and Progesterone levels (4.11).
(a) MH (2016) (b) MH adapted from Chicca et al Journal of Neuroin ammation
(a) (b)
Figure 4.7: Brain: AEA(a), 2AG(b)
(a) (b) (c)
Figure 4.9: Brain: AA(a), PGE2(b), PGD2(c)
(a) (b) (c)
Figure 4.10: Brain: PEA(a), OEA(b), LEA(c)
(a) (b)
4.1.3 Liver
Noticeable, like in the kidney, 2AG and AEA levels signi cantly increased at 10mg/kg and 20mg/kg of MH (Figure 4.12, 4.13a and 4.13b). A MH eect could not be concluded with these values. Noladin-ether showed a tendency to increase similar to AEA and 2AG. In particular, it was mostly non-detectable in the control animals as well as mice trated with 3mg/kg of MH. In line with the data obtained in kidney, NE levels became robustly detectable upon treatment with MH at 10 and 20 mg/kg (Figure 4.13c).
In the liver, a signi cant increase of AA levels could be noticed at 10mg/kg of MH with a tendency to increase at 20mg/kg of MH (Figure 4.14a). PGE2 level was below limit of quanti cation while PGD2 levels only marginally detectable without changes induced by the treatments (Figure 4.14b).
Here, like in the kidney, when 2AG and AEA increased, a signi cant increase of LEA respect to control was observed, while PEA and OEA levels did not change (Figure 4.15). No signi cant changes could be found in Corticosterone and Progesterone (Figure 4.16).
Figure 4.12: Liver: MH
(a) (b) (c)
(a) (b)
Figure 4.14: Liver: AA(a), PGD2(b)
(a) (b) (c)
Figure 4.15: Liver: PEA(a), OEA(b), LEA(c)
(a) (b)
4.1.4 Lung
MH was quanti able also in the lung and values are showed in Figure 4.17. The levels of AEA and 2AG did not signi cant change compared to control despite a tendency to increase upon treatment with MH at 10 and 20mg/kg (Figure 4.18). NE was not detectable.
MH treatment did not aect the levels of AA, while it strongly reduced PGE2 and PGD2 levels at all three tested doses (Figure 4.19). Our data do not indicate whether this eect was mediated via COX-2 inhibition because AA, 2AG and AEA levels were not aected. Furthermore, the decrease of both PGs was visible also at 3mg/kg of MH treatment while in the other organs, this dose did not show any signicant eects as compared to the control.
PEA and OEA levels did not signi cantly change upon MH treatment, while LEA increased (Figure 4.20). The latter was not associated with an increase of AEA as observed in liver and kidney. Corticos-terone and ProgesCorticos-terone levels were not aected by MH treatment as compared to control also if a large variability, especially in the Progesterone level, was observed (Figure 4.21).
(a) (b)
Figure 4.18: Lung: AEA(a), 2AG(b)
(a) (b) (c)
(a) (b) (c)
Figure 4.20: Lung: PEA(a), OEA(b), LEA(c)
(a) (b)
4.1.5 Spleen
The highest levels of MH were found in the spleen (around 5-10 times higher than the other tissues) (Figure 4.22). Due to the elevate concentration of immune cells which express CB2 receptors, it is possible that MH reaches the spleen bound to the receptors rather than distribute in the tissue stroma. In fact, AEA levels did not change while 2AG exhibited only a tendency to increase at the higher doses (Figure 4.23). NE was not detectable.
The AA level in the spleen was not aected, while concentration-dependent increase of both PGE2 and PGD2 levels was observed (Figure 4.24). This behavior seems to be in contrast with the COX-2 SSI eect, where PG levels should not be aected and with the activation of CB2 receptors which would inhibit COX-2 expression. In addition, the PGs increase was not associated with a change in the AA level. The mechanism of these eects should be further investigates to understand the possible roles of CB2 and/or COX-2.
PEA, OEA and LEA levels did not signi cantly change and this eect was coupled with no change in endocannabinoid levels, like in the brain (Figure 4.25). Corticosterone levels seemed to be not aected by MH, despite some variability in 10mg/kg and 20mg/kg values (Figure 4.26). Progesterone was not detectable.
(a) (b) (c)
Figure 4.24: Spleen: AA(a), PGE2(b), PGD2(c)
(a) (b)
(a) (b) (c)
Figure 4.25: Spleen: PEA(a), OEA(b), LEA(c)
4.2 CB2 receptor binding assay
In order to con rm the binding activity of MH to a CB2 receptor, some binding assays were repeated. The Ki value was calculated using the Cheng-Pruso equation (Ki = IC50/ 1 + [L]/Kd) and resulted in 660nM (95% CI, 549-792) for MH and 2.3nM (95% CI, 1.8-2.9) for (R)-WIN55,212-2. The Ki value obtained for MH is comparable with the a nity reported in the literature and previously measured in the group of Prof. J.Gertsch [49] (Figure 4.27).
Figure 4.27: CB2 receptor binding for MH and (R)-WIN55,212-2
4.3 Overview
MH acts as a CB2 receptor agonist and as a COX-2 SSI. In order to understand whether some of the eects showed in the experiment were due to one or both targets, additional experiments will be performed. For example, one experiment should be performed in mice treated with MH, a CB2 agonist, a COX-2 inhibitor and a combination of both. The comparison of the dierent treatments, could provide compelling evidence to better understand the mechanisms of action of MH.
Bibliography
[1] W A Devane, F A Dysarz, M R Johnson, L S Melvin, and A C Howlett. Determination and characterization of a cannabinoid receptor in rat brain. Molecular pharmacology, 34(5):60513, nov 1988.
[2] S Munro, K L Thomas, and M Abu-Shaar. Molecular characterization of a peripheral receptor for cannabinoids. Nature, 365(6441):615, sep 1993.
[3] Natalia Battista, Monia Di Tommaso, Monica Bari, and Mauro Maccarrone. The endocannabinoid system : an overview. 6(March):17, 2012.
[4] Vincenzo Di Marzo, Nephi Stella, and Andreas Zimmer. Endocannabinoid signalling and the deteri-orating brain. Nature Publishing Group, 16(1):3042, 2015.
[5] V Di Marzo, A Fontana, H Cadas, S Schinelli, G Cimino, J C Schwartz, and D Piomelli. Formation and inactivation of endogenous cannabinoid anandamide in central neurons. Nature, 372(6507):686 91, dec 1994.
[6] Daniele Piomelli. More surprises lying ahead. The endocannabinoids keep us guessing. Neurophar-macology, 76 Pt B:22834, jan 2014.
[7] Bradley E Alger. Retrograde signaling in the regulation of synaptic transmission : focus on endo-cannabinoids, volume 68. 2002.
[8] Daniele Piomelli. THE MOLECULAR LOGIC OF ENDOCANNABINOID SIGNALLING. 4(Novem-ber), 2003.
[9] W A Devane, L Hanus, A Breuer, R G Pertwee, L A Stevenson, G Gri n, D Gibson, A Mandelbaum, A Etinger, and R Mechoulam. Isolation and structure of a brain constituent that binds to the cannabinoid receptor. Science (New York, N.Y.), 258(5090):19469, dec 1992.
[10] R G Pertwee, A C Howlett, M E Abood, S P H Alexander, V Di Marzo, M R Elphick, P J Greasley, H S Hansen, and G Kunos. International Union of Basic and Clinical Pharmacology . LXXIX . Cannabinoid Receptors and Their Ligands : Beyond CB 1 and CB 2. 62(4):588631, 2010.
[11] Vincenzo Di Marzo. perspective Endocannabinoid signaling in the brain : biosynthetic mechanisms in the limelight. Nature Publishing Group, 14(1):915, 2011.
[12] R Mechoulam, S Ben-Shabat, L Hanus, M Ligumsky, N E Kaminski, A R Schatz, A Gopher, S Almog, B R Martin, and D R Compton. Identi cation of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. Biochemical pharmacology, 50(1):8390, jun 1995.
[13] T Sugiura, S Kondo, A Sukagawa, S Nakane, A Shinoda, K Itoh, A Yamashita, and K Waku. 2-Arachidonoylglycerol: a possible endogenous cannabinoid receptor ligand in brain. Biochemical and biophysical research communications, 215(1):8997, oct 1995.
[14] Tiziana Bisogno, Fiona Howell, Gareth Williams, Alberto Minassi, Maria Grazia Cascio, Alessia Ligresti, Isabel Matias, Aniello Schiano-moriello, Praveen Paul, Emma-jane Williams, Uma Gangad-haran, Carl Hobbs, Vincenzo Di Marzo, and Patrick Doherty. Cloning of the rst sn1-DAG lipases points to the spatial and temporal regulation of endocannabinoid signaling in the brain The Journal of Cell Biology. 163(3):463468, 2003.
[15] Giulio G Muccioli. Endocannabinoid biosynthesis and inactivation , from simple to complex. Drug Discovery Today, 15(11-12):474483, 2010.
[16] Andrea Chicca, Janine Marazzi, Simon Nicolussi, and Jürg Gertsch. Evidence for bidirectional endo-cannabinoid transport across cell membranes. The Journal of biological chemistry, 287(41):3466082, oct 2012.
[17] Fabio Arturo Iannotti, Vincenzo Di Marzo, and Stefania Petrosino. Endocannabinoids and endocannabinoid-related mediators: Targets, metabolism and role in neurological disorders. Progress in lipid research, 62:10728, apr 2016.
[18] Jacqueline L Blankman, Gabriel M Simon, and Benjamin F Cravatt. NIH Public Access. 14(12):1347 1356, 2009.
[19] Daniele Piomelli, Giorgio Tarzia, Andrea Duranti, Andrea Tontini, Marco Mor, Timothy R Compton, Olivier Dasse, Edward P Monaghan, Je A Parrott, and David Putman. Pharmacological Pro le of the Selective FAAH Inhibitor KDS-4103 ( URB597 ). 12(1):2138, 2006.
[20] Jonathan Z Long, Weiwei Li, Lamont Booker, James J Burston, Steven G Kinsey, Joel E Schlosburg, Franciso J Pavón, Antonia M Serrano, Dana E Selley, Loren H Parsons, Aron H Lichtman, and Benjamin F Cravatt. Selective blockade of 2-arachidonoylglycerol hydrolysis produces cannabinoid behavioral eects. Nature chemical biology, 5(1):3744, jan 2009.
[21] A M Y C Porter, John-michael Sauer, Michael D Knierman, Gerald W Becker, Michael J Berna, Jingqi Bao, George G Nomikos, Petra Carter, Frank P Bymaster, Andrea Baker Leese, Christian C Felder, Neuroscience Division A C P, and Drug Disposition J S. Characterization of a Novel Endocannabinoid , Virodhamine , with Antagonist Activity at the CB1 Receptor. 301(3):10201024, 2002.
[22] Filomena Fezza, Tiziana Bisogno, Alberto Minassi, Giovanni Appendino, Raphael Mechoulam, and Vincenzo Di Marzo. Noladin ether, a putative novel endocannabinoid: inactivation mechanisms and a sensitive method for its quanti cation in rat tissues. FEBS letters, 513(2-3):2948, feb 2002.
[23] Susan M Huang, J Michael Walker, M Susan, and J Michael Walker. Enhancement of Spontaneous and Heat-Evoked Activity in Spinal Nociceptive Neurons by the Endovanilloid / Endocannabinoid N -Arachidonoyldopamine. pages 12071212, 2006.
[24] Maria Grazia Cascio. PUFA-derived endocannabinoids: an overview. The Proceedings of the Nutrition Society, 72(4):4519, nov 2013.
[25] Barbara Costa, Francesca Comelli, Isabella Bettoni, Mariapia Colleoni, and Gabriella Giagnoni. The endogenous fatty acid amide , palmitoylethanolamide , has anti-allodynic and anti-hyperalgesic eects in a murine model of neuropathic pain : involvement of CB 1 , TRPV1 and PPAR c receptors and neurotrophic factors. Pain, 139(3):541550, 2008.
[26] S Hjorth, N-o Hermansson, J Leonova, T Elebring, E Ryberg, N Larsson, S Sjo, K Nilsson, T Drmota, and P J Greasley. The orphan receptor GPR55 is a novel cannabinoid receptor. (September):1092 1101, 2007.
[27] Miki Igarashi, Nicholas V DiPatrizio, Vidya Narayanaswami, and Daniele Piomelli. Feeding-induced oleoylethanolamide mobilization is disrupted in the gut of diet-induced obese rodents. Biochimica et biophysica acta, 1851(9):121826, sep 2015.
[28] Andrea Sarro-Ramírez, Daniel Sánchez-López, Alma Tejeda-Padrón, Carmen Frías, Jaime Zaldívar-Rae, and Eric Murillo-Rodríguez. Brain molecules and appetite: the case of oleoylethanolamide. Central nervous system agents in medicinal chemistry, 13(1):8891, mar 2013.
[29] Vincenzo Di Marzo, Maurizio Bifulco, and Luciano De Petrocellis. The endocannabinoid system and its therapeutic exploitation. Nature Reviews Drug Discovery, 3(9):771784, sep 2004.
[30] Miles Herkenham, Allison B Lynn, Mark D Litrle, M Ross Johnsont, Lawrence S Melvin, Brian R D E Costa, and Kenner C Riceo. Cannabinoid receptor localization in brain. 87(March):19321936, 1990.
[31] Nephi Stella. Cannabinoid and cannabinoid-like receptors in microglia, astrocytes, and astrocytomas. Glia, 58(9):101730, jul 2010.
[32] A C Howlett, F Barth, T I Bonner, G Cabral, P Casellas, W A Devane, C C Felder, M Herkenham, K Mackie, B R Martin, R Mechoulam, and R G Pertwee. International Union of Pharmacology . XXVII . Classi cation of Cannabinoid Receptors. 54(2):161202, 2002.
[33] Sylvaine Galgglje, Sophie Mary, Jean Marchand, Danielle Dussossoy, Dominique Carrikre, Pierre Camyon, Monsif Bouaboula, David Shirez, Gcrard L E Fur, and Pierre Casellas. Expression of central and peripheral cannabinoid receptors in human immune tissues and leukocyte subpopulations. 61:5461, 1995.
[34] I Mancini, R Brusa, G Quadrato, C Foglia, P Scandroglio, L S Silverman, and D Tulshian. THEMED ISSUE : GPCR Constitutive activity of cannabinoid-2 ( CB 2 ) receptors plays an essential role in the protean agonism of ( + ) AM1241 and L768242 Abbreviations :. (December 2008):382391, 2009.
[35] Francesca Aiello, Gabriele Carullo, Mariateresa Badolato, and Antonella Brizzi. TRPV1-FAAH-COX: The Couples Game in Pain Treatment. ChemMedChem, 11(16):168694, aug 2016.
[36] William L Smith, David L Dewitt, and R Michael Garavito. C YCLOOXYGENASES : Structural , Cellular , and Molecular Biology. 2000.
[37] Garret A FitzGerald. COX-2 and beyond: Approaches to prostaglandin inhibition in human disease. Nature reviews. Drug discovery, 2(11):87990, nov 2003.
[38] David L Dewitt and William L Smith. Primary structure of prostaglandin G / H synthase from sheep vesicular gland determined from the complementary DNA sequence. 85(March):14121416, 1988.
[39] D A Kujubu, B S Fletcher, B C Varnum, R W Lim, and H R Herschman. TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin syn-thase/cyclooxygenase homologue. The Journal of biological chemistry, 266(20):1286672, jul 1991.
[40] L Daniel. Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. 88(April):26922696, 1991.
[41] J B C Papers, M Doi, Kevin R Kozak, Scott W Rowlinson, and Lawrence J Marnett. Oxygenation of the Endocannabinoid , 2-Arachidonylglycerol , to Glyceryl Prostaglandins by Cyclooxygenase-2 *. 275(43):3374433749, 2000.
[42] Daniel J Hermanson, Nolan D Hartley, Joyonna Gamble-George, Naoko Brown, Brian C Shonesy, Phillip J Kingsley, Roger J Colbran, Jerey Reese, Lawrence J Marnett, and Sachin Patel. Substrate-selective COX-2 inhibition decreases anxiety via endocannabinoid activation. Nature Neuroscience, 16(9):12911298, aug 2013.
[43] Fabiana Piscitelli and Vincenzo Di Marzo. "Redundancy" of endocannabinoid inactivation: new challenges and opportunities for pain control. ACS chemical neuroscience, 3(5):35663, may 2012.
[44] Kevin R Kozak, Brenda C Crews, Jennifer L Ray, Hsin-hsiung Tai, Jason D Morrow, and Lawrence J Marnett. Metabolism of Prostaglandin Glycerol Esters and Prostaglandin Ethanolamides in Vitro and in Vivo *. 276(40):3699336998, 2001.
[45] O Sagredo, V Di Marzo, S Valdeolivas, T Bisogno, F Piscitelli, F A Iannotti, and M Allara. The inhi-bition of 2-arachidonoyl-glycerol ( 2-AG ) biosynthesis , rather than enhancing striatal damage , pro-tects striatal neurons from malonate-induced death : a potential role of cyclooxygenase-2-dependent metabolism of 2-AG. 2013.
[46] Kelsey C Duggan, Daniel J Hermanson, Joel Musee, Jeery J Prusakiewicz, Jami L Scheib, Bruce D Carter, Surajit Banerjee, J A Oates, and Lawrence J Marnett. (R)-Profens are substrate-selective inhibitors of endocannabinoid oxygenation by COX-2. Nature chemical biology, 7(11):8039, nov 2011.
[47] Wolfgang Schuehly, Juan Manuel Viveros Paredes, Jonas Kleyer, Antje Huefner, Sharon Anavi-Goer, Stefan Raduner, Karl-Heinz Altmann, and Jürg Gertsch. Mechanisms of osteoclastogenesis inhibition by a novel class of biphenyl-type cannabinoid CB(2) receptor inverse agonists. Chemistry & biology, 18(8):105364, aug 2011.
[48] Yong Kyung, Lee Æ Dong, Yeon Yuk, Æ Tae Il, Beom Jun, Lee Æ Sang-yoon Nam, and Æ Jin Tae. Protective eect of the ethanol extract of Magnolia o cinalis and 4- O -methylhonokiol on scopolamine-induced memory impairment and the inhibition of acetylcholinesterase activity. pages 274282, 2009.
[49] Andrea Chicca, Maria Salomé Gachet, Vanessa Petrucci, Wolfgang Schuehly, Roch-Philippe Charles, and Jürg Gertsch. 4'-O-methylhonokiol increases levels of 2-arachidonoyl glycerol in mouse brain via selective inhibition of its COX-2-mediated oxygenation. Journal of neuroin ammation, 12:89, may 2015.
[50] Wolfgang Schühly, Antje Hüfner, Eva M Pferschy-wenzig, Elke Prettner, Michael Adams, Antje Bo-densieck, Olaf Kunert, Asije Oluwemimo, Ernst Haslinger, and Rudolf Bauer. Bioorganic & Medicinal Chemistry Design and synthesis of ten biphenyl-neolignan derivatives and their in vitro inhibitory potency against cyclooxygenase-1 / 2 activity and 5-lipoxygenase-mediated LTB 4 -formation. Bioor-ganic & Medicinal Chemistry, 17(13):44594465, 2009.
[51] Young-Jung Lee, Dong-Young Choi, Im Seop Choi, Ki Ho Kim, Young Hee Kim, Hwan Mook Kim, Kiho Lee, Won Gil Cho, Jea Kyung Jung, Sang Bae Han, Jin-Yi Han, Sang-Yoon Nam, Young Won Yun, Jae Hwang Jeong, Ki-Wan Oh, and Jin Tae Hong. Inhibitory eect of 4-O-methylhonokiol on lipopolysaccharide-induced neuroin ammation, amyloidogenesis and memory impairment via in-hibition of nuclear factor-kappaB in vitro and in vivo models. Journal of neuroin ammation, 9:35, 2012.
[52] Yong Kyung Lee, Im Seop Choi, Jung Ok Ban, Hwa Jeong Lee, Ung Soo Lee, Sang Bae Han, Jae Kyung Jung, Young Hee Kim, Ki Ho Kim, Ki Wan Oh, and Jin Tae Hong. 4-O-methylhonokiol attenuated ??-amyloid-induced memory impairment through reduction of oxidative damages via in-activation of p38 MAP kinase. Journal of Nutritional Biochemistry, 22(5):476486, 2011.
[53] Huishan Han and Jae Kyung Jung. Anxiolytic-Like Eects of 4-O-Methylhonokiol Isolated from Magnolia o cinalis Through Enhancement of GABAergic Transmission and Chloride In ux. 14:724 731, 2011.
[54] Hyung Eun Yu, Soo Jin Oh, Je Kyung Ryu, Jong Soon Kang, Jin Tae Hong, Jae-kyung Jung, Sang-bae Han, Seung-yong Seo, Young Heui Kim, Song-kyu Park, Hwan Mook Kim, and Kiho Lee. Pharmacokinetics and Metabolism of 4- O -Methylhonokiol in Rats. 578(February 2013):568578, 2014.
[55] Matthew D Breyer and Raymond C Harris. Cyclooxygenase 2 and the kidney. 2001.
[56] Philip J Kingsley and Lawrence J Marnett. Analysis of endocannabinoids by Ag+ coordination tandem mass spectrometry. 314:815, 2003.
