Inhibition of glycosidases has great potential in the quest for highly potent and specific drugs to treat diseases such as diabetes, cancer, and viral infections. One of the most effective ways of designing such compounds is by mimicking the transition state. Here we describe the structural, kinetic, and thermodynamic dissection of binding of two glucoimidazole-derived compounds, which are among the most potent glycosidase inhibitors reported to date, with two family 1 beta-glycosidases. Provocatively, while inclusion of the phenethyl moiety improves binding by a factor of 20-80-fold, this does not appear to result from better noncovalent interactions with the enzyme; instead, improved affinity may be derived from significantly better entropic contributions to binding displayed by the phenethyl-substituted imidazole compound.

It was shown that retaining b-glucosidases and galactosidases of families 1 ± 3 feature a strong interaction between C(2)OH of the substrate and the catalytic nucleophile. An analogous interaction can hardly take place for retaining b-mannosidases. A structureÀactivity comparison between the inhibition of the b-glucosidase from Caldocellum saccharolyticum (family 1) and b-glucosidase from sweet almonds by the gluco-imidazoles 1 ± 6, and the inhibition of snail b-mannosidase by the corresponding manno-imidazoles 8 ± 13 does not show any significant difference, suggesting that also the mechanisms of action of these glycosidases do not differ significantly. For this comparison, we synthesized and tested the manno-imidazoles 9 ± 13, 28, 29, 32, 35, 40, 41, 43, 46, 47, and 50. Among these, the alkene 29 is the strongest known inhibitor of snail b-mannosidase (K i 6 nm, non-competitive); the aniline 35 is the strongest competitive inhibitor (K i 8 nm).Introduction. ± The strong inhibition of b-glucosidases by imidazoles of type 1 [1 ± 3] has been rationalized by the similarity of shape of the inhibitor and of the putative reactive intermediate, an oxycarbenium cation, and by the cooperative interaction of the imidazole with the catalytic nucleophile and acid [4]. A correlation between the inhibition constant and the pK value of the C(2)-and C(3)-acetamido imidazoles 7 and 14 ± 16, and by related azoles has established that substituents at C(3) lower the inhibitory activity [5]. The structureÀactivity relation (SAR) resulting from varying the C(2)-substituents has been studied in detail [6]. It was shown that the HOCH 2 group at C(2) in 2, and particularly the flexible hydrophobic PhCH 2 CH 2 group in 3 lead to an improved inhibition, with K i values as low as 0.1 nm (against Caldocellum saccharolyticum b-glucosidase) 1 ). The C(2)-substituents affect both the strength and the type of the inhibition (competitive or mixed, with a varying between 2.5 and 15).Legler and Withers evidenced that the C(2)OH group of b-glucosides and bgalactosides interacts with the catalytic nucleophile of the retaining b-glucosidases and b-galactosidases of families 1 [8], 2 [9], and 3 [10] [11]. 2-Deoxy-and 2-deoxy-2-fluorob-d-glucosides and -galactosides are cleaved much less readily than the parent substrates, the rate-determining step being deglycosylation of the enzyme. The transition state for this reaction is considered very similar to that of the enzyme glycosylation [9], and the most important interaction in the transition state was considered with the C(2)OH 2 ). That 2-deoxyglucosides are cleaved less readily than the parent substrates is surprising, as the OH group at C(2) is known to destabilize an

The gluco-configured analogue 15 of nagstatin (1) and the methyl ester 14 were synthesized via condensation of the thionolactams 17 or 18 with the b-amino ester 19. The silyl ethers 20 and 21 resulting from 17 were desilylated to 22 and 23; these alcohols were directly obtained by condensing 18 and 19. The attempted substitution of the C(8)ÀOH group of 22 by azide under Mitsunobu conditions led unexpectedly to the deoxygenated a-azido esters 24. The desired azide 25 was obtained by treating the manno-configured alcohol 23 with diphenyl phosphorazidate. The azide was transformed to the debenzylated acetamido ester 14 that was hydrolyzed to the nagstatin analogue 15. The imidazole-2-acetates 14 and 15 are nanomolar inhibitors of the Nacetyl-b-glucosaminidases from Jack beans and from bovine kidney, submicromolar to micromolar inhibitors of the b-glucosidase from Caldocellum saccharolyticum, and rather weak inhibitors of the snail b-mannosidase. In all cases, the ester was a stronger inhibitor than the corresponding acid. As expected from their glucoconfiguration, both imidazopyridines 14 and 15 are stronger inhibitors of the b-N-acetylglucosaminidase from bovine kidney than nagstatin.Introduction. ± Nagstatin (1), a strong inhibitor of several hexosaminidases [1 ± 4], is a N-acetylgalactosamine-derived tetrahydropyridoimidazole-2-acetic acid [5]. Its inhibitory activity is essentially associated with the imidazole ring and not with the carboxymethyl substituent [2], although substituents on the imidazole ring may strongly affect the inhibition of b (and a-)-glycosidases [6 ± 8]. In the preceding paper [9], we described the influence of the hydrophobic character of C(2)-methyl ester and carboxylic acid substituents on the inhibition of b-glycosidases. Imidazole-2-propionates 10 ± 13 are stronger inhibitors of the b-glucosidase from Caldocellum saccharolyticum and of the b-mannosidase from snail than imidazole-2-acetates 6 ± 9, and these are stronger than imidazole-2-carboxylates 2 ± 5. There is a parallel sequence of inhibitory activity for the methyl esters and the corresponding acids, with the esters being stronger inhibitors.

The acetylcholinesterase inhibitor (-)-huperzine A was synthesized from (S)-4-hydroxycyclohex-2-enone in 17 steps by a route that involved two cyclobutane fragmentations. The first of these employed a retro-aldol cleavage to generate the α-pyridone ring of huperzine A, and the second invoked a novel intramolecular aza-Prins reaction in tandem with stereocontrolled scission of a cyclobutylcarbinyl cation to create the aminobicyclo[3.3.1]nonene framework of the natural alkaloid.

No abstract

The gluco-and manno-tetrahydropyridoimidazole-2-acetates and -acetic acids 16 and 17, and 20 and 21, respectively, were synthesized by condensation, in the presence of HgCl 2 , of the known thionolactam 26 with the b-amino ester 25 that was obtained by addition of AcOMe to the imine 22, followed by debenzylation. The resulting methyl esters 16 and 20 were hydrolyzed to the acetic acids 17 and 21. The (methoxycarbonyl)-imidazole 14 and the acid 15 were obtained via the known aldehyde 29. The imidazoles 14 ± 17, 20, and 21 were tested as inhibitors of the b-glucosidase from Caldocellum saccharolyticum, the a-glucosidase from brewers yeast, the b-mannosidase from snail, and the a-mannosidase from Jack beans (Tables 1 ± 3). There is a similar dependence of the K i values on the nature of the C(2)-substituent in the gluco-and manno-series. With the exception of 19, manno-imidazoles are weaker inhibitors than the gluco-analogues. The methyl acetates 16 and 20 are 3 ± 4 times weaker than the methyl propionates 5 and 11, in agreement with the hydrophobic effect. The gluco-configured (methoxycarbonyl)-imidazole 14 is 20 times weaker than the methyl acetate 16, reflecting the reduced basicity of 14, while the manno-configured (methoxycarbonyl)-imidazole 18 is only 1.2 times weaker than the methyl acetate 20, suggesting a binding interaction of the MeOCO group and the b-mannosidase. The carboxylic acids 6, 12, 15, 17, 19, and 21 are weaker inhibitors than the esters, with the propionic acids 6 and 12 being the strongest and the carboxy-imidazoles 15 and 19 the weakest inhibitors. The manno-acetate 21 inhibits the b-mannosidase ca. 8 times less strongly than the propionate 12, but only 1.5 times more strongly than the carboxylate 19, suggesting a compensating binding interaction also of the COOH group and the b-mannosidase. The a/b selectivity for the gluco-imidazoles ranges between 110 for 15 and 13.4´10 3 for 6; the manno-imidazoles are less selective. The methyl propionates proved the strongest inhibitors of the a-glucosidase (IC 50 (5) 25 mm) and the a-mannosidase (K i (11) 0.60 mm). . The carboxymethyl group at C(2) does not appear to be important for the inhibitory activity, since debranched nagstatin analogues and related C(2)-unsubstituted glucose-and mannose-derived tetrahydroimidazopyridines are also potent inhibitors of b-glycosidases [5 ± 7]. These observations and the rationalisation of the inhibitory activity of these and analogous azoles ± their similarity with the putative reaction intermediate, and the cooperative interaction of the azole ring with the catalytic acid and nucleophile ± indicated a negligible role for the substituents at C(2) [8 ± 10]. Subsequently, however, it was shown that C(2)-substituents of tetrahydropyridoimidazoles (and corresponding substituents of related azoles and pyrroles) can strongly affect the inhibitory properties, either indirectly, by competing with the catalytic acid for H-bonding to the glycosidic heteroatom, and/or

An asymmetric total synthesis of the nootropic alkaloid (-)-huperzine A was completed using a cascade sequence initiated by an intramolecular aza-Prins reaction and terminated by a stereoelectronically guided fragmentation of a cyclobutylcarbinyl cation as the key step in assembling the bicyclo[3.3.1]nonene core of the natural product. Intramolecular [2 + 2]-photocycloaddition of the crotyl ether of (S)-4-hydroxycyclohex-2-enone afforded a bicyclo[4.2.0]octanone containing an embedded tetrahydrofuran in which the cyclohexanone moiety was converted to a triisopropylsilyl enol ether and functionalized as an allylic azide. The derived primary amine was acylated with α-phenylselenylacrylic acid, and the resulting amide was reacted with trimethylaluminum to give a [2 + 2]-cycloadduct, which underwent retroaldol fission to produce a fused α-phenylselenyl δ-lactam. Periodate oxidation of this lactam led directly to an α-pyridone, which was converted to a fused 2-methoxypyridine. Reductive cleavage of the activated "pyridylic" C-O bond in this tetracycle and elaboration of the resultant hydroxy ketone to a diketone was followed by chemoselective conversion of the methyl ketone in this structure to an endo isopropenyl group. Condensation of the remaining ketone with methyl carbamate in the presence of acid initiated the programmed cascade sequence and furnished a known synthetic precursor to huperzine A. Subsequent demethylation of the carbamate and the methoxypyridine, accompanied by in situ decarboxylation of the intermediate carbamic acid, gave (-)-huperzine A.

The D‐manno‐tetrahydroimidazopyridine‐2‐phosphonate 11 was prepared via a high‐yielding Pd(PPh3)4‐catalysed diphenylphosphonylation of the manno‐iodoimidazole 12, followed by transesterification to the diethyl phosphonate 14 and dealkylation, providing 11 in eight steps from the thionolactam 1 and in an overall yield of 15%. Alternatively, a more highly convergent synthesis based on the HgCl2/Et3N‐promoted condensation of the thionolactam 1 with the α‐aminophosphonate 24 in THF led to 11 in four steps and in the same overall yield. In the presence of HgCl2/Et3N, the thionolactam 1 reacted at 80° with 2‐methoxyethanol to provide 66% of a 64 : 36 mixture of the gluco‐ and manno‐iminoethers 29/30. Performing the reaction at 22° yielded preferentially the gluco‐isomer 29 (86%, 84 : 16).