All reagents and solvents used in this study were of the highest commercially available grade and were used as obtained. (CH₃COO)₂Zn・2H₂O, ZnCl₂, salicylaldehyde, β-alanine, ethylenediamine, 2-hydroxy-1-naphtalde- hyde, LiOH・H₂O, HEPES, NaOH, KH₂PO₄, dithiothreitol, α-glucosidase (from Saccharomyces sp.), maltose, d- (+)-glucose, powdered acacia, and a Glucose C-II Test kit were purchased from Wako Pure Chemical Industries (Osaka, Japan). O-phenylenediamine was purchased from Kanto Chemical Co., Inc. (Tokyo, Japan). N, N-dime- thylethylenediamine was purchased from Tokyo Chemical Industry Co. (Tokyo, Japan). EDTA・2Na・2H₂O and Triton X-100 were purchased from Nacalai Tesque, (Kyoto, Japan). Rat small intestine acetone powder was obtained from Sigma Chemical Co. (St. Louis, MO, USA). Twelve-week-old ddYmice, which are non-inbred mice maintained in a closed colony, were purchased from Shimizu Laboratory Supplies Co. (Kyoto, Japan). All the mice were maintained on a 12 h light/dark cycle in our temperature-controlled central animal facility for breeding under fixed condition. The animal study was approved by the Experimental Animal Research Committee at the Kyoto Pharmaceutical University (KPU) and was performed according to the Guidelines for Animal Experimentation.

We synthesized four Schiff base ligands and their Zn(II) complexes (Schemes 2-5). The syntheses of these ligands were performed as described in previous reports [9] [21] [22] . The intended Zn(II) complexes were readily

Scheme 1. Chemical structure of Schiff base.

Scheme 2. Synthesis of N-sβ and [N-sβ-Zn].

Scheme 3. Synthesis of N-bsE and [N-bsE-Zn].

Scheme 4. Synthesis of N-bsP and [N-bsP-Zn].

prepared by adding (CH₃COO)₂Zn・2H₂O or ZnCl₂ to methanolic solutions of the various ligands at room temperature. The Zn(II) complexes were purified with water, methanol, and acetone to obtain [N-sβ-Zn], [N-bsE-Zn], [N-bsP-Zn], and [Zn₂{(DMN)₂Cl₂}].

The rat small intestine acetone powder was suspended in a 10 mM sodium phosphate buffer (pH 6.8). The suspension was homogenized, sonicated for 30 min and followed by the addition of 2% Triton X buffer (pH 7.0)

Scheme 5. Synthesis of DMN and [Zn₂{(DMN)₂Cl₂}].

containing 3 mM EDTA-2Na and 1 mM dithiothreitol (DTT) in the ice bath and centrifugation for 60 min at 20,000 g (4˚C). The supernatant was subjected to the ammonium sulfate precipitation. The precipitates were collected and dialyzed. The resulting solution was used for the assay.

Solutions of ligands or Zn(II) complexes at various concentrations were prepared, and their α-glucosidase inhibitory effects were evaluated using a modified Dahlqvist method [23] . The substrate and test solution were mixed and incubated at 37˚C for 5 min. A solution containing α-glucosidase enzyme (5 unit/ml) from Saccharomyces sp. or rat small intestine was added continuously and incubated at 37˚C for 1 h. After incubation, the reactions were terminated by heating at 90˚C for 5 min. The glucose concentration was determined by using a Glucose C-II Test kit.

Twelve-week-old ddY mice were fasted for 6 h and were then orally administered one of the test solutions. After 30 min, a 5% acacia solution of maltose or glucose was orally administered [24] . Blood samples were obtained from the tail vein at 0, 15, 30, 60, and 90 min. The blood glucose levels were measured using a glucose oxidase method (Glucocard; Arkray, Kyoto, Japan).

The mode of α-glucosidase inhibition by [Zn₂{(DMN)₂Cl₂}] was determined as reported elsewhere [25] [26] . The substrate and test solution were mixed and incubated at 37˚C for 5 min. Next, an α-glucosidase enzyme solution was added and incubated at 37˚C for 30 min. After incubation, the reactions were terminated by heating at 90˚C for 5 min. The glucose concentration was determined by using a Glucose C-II Test kit. Substrate solutions containing sucrose at concentrations of 75, 150, and 300 mM (Saccharomyces sp. experiment) and of 15, 30, and 60 mM (rat small intestine experiment), and were prepared. [Zn₂{(DMN)₂Cl₂}], which showed the highest activity in vitro, was used the IC50 value in in vitro study.

All experimental data are expressed as the mean ± standard derivation (SD). Statistical analyses of the in vitro data were performed using Student’s t-test. Statistical analyses of the in vivo data were performed using one-way analysis of variance (ANOVA), followed by Dunnett multiple comparison post-hoc tests. Differences were considered to be statistically significant when p values were < 0.01 or < 0.05, as noted.

Four Schiff base ligands and their Zn(II) complexes were characterized by several physicochemical methods. These data are shown in Table 1 and Table 2. For elemental analysis, both calculated and measured values of the percent concentration of C, H, and N were identical and within the estimated range of experimental error. In the IR spectra, we observed the frequencies due to the ν_C=N of Zn(II) complexes with four Schiff bases. And we observed the parent peak of these Zn(II) complexes in the Mass Spectra. When we measured NMR spectrometry, the Zn(II) ion induced the ¹³C NMR chemical shift changes of the N-sβ, as shown in Table 2. [N-sβ-Zn] indicated a downfield shift ∆δ = 4.76, 10.57, 4.71, for C(1), C(5), C(7), respectively. This results revealed that the Zn(II) ion binds exclusively with the O and N atoms of N-sβ (Table 2). From these data, we concluded that the structure of Zn(II) complexes with four Schiff bases were likely (Schemes 2-4).

In in vitro experiments involving yeast α-glucosidase, the Zn(II) complexes exhibited α-glucosidase inhibitory effects in the following order: (CH₃COO)₂Zn・2H₂O (2.41 ± 0.19 μM) > [N-sβ-Zn] (2.89 ± 0.91 μM) > [N- bsE-Zn] (3.10 ± 0.74 μM) > zinc gluconate (3.94 ± 0.34 μM) > [Zn₂{(DMN)₂Cl₂}] (4.06 ± 0.45 μM) > [N-bsP- Zn] (16.1 ± 2.11 μM). On the other hand, the Schiff base ligands did not show α-GI effects by themselves (Figure 2(a) and Figure 3(a)).

In in vitro experiments involving rat intestinal α-glucosidase, the Zn(II) complexes exhibited α-glucosidase inhibitory effects in the following order: [Zn₂{(DMN)₂Cl₂}] (86.0 ± 26.4 μM) > [N-sβ-Zn] (126 ± 39.1 μM) > (CH₃COO)₂Zn・2H₂O (139 ± 36.6 μM) > [N-bsE-Zn] (189 ± 57.3 μM) > zinc gluconate (238 ± 69.1 μM). In contrast, [N-bsP-Zn] and the Schiff base ligands did not show α-GI effects by themselves (Figure 2(b) and Figure 3(b)).

[Zn₂{(DMN)₂Cl₂}], which showed a particularly high inhibitory effect against rat intestinal α-glucosidase, and its Schiff base ligand, DMN, were selected for oral maltose and glucose tolerance testing. In the maltose tolerance test, the postprandial blood glucose levels in the [Zn₂{(DMN)₂Cl₂}] group were significantly lower than those in control groups. In the glucose tolerance test, the postprandial blood glucose levels in the DMN group and the (CH₃COO)₂Zn・2H₂O group were lower than those in the control groups. The values of the area under the curve (AUC) also showed similar transitions in each test (Figures 4(a)-(b) and Figures 5(a)-(b)).

The modes by which [Zn₂{(DMN)₂Cl₂}] inhibited yeast and rat intestinal α-glucosidase were studied. Lineweaver―Burk plots showed that [Zn₂{(DMN)₂Cl₂}] acted as a non-competitive inhibitor of the yeast and rat intestinal α-glucosidases (Figures 6(a)-(b)).