α-Glucosidase and Bacterial β-Glucuronidase Inhibitors from the Stems of Schisandra sphaerandra Staph

α-Glucosidase (AGS) is a therapeutic target for Type 2 diabetes mellitus (T2DM) that tends to complicate with other diseases. Some medications for the treatment of T2DM complications have the risk of inducing severe adverse reactions such as diarrhea via the metabolism of intestinal bacterial β-glucuronidase (BGUS). The development of new AGS and/or BGUS inhibitors may improve the therapeutic effects of T2DM and its complications. The present work focused on the isolation and characterization of AGS and/or BGUS inhibitors from the medicinal plant Schisandra sphaerandra. A total of eight compounds were isolated and identified. Sphaerandralide A (1) was obtained as a previously undescribed triterpenoid, which may have chemotaxonomy significance in the authentication of the genus Schisandra and Kadsura. 2′-acetyl-4′,4-dimethoxybiphenyl-2-carbaldehyde (8) was obtained from a plant source for the first time, while compounds 2–7 were isolated from S. sphaerandra for the first time. In the in vitro assay, compounds 1–5 showed potent to moderate activity against AGS. Interestingly, compound 3 also exhibited significant BGUS inhibitory activity, demonstrating the potential of being developed as a bifunctional inhibitor that may find application in the therapy of T2DM and/or the diarrhea induced by medications for the treatment of T2DM complications.


Introduction
Diabetes mellitus (DM), a chronic metabolic disease characteristic of prolonged high blood sugar levels, has caused widespread concerns around the world. By 2040, the prevalence of DM is expected to reach 642 million [1]. Type 2 diabetes mellitus (T2DM), caused by impaired insulin secretion and insulin resistance, accounts for more than 90% of all the incidences and is the most prevalent form of DM [2]. α-Glucosidases (AGS) hydrolyze the α-glucopyranosidic bond in complex carbohydrates to release glucose and other monosaccharides, leading to elevated blood sugar levels [3]. The inhibition of AGS can delay the digestion of carbohydrates and diminish the absorption of monosaccharides, which renders it as an ideal target for the management of T2DM. In fact, the use of AGS inhibitors has been proven to be the most efficient remedy for the control of postprandial hyperglycemia in T2DM [3]. Currently, three AGS inhibitors, namely acarbose, miglitol and voglibose, are used in clinic. However, regular consumption of these drugs has been reported to cause various side effects [4]. Hence, researchers are still engaged in the discovery of novel bioactive AGS inhibitors.
The isolation and identification of 1-8 further expanded the chemical diversity of S. sphaerandra. According to the literature [29,[37][38][39], heteroclitalactone D has been solely isolated from the genus Kadsura that was closely related to the genus Schisandra. Acquisition of the new triterpenoid sphaerandralide A from S. sphaerandra in the present study put forward the possibility that these two epimers may be of significance in the chemotaxonomy of Schisandra and Kadsura. Compounds 2-8 were all isolated from S. sphaerandra for the first time. In particular, it was the first time that compound 8 was obtained from a plant source.

AGS and BGUS Inhibitory Activity
Compounds 1-8 were evaluated for the in vitro AGS and BGUS inhibitory activities, respectively. In the AGS inhibitory assay, compounds 1-5 showed potent to moderate activity with IC 50 values ranging from 14.08 ± 0.29 to 74.45 ± 1.13 µM (Table 2), as compared with acarbose (IC 50 = 422.3 ± 8.44 µM). Compounds 1-3 possessed rearranged triterpenoid skeletons that were frequently encountered in plants of the genus Schisandra and Kadsura. Preliminary structure-activity relationship was also summarized. Specifically, compound 3 exhibited more potent activity than that of 1 and 2, suggesting that the sixmembered ring B and the fused cyclopropane ring may play a vital role in retaining the AGS inhibitory activity of this series of triterpenoids. Compound 4 was a 3,4-seco-cycloartane triterpenoid, while 5 and 6 belonged to 3,4-seco-cycloartane octanortriterpenoids. By comparing the activity of 4-6, it was concluded that the existence of the side chain at C-17 contributed to the maintenance of the AGS inhibitory activity as in the case of 4 that exhibited potent activity. For the two 3,4-seco-cycloartane octanortriterpenoids, hydrolysis of the methyl ester at C-3 led to the loss of the activity.

Analysis of Inhibition Kinetics
The inhibition kinetic mechanisms of 1-5 against AGS were studied using the Lineweaver-Burk plots ( Figure 4). The inhibition constant of the enzyme K i and the inhibition constant of the enzyme-substrate complex K i' of the inhibitors (Tables 3 and 4) were obtained by secondary plots of "slope versus [I]" and "Y-intercept versus [I]", respectively. As shown in Figure 4, data lines of 1 intersected in the second quadrant, while those of 3 had intersections in the third quadrant. In addition, K m and V max values of both 1 and 3 changed with the increased concentration of inhibitors. These results suggested that both 1 and 3 were mixed-type inhibitors of AGS, indicating they were able to bind either the free AGS or the AGS-substrate complex. As for 1, the inhibition constant K i (35.78 µM) was smaller than K i' (56.99 µM), demonstrating that it bound more easily and tightly to the free AGS than the AGS-substrate complex. On the contrary, 3 tended to bind more preferably to the AGSsubstrate complex as indicated by a larger K i (10.47 µM) than K i (7.93 µM). In addition, the smaller K i and K i' of 3 (10.47 and 7.93 µM, respectively) indicated better inhibitory potency against AGS as compared with that of 1 (35.78 and 56.99 µM, respectively), which was consistent with the IC 50 values. Intersection of the data lines on the y-axis indicated that both 2 and 4 were competitive inhibitors of AGS, which was supported by the increased K m and constant V max values. By comparison with the Ki values of 4 (7.97 µM) and 2 (45.15 µM), 4 was proposed to be a more potent inhibitor of AGS, which was verified by the IC 50 values in the AGS inhibition assay. The inhibition behavior of 5 could not be well defined since the Lineweaver-Burk plots intersected in the first quadrant. As a result, other kinetic parameters of 5 remained undetermined, except for the K m values.
Compound 3 was found to be a mixed-type inhibitor of BGUS, since the straight lines on the Lineweaver-Burk plots intersected in the third quadrant ( Figure 5). This inhibition mode against BGUS was also validated by the varied K m and V max values following concentration changes of 3 (Table 4). However, the replots of slope and Y-intercept versus the concentration of 3 were not linearly fitted, which limited the application of Equations (3) and (4). Consequently, the K i and K i of 3 remained not calculated.

Molecular Docking Studies
Given the significant activity of 3 and 4, docking studies were performed to illustrate the molecular determinants of these two compounds in inhibiting AGS or BGUs. The interactions between the ligands and AGS or BGUS were studied using MOE. As shown in Figure 5, both 3 and 4 could be well docked into the active site of AGS and/or BGUS; 3 formed a hydrogen bond interaction with Asn241 of AGS via the carbonyl of the six-membered lactone ring with a length of 1.96 Å. In addition, the complex of AGS and 3 was also stabilized by the hydrophobic interactions with residues such as His279, Pro309, Phe157, Arg312, and His239 ( Figure 5A,D). Likewise, a significant hydrogen bond interaction between 3 and Met447 (3.29 Å) of BGUS could be observed in the complex of 3 and BGUS ( Figure 5B,E), as well as hydrophobic interaction between 3 and residues Leu362, Ile363, Asp163, Ser557, and Tyr472 in BGUS. These interactions were proposed to contribute to the bifunctional inhibitory activity of AGS and BGUS by 3.
In the case of 4, it formed hydrogen bond interactions with Asp214 (2.22 Å) and Arg439 (2.33 Å) in AGS via the carboxyl group in the side chain. Furthermore, a hydrogenπ interaction was detected between Phe157 and the methoxy group of 4. The presences of hydrophobic interactions with residues Arg312, Phe300, and His239 also stabilized the complex of AGS and 4 ( Figure 5C,F).
In fact, S. sphaerandra has scarcely been subjected to modern pharmacological study. This was the first time that triterpenoids from S. sphaerandra were identified as AGS and/or BGUS inhibitors. Nevertheless, the in vivo efficacy, mode of action, toxicology, and pharmacokinetics of these bioactive triterpenoids remain to be explored, in spite of their encouraging in vitro enzyme inhibition activities. They are expected to improve postprandial hyperglycemia in vivo after an oral sucrose tolerance test, such as other natural terpenoids exhibiting in vitro AGS inhibitory activity and in vivo antidiabetic effects [40][41][42]. In addition to the potent AGS inhibition activity, 3 also significantly inhibited BGUS, demonstrating the potential to be developed as a bifunctional inhibitor. Since AGS and BGUS co-localized in the intestinal tract, it is reasonable to postulate that 3 may work as a bifunctional inhibitor by targeting both enzymes simultaneously in vivo. In view of the wide prevalence of T2DM, as well as the high incidence of T2DM complications and the relevant risk of drug-induced diarrhea, the discovery and development of AGS and BGUS bifunctional

Plant Material
The plant was collected in 2018 at Liangwang mountain of Yunnan Province, China. It was identified as S. sphaerandra by Dr. Jun Zhang from Kunming Zhifen Biotech. Sam-

Plant Material
The plant was collected in 2018 at Liangwang mountain of Yunnan Province, China. It was identified as S. sphaerandra by Dr. Jun Zhang from Kunming Zhifen Biotech. Samples of the plant material (ID ZJUT-WWZ2018-01) were deposited at the College of Pharmaceutical Science, Zhejiang University of Technology, Hangzhou, China.
ples of the plant material (ID ZJUT-WWZ2018-01) were deposited at the College of Pharmaceutical Science, Zhejiang University of Technology, Hangzhou, China.

Single-Crystal X-ray Diffraction
Single crystals of 1 were obtained from a mixture of n-hexane/EtOAc/CH 2 Cl 2 (2:1:1, v/v) at 4 • C. The structures were solved with the ShelXT [43] program and refined with the ShelXL [44] refinement package using least squares minimization. Data were collected using Olex2 [45]. The crystallographic data for 1 (CCDC: 2130778) were deposited at the Cambridge Crystallographic Data Centre. The crystal structure of 1 was depicted in Figure 3.

Inhibition Kinetics Analysis
The mode of inhibition for the active compounds was determined by the Lineweaver-Burk plot fitted by GraphPad Prism 8.0 software. The K m value and the V max value were obtained from the slope and Y-axis intercept of the Lineweaver-Burk plot based on the following equation: The K i and K i' values were available by secondary plotting of the slope and Y-intercept on the Lineweaver-Burk plot versus the inhibitor [I] through the following equations:

Molecular Docking Simulation
Inhibitors with significant potency were subjected to docking simulation, with the aim of revealing the probable molecular determinants underlying the inhibitory activity against AGS or BGUS. The X-ray crystal structure of BGUS (PDB ID: 3LPF) was obtained from the Protein Data Bank (PDB) database. Since the X-ray crystal structure of AGS of Saccharomyces cerevisiae has not been reported, a homology model of the enzyme was constructed by employing the crystal structure of isomaltase as a template (PDB ID: 3AJ7) on the SWISS MODEL webserver [46]. Then, the X-ray crystal structure of AGS built by homology modeling or BGUS was prepared using MOE (Version 2014. 09, Chemical Computing Group Inc., Montreal, Canada). The target compounds were docked into the active sites of AGS or BGUS using the Triangular Matching docking method. A total of 30 conformations for each ligand-protein complex were generated. Finally, the 2D and 3D plots were depicted for analysis of the interactions among inhibitors and the amino acid residues in the binding pocket.

Conclusions
Phytochemical studies on the stems of S. sphaerandra resulted in the isolation and identification of eight compounds. Sphaerandralide A (1) was obtained as a new triterpenoid, which may have chemotaxonomy significance in the authentication of the genus Schisandra and Kadsura. 2 -acetyl-4 ,4-dimethoxybiphenyl-2-carbaldehyde (8) was obtained from a plant source for the first time. Compounds 2-7 were discovered from S. sphaerandra for the first time. In the in vitro AGS inhibition assay, compounds 1-5 showed potent to moderate activity. Inhibition kinetic studies revealed that 1 and 3 were mixed-type inhibitors, while 2 and 4 were competitive inhibitors against AGS. In particular, 3 also significantly inhibited the activity of BGUS in vitro in a mixed-type inhibition mode, demonstrating the potential to be developed as a bifunctional inhibitor that may find application in the therapy of T2DM and/or the diarrhea induced by medications for the treatment of T2DM complications.