Current Medicinal Chemistry, 2000, 7, 355-374 355 Huperzine A, A Potential Therapeutic Agent for Treatment of Alzheimer’s Disease D.L. Bai*, X.C. Tang and X.C. He Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 294 Taiyuan Road, Shanghai 200031, China Abstract: HupA is a potent, reversible and selective inhibitor of AChE with a rapid absorption and penetration into the brain in animal tests. It exhibits memory-enhancing activities in animal and clinical trials. Compared to tacrine and donepezil, HupA possesses a longer duration of action and higher therapeutic index, and the peripheral cholinergic side effects are minimal at therapeutic doses. This review article deals with a comprehensive survey of the progress in chemical and pharmacological studies of HupA including the isolation and structure elucidation, pharmacological actions, total synthesis, SAR studies and the future development of HupA. Recently, it has been reported that HupA could reduce neuronal cell death caused by glutamate. The dual bio-activities of HupA would further enhance its value and potentiality as the therapeutic agent for Alzheimer’s disease. Introduction In recent decades, some few of bio-active ingredients and promising lead compounds have been discovered from traditional Chinese medicinal plants. artemisinin and huperzine A (1 , HupA) are the most famous ones among them [1]. Alzheimer’s disease (AD) is one of the major diseases affecting the elderly people in the world. The disease is characterized by a gradual and progressive memory impairment. There is considerable evidence that the memory deficits are due to the impairment of cholinergic neurotransmission in the central nervous system. Therefore, the cholinergic enhancement strategy has been the major efforts to palliate the cognitive symptoms. The well-known naturally occurring acetylcholinesterase (AChE) inhibitors physostigmine and galanthamine have already shown their utilities in the symptomatic treatment of AD. Pharmacological studies demonstrated that HupA is a potent, selective and reversible inhibitor of AChE, and showed memory-enhancing effects on a range of behavioral models in animals. Now, it has been approved and clinically used as a palliative agent for AD in China. *Address correspondence to this author at the Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 294 Taiyuan Road, Shanghai 200031, China 0929-8673/00 $19.00+.00 Most recently, it was reported that HupA could reduce neuronal cell death caused by glutamate. The pharmacologically dual actions would further enhance the value of HupA as the therapeutic agent for the treatment of AD. A number of brief accounts and reviews dealing with HupA covering the period 1986-1993 by Kozikowiski [2,3], Ayer [4], Ji [5] and Tang [6] have been published. In this review, we would like to comprehensively survey the progress of chemical and pharmacological studies of HupA, and the literature in Chemical Abstracts is covered up to the end of 1998. Isolation and Structure Elucidation The new Lycopodium alkaloid HupA was isolated from Huperzia serrata (Thunb) Trev. by Liu and his coworker in the early 1980’s. This club moss was used as a folk medicine in China for the treatment of contusion, strain, swelling and schizophrenia. The dried powdered herb was extracted with 95% ethanol and the crude alkaloids were extracted with 1% aqueous solution of sodium hydroxide. The alkali solution was then neutralized with hydrochloric acid and brought back to pH>10 with concentrated ammonia. The aqueous solution was extracted with chloroform and the residue of the phenolic alkaloids from the extracts was chromatographed on basic silica gel column eluted with chloroform-methanol. The © 2000 Bentham Science Publishers B.V. 356 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. Me 7 H 10 N 8 9 1 11 Me O 5 13 NH 2 4 Huperzine A [5R-(5α,9β,11E)]-5-amino-11-ethlidene-5,6,9,10-tetrahydro7-methyl-5,9-methanocycloocta [b] pyridin-2(1H)-one 3 1 crude HupA was crystallized from acetone to give HupA d with a mp 229-30C and [α]25 –150.4 (c 0.498, MeOH). The average yield of HupA in plants is 0.011% [7, 8, 9]. The structure and absolute configuration of HupA were determined on the basis of NMR, IR, UV, CD data and chemical transformations, and it was further confirmed by X-ray crystallographic analysis [10]. The structure of HupA is very similar in many aspects to that reported for selagine and isoselagine. Reinvestigation of these structures has revealed that the earlier structural assignments for selagine and isoselagine are incorrect. In fact, they are identical with HupA [11]. Pharmacological Actions Effects on Cholinesterase Activity The cholinesterase (ChE) inhibition of HupA was evaluated in vitro and in vivo using a spectrophotometric method [12] with slight modification. HupA inhibited the activity of AChE in the rat cortex with the effect being initiated at 10 nM. The concentration of inhibitor yielding 50% inhibition of enzyme activity (IC50) of HupA on AChE and butyrylcholinesterase (BuChE) compared with other ChE inhibitors are listed in Table 1 . The inhibition of AChE activity induced by HupA was more potent than that of tacrine and galanthamine, but less pronounced than that of donepezil in vitro [13,14,15]. In contrast, HupA inhibited BuChE at much higher concentration than needed for inhibition of AChE compared with Table 1. donepezil. HupA has the highest specificity for AChE. A Lineweaver-Burke plot for HupA indicated a pattern of AChE inhibition of the mixed competitive type, as the intersection of the lines occurred in the second quadrant [13,15]. The Ki value of HupA was 24.9 nM. HupA was about 4-fold and 9-fold as potent as tacrine and galanthamine, respectively, and was about 2-fold less potent compared with donepezil [15]. The AChE activity did not exhibit progressive decrease with the prolongation of incubation with HupA in vitro, and the AChE activity recovered to 94% of the control after 5 times washing, indicating the inhibitory manner of HupA was reversible [13,14]. Significant inhibition of the brain AChE activity was demonstrated in rat following administration of HupA at doses of 0.12 - 0.5 mg/kg in a dose-dependent manner[16-19]. In contrast to the AChE inhibition in vitro, the relative inhibitory effect of oral HupA on cortex AChE was found to be about 24-fold and 180-fold, on an equimolar basis, potent than donepezil and tacrine, respectively. [18] HupA injected intraperitoneally exerted similar efficacy of AChE inhibition in rats as observed following oral administration, while tacrine and donepezil resulted in a greater inhibition not only on AChE activity, but also on serum BuChE [18]. The inhibitory potency of HupA on brain AChE was less than that of donepezil after the intraventricular injection[19]. These findings indicated that HupA has higher bio-availability and penetrates the blood brainbarrier easier than that of donepezil and tacrine. Maximal AChE inhibition in rat cortex and whole brain Anticholinesterase Effects of Huperzine A, Donepezil, Tacrine and Galanthamin In vitro IC50 µM Cholinesterase inhibitors AChE (rat cortex) BuChE (rat serum) Inhibitory pattern Ki * (nM) Huperzine A 0.082 74.43 mixed competition 24.9 Donepezil 0.010 5.01 noncompetition 12.5 Tacrine 0.093 0.074 noncompetition 105.0 Galanthamine 1.995 12.59 competition 210.0 *assayed with rat erythrocyte membrane AChE. Treatment of Alzheimer’s Disease reached at 30 - 60 min and maintained for 360 min following administration of HupA (Fig.1 ) [16,17,18]. Repeated doses of HupA showed no significant difference on the AChE inhibition as compared to that of single dose, indicating no tolerance to HupA occurred [18,20]. Current Medicinal Chemistry, 2000, Vol. 7, No. 3 357 parietal cortex [16,17,18]. Considering that ACh level is particularly low in the cerebral cortex of patients with AD, this particularly regional specificity produced by HupA may constitute a therapeutic advantage. HupA also caused significant increase of brain norepinephrine and dopamine levels following either systemic administration or local administration through the microdialysis [21]. These effects may involved in memory improvement of HupA, since there was evidence of interaction between cholinergic and monoaminergic system in the control of cognition [23]. Effects on Brain Receptor Pretreatment of cultured brain neurons with HupA (100 µM) reduced neuronal cell death caused by toxic levels of glutamate, and reduced glutamate-induced calcium mobilization, but did not affect the increase in intracellular free calcium channel induced by exposure to high KCl or a calcium activator Bay-K-8644[24], suggesting that HupA might act on glutamate receptors to exert its neuroprotective effects . HupA (0.1 - 300 µM) reversibly inhibited NMDA (100 µM)-induced current in acutely dissociated hippocampus pyramidal neurons in a concentrationdependent manner with IC50 of 0.49 µM. The results in binding assay clearly demonstrated that HupA acted directly on NMDA receptor [25]. The studies on displacement of 3H-QNB and 3H(-)nicotine specific binding showed that HupA generally had little direct effect on cholinergic receptors as compared to other ChE inhibitors [16,22]. Enhancement of Learning and Memory Fig. (1). Time course of ChE inhibition following oral administration of HupA (1.5 µmol/kg), donepezil (16 µmol/kg), and tacrine (120 µmol/kg) in rats. Values are expressed as % inhibition (vs saline control). n = 4 – 6. Effects on Neurotransmitter Level HupA caused a significant increase in acetylcholine (ACh) level in rat brain. The rats treated with HupA at doses of 0.3, 0.5 or 2 mg/kg increased ACh levels for about 6 hr after administration [16,17,21]. HupA produced a more prolonged increase of ACh levels in whole brain than that of tacrine, heptylphysostigmine, physostigmine and metrifonate, respectively[22]. There was considerable regional selectivity in the degree of ACh elevation after HupA administration, the maximal increase was seen at 60 min in frontal and HupA has been found to be an effective cognitive enhancer in a number of different animal species. Beneficial effects were seen not only in intact adult rodents [14], aged rodents and monkeys (Fig.2 ) [26,27], but also in cognitively impaired rodents (Fig.3 ) and monkeys [26-31]. The duration of improving effects with oral HupA on learning and memory retention were longer than that of physostigmine, galanthamine and tacrine, respectively[32]. The improving effect on memory of aged monkey remained for about 24 hr after a single injection of HupA 10 µg/kg [27]. The magnitudes of improving effects produced by oral or intraperitoneal administration of HupA at a dose of 0.2 mg/kg were nearly equivalent, indicating HupA had a higher efficacy with oral route than that of tacrine and donepezil, respectively[29]. HupA induced no significant tolerance in the cognitive improvement after oral once daily for 8 d [31]. HupA did not show any significant affinity for muscarinic and nicotinic receptors [16], was devoid of pre- and post-synaptic actions [33], 358 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 as well as was devoid of inhibition on choline acetyltransferase activity [17,20], indicating the improving effect of HupA on cognition was due primarily to its inhibition of brain AChE. Bai et al. Total Synthesis The pronounced bioactivity of HupA and its low yields in plants have provided the impetus for renewed interests in the synthesis of this target molecule. Several groups have devoted intensive efforts to the chemical synthesis of HupA. HupA possesses the rigid bicyclo [3.3.1]nonene skeleton and fused pyridone ring. The retro-synthetic analysis of this molecule is shown in Scheme 1 . tandem Michael-aldol reaction and dehydration Me Wittig olefination Me H N H O Me N NH 2 O Curtius rearrangement 1 Fig. (2). Effects of huperzine A in aged monkeys. Saline or huperzine A was administered intramuscularly 20 min before testing. Huperzine A produced a dose-related improvement in the delayed response performance of aged monkeys (n = 4). Values represent mean + S.E.M. number of trials correct out of a possible 30 trials. Clinical trials have been demonstrated that HupA significantly improves cognition in junior middle school students complaining of memory inadequacy [34], patients suffering from aged related memory dysfunction [35] or AD [36]. Peripheral cholinergic side effects at cognitively efficacious dose of HupA are not detected. O COOMe O Me H N N OMe OMe O H O O OMe COOMe 2 3 Scheme 1. The β−keto ester functionalities in key intermediate 2 would provide not only an activating group for the construction of three-carbon bridge by tandem Fig. (3). Effects of ChE inhibitors on AF64A-induced working memory deficit in a partially baited radial maze paradigm in rats. ** P < 0.01 vs non-lesioned group, ++ P < 0.01 vs AF64A-lesioned, saline drug control. Treatment of Alzheimer’s Disease Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Me 359 O Me R R2 O O NH Me O O NH 2 R1 O O R 4 1 5 Scheme 2. Michael-aldol reaction, but also the latent groups for the formation of both ethylidene substitution by Wittig reaction and amino group via Curtius rearrangement of the acid from ester 2 . Endocyclic double bond may be formed by dehydration of the aldol product. The pyridone ring could be protected as a stable methoxy pyridine. Based on the synthetic strategy in Scheme 1 , the total synthesis of racemic and natural HupA was accomplished starting from β−keto ester 2 which usually occurs in enol form 3 . further employed to prepare some pyridone ring opening analogues. However, there is no report on the total synthesis of HupA itself and the preparation of the desired heterocyclic analogues using functionalized bicyclo [3.3.1]nonane derivatives as key intermediates. Synthesis of β−Keto Ester A number of methods have been developed for the preparation of the key intermediate, β−keto ester 2 , from different starting materials. Another approach to HupA is to construct the functionalized bicyclo [3.3.1]nonane derivative 4 first, Me Me Me CH 2 HO O O O O O Me O Me Me O NHCOOMe COOMe 6 SOPh 7 COOMe 8 followed by the formation of the fused pyridone ring (Scheme 2 ). Compound 4 could be obtained from α− carbomethoxy- or phenylsulfinyl cyclohexane-1,4dione monoketal 5 . A couple of bicyclo [3.3.1]nonane or nonene compounds 6, 7, 8 and 9 were prepared [37, 38, 39, 40]. These intermediates have been 9 From 1,4-Cyclohexanedione Ketal 10 Mono-ethylene The fused 2−pyridone l2 was prepared by heating the enamine1 1 derived from mono-ethylene ketal 1 0 with acrylamide followed by hydrolysis. The resulting H O O HC N CCOOCH 3 N O OMe Ag2CO3 O NH3 / MeOH O MeI O O O 13 12 10 I, H+ pyrrolifdine p-TsOH HC 2, (MeO) 2 CO KH CCONH2 N OMe N O 11 O Scheme 3. O COOMe 2 360 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. H Me N O Me 1, MeI/Ag2 CO3 N OMe 2, LAH EtOOC N 1, PhLi/HCHO 2, SOCl2 HOCH2 OMe ClCH2 ClCH2 14 MeO 2C 1, NaCN/DMSO N OMe N MeO 2C 2, HCl/MeOH THF O N N O COOMe O3 /NaOH 15 Me OMe NaH 2 Br N OMe OMe N Br Me O N OMe 1, LDA/AllyIBr 2, n-BuLi/CuI/AllyIBr Br 17 16 Scheme 4. mixture of unsaturated lactams was N-benzylated and then dehydrogenated by α−selenation and oxidative elimination, yielding N-benzylpyridone which was debenzylated by hydrogenolysis to afford pyridone 1 2 . The pyridone ring in 1 2 proved to be sensitive in subsequent steps, and was protected by conversion to methoxypyridine 1 3 with methyl iodide and silver carbonate. Hydrolysis of the ketal and α− carbomethoxylation with potassium hydride and dimethyl carbonate produced β− keto ester 2 [41]. This route was somewhat laborious. Therefore, Kozikowski and his coworkers examined alternate routes to 2− pyridone l2 and found that a one-pot, threecomponent condensation was the best one. By simply admixing mono-ethylene ketal 10 and methyl propiolate in ammonia-saturated methanol in a Parr reactor at 100 C, 2−pyridone l2 was obtained in a yield of 70% (Scheme 3 ) [42]. Compound 1 2 was also produced by the reaction of enamine 1 1 and propynamide in THF or DMF at 80C in a yield of 75-80% [43]. which reacted with phenyllithium and formaldehyde yielding diol followed by conversion of diol into dichloride. Treatment of dichloride with sodium cyanide and methanolysis of the resulting dicyanide gave diester 1 5 . Dieckmann condensation of 1 5 afforded the key intermediate 2 . The overall yield was 39% via 7 step sequence of reactions (Scheme 4 ) [44]. The key step in this approach is to construct two appending carbomethoxy side chains. 2-Methoxy-6methylpyridine 1 6 was also used as a starting material for the preparation of diester 1 5 . Bromination of 1 6 with 1, 3-dibromo-5.5-dimethyl hydantoin gave desired bromide. Twice allylation of bromide formed compound 1 7 . An oxidative cleavage of two double bonds in 1 7 followed by esterification gave rise to diester 1 5 in an overall yield of 41% (Scheme 4 ) [45]. From Dimethyl 4-oxopimelate Yang et al developed a four step approach to β−keto ester 2 from readily available dimethyl 4-oxopimelate 1 8 (Scheme 5 ). The enamine 1 9 was directly condensed with propynamide to give pyridone 2 0 with two desirable carbomethoxy side chains. After Omethylation, Dieckmann condensation of diester 1 5 gave 2 in 39% overall yield [46]. From Pyridine Derivatives Ji group reported that β−keto ester 2 was obtained from ethyl 2-methyl-6-hydroxy-nicotinate 1 4 . Selective O-methylation and reduction of 1 4 afforded alcohol, O pyrrolidine MeOOC COOMe MeOOC H N Ag2CO3 MeI MeO 2C 20 Scheme 5. O CCONH2 TiCl4 18 MeO 2C HC N MeO 2C 19 N MeO 2C OMe COOMe N OMe NaH O 15 COOMe 2 Treatment of Alzheimer’s Disease Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Me Me N OMe 1, methacrolein MeONa RO AcONa AcOH N N 2, MsCl Et3N O 361 130o C OMe OMe O O COOMe COOMe COOMe 21 R = H 22 R = Ms 2 23 Me Me Ph3p +C2 H5 BrBuLi KOH N Me OMe COOMe MeOH OMe COOH 24 Me (PhO)2 P(O)N 3 N Me 25 1, Me3SiCl, Nal CH3 CN (±)-1 EtOH N Me 2, KOH, toluene 18-crown-6 ether OMe NHCO 2Et 26 Scheme 6. Synthesis of Racemic HupA and only E-isomer could be hydrolyzed to acid 2 5 . Curtius rearrangement of acid 2 5 gave urethane 2 6 . (±)-HupA was obtained after deprotection. Total synthesis of (±)-HupA was first accomplished independently by both Ji and Kozikowski groups in 1989. Almost the same synthetic strategy was adopted starting from β−keto ester 2 (Scheme 1 ). The earlier works of Kozikowski group have been surveyed [2-4]. The following is the synthetic approach to (±)-HupA reported by Ji et al. [44,47] (Scheme 6 ). The yields of E-isomer of 2 4 was greatly improved by Kozikowski et al. [48] through the isomerization of the mixture by heating with thiophenol and 2,2’azobisisobutyronitrile, giving E- and Z- alkene in 9:1 ratio. Furthermore, in the study of pyrimidone analogues of HupA, Kozikowski et al. utilized the Danheiser methodology to obtain solely E-olefine (Scheme 7 ) [49]. Ketone 2 7 was transformed into a sterically less crowded β−lactone 2 8 with complete stereoselectivity by the addition of the anion of thiophenol ester of propionic acid. Upon heating of β− lactone 2 8 in the presence of silica gel in toluene, [2+2] cycloreversion reaction occurred to afford the desired E-olefinic product 2 9 in a total yield of 70%. This stereoselective construction of ethylidene moiety The tandem Michael-aldol reaction of β−keto ester 2 and methacrolein was catalyzed with sodium methoxide in methanol to generate stereoisomeric aldol mixture 2 1 , which was converted into corresponding mesylates 2 2 . Endocyclic alkene 2 3 was given in 30% low yield, possibly only axial mesyl group could be eliminated. Wittig reaction of 2 3 produced a Z- and Ealkene mixture 2 4 in a ratio of 19:1 to 17:3, which was treated with potassium hydroxide in refluxing methanol CH 2 CH 2 EtC(O)SPh LDA N CH 2 Me OMe SiO2 N H O N Me OMe OMe toluene O COOMe 27 Scheme 7. O COOMe 28 COOMe 29 362 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. R1 R 30 R = OH, R1 = H 31 R = H, R1 = OH 33 R = OAc, R1 = H N OMe N OMe O COOMe O COOMe 32 via [2+2] cycloreversion reaction may represent a practical method instead of Wittig reaction and double bond isomerization in the synthesis of HupA. the mixture of mesylates of resultant aldol adducts obtained by this reaction in the presence of different bases was subjected to the elimination respectively, the endocyclic olefine was obtained in a wide range of 19-68% yields. The big difference in yields probably depends upon the different ratio of four diastereomers in aldol adducts formed under different reaction conditions [51]. The low yield of endocyclic alkene 2 3 is related to the configurations of mesylates 2 2 , since the mesyloxy group and the adjacent hydrogen in 2 2 may not adopt a trans-diaxial orientation for base-induced elimination to alkene2 3 [38]. In the preparation of HupA analogues by author’s group, acrolein was used in Michael-aldol reaction, two aldol products were separated by column chromatography to produce axial hydroxyl compound 3 0 and equatorial hydroxyl compound 3 1 . They were converted into mesylates, which were eliminated in acetic acid with sodium malonate at 130 C respectively. The expected product 3 2 was only formed from axial mesylate. However, SN2 reaction occurred with equatorial mesylate to give axial acetate 3 3 [50]. Now it is evident that only axial mesylate in the mixture of 2 2 could be eliminated to form olefine 2 3 , and the equatorial mesylate would be substituted in acetic acid to form axial acetate. Based on the elimination mechanism, Yang et al. reported that when mesylates 2 2 were treated with 1,8-diazabicyclo [5.4.0] undec-7-ene (DBU) and sodium iodide in DMF at 100-110C, the yields of the desired alkene 2 3 was increased to 47-57%[52]. To avoid the low yielding step of elimination in Scheme 6 , a palladium-catalyzed bicycloannulation of 1-carbomethoxy-2-tetralone with a bifunctional allylic agent, 2-methylene-1, 3-propanediol diacetate 3 4 was reported by Gravel et al. [53]. The regioselective double bond migration completed the construction of three-carbon bridge in 70-80 yields. On the basis of this report, the palladium-catalyzed bicycloannulation of β−keto ester 2 using tetramethylguanidine (TMG) as a base and 2- In the course of the studies of stereoselective synthesis of HupA via Michael-aldol reaction, a variety of chiral bases were tested by Terashima et al.. When N OMe AcO CH 2 CH 2 CH 2 OAc 34 N N OMe O COOMe Pd(OAc) 2, Ph3 P TMG OMe Me O NH 2 COOMe 2 35 27 TfOH TfOH Me (±)-1 N OMe O COOMe 23 Scheme 8. Treatment of Alzheimer’s Disease Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Me Me 1, Et3N (S)-MTPA-Cl N 1, Jones oxidation N MeO 2, column chromatography 3, LAH 2, MeI, DBU CH 3CN MeO CH 2OH CH 2OH Me 363 Me (Z)-36 (Z)-(±)-36 Me Me Me 1, olefin isomerization N N 2, hydrolysis MeO MeO H Me N COOMe COOH Me (Z)-37 O Me NH 2 (E) -25 (+)-1 Scheme 9. methylene-1,3-propanediol diacetate 3 4 as the biselectrophile in the presence of tetrakis (triphenylphosphine)-palladium(0) in refluxing dioxane was first achieved by Kozikowski et al. (Scheme 8 ). The resulting methylene-bridged compound 27 was in 92% yield. The double bond migration was pursued with triflic acid in dioxane at 93C, affording endocyclic olefine 2 3 in 90% yield. Since the double bond migration could be performed in last step, an isomer of HupA, compound 3 5 , was obtained. The overall yield of (±)-HupA was 40% from β−keto ester 2 [54, 55]. Synthesis of Enantiomer Natural (-)-HupA and its For the preparation of natural (-) HupA and its enantiomeric (+) HupA, three methodologies have been developed by now. via Diastereomeric Separation Kozikowski et al. reported the diastereomeric separation to obtain (+)-HupA [56] (Scheme 9 ). Z - (±) Alcohol 36 reacted with (S)-α-methoxy-α(trifluoromethyl) phenylacetyl chloride (MTPA-Cl) to form diastereomeric esters, which were separated by column chromatography. The resulting optically active ester was reduced with lithium aluminium hydride (LAH) to give 3 6 . Jones oxidation of 3 6 followed by esterification afforded ester 3 7 , which was isomerized to E-olefine and then hydrolyzed to optically pure Eacid 2 5 . E-Acid 2 5 was further transformed into (+)HupA according to the reaction sequence as shown in Scheme 6 . via Stereoselective Michael-aldol Reaction Using Chiral Auxiliary The first molecular modeling-based synthesis of natural (-)-HupA was achieved by Kozikowski group in 1991 (Scheme 1 0 ). β-Keto ester 2 was transesterified with (-)-8-phenylmenthol, giving ester 3 8 , which was an effective substrate for the stereoselective Michael-aldol addition of methacrolein in the presence of TMG at –20 C. After dehydration, the ratio of two diastereomers was 9:1 with the major isomer possessing the desired absolute configuration. The chromatographically separable major compound 3 9 was subjected to further transformation. Reduction of ester 4 0 with LAH and re-oxidation of the resulting alcohol gave (+)- acid 2 5 . Subsequent Curtius rearrangement of (+)-2 5 and deprotection led to the natural (-)-HupA [56]. Using Chiral Catalyst Yang et al. first reported the stereoselective Michael-aldol reaction of β-ketoester 2 catalyzed by (-)quinine at room temperature for 10 days. After elimination of adducts, alkene 2 3 was obtained in 52% ee. The final product was afforded in 62.8% optical purity in comparison with the natural (-)-HupA [52]. Chiral guanidines 4 1 and 4 2 were also tested for the stereoselective Michael-aldol addition of β-keto ester 2 with methacrolein. Disappointedly, the ee values of the products were insignificant [57]. Terashima et al. also reported the (-)-cinchonidine promoted stereoselective Michael-aldol reaction of βketoester 2 with methacrolein. The reaction took place 364 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. Ph N N OMe 1, Michael-aldol reaction, TMG R*OH R*= O H+ O O Me 2 N 2, dehydration, 3, chromatographic separation O H COOMe Me OMe OR* OMe COOR* 38 39 Me Me 1, LAH 1, Wittig reaction N 2, isomerization Me OMe 2, Jones reagent N Me OMe (-)-1 COOR* COOH 40 (+)-25 Scheme 10. smoothly at –10 C for 10.5 days . After dehydration of the adducts, endo-olefine (+)-2 3 was obtained in 64% ee. The enantiomerically pure (+)-2 3 was readily obtained by recrystallization of the product from hexane, and converted into (-)-HupA according to the reaction sequence reported for (±)-HupA. Enantiomer (-)-2 3 was similarly prepared in 61% ee using (+)cinchonine as chiral catalyst [51, 58]. via Enantioselective Bicycloannulation ligand (R)-(S)-4 3 and TMG in 1,2-dimethoxyethane at – 30C followed by gradual warming to 15C. The bridged compound (+)-27 was yielded in 64% ee and 92% yield. The enantiomer (-)-2 7 with 63% ee was also afforded using the enantiomeric ligand (S)-(R) –4 3 . Both enantiomers were converted to the endocyclic olefines 2 3 by double bond migration with triflic acid. Enantiomeric enrichment was performed by recrystallization of optically crude product from hexane to give optically pure (+)-2 3 and (-)-2 3 respectively. Following the reaction steps reported for the preparation of (±)-HupA, the natural (-)-HupA and unnatural (+)-HupA were finally furnished [51,59]. Palladium-catalyzed Terashima and his coworkers investigated the bicycloannulation of the β-keto ester 2 with 2methylene-1,3-propanediol diacetate 34 in the presence of palladium catalyst with chiral ligands (Scheme 1 1 ). They first examined various bisphosphines as chiral ligands of palladium, but no promising results were observed. When a couple of ferrocenylphosphine ligands carrying appropriate linker chains were tested, they found the most effective ligand (R)-(S)-4 3 . The reaction of 2 and 3 4 was carried out in the presence of palladium diacetate, the chiral The author’s group has been also interested in the synthesis of (-)-HupA by palladium-catalyzed asymmetric bicycloannulation. Different bifunctional allylic esters 4 4 and chiral ligands were tested. A 52% ee of 2 7 induced by (S)-2,2’-bis(diphenylphosphino)1,1’- binaphthyl (BINAP) was observed [60]. More recently, a few of new chiral ferrocenylphosphine ligands have been developed for the allylic bicycloannulation. Compound (+)-2 7 has been NH OH R R N N N OR1 RO Me Ph Ph 41 R = Me 42 R = 2,4,6-trimethyl-benzyl- Fe 44 PPh2 PPh2 R1 (R)-(S)-43 R= = COPh R = Ac, R1 = CO2 Me Treatment of Alzheimer’s Disease N Current Medicinal Chemistry, 2000, Vol. 7, No. 3 AcO OAc 34 O 1, TfOH N Pd(OAc)2, TMG (R)-(S)-43 COOMe Me CH 2 CH 2 OMe 365 OMe N OMe 2, recryst. O O COOMe COOMe 2 (+)-27 (+)-23 (-)-1 Scheme 11. obtained in 80% ee and 90% chemical yield. Optically pure (+)-2 3 was furnished in 70% yield after recrystallization. The synthetic (-)-HupA obtained from (+)-2 3 possesses the same anti-AChE potency as natural HupA [61]. Studies on Relationships the Structure-activity The stereoselectivities of the inhibition of rat cortical AChE by two enantiomers of Hup A were first determined by McKinney et al. [62]. The natural (-)-Hup A is the more potent one with a Ki value of 8nM, and (+)-HupA is 38-fold less potent with a Ki value of 300 nM. Racemic Hup A is about 2-fold less potent than (-)Hup A. The comparison of the inhibitory effects of two enantiomers and racemate of HupA on AChE was made by Tang et al. [17]. The (±)-Hup A is 3 times less potent than (-)-HupA in vitro using rat hippocampal crude homogenates as the enzyme resource. The ratio of potencies for (-), (±), and (+)-HupA is 70:3:1, which is comparable to those reported by MicKinney et al. As a promising lead compound, scientists have been much interested in the chemical modification of Hup A for the search of new analogues or derivatives, which may possess high activity, longer duration of action, less toxicity, and could be prepared by simpler and efficient approaches as compared with HupA itself. In recent 10 years, major efforts have been devoted to H N R N O the preparation of both structurally simplified analogues and derivatives with the tricyclic skeleton of Hup A. In comparison with the flexible chain structure of ACh, there is no conformational isomers of HupA because of the rigidity of the bridged ring structure. However, the computer-generated superposition of Hup A and ACh suggested that HupA possessed the basic structural features of ACh. The amino nitrogen atom in HupA would attract a proton in vivo to form a positively charged center, it is as distant from the carbonyl of the pyridone ring as the quaternary nitrogen atom is from the ester carbonyl in ACh. It is apparent that a reasonable structural similarity can be found between the nitrogen, oxygen and carbonyl in ACh with the corresponding amino nitrogen, nitrogen and carbonyl in HupA.. Therefore, 5-aminomethyl-2 (1H)-pyridone part of HupA may constitute a pharmacophoric moiety [1,48]. Structurally Simplified Analogues of HupA Several types of the simplified analogues were designed and prepared. All of them possess the supposed pharmacophoric moiety of HupA. 5-Substituted aminomethyl-2(1H)-pyridones 4 5 [3,48,63,64], 5-substituted amino-5,6,7,8tetrahydroquinolinones 4 6 , 47 [63,64,65,66]and 5substituted aminoquinolinones 4 8 [67] were prepared and tested successively. None of the conformationally O CH 2CH=CMe2 N N R3 R R1 R = H, alkyl R1 = H, alkyl 45 N R2 46 R1 R = H, alkyl R1 = H, alkyl R2 = H, alkyl, arylakyl R3 = H, alkyl O 46a: Ar=3,4-dichlorophenyl 46b: Ar=2-chlorophenyl 46c: Ar=3-chlorophenyl 46d: Ar=4-hydroxyphenyl NH(CH2) 2 Ar 366 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 H N Bai et al. R N O Me O R Me CH2 Me R1 Me R2 NHR1 NH 2 47 49a: R 1=R 2=H, R=OAc 49b: R1 =R2 =H, R=NH2 49c: R 1=R 2=Me, R=nicotinoyloxy 49d: R1 =H, R2 =C(O)Bu-n, R=OCONMe2 48 R = H, CH 2Ph R1 = (CH 2) nAr more flexible aminomethypyridones 4 5 was found to be an AChE inhibitor. Analogues 4 6 were designed as simplified analogues of HupA without bridge moiety. Among them compounds 4 6 a -d inhibited AChE in vitro markedly, whereas they were less active than (±)HupA. Compound 4 6 a and 4 6 d also showed the ability to reverse scopolamine-induced memory impairment in mice. Because these compounds lack the unsaturated three-carbon bridge of HupA, and both the quinolinone nitrogen atom and the amino group must be substituted, it is unlikely that they bind to AChE in a similar fashion as HupA [65]. The single ring analogues 4 9 derived from HupA by the removal of the bridge and the opening of pyridone ring, and by the replacement of amide group with the isosteric ester group as well were also reported [68]. The anti-AChE activities of these compounds were at least 1000 times less active than (-)-Hup A. Compound 4 7 is derived from the removal of the three-carbon bridge of HupA, and was at least 4727fold less active than (±)-HupA [66]. It clearly demonstrates the necessity of the bridge ring in Hup A, since compound 4 7 mimics the quinolinone portion of HupA quite well. The aminoquinolinones 48 showed no inhibitory activity at all. All these structurally extensively simplified analogues mentioned above were found to be inactive or less active in the inhibition of AChE. It means that the conformational constraints, hydrophobic binding, steric and electrostatic fields provided by the unsaturated bridge and fused pyridone ring in HupA must involve in its AChE inhibitory activity. In addition, several bridged ring analogues 50a-c [40,37] derived from the opening of pyridone ring of HupA were prepared, and the inhibitory activities of 5 0 a and 5 0 b dropped down dramatically. The bioactivity of 5 0 c is not reported yet. Me CH 2 OCONMe2 R Me CH2 NMe2 50a: R=nicotinoyloxy 50b: R=nicotinamido NH 2 50c Treatment of Alzheimer’s Disease Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Derivatives from Natural (-)-Hup A Some dozens of derivatives using (-)-Hup A as the starting material were prepared, i.e. alkylamino-, acylamino-,1-alkyl-,3-bromo-,dihydro-and tetrahydroderivatives, and Schiff base of Hup A. (Fig. 4 ) AChE All these optical active derivatives were tested for their inhibitory activities of AChE from rat erythrocyte membrane over a concentration range from 1 nmol. L-1 to 10 mmol. L-1. The IC50 for AChE by these compounds revealed that most of them were less active than (-)-HupA itself, except a few of Schiff bases AChE BuChE 25.1 / 1.58 56.2 0.631 200 O 1.58 / O 0.891 398 1.00 100 BuChE NH NH O O 0.0631 63.1 NH 2 NHCOCH2 O NH NH O N 158 11000 O CH 3 NHCH2 CH 3 CH 3 NH N O N 316 11000 CH 3 O N CH C3H7 CH 3 NH NH O 0.2 15.8 NH 2 NH H2C O NH NH O 3.16 NH 2 50.1 NH CH 2OH CH 2 NH O 3160 11000 N CH 2 HO NHCOCH3 NH O NH O 7.94 166 NHCH2 N NH 2 367 Br Fig. (4). IC50 (µM) of anti-ChE activities by (-)-HupA analogues on rat erythrocyte membrane AChE and rat serum BuChE. 368 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. 5 1 . The anti-AChE activities and selectivities between AChE and BuChE of 5 1 a and 5 1 b are comparable to (-)-Hup A [69,70], Hydrogenation of the exocyclic double bond or both exo- and endocyclic double bonds in Hup A led to a diminution of anti-AChE activity. Me Me NH O Me Me N=CHR NH 2 51a: R=C3 H7 52 51b: R= Based on the above-mentioned results, it suggests that the structural requirements for high anti-AChE activity of Hup A appear to include a free amino group or protected amino group which can be easily regenerated, two double bonds and a unsubstituted nitrogen atom in pyridone ring. ring skeleton and only to alter or remove certain atoms or groups in Hup A. The first example was a benzene isostere 5 2 , which was at least 1000-fold less potent than (-)-Hup A. (±)-Z-HupA 5 3 , a geometric isomer of HupA was 60-fold less potent than (-)-Hup A [71,72]. In order to make an evaluation of the contribution of the endocyclic double bond, C-14 and C-12 methyl groups to the activity of HupA, compounds 5 4 a -f and 5 5 a -i were prepared by Kozikowski group and author’s group, respectively [48,50,55]. Unfortunately, none of these compounds is an effective AChE inhibitor. The dinor HupA 5 5 a and nor HupA 5 5 b are 3500- and 30fold less potent than (-)-HupA respectively. The ethylidene moiety modifiers 55c, 55g and 5 5 h were much less active too, and 5 5 f showed a 40-fold lower potency than (±) Hup A. The 12-phenyl 5 5 d , dimethylamino 5 5 e and aminomethyl 5 5 i analogues were all poorly active. R1 Me NH NH O R2 O Me NR 2 Analogues Bearing Skeleton of Hup A the Bridged Tricyclic 1 55a: R=R =R =H 55b: R=R 2=H, R 1 =Me 1 Although a number of the chemical modification to Hup A and the anti-AChE activities of these compounds have been reported including alterations of amino group, three-carbon bridge, ethylidene side chain, and the endocyclic double bond, no analogues thus obtained has achieved the anti-AChE potency as parent compound HupA. It is unlikely that a more potent analogue could be found by means of the extensively structural simplification of HupA. R Me NH NH O Me O R1 55i 2 55c: R=H, R =Me, R =Et 55d: R=H, R 1 = Ph, R2=Me 55e: R=R1 =R2 =Me 55f: R=H, R 1=Me, R2 =CH2 OH 55g: R=H, R 1 =Me, R2= CO2 Et 55h: R=H, R 1 =Me, R2= CN The demethylated and hydroxylated analogues 5 6 a , 56b [73], and 1-methyl pyridone analogues 5 7 a and 5 7 b were much less effective on anti-AChE activity than (-)-HupA [74]. 1-Methyl substitution of HupA would impede the formation of H-bond of the pyridone nitrogen atom to the protein residue. Actually, 1-methyl- (-)-HupA was less active by two order of magnitude than (-)-HupA [75]. NH 2 NH 2 NH 2 2 R HO 1 53 54a: R=ax Me, R =Me 54b: R=eq Me, R 1=Me NH 54c: R=H, R 1=Me 54d: R=R =H 54e: R=ax NH2 , R1=Me 54f: R=eq NH2, R 1=Me O R1 CH2 NH 2 56a: R=ax OH 56b: R=eq OH In the meantime, many attempts have been made in the chemical modification of Hup A to retain the tricyclic NMe O 1 NH 2 57a: R=R1 =H 57b: R=Me, R1 =H 57c: R=R1 =Me Treatment of Alzheimer’s Disease Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Me CH 2 NMe NH O NH O Me O Me 369 NH 2 NH 2 58 59 60 A cyclopropane bearing analogue 5 8 was synthesized, but its biological activity is not reported yet [76]. Diamine 5 9 and di-exo-double bonds analogue 6 0 were 88- and 22-fold less potent than (±)HupA respectively [55]. surprising to find that they are less active than (±)HupA. However, it is unexpected that 6 1 is even more less active than 6 2 . In fact, the structure of 6 1 is more closely related to HupA. The studies on the pyrimidone analogues revealed that even a minor alteration of the HupA structure via the replacement of CH in pyridone ring by N caused a major decrease in anti-AChE activity. A thiazolone analogue 6 4 was ineffective either [78]. The replacement of the pyridone ring in HupA by a pyrimidone or other heterocyclic ring may lead to an analogue having improved AChE inhibitory potency due to its ability to form an additional H-bond. Kozikowski group prepared compound 6 1 , 6 2 , the The phenol and catechol analogues of HupA, compounds 6 5 a , 6 5 b , were recently reported by Me Me Me NH H2N NH O Me NH H2N O N N N NH 2 Me Me 63 62 61 pyrimidone analogues of both HupA and iso-HupA, and 6 3 , the pyrazole analogues of iso-HupA as well [49,77]. None of these analogues was found to be an active AChE inhibitor. In view of the relatively dramatic structural alterations in compound 6 2 and 6 3 , it is not Kozikowski group [79]. Surprisingly, no inhibition of both compounds was observed. It is well recognized that introduction of fluorine atom into bio-active compounds frequently improves their therapeutic profiles. Five fluorinated analogues R1 Me Me R NH O Me S NH 2 64 AChE R1 Me NH 2 65a: R=H, R 1=OH 65b: R=R 1=OH NH O R NH 2 66a: R=Me, R1 =CH2F 66b: R=Me, R1 =CF3 66c: R=CF3 , R1 =Me 66d: R=CF 3, R 1=CF3 66e: R=CH2 F, R1=Me 370 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Bai et al. R Me N N Me NH 2 NH 2 68a: R=Me 68a: R=Et 67 6 6 a - e were reported by Terashima group and Ji group respectively [80,81,82,83,84]. However, all the fluorinated analogues had the inhibitory activities inferior to that of HupA. 69a was obtained from 2.2-dimethyl-1.4cyclohexanedione mono-ethylene ketal via the reaction sequence reported for the preparation of HupA. Using AChE purified from FBS, the kinetic and inhibition parameters demonstrated that 10.10dimethyl analogue of HupA was comparable in the inhibitory activity to (±) HupA, but less active than natural (-) HupA [87] Seventeen polycyclic compounds structurally related to tacrine and HupA were prepared as the hybrids of both molecules by Camps et al. [85]. Hybrid compound 6 7 exhibited an AChE inhibitory activity approximately 2.5 times lower than tacrine. The configuration of the ethylidene group is critical, since the Z-isomer of 6 7 showed a much lower activity. Compound 6 8 a and 6 8 b derived from 6 7 by elimination of the ethylidene unit were 2 and 4 times more potent than tacrine respectively. Their inhibitory activity was not directly compared with HupA. Camps et al. [86] have also developed an enantioselective approach to these molecules, but the bio-assay data of the enantiomers of compound 6 8 a and 6 8 b have not been reported yet. The C-10 axial methyl compound 6 9 c showed 8fold potency for AChE inhibition, and the equatorial methyl compound 6 9 b is about 1.5-fold less active than (±) HupA (Table 2). The axial alkyl groups larger than methyl, for example, the ethyl analogue was 100fold less active than (±) HupA. The C-10 n-propyl analogue was inactive at all [88]. The C-10 methyl HupA was prepared from 2-methyl-1.4-cyclohexanedione mono-ethylene ketal via the previously reported route to HupA. The axial and equatorial methyl isomers could be separated by fractional crystallization of the acid 7 0 [88]. Although quite a number of analogues and derivatives of HupA have been successively reported during last 10 years, but none of these compounds is able to compete with natural (-)-HupA in its AChE inhibitory activity until 1996. In exploring further modification at this position of HupA, a 10-spirocyclopropyl analogue 7 1 was prepared in an optically active form [89]. Compound (-)7 1 was found to be nearly as active as natural (-)-HupA (Table 3 ). According to the synthetic route to (-)-HupA, (-)-10-spirocyclopropyl HupA 7 1 was synthesized from 2-spirocyclopropyl-1,4-cyclohexanedione monoethylene ketal by a diastereoselective Michael-aldol reaction, using (-)-8-phenylmenthol as a chiral auxiliary. Recently, in the continuation of structure-activity relationships studies on HupA, Kozikowski and his coworkers have identified several C-10 methyl analogues that showed the similar or even better antiAChE activities than (±) HupA. 10.10-Dimethyl HupA Table 2. Ki Values for the Inhibition of FBS AChE and Equine BuChE by HupA and Its Analogues Compound Ki*(Μ) AChE Ki(Μ) BuChE (+)-HupA 0.024 24 ax Me 69c 0.003 5.8 eq Me 69b 0.035 5.5 dimethyl 69a 0.017 9.5 *dissociation constant C-10 Methyl Treatment of Alzheimer’s Disease Table 3. Current Medicinal Chemistry, 2000, Vol. 7, No. 3 Ki Values for the Inhibition of AChE by HupA and Its Analogues FBS C-10 Compound Ki*/nM (-)-HupA 3.9 (+)-10,10-dimethyl HupA 13.2 (+)-10-spirocyclopropyl HupA 14.0 (-)-10-spirocyclopropyl HupA 6.4 * Ki= koff / kon In order to understand the differences in bioactivities between axial methyl compound 6 9 c and equatorial methyl compound 6 9 b , Kozikowski and his coworkers carried out molecular modeling studies using the refined binding site for HupA in TcAChE by including crystallographic waters [88,90]. The modeling studies of C-10 methylated analogues of HupA demonstrated that either one or two methyl groups at this position did not change their binding mode nor cause any significant conformational change of the protein residues. 371 The interaction between fluorinated analogues 6 6 b and the active site of TcAChE was simulated by molecular dynamic simulation method. It was found that C-12 methyl group in HupA could form weak H-bonds with the phenol oxygen of Tyr 121 and the main-chain oxygen of Gly 118 in the active site of AChE [92]. The difference of structure between HupA in TcAChE-(-)-HupA complex and in gas phase was studied by Jiang et al. using ab inito method of quantum chemistry [93]. The results indicated that the binding conformer of HupA did not adopt its lowest energy one. Because of the formation of H-bonds between HupA and AChE in complex, great changes took place in the total atomic charges, bond lengths and the orientation of the atoms and bonds of HupA. Ashani et al. reported the biochemical constants K on, K off and Ki of complexes formed between HupA and mutants of rHuAChE, which were compared with wild-type rHuAChE and TcAChE. The results demonstrated that inhibition of AChE by HupA occurred via association with residues located inside the active site gorge rather than at the rim of the gorge. Tyr 337 was essential for inhibition of rHuAChE by HupA, an aromatic lining constituted from Tyr 337, Phe Me Me Me R NH Me R1 NH 2 O NH Me NH Me 69a: R=R1 =Me 69b: R=Me, R1 =H 69c: R=H, R 1=Me The molecular modeling and docking studies of the C-10 methyl analogues to the X-ray crystal structure of TcAChE revealed that the C-10 axial methyl pointed to a hydrophobic region of the residues, while the equatorial methyl was directed to a less favorable hydrophilic region. The hydrophobic-hydrophilic contact of C-10 equatorial methyl group was not beneficial to the overall binding energy. A 3D-QSAR analysis for 10 analogues of HupA by molecular modeling and comparative molecular field analysis method indicated a good conventional statistical correlation between the 3D-structures of these compounds and their anti-AChE activities [91]. O Me O NH 2 COOH 70 71 295 and probably Trp 86 was likely to offer a multicontact sub-site that interacted with the ammonium group and both exo- and endocyclic double bonds of HupA [94]. Hitherto, a variety of HupA analogues have been designed and prepared so far using trivial and rational structural modification by several groups, the docking studies of these analogues to the X-ray crystal structure of TcAChE by Kozikowski et al. demonstrated the importance of lipophilic substituents capable of providing additional hydrophobic contacts, which could increase the anti-AChE potency of the analogues [88]. 372 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 More recently, Sussman and his coworkers [95] solved the 3-D structure of a TcAChE-(-)-HupA complex by X-ray diffraction at 2.5 Å resolution. The results showed an unexpected orientation of HupA with surprisingly few strong direct interactions with the protein residues in the active site gorge of the enzyme. The principle protein-ligand interactions include: 1) a strong H-bond of the carbonyl of (-)-HupA to the hydroxyl oxygen of Tyr130 ; 2) in the active-site gorge, the interactions of (-)-HupA and protein residues are mediated by one or two molecules of water ; 3) interaction of the positively charged amino group of the ligand with the aromatic rings of Trp 84 and Phe 330 ; 4) an unusually short H-bond between the ethylidene methyl group and the main chain oxygen of His 440 ; 5) several hydrophobic contacts of (-)-HupA with the sideand main-chain atoms of Trp 84, His 440 and with residues Gly118 through Ser 122 ; 6) a significant change in the main-chain conformation of the protein. Bai et al. Acknowledgment We are grateful to the State Key Laboratory of Drug Research for financial support. List of Abbreviations HupA = Huperzine A AD = Alzheimer’s disease ChE = Cholinesterase ACh = Acetylcholine AChE = Acetylcholinesterase BuChE = Butyrylcholinesterase rHuAChE = Recombinant human AChE TcAChE = Torpedo californica AChE FBS = Fetal bovine serum IC50 = Concentration of inhibitor yields 50% inhibition of enzyme activity NMDA = N-Methyl-D-aspartate TMG = Tetramethylguanidine At present, many analogues of HupA have been prepared, but neither the simplified analogues nor the derivatives from the natural HupA possess the antiAChE potency as HupA itself. Although C-10 axial methyl, C-10 dimethyl and C-10 spirocyclopropyl analogues of HupA have been found to have comparable or somewhat more potent anti-AChE activities than HupA, the preparation of these analogues are quite laborious, and the costs will be even more expensive than natural HupA. MTPA-Cl = α-Methoxy-α−(trifluoromethyl)phenylacetyl chloride LAH = Lithium aluminium hydride DBU = 1,8-Diazabicyclo [5.4.0] undec-7-ene ee = Enantiomer excess BINAP = 2, 2’- Bis(diphenylphosphino)-1, 1’binaphthyl The structure-based SAR studies of HupA should be valuable for the rational design of new analogues of HupA with improved therapeutic profile in future. R = rectus (right) S = sinister (left) Several stereoselective syntheses of natural (-)HupA have been developed, however they have to be further improved in many aspects such as ee, chemical yield, reaction duration, and operational simplicity. The stereoselective bicycloannulation methodology utilizing chiral palladium catalysts seems to be more feasible. E = Entgegen (opposite) Z = Zusammen (together) [1] Bai, D.L. Pure & Appl. Chem., 1993, 65, 1103. In view of the potent, highly selective anti-AChE activity and its chemical stability, HupA is likely to be used as a safe and long-lasting prophylactic treatment against organophosphate nerve agents in human [96]. In United Sates, HupA will be accessible to AD patients as a nutraceutical or dietary supplement [97]. [2] Kozikowski, A.P. J. Heterocyclic Chem., 1990, 27, 97. [3] Kozikowski, A.P.; Thiels, E.; Tang, X.C.; Hanin, In Advances in Medicinal Chemistry ; Maryanoff, B.E.; Maryanoff, C.A., Eds.; JAI Press: Greenwich, 1992, Vol. 1, pp175-205. A correlation of the affinities of some analogues with their interactions with the protein by docking studies suggested the importance of individual hydrophobic interactions between (-)-HupA and aromatic residues in the active-site gorge of the enzyme. Future Directions References Treatment of Alzheimer’s Disease [4] Ayer, W.A.; Trifonov, L.S. In The Alkaloids, Chemistry and Pharmacology; Cordell, G.A.; Brossi, A., Ed.; Academic Press: San Diego, 1994; Vol. 45, pp233-266. [5] Ji, R.Y. Med. Chem. Res.,1995, 5, 587. [6] Tang, X. C. Acta Pharmacol. Sin. 1996, 17, 481. [7] Liu, J.S.; Yu, C.M.; Zhou, Y.Z. Han, Y. Y. ; Wu, F. W.; Qi, B. F.; Zhu, Y. L. Acta Chim. Sin., 1986, 44, 1035. [8] [9] Current Medicinal Chemistry, 2000, Vol. 7, No. 3 373 [31] Xiong, Z.Q.; Tang, X.C. Pharmacol. Biochem. Behav., 1995, 51, 415. [32] Tang, X.C.; Xiong, Z.Q.; Qian, B.C.; Zhou, Z.F.; Zhang, C.C. In Alzheimer Therapy: Therapeutic Strategies; Giacobini, E., Becker, R., Eds.; Birkhuser: Boston, 1994, pp.113-119. [33] Lin, J.H.; Hu, G.Y.; Tang, X.C. Acta Pharmacol. Sin., 1997, 18, 6. [34] Sun, Q.Q.; Pan, J.L.; Guo, H.M.; Cao, W.Q.; Xu, S.S. Acta Pharmacol. Sin.,1999, 20, 601. Liu, J.S.; Zhu, Y.L.; Yu, C.M.; Zhou, Y.Z.; Han, Y.Y.; Wu, F.W.; Qi, B.F. Can. J. Chem., 1986, 64. 837. [35] Xu, S.S.; Xie, H.B.; Du, Z.W.; Tong, Z.H.; Shi, Q.C.; Lu, K.M.; Li, S.L.; Lin, B. Chin. J. Clin. Pharmacol. Ther., 1997, 2, 1. Yu, C. M.; Tang, C. X.; Liu, J. S.; Han, Y. Y., U.S. Patent, 5, 177, 082, Jan 5. 1993. [36] Xu, S.S.; Gao, Z.X.; Wang, Z.; Du, Z.M.; Xu, W.A.; Yang, J.S.; Zhang, M.L.; Tong, Z. H.; Fang, Y.S.; Chai, X.S.; Li, S.L. Acta Pharmacol. Sin., 1995, 16, 391. [37] Kozikowski, A.P.; Campiani, G.; Tuckmantel, W. Heterocycles, 1994, 39, 101. [10] Geib, S.J.; Tuckmantel, W.; Kozikowski, A.P. Acta Cryst., 1991, C47: 824. [11] Ayer, W.A.; Browne, L.M.; Orszanska, H.; Velenta, Z.; Liu, J. S. Can. J. Chem., 1989, 67, 1538. [38] Camps, P.; Contreras, J. Synth. Commun., 1996, 26, 9. [12] Ellman, G.L.; Courtney, K.D.; Andre, V. Fr.; Featherstone, R.M. Biochem. Pharmaco.l 1961, 7, 88. [39] Kraus, G.A.; Hansen, J.; Vines, D. Synth. Commun., 1992, 22, 2625. Wang, Y.E.; Yue, D.X.; Tang, X.C. Acta Pharmacol. Sin. 1986, 7, 110. [40] Wu, B.G.; Bai, D.L. Chin. Pharm. J., 1995, 30(supplement), 63. [41] Xia, Y.; Kozikowski, A.P. J. Am. Chem. Soc.; 1989, 111, 4116. [42] Kozikowski, A.P.; Reddy, E.R..; Miller, C.P. J. Chem. Soc. Perkin Trans. 1, 1990, 195. [43] Chen, W.P.; Yang, F.Q. Chin. J. Pharmaceut., 1991, 22, 256. [44] Qian, L.G.; Gu, K.J.; Ji, R.Y. Chin. J. Med. Chem., 1992, 2, 1 [13] [14] Tang, X.C.; Zhu, X.D.; Lu, W. H. In Current Research in Alzheimer Therapy; Giacobini, E.; Becker, R., eds.; Taylor & Francies: New York, 1988, 289-293. [15] Cheng, D.H.; Ren, H.; Tang, X.C. NeuroReport, 1996, 8, 97. [16] Tang, X.C.; De Sarno, P.; Sugaya, K.; Giacobini, E. J. Neurosci. Res., 1989, 24, 276. [17] Tang, X.C.; Kindel, G.H.; Kozikowski, A.P.; Hanin, I. J. Enthnopharmacol., 1994, 44, 147. [45] Chassaing, C.; Haudrechy, A.; Langlois, Y. Synth. Commun., 1997, 27, 61. [18] Wang, H.; Tang, X.C. Acta Pharmacol. Sin., 1998, 19, 27. [46] Chen, W.P.; Yang, F.Q. Chin. J. Med. Chem., 1992, 2(1), 34. [19] Cheng, D.H.; Tang, X.C. Pharmacol. Biochem. Behav., 1998, 60, 377. [47] Qian, L.G.; Ji, R.Y. Tetrahedron Lett., 1989, 30, 2089. [48] [20] Laganiere, S.; Coray, J.; Tang, X.C.; Wlfer, E.; Hanin, I. Neuropharmacol. 1991, 30, 763. Kozikowski, A.P.; Xia, Y.; Reddy, E.R..; Tuckmantel, W.; Hanin, I.; Tang, X.C. J. Org. Chem., 1991, 56, 4636. [49] [21] Zhu, X.D.; Giacobini, E. J. Neurosci. Res,. 1995, 41, 828. Kozikowski, A.P.; Campiani, G.; Nacci, V.; Sega, A.; Saxena, A.; Doctor, B.P. J. Chem. Soc. Perkin Trans. 1, 1996, 1287. [22] De Sarno, P.; Pomponi, M.; Giacobini, E.; Tang, X.C.; Williams, E. Neurochem. Res., 1989, 14, 971. [50] Xu, Z.R.; He, X.C.; Bai, D.L. Acta Pharm. Sin., 1996, 31; 258. [51] [23] Decker, M.W.; McGaugh, T.L. Synapse 1991, 7, 151. Kaneko, S.; Yoshino, T.; Katoh, T.; Terashima, S. Tetrahedron, 1998, 54, 5471. [24] Ved, H.S.; Koenig, M.L.; Dave, J.R.; Doctor, B.P. NeuroReport, 1997, 8, 963. [52] Chen, W.P.; Yang, F.Q. Chin. J. Med. Chem., 1995, 5,10. [53] [25] Wang, X.D.; Zhang, J.M.; Yang, Y.H.; Hu, G.Y. Acta Pharmacol. Sin., 1999, 20, 31. Gravel, D.; Benoit, S.; Kumanovic, S.; Sivaramakrishnan, H. Tetrahedron Lett., 1992, 33, 1407. [54] [26] Lu, W.H.; Shou, J.; Tang, X.C. Acta Pharmacol. Sin., 1988, 9, 11. Kozikowski, A.P.; Campiani, G.; Aagaard, P.; McKinney, M. J. Chem. Soc. Chem. Commun., 1993, 860. [27] Ye, J.W.; Cai, J.X.; Wang, L.M.; Tang, X.C. J. Pharmacol. Exp. Ther., 1999, 288, 814. [55] Campiani, G.; Sun, L.Q.; Kozikowski, A.P.; Aagaard, P.; McKinney, M. J. Org. Chem.. 1993, 58, 7660. [28] Han, Y.F.; Tang, X.C. In Alzheimer Disease: From Molecular Biology to Therapy; Becker, R., Giacobini, E., Eds.; Birkhuser: Boston, 1996, 245-250. [56] Yamada, F.; Kozikowski, A.P.; Reddy, E.R.; Pang, Y.P.; Miller, J.H.; McKinney, J. Am. Chem. Soc., 1991, 113, 4695. [57] [29] Wang, T.; Tang, X.C. Eur. J. Pharmacol., 1998, 349, 137. Chen,W. P. Ph.D. Dissertation, Shanghai Institute of Pharmaceutical Industry, 1993. [30] Xiong, Z.Q.; Han, Y.F.; Tang, X.C. NeuroReport, 1995, 6, 2221. 374 Current Medicinal Chemistry, 2000, Vol. 7, No. 3 [58] Kaneko, S.; Yoshino, T.; Katoh, T.; Terashima, S. Heterocycles, 1997, 46, 27 [59] Kaneko, S.; Yoshino, T.; Katoh, T.; Terashima, S. Tetrahedron: Asymmetry, 1997, 8, 829. Bai et al. [79] Campiani,G.; Kozikowski, A.P.; Wang, S.; Ming, L.; Nacci, V.; Saxena, A.; Doctor, B. P. Bioorg. Med. Chem. Lett., 1998, 8, 1413. [80] Kaneko, S.; Nakajima, N.; Shikano, M.; Katoh,T.; Terashima, S. Tetrahedron, 1998, 54, 5485. [81] Kaneko, S.; Shikano, M.; Katoh, T.; Terashima, S. Synlett, 1997, 447. [82] Kaneko, S.; Nakajima, N.; ShiKano, M.;Katoh, T. Terashima, S. Bioorg. Med. Chem. Lett., 1996, 6, 1927. [83] Kaneko, S.; Nakajima, N.; Katoh, T.; Terashima, S. Chem. Pharm. Bull., 1997,45, 43. [84] Zeng, F.X.; Jiang, H.L.; Tang, X.C. Chen, K.X.; Ji, R.Y. Bioorg. Med. Chem. Lett., 1998, 8, 1661. [85] Badia, A.; Banos, J.E.; Camps, P. Contreras, J.; Gorbig, D. M.; Munoz-torrero, D.; Simon, M,; Vivas, N.M. Bioorg. Med. Chem., 1998, 6, 427. [60] He, X.C.; Wang, B.; Bai, D.L. Tetrahedron Lett., 1998, 39, 411. [61] Liu, J.; Zhang, H.Y.; Tang X.C.; Wang, B.; He, X.C,; Bai, D.L. Acta Pharmacol. Sin., 1998, 19(5), 413. [62] McKinney, M.; Miller, J.H.; Yamada, F.; Tuckmantel, W.; Kozikowski, A.P. Eur.J. Pharmacol.,1991, 203, 303. [63] He, X.C.; Wang, Z.; Li, Y.L.Xu, Z.R.; Bai, D. L. Chin. Chem. Lett., 1993, 4, 597. [64] He, X.C.; Wang, Z.; Li, Y.L.; Bai, D.L. Chin. J. Med. Chem., 1994, 4, 257. [65] Fink, D.M.; Bores, G.M.; Effland, R.C.; Huger, F.P.; Kurys, B.E.; Rush, D.K.; Selk, D.E. J.Med.Chem., 1995, 38, 3645. [66] Kozikowski, A.P.; miller, C.P.; Yamada, F.; Pang, Y.P.; Miller, J.H.; Mckinney, M.; Ball, R.G. J. Med. Chem., 1991, 34, 3399. [86] Camps, P.; Contreras, J.; Font-Bardia, M.; Morral J.; MunozTorrero D.; Solans X. Tetrahedron: Asymmetry, 1998, 9, 835. [67] Wu, B.G. Ph. D. Dissertation, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 1996. [87] Kozikowski, A.P.; Ding, Q.; Saxena, A.; Doctor, B. P. Bioorg. Med. Chem. Lett., 1996, 6, 259. [68] Wu, B.G.; Zhen W.P.; Bo, Y.X.; He,X.C.; Bai, D. L. Chin. Chem. Lett., 1995, 6,193. [88] Kozikowski, A.P.; Campiani, G.; Sun, L.Q.; Wang, S.; Saxena,A; Doctor, B. P. J. Am. Chem. Soc., 1996, 118, 11357. [69] Tang, X.C.; Xu, H.; Feng, J.; Zhou,T.X.; Liu, J, S. Acta Pharmacol. Sin., 1994, 15, 107. [89] Kozikowski, A.P.; Prakash, K.R.C.; Saxena, A.; Doctor, B. P. J. Chem. Soc. Chem. Commun., 1998, 1287. [70] Xiong, Z.Q.; Tang, X.C. ; Lin, J. L.; Zhu, D. Y. Acta Pharmacol. Sin., 1995, 16, 21. [90] Pang, Y. P.; Kozikowski, A.P. J. Computer-Aided Mol. Design, 1994, 8, 669. [71] Xia, Y.; Reddy, E.R.; Kozikowski, A.P. TetrahedronLett., 1989, 25, 3291. [91] Jiang, H. L.; Chen, K. X.; Chen, J. Z. Tang, Y.; Wang, Q. M.; Li, Q.; Shen, X.; Ji, R. Y. Chin. Chem. Lett., 1996, 7, 253. [72] Kozikowski, A.P.; Yamada, F.; Tang, X.-C.; Hanin, I. Tetrahedron Lett., 1990, 31, 6159. [92] Liu, D. X.; Jiang, H. L.; Wang, Q. M.; Chen, K. X.; Ji, R. Y. Bioorg. Med. Chem. Lett., 1998, 8, 419. [73] Li, Y. L.; He, X. C.; Bai, D. L. Chin. J. Med. Chem., 1996, 6, 185. [93] [74] Xu, Z. R.; He, X. C.; Bai, D. L. Acta Pharm. Sin., 1996, 31, 364. Zhu, W. L.; Jiang, H. L.; Chen, J.Z.; Gu, J. D.; Lin, D. X.; Lin, M. W.; Chen, K X.; Ji, R. Y. Acta. Chim. Sin.. 1998, 56, 233. [75] Wang, B.; He, X.C.; Bai, D.L. Acta Pharm. Sin., 1999, in press. [94] Ashani Y.; Grunwald J.; Kronman C.; Velan B.; Shafferman A. Mol. Pharmacol., 1994, 45, 555. [76] Kozikowski, A.P.; Yamada, F.; Pang Y. P. Tetrahedron Lett., 1992, 33, 2653. [95] Raves, M. L.; Harel, M.; Pang, Y. P.; Sil\man, I.; Kozikowski, A.P.; Sussman, J. L. Nature Struct. Biol ., 1997, 4, 57. [77] Kozikowski, A.P.; Campiani, G.; Saxena, A.; Doctor, B. P. J. Chem .Soc. Chem. Commun., 1995, 283. [96] Grunwald, J.; Raveh, L.; Doctor, B. P.; Ashani, Y. Life Sci.,1994, 54, 991. [78] Kozikowski, A.P.; Tuckmantel, W.; Saxena, A.; Doctor, B. P. Helv. Chim. Acta., 1994, 77, 1256. [97] Bormau, S. C&E News, 1998, June 1, 45.