Cannabinomimetric Lipids: From Natural Extract to Artificial Synthesis

Endocannabinoid system is related with various physiological and cognitive processes including fertility, pregnancy, during pre- and postnatal development, pain-sensation, mood, appetite, and memory. In the latest decades, an important milestone concerning the endocannabinoid system was the discovery of the existence of the cannabinoid receptors CB1 and CB2. Anandamide was the first reported endogenous metabolite, which adjusted the release of some neurotransmitters through binding to the CB1 or CB2 receptors. Then a series of cannabinomimetric lipids were extracted from marine organisms, which possessed similar structure with anandamide. This review will provide a short account about cannabinomimetric lipids for their extraction and synthesis.


Introduction
In the past decades, pharmacologists devoted more interest to the study of endocannabinoid system due to its relation with various physiological and cognitive processes including fertility, pregnancy, during pre-and postnatal development, pain-sensation, mood, appetite, and memory [1]. Originally, the endocannabinoid system was discovered while scientists tried to understand the physical and psychological effects of cannabis, thereby named it the endocannabinoid system for this reason. An important milestone concerning the endocannabinoid system was the discovery of the existence of the cannabinoid receptors (CB 1 and CB 2 ) in central and peripheral mammalian tissues [2][3][4][5]. Both receptors CB 1 and CB 2 belong to the large family of G-protein-coupled receptors (GPCR). CB 1 receptor exhibits a widespread distribution in the mammalian brain and are responsible for the psychological and anticonvulsive effects produced by marijuana [2][3][4][5], while CB 2 receptor is most abundant in the immune and hematopoietic system and is involved in the anti-inflammatory and possibly other therapeutic effects of cannabis [4,5]. The discovery of cannabinoid receptors (CB 1 and CB 2 ) has launched the quest for endogenous ligands of these receptors. Based on the assumption that the endogenous cannabinoid ligand was a lipid soluble compound, a lipid derivative was first isolated from chloroform-methanol extracts of porcine brain and christened anandamide by Mechoulam et al. in 1992 (Fig. 1) [6]. This endogenous metabolite bound to both CB 1 and CB 2 receptors and was found in nearly all tissues in a wide range of animals [7]. Then a series of alkyl amides were extracted from marine organisms, which resembled structurally some aspects of anandamide and had been termed cannabimimetic lipids. In the biological activity tests, they showed the ability to bind and activate at least one cannabinoid receptor [8]. This review will provide a short account of cannabinomimetric lipids for their natural extract and artificial synthesis.

Grenadamide
Grenadamide was isolated from the organic extract of a Grenada collection of the marine cyanobacterium Lyngbya majuscula by Gerwick et al. in 1998. It exhibited brine shrimp toxicity (LD 50 = 5 lg/mL) and modest cannabinoid receptor binding activity (K i = 4.7 lM) [9]. Gerwick et al. verified the structure and the relative stereochemistry of grenadamide, which was a trans-cyclopropyl-containing fatty acid-derived metabolite.

Mooreamide A
Mooreamide A was extracted from cyanobacterium Moorea bouillonii by Gerwick et al. from Papua New Guinea, and showed strong and selective affinity to CB 1 ligand [10].

Serinolamides
Serinolamide A was isolated from marine cyanobacteria Lyngbya majuscule collected in Papua New Guineal. It displayed a moderate agonist effect and selectivity for the CB 1 cannabinoid receptor [11]. Serinolamide B, a closely related analogue of serinolamide A, was isolated from a Lyngbya sample from the piti Bomb Holes in Guam by Luesch et al. [12]. It showed moderate affinities to both CB 1 and CB 2 , while exhibited a higher selectivity for CB 2 (K i = 5.2 lm) over CB 1 (K i = 16.4 lm)

Semiplenamides
Semiplenamides A to G were isolated from the marine cyanobacterium Lyngbya semiplena collected from Papua New Guinea by Gerwick et al. [13]. In the test of their affinity to cannabinoid receptors of the rat brain membranes, only semiplenamides A, B and G worked. In the test of fatty acid amide hydrolase (FAAH), the semiplenamides A to G were not found appreciable inhibitory effect.

Malyngamides
Malyngamides include over 30 members, characterized by different N-substitution groups of amides. They were isolated from Marine cyanobacterium Lyngbya majuscule, and showed a wide range of biological activities, such as antifeedant activity, ichthyotoxicity, toxicity to other marine animals, cytotoxicity to cancer cells, anti-HIV, antileukemic, and anti-tumor activity [14][15][16]. The finding that malyngamide B possesses cannabimimetic properties provides new insight into the biological activities of malyngamides. The extraction information of malyngamides was illustrated in Table 1.
Hermitamides resemble the malyngamide-type compound in structure and were still isolated from the marine marine cyanobacterium L. majuscula of other species of Gracilaria [17][18][19]. Hermitamides were evaluated for their biological activity in several systems. Hermitamides A (1) and B (2) showed LD 50 values of 5 lM and 18 lM respectively in the brine shrimp (Artemia salina) toxicity assay, and showed IC 50 values of 2.2 lM and 5.5 lM respectively to Neuro-2a neuroblastoma cells in tissue culture.

Synthesis of Grenadamide
In 2004, Baird and co-workers reported the synthesis of grenadamide and confirmed its absolute stereochemistry (Scheme 1) [9]. The synthesis started from the aldehyde 1, which was converted to olefin 2 through Wittig reaction the following ester hydrolysis. Then removal of the double bond gave 3. Oxidation of alcohol 3 got aldehyde 4 and epimerisation of 4 using sodium methoxide in methanol afforded the epimer 5. Then 5 underwent HWE reaction with ethoxycarbonyl triphenylphosphosphorane to give the ester 6, which was removed the double bond with dipotassium azodicarboxylate and hydrolysed with KOH to afford acid 7. The compound 7 was converted into the corresponding chloride, then treated with 2-phenylethylamine to give the amide 8, which had an equal and opposite absolute rotation compared with natural grenadamide. So the synthetic sample was the enantiomer of the natural product grenadamide.
One year later, Bull and co-workers reported an asymmetric synthesis of grenadamide in 6 steps using (R)-5,5dimethyl-oxa-zolidin-2-one as a chiral auxiliary (Scheme 2) [46]. The starting material 9 as a chiral auxiliary was acetylated with chloroacetyl chloride to give 10. Then treatment of 10 with 9-BBN-OTf and i-Pr 2 NEt and following reaction with a,b-unsaturated aldehyde afforded syn-aldol product 11 in 92% de. Cyclopropanation of 11 with Et 2 Zn and CH 2 I 2 afforded 12 with high stereoselectivity. Then replacing the oxazolidin-2-one fragment with phenylethylamine gave 13, which was treated with SmI 2 resulted in clean elimination reactions to afford (E)-a,bunsaturated amide. Finally reduction of the double bond with NaBH 4 and CoCl 2 afford grenadamide.
To the further study, the author obtained the two enantiomers of grenadamide. Fatty acid 22 was coupled with Evans' auxiliary and chromatographically separating the diastereomers 23 and 24, which were removed of the auxiliary to give (-)-22 and (?)-22 respectively. Then the enantiomerically pure fatty acids were subjected to the Arndt-Eistert protocol to give (-)-grenadamide and (?)grenadamide.
In 2007, Piva and co-workers reported the synthesis of racemic grenadamide through a sequential cross-metathesis/Simmons-Smith cyclopropanation (Scheme 4) [48]. Cross-metathesis of 26 with 1-nonene 25 catalyzed by Grubbs type catalyst 27 delivered 28 as mixture of E and Z isomers. Then cyclopropanation of the mixture of 28 afforded grenadamide.
In 2010, Boysen and co-workers reported an asymmetric synthesis of (?)-grenadamide, an enantiomer of the natural product (-)-grenadamide (Scheme 5) [49]. The cyclopropyl carboxylic ester 29 was transformed into the corresponding aldehyde 30 by reduction with lithium aluminium hydride to alcohol, followed by Swern oxidation. Then aldehyde 30 underwent Wittig olefination to give a,b-unsaturated ester 31. Reduction of 31 and followed by hydrolysis afforded acid 32, which was coupled with phenethylamine gave (?)-grenadamide.

The Synthesis of Serinolamide A
In 2013, our group reported the first total synthesis of (?)serinolamide A in nine steps from L-serine with 30% overall yield (Scheme 6) [50]. The synthesis of (?)-

Synthesis of Semiplenamides
In 2005, Bull and co-workers developed an efficient method for the synthesis of (E)-a,b-unsaturated amide and applied the methodology for the synthesis of semiplenamide C (Scheme 8) [52]. L-Alanine methyl ester 49 was chosen as the starting material. Reduction of 49 with LiAlH 4 and then protection with diethyl carbonate afford 50. Subsequent treatment of 50 with n-BuLi and propionyl chloride gave 51. Then pretreatment of 51 with 9-BBN-OTf and i-Pr 2 NEt underwent an aldol reaction with tetradecanal to give 52 in [ 95% de. Finally, deprotection of 52 with KO t Bu afforded semiplenamide C. In 2009, Das and co-workers reported the synthesis of semiplenamides C and E through the Baylis-Hillman adducts [53]. (Scheme 9) The Baylis-Hillman adducts 53 and 54 were treated with PPh 3 /CBr 4 afforded the corresponding allyl bromides 55 and 56, which were subsequently treated with Zn and CH 3 COOH to give 57, 58 respectively. The esters 57 and 58 were then hydrolyzed with KOH to give the corresponding acids 59 and 60, which were condensed with (S)-alaninol to form semiplenamide C (61) and 62 respectively. Compound 62 was further acetylated with acetic anhydride to furnish semiplenamide E.

Synthesis of Malyngamides
In 2006, Piva and co-workers reported both the racemic synthesis and a formal enantioselective synthesis of hermitamides A and B (Scheme 10) [54]. Homoallylic ether 64 was prepared through the Grignard reaction from the octanal 63 and allyl magnesium bromide and the following protection of the hydroxyl group by MeI in the presence of NaH. Then 64 and butenoic acid proceeded RCM reaction to afford 65 with the E/Z ratio of 95/5. Finally, racemic hermitamide A and B were synthesized through the condensation of 65 with the phenethylamine and 3-indolyl- In 2011, Paige and co-workers reported an asymmetric synthesis of hermitamides A and B (Scheme 12) [56].
Asymmetric allylation of octanal with allyltributylstannane mediated by a titanium-binol complex gave homoallylic alcohol followed by the methylation with MeI to afford ether 77. Oxidative cleavage of the terminal double bond yielded aldehyde 78, which was reacted with vinylmagnesiun bromide to afford allylic alcohol 79. The compound 79 was then generated Johnson-Claisen rearrangement with the addition of trimethylorthoacetate in the presence of catalytic amount of propionic acid to afford methyl ester of lyngbic acid followed by the saponification with lithium hydroxide to give 65. Then acid 65 was coupled with phenethylamine or tryptamine to afford hermitamides A and B respectively.

Scheme 13 Synthesis of hermitamides A and B reported by Narender [57]
Cannabinomimetric Lipids: From Natural Extract to Artificial Synthesis 11 In 2014, Narender and co-workers reported a concise total synthesis of hermitamides A and B in a high enantioselectivity (Scheme 13) [57]. The synthesis was commenced from octanal 63. Vinylation with vinyl magnesium bromide gave allylic alcohol 80, which was oxidated by IBX to afford enone 81. Then asymmetric reduction of ketone 81 catalyzed by CBS gave the chiral allylic alcohol with high enantioselectivity, which was treated with Meerwein's reagent to afford the methylation product (S)-82. Subsequently, hydroboration-oxidation of 82 afforded the primary alcohol 83. Then 83 with 1-phenyltetrazole-5thiol (84) underwent Mitsunobu reaction to give aryl sulfide 85, which was further oxidated by ammonium molybdate and hydrogen peroxide to give sulfone 86. Coupling of alkyl sulfone 86 with aldehyde 87 via Julia-Lythgoe olefin provided the corresponding olefin 88 with E-geometry exclusively. Then removal of benzyl obtained the primary alcohol following oxidation of the hydroxyl group to furnish lyngbic acid 65. Acid 65 was coupled with 2-phenylethyl amine or 3-indolylethylamine to provide hermitamides A and B, respectively.
In 2006, Cao and co-workers reported the synthesis of serinol-malyngamide for the first time (Scheme 14) [58]. The molecular was divided into two parts, the fatty acid 97 and the derivative of serinol 100. The stereoselective synthesis of fatty acid 97 had been reported previous [59], which was starting from 1-tetradecanol. Oxidation of 89 afforded the aldehyde 90, which further reacted with allyltributyltin catalyzed by bis-(R)-Ti(IV) oxide (91) to afford allyl alcohol 92. Methylation of hydroxyl group and followed by oxidation of terminal olefin afforded aldehyde, subsequently reacted with PPh 3 /CBr 4 to produce dibromide 93. Then alkyne 94 was obtained after the addition of n-BuLi. Alkyne coupled with 95 gave 96, reduction of which obtained olefin with E-configuration. Sequent deprotection of THP group and oxidation of the obtained primary alcohol afforded fatty acid part 97. The other part of the derivative of serinol 100 would be synthesized from chiral starting material D-serine by the method of Meyers et al. Finally the acid 97 and the derivative of serinol 100 were condensed under the DCC conditions to afford serinolmalyngamide.
In 2007, Isobe and co-workers reported the synthesis of malyngamide X [60], which possessed an unusual tripeptide portion connecting to a methoxylated fatty acyl chain (Scheme 15). Malyngamide X still was divided into two parts tripeptide portion 108 and fatty acid portion 65. The synthesis of tripeptide portion was started from  [58] commercially available N-Boc-L-valine (101), which was coupled with Meldrum's acid (102)   on the work of Adams provide the key intermediate 120.
Aldol condensation of 120 with amido-aldehyde 115 afforded two epimers 121 and 122. The configuration of 122 was in accordance with malyngamide U. Thus methylation of 122 with MeI gave 123, which was also obtained by Mitsunobu reaction of 121. Finally, removal of the allyl protecting group completed the synthesis of malyngamide U. In 2009, Cao and co-workers reported a convergent route for the total synthesis of malyngamides O, P, Q, and R (Scheme 17) [63]. Preparation of key intermediate 131 began with ethyl 4-chloro-3-oxobutanoate 124. Azidation of 124 afforded azide 125, which was subsequently hydrogenated by H 2 in the presence of di-tert-butyl dicarbonate gave the Boc-protected amine 126. Reduction of both keto and ester carbonyl groups in ester 126 with diisobutylaluminum hydride (DIBAL-H) afforded the corresponding diol, followed by monoprotection of the primary hydroxy group with tert-butyldiphenylsilyl chloride (TBDPSCl) to give the corresponding silyl ether 127. Then oxidation of secondary alcohol 127 with 2-iodoxybenzoic acid (IBX) provided the corresponding ketone 128, which was subjected directly to Wittig olefination with chloromethyltriphenylphosphonium iodide (129) to give the vinyl chloride as a mixture of Z-and E-isomers (Z:E = 3:1). The Z-configuration of the vinyl chloride was consistent with that in natural malyngamides O and P. Then N-methylation of 130 provided the key vinyl chloride 131. Thus, deprotection of the TBDMS group of 131 with TBAF, followed by oxidation of alcohol with IBX afforded aldehyde 132. Then aldehyde 132 reacted with the enolate derived from methyl acetate in THF to give 133. Racemic alcohol 133 was immediately submitted to deprotection of the Boc group to generate the corresponding amine, which was directly condensed with the carboxylic acid 65 to afford amide 134. Finally, oxidation of 134 with Dess-Martin periodinane gave malyngamide P. Deprotionation and the following methylation provided malyngamide O. Then the authors continued to synthesize malyngamides Q and R, bearing the more challenging structure. The acetamide 138 bearing the pyrrolidone ring was prepared from L-serine 135.  Cannabinomimetric Lipids: From Natural Extract to Artificial Synthesis 15 In 2010, Cao and co-workers reported the stereoselective synthesis of malyngamide M [64], which was still divided into two parts (Scheme 18). The lyngbic acid part 65 was achieved through the common methodology    [65] group with DDQ and subsequent oxidation of the alcohol with DMP afford amide 164. Finally, removal of the TBS moiety gave malyngamide W.
In 2011, Cao group reported the total synthesis of malyngamides K (Scheme 20), L, and 5 00 -epi-C and confirmed absolute configuration of malyngamide L [66]. For the synthesis of malyngamide K, boronic acid part 167 began from 2-cyclohexen-1-one (165). Thus, bromination of enone 165 with bromine generated the bromoenone, followed by protection of the carbonyl group with ethylene glycol to afford the ketal 166. Ketal 166 was easily transformed to boronic acid 167 by treatment with trimethyl borate in the presence of n-butyllithium, and subsequent treatment with hydrogen chloride. The other part amides 172 and 174 began with ethyl propiolate 168, which was converted to ester 169 in the presence of n-tetrabutylammonium iodide. Then reduction of ester gave the intermediate alcohol 170 as a mixture of E-and Z-isomers (E:Z = 1.1:1). The E-isomer could be converted to the desired Z-isomer by irradiation with UV light ([ 350 nm) in DCM. Thus, configuration of Z-170 was consistent with that in natural malyngamides K, L, and epi-C. Then protection of the hydroxyl group with p-toluenesulfonyl group afforded the tosylate 171, which was transformed into the intermediates 172 and 174, respectively. The synthesis of amide 172 was also achieved by treatment of tosylate 171 with an excess of aqueous methylamine and followed by amidation with acid 65. Azidation of 171 and reduction of the obtained azide 173, followed condensation with the acid 65 afforded 174. Then malyngamides K was achieved by the Suzuki coupling reaction of 174 and 167.
Then the authors completed the more complex malyngamides L, and 5 00 -epi-C (Scheme 21) [66]. For the synthesis of malyngamides 5 00 -epi-C, the key was the synthesis of boronic acid part 182. Oxidation of 175 using nitrosobenzene gave the corresponding hydroxylamine, which was reduced to afford a single alcohol under Luche conditions, followed by reductive cleavage of the N-O bond to afford diol 176. Then deketalization and elimination of diol 176 with hydrogen chloride in THF/water (1:1) gave enone 177. Protection of the hydroxyl group of 177 with tert-butyldimethylsilyl chloride and followed by bromination of the corresponding silyl ether gave bromoenone 178. Then ketalization of ketone 178 gave 179 and 180. Deprotection of ketal 179 also afforded ketal 180. Then protection of the hydroxyl group of 180 with allyl bromide afforded the allyl ether 181. Then the boronic acid 182 was prepared by a procedure similar to that for the Scheme 20 Synthesis of malyngamide K reported by Cao [66] preparation of boronic acid 167. Thus, the skeleton of 5 00epi-malyngamide C could be constructed via Suzuki crosscoupling reaction with 182 and previous prepared 172.
Then the allyl ether was converted to silyl ether, followed by stereoselective epoxidation of silyl ether with hydrogen peroxide and benzyltrimethylammonium hydroxide corresponding epoxide 184. Finally, removal of the TBS protecting group with tetrabutylammonium fluoride (TBAF) provided the 5 00 -epi-malyngamide C, which would be convert to malyngamide C via the Mitsunobu reaction [42]. Then malyngamide L was prepared via the similar methodology (Scheme 22) [66]. The authors initially began with (R)-(-)-carvone, which finally provided 3 00 ,4 00 -epimalyngamide K. Therefore, (S)-(-)-carvone was chosen as the starting material instead. The preparation of the enone 185 underwent a similar sequence showed in Scheme 18 in the preparation of malyngamide W. Then protection of the hydroxyl group of 185 with MOMCl gave the corresponding ether, followed by bromination to afford the bromoenone 186. Then boronic acid part 190 was prepared through the sequence as the preparation of boronic acid 182. Then Suzuki cross-coupling reaction with 190 and 172 afforded 191, which was removed allyl protecting group to finish malyngamide L.

Conclusion
This review illustrated a series of lipids which resemble anandamide in structure. At present, only several marine cyanobacterial fatty acid amides have been reported with binding affinities to the cannabinoid receptors, which were grenadamide, mooreamide, semiplenamides A, B, and G, serinolamides A, B and malyngamide B. Others, due to absence of functional assays test, only have the possibility to interact with CB 1 and CB 2 . Additionally, the metabolites act as receptor agonists implying that they can mediate certain physiological effects through this pathway, which would open more research avenues. Further, a number of total synthesis and well-established synthetic routes have been available; these can assist structural optimization efforts towards more potent analogues, which would be of benefit for understanding the pharmacological mechanisms of cannabinoids and their receptors.