Simple Synthesis of Sulfonyl Amidine-Containing Glucosidase Inhibitors by a Chemoselective Coupling Reaction Between D-Gluconothiolactam and Sulfonyl Azides

In this report, we describe a simple synthesis of gluconoamidinylsulfones as a new class of potential inhibitors toward glycan processing enzymes. Gluconoamidinylsulfones have a glucose-based sulfonyl amidine skeleton, thus would form a distorted half-chair conformation with positive charge, which is analogous to transition state in the enzymatic process. A chemoselective coupling reaction between thioamide and sulfonyl azide enabled one-step synthesis of the iminosugar derivatives from commercially available D-gluconothiolactam in a protection-free manner. The phenylsubstituted gluconoamidinylsulfone displayed high inhibitory ability toward and -glucosidases with Ki values of 13.9 and 8.2 M, respectively, resulting that gluconoamidinylsulfones would be expected to entry in a new class of promising potential inhibitors toward various glycan-processing enzymes. Short Research Article


INTRODUCTION
Iminosugars have been developed as small molecule inhibitors for various types of glycan processing enzymes [1,2]. These inhibitors are crucial for detailed understanding of glycobiology as well as for development of new promising medicines prescribed in the treatmen diabetes, Gaucher's disease, influenza infection, HIV, hepatitis, cancer, etc., involving glycoside bond cleavage pathway [3]. Chemical structures of glycosidase-inhibiting natural products afford valuable inspirations to design their synthetic analogues. Miglitol [4,5] and miglustat [6,7] are representative therapeutic iminosugars for diabetes and type-1 Gaucher's disease (GD1), respectively, derivatized from a natural glucosidase inhibitor, 1-deoxynojirimycin [8] (Fig. 1a). On the other hand, rationally designed iminosugars based on the mechanism of the enzymatic hydrolysis reaction have also been synthesized as candidates of lead compounds for glycosidase inhibition [3,9]. A mechanism based general strategy is transition analogue of the enzymatic process. Shape and charge are the key points to mimic the transition state, that is, a partially planar structure in distorted half-chair conformations and delocalized positive charge around the anomeric center of the glycosyl oxocarbenium intermediates (Fig.  1b), the species considered close to the transition state of the have been developed as smallmolecule inhibitors for various types of glycan processing enzymes [1,2]. These inhibitors are crucial for detailed understanding of glycobiology as well as for development of new promising medicines prescribed in the treatment of diabetes, Gaucher's disease, influenza infection, HIV, hepatitis, cancer, etc., involving glycosidebond cleavage pathway [3]. Chemical structures inhibiting natural products afford valuable inspirations to design their synthetic ues. Miglitol [4,5] and miglustat [6,7] are representative therapeutic iminosugars for 1 Gaucher's disease (GD1), respectively, derivatized from a natural deoxynojirimycin [8] nd, rationally designed iminosugars based on the mechanism of the enzymatic hydrolysis reaction have also been synthesized as candidates of lead compounds for glycosidase inhibition [3,9]. A mechanismbased general strategy is transition-state the enzymatic process. Shape and charge are the key points to mimic the transition state, that is, a partially planar structure in chair conformations and delocalized positive charge around the anomeric center of the glycosyl oxocarbenium termediates (Fig.  1b), the species considered close to the transition state of the hydrolysis process, are important common factors for designing potent glycosidase inhibitors. One of the successful inhibitors designed in this way is gluconoamidines (Fig. 1c) that have both a partially flattened geometry and positive charge at a basic amidine center under physiological conditions, which exhibited remarkable inhibition abilities toward glycosidases [10][11][12].
We recently reported a chemoselectiv between thioamides and sulfonyl azides to yield sulfonyl amidines without side reaction even under the coexistence of hydroxy, amino, and carboxy groups [13,14] (Scheme 1a). The reaction proceeds in various solvents without any activation additives. In the previous report [13], a six-membered cyclic thioamide, 2with methyl and phenyl sulfonyl azides showed good reactivity in EtOH or H 2 O (Scheme 1b). Inspired by the structural similarity of 2 thiopiperidone with iminosugars, we that commercially available D-gluconothiolactam would afford sulfonyl amidine derivatives of D gluconolactam, namely gluconoamidinylsulfones ( Fig. 1d), by the coupling reaction with sulfonyl azides. Here we report simple synthesis of gluconoamidinylsulfones 1 and 2 as a new class of potential inhibitors for glycosidase. In addition, conventional inhibitory assay of these compounds toward -and -glucosidases is also described. hydrolysis process, are important common factors for designing potent glycosidase inhibitors. One of the successful inhibitors midines (Fig. 1c) that have both a partially flattened geometry and positive charge at a basic amidine center under physiological conditions, which exhibited remarkable inhibition abilities toward We recently reported a chemoselective reaction between thioamides and sulfonyl azides to yield sulfonyl amidines without side reaction even under the coexistence of hydroxy, amino, and carboxy groups [13,14] (Scheme 1a). The reaction proceeds in various solvents without any ves. In the previous report [13], a -thiopiperidone, with methyl and phenyl sulfonyl azides showed O (Scheme 1b). Inspired by the structural similarity of 2thiopiperidone with iminosugars, we considered gluconothiolactam would afford sulfonyl amidine derivatives of Dgluconolactam, namely gluconoamidinylsulfones 1d), by the coupling reaction with sulfonyl azides. Here we report simple synthesis of as a new class of potential inhibitors for glycosidase. In addition, conventional inhibitory assay of these glucosidases is also

1
H and 13 C NMR spectra were obtained at 400 and 100 MHz, respectively, on a JEOL ECX 400P spectrometer. ESI-HRMS analyses were conducted on a Thermo LTQ Orbitrap XL ETD mass spectrometer.

Materials
D-Gluconothiolactam 3 is commercially available from FCH Group (order number: FCH3937573) but takes long time around 8 weeks to arrive, thus 3 was synthesized by simple procedures shown in Scheme 3. D-Gluconolactam phenyl sulfonyl azide [14], and mesyl azide were prepared according to literature procedures. Other materials including dehydrate grade solvents were all commercially available (Wako Pure Chemical Industries, Ltd. and Tokyo Chemical Industry Co., Ltd.).

AND METHODS
C NMR spectra were obtained at 400 and 100 MHz, respectively, on a JEOL ECX-HRMS analyses were conducted on a Thermo LTQ Orbitrap XL ETD is commercially available p (order number: FCH3937573) but takes long time around 8 weeks to arrive, was synthesized by simple procedures Gluconolactam 5 [15,16], phenyl sulfonyl azide [14], and mesyl azide [17] were

N-Sulfonylmethyl D-gluconoamidine 2
A mixture of 3 (5.8 mg, 0.03 mmol azide (17.8 mg, 0.15 mmol) in distilled water (1 mL) was vigorously stirred at room temperature for 72 h. After removal of the solvent, the residue was purified by silica gel PTLC with a development solvent of CHCl 3 :MeOH = 2:1 to give 2 as a colorless solid (6.2 mg, 82%

D-Gluconothionolactam 3
AcCl (221 mg, 2.82 mmol) was added slowly and dropwisely to a MeOH (6 mL) solution of 7 (153 mg, 0.235 mmol) at room temperature and the reaction mixture was stirred for 6 h at that temperature. After removal of the solvent, the residue was dissolved in water.

Enzymatic Assays
Glucosidase inhibitory assays were carried out for 1, 2, and 1-deoxynojirimycin by means of a conventional spectrometric method [18] at 37°C using 0.01 M KH 2 PO 4 /K 2 HPO 4 buffer solution (pH 6.8). As a substrate, 4-nitrophenyl--glucopyranoside (-NGP) toward glucosidase from yeast Saccharomyces cereviceae or 4-nitrophenyl--glucopyranoside (-NGP) toward β−glucosidase from sweet almonds was selected. In advance to the assays, stock solutions of enzymes (2U/mL for glucosidase and 6U/mL for -glucosidase), substrates (1 mM), and inhibitors (1, 5, 10, 25, 50, 100, and 1000 M) were prepared by diluting with the buffer. Mixed solutions of an appropriate amount of -NGP or -NGP substrate with various concentrations of each inhibitor solution were poured in cells of a 96-well microplate. Addition of the enzyme solution to the cells immediately progressed the enzymatic reaction, affording 4-nitrophenolate anion as a cleavage product that can be monitored at 405 nm of absorbance by a microplate reader (FilterMax F5 TM ). The absorption data were simultaneously collected from 1 to 25 min at 2 min intervals. In these assays, the final concentrations of the substrates were 16.7, 33.3, 50.0, 66.7, and 83.3 M for -glucosidase whereas 30.0, 60.0, 90.0, 120, and 150 M in absence or presence of the inhibitors. The inhibition constants (K i ) were determined by using the slopes of Lineweaver−Burk plots and double reciprocal analysis. All experiments were conducted in duplicate and obtained data were averaged.

RESULTS AND DISCUSSION
One-step syntheses of gluconoamidinylsulfones 1 and 2 were performed by simply mixing of commercially available D-gluconothiolactam 3 with phenyl sulfonyl azide or mesyl azide in water (Scheme 2). The reaction mixture was vigorously stirred for an appropriate reaction time, affording gluconoamidinylsulfones 1 and 2 in 68% and 86% isolated yields, respectively. These compounds were characterized by means of 1 H NMR, 13 C NMR and ESI-HRMS (electrospray ionization high resolution mass spectrometry) (Fig. 2). Because regio-and stereoselective sugar derivatization generally tends to be a complicated multistep synthesis involving protection and deprotection steps, it is noteworthy that, by using this coupling reaction, one-step synthesis in a protection-free manner generated a new class of potential inhibitors directly. is commercially available, however, it seems to take about two months to separately according to literature procedures [10,15,16] with minor modification (Scheme 3). Gluconolactam initially synthesized from a readily purchasable in a manner similar to those reported [15,16]. Unfortunately, direct thioamidation of modification (Scheme 3). Gluconolactam 5 was initially synthesized from a readily purchasable 4 in a manner similar to those reported [15,16]. Unfortunately, direct thioamidation of 5 by Lawesson's reagent (LR) was failed to produce 3. Silylation of 5 with TBDMSCl followed by thioamidation with LR was a successful route, affording a thiolactam derivative 7 the silyl-protection with AcCl in MeOH generated D-gluconothiolactam 3 quantitatively.
Next, enzymatic assay was performed for gluconoamidinylsulfones toward glucosidases by means of a conventional spectroscopic methodology with 4-nitrophenyl glucopyranoside as a substrate.  Lawesson's reagent (LR) was failed to produce with TBDMSCl followed by thioamidation with LR was a successful route, 7. Removal of protection with AcCl in MeOH generated uantitatively.
Next, enzymatic assay was performed for gluconoamidinylsulfones toward glucosidases by means of a conventional spectroscopic nitrophenyl -or glucopyranoside as a substrate. Table 1 8], the obtained K i values in this study fall within the range of these values, indicating that our assay condition and analysis procedure would be appropriate and reliable. Phenyl exhibited strong inhibition against both glucosidases with K i of 13.9 a respectively, being more high inhibitory ability than that of 1-deoxynojirimycin. On the other hand, methyl-substitution showed weak inhibition against both -and -glucosidases.
Glycosyl hydrases have several subsites in t binding pocket such as glycon and aglycon binding sites [18]. The aglycon subsites are usually made up of several hydrophobic residues such as phenylalanine, tyrosine, and tryptophan surrounding a ligand saccharide. Therefore, as an old trick, connecting a hydrophobic glycon analogue to aglycon mimics like iminosugars has been used for glycol-modification [19,20]. The strong glucosidase inhibition of the gluconoamidinylsulfone 1 seemly caused by the aglycon phenyl-group that might fit with the me 3. Synthesis of D-gluconothiolactam Reagents  values in this study fall within the range of these values, indicating that our assay condition and analysis procedure would be appropriate and reliable. Phenyl-substituted 1 exhibited strong inhibition against both -and of 13.9 and 8.2 μM, respectively, being more high inhibitory ability deoxynojirimycin. On the other substitution showed weak inhibition glucosidases.
Glycosyl hydrases have several subsites in their binding pocket such as glycon and aglycon binding sites [18]. The aglycon subsites are usually made up of several hydrophobic residues such as phenylalanine, tyrosine, and tryptophan surrounding a ligand saccharide. Therefore, as ing a hydrophobic glycon analogue to aglycon mimics like iminosugars has modification [19,20]. The strong glucosidase inhibition of the seemly caused by the group that might fit with the aglycon subsite of -and -glucosidases. From the analysis of amino acid sequence and threedimensional structure, there are several phenylalanine residues around aglycon subsite of glucosidases [21]. In addition to the geometrically and electrostatically well-fitting of the gluconoamidine skeleton at the glycon subusite, hydrophobic and - interactions at the aglycon site might affect effectively, at least in part, to the inhibition activity of phenyl-substituted 1, while the methyl group in 2 exhibited weak interaction with the subsite. Although only two gluconoamidinylsulfones and glucosidases were used in this report, the results indicate that further study may reveal the potential of this skeleton as a new class of inhibitors toward various glycan-processing enzymes such as mannosidases, galactosidases, and transferases, by taking advantage of the simple synthetic approach.

CONCLUSION
The chemoselective coupling reaction of Dgluconothiolactam and sulfonyl azides successfully generated gluconoamidinylsulfones in a simple synthetic manner. The phenylsubstituted gluconoamidinylsulfone showed high inhibitory ability toward α-and β-glucosidases so that gluconoamidinylsulfones would be expected to entry in a new class of promising potential inhibitors toward various glycosidases. Making the best use of the synthetic advantage, expansion of the compound library of gluconoamidinylsulfones and the following enzyme assays toward wide variety of glycosyl hydrases and transferases are currently in progress.