Evaluation of Traditional Indian Antidiabetic Medicinal Plants for Human Pancreatic Amylase Inhibitory Effect In Vitro

Pancreatic α-amylase inhibitors offer an effective strategy to lower the levels of post prandial hyperglycemia via control of starch breakdown. Eleven Ayurvedic Indian medicinal plants with known hypoglycemic properties were subjected to sequential solvent extraction and tested for α-amylase inhibition, in order to assess and evaluate their inhibitory potential on pancreatic α-amylase. Analysis of 91 extracts, showed that 10 exhibited strong Human Pancreatic Amylase (HPA) inhibitory potential. Of these, 6 extracts showed concentration dependent inhibition with IC50 values, namely, cold and hot water extracts from Ficus bengalensis bark (4.4 and 125 μgmL−1), Syzygium cumini seeds (42.1 and 4.1 μgmL−1), isopropanol extracts of Cinnamomum verum leaves (1.0 μgmL−1) and Curcuma longa rhizome (0.16 μgmL−1). The other 4 extracts exhibited concentration independent inhibition, namely, methanol extract of Bixa orellana leaves (49 μgmL−1), isopropanol extract from Murraya koenigii leaves (127 μgmL−1), acetone extracts from C. longa rhizome (7.4 μgmL−1) and Tribulus terrestris seeds (511 μgmL−1). Thus, the probable mechanism of action of the above fractions is due to their inhibitory action on HPA, thereby reducing the rate of starch hydrolysis leading to lowered glucose levels. Phytochemical analysis revealed the presence of alkaloids, proteins, tannins, cardiac glycosides, flavonoids, saponins and steroids as probable inhibitory compounds.


Introduction
Diabetes mellitus is a carbohydrate metabolism disorder of endocrine system due to an absolute or relative deficiency of insulin secretion, action, or both [1]. The disorder affects more than 100 million people worldwide and by 2030 it is predicted to reach 366 million. The most prevalent form both in the global and Indian scenario is the noninsulin dependent diabetes mellitus (NIDDM type 2) which is associated with elevated postprandial hyperglycemia [2].
Pancreatic α-amylase (E.C. 3.2.1.1), is a key enzyme in the digestive system and catalyses the initial step in hydrolysis of starch to maltose and finally to glucose. Degradation of this dietary starch proceeds rapidly and leads to elevated post prandial hyperglycemia (PPHG). It has been shown that activity of Human Pancreatic α-amylase (HPA) in the small intestine correlates to an increase in post-prandial glucose levels, the control of which is therefore an important aspect in treatment of diabetes [3]. Hence retardation of starch digestion by inhibition of enzymes such as αamylase would play a key role in the control of diabetes. However, the discovery of specific high-affinity inhibitors of pancreatic α-amylase for the development of therapeutics has remained elusive. Inhibitors currently in clinical, use for example, acarbose, miglitol, and voglibose, are known to inhibit a wide range of glycosidases such as α-glucosidase and α-amylase. Because of their nonspecificity in targeting different glycosidases, these hypoglycemic agents have their limitations and are known to produce serious side effects. Therefore, the search for more safer, specific, and effective hypoglycemic agents has continued to be an important area of investigation with natural extracts from readily available traditional medicinal plants offering great potential for discovery of new antidiabetic drugs [4,5].
While plant derivatives with purported hypoglycemic properties have been used in folk medicine and traditional 2 Evidence-Based Complementary and Alternative Medicine healing systems, very few of these traditional anti-diabetic plants have received proper scientific or medical scrutiny despite recommendations by World Health Organization (WHO). Ayurveda and other Indian traditional approaches have described more than 800 plants in the Indian subcontinent, known to possess antidiabetic potential. These require to be effectively studied and in fact only few of them have been characterized for their mechanistic actions [6][7][8][9]. Plants such as Barringtonia racemosa, Phyllanthus amarus, and so forth, have been tested for the presence of pancreatic αamylase inhibitors from various regions of the world [10][11][12][13][14][15]. Most of them have been tested on porcine pancreatic αamylase (PPA) and salivary amylase while reports on their effect on human pancreatic amylase (HPA), if any, are scarce.
In this study, we have selected medicinal plants from the Indian Ayurvedic system with known hypoglycemic or antidiabetic activity, in an attempt to screen for new inhibitors for HPA. The selections of these plants were based on their traditional usage and also taking into consideration previous studies that have demonstrated their antidiabetic properties (Table 1), thus providing a preliminary screening assessment [16][17][18][19][20][21][22][23][24][25][26]. However, no known reports of these plants as HPA inhibitors exist. Sequential solvent extracts of Azadirachta indica A. Juss., Bixa orellana L., Bougainvillea spectabilis Willd., Cinnamomum verum J. S. Presl, Curcuma longa L., Ficus bengalensis L., Ficus racemosa L., Momordica charantia L., Murraya koenigii L. Spreng., Syzygium cumini L. Skeels, and Tribulus terrestris L. were tested for the presence of PPA and HPA inhibitors. The lead extracts have also been subjected to preliminary kinetics on HPA to ascertain the type of inhibition as well as for phytochemical analysis to determine the probable inhibitory compounds present.

Plant Material.
In this study, anti-diabetic plants from the Indian sub-continent were selected on the basis of known ethno botanical/traditional Ayurvedic literature. Plants were obtained from Pune city and nearby areas of the Western Ghats, India, in the months of August-January. The plants, their voucher numbers (V. No.), the part(s) used, and the hypoglycemic properties as seen from literature search are listed in Table 1. A specimen of each plant was submitted to Botanical Survey of India (BSI), Pune, India, for authentication. The plants were separated into their various parts, namely., roots, stem, leaves, seeds and so forth, washed with tap water, air-dried in shade, and processed immediately as mentioned below.

Preparation of Plant Extracts.
The air-dried plant material (60-100 g) was crushed with liquid nitrogen, powdered, and successively extracted in polar to nonpolar solvent on an increasing degree of non-polarity [27]. The different extracts obtained sequentially were with cold water, hot water, methanol, isopropanol, acetone, methyl-butyl-tertiary ether, and cyclohexane. This kind of sequential extraction was performed taking into consideration the fact that traditional methods of preparing herbal formulations are mainly aqueous. Also, aqueous extracts contain peptides, proteins, or glycans, which would otherwise be denatured by organic solvents and high-temperature extraction. Distilled water was added to the crushed material in a ratio of 1 : 4 (w/v) and kept at 30 • C (24 h) and 55 • C (2 h) at 130 rpm for cold-and hot-water extracts, respectively. For each solvent, the extract was filtered, centrifuged, and the residue collected for subsequent solvent extraction. The organic solvents were added in a ratio of 1 : 3 (w/v) and refluxed with the residue for 3 h at their respective boiling temperatures. Each extract was filtered and stored at −20 • C.

α-Amylase
Inhibition and Kinetic Studies. PPA was used for preliminary screening of α-amylase inhibitors from the extracts. The inhibition assay was performed using the chromogenic DNSA method [28]. The total assay mixture composed of 500 μL of 0.02 M sodium phosphate buffer (pH 6.9 containing 6 mM sodium chloride), 0.04 units of PPA solution, and extracts at concentration from 0.1-1.5 mg mL −1 (w/v) were incubated at 37 • C for 10 min. After pre-incubation, 500 μL of 1% (v/v) starch solution in the above buffer was added to each tube and incubated at 37 • C for 15 min. The reaction was terminated with 1.0 mL DNSA reagent, placed in boiling water bath for 5 min, cooled to room temperature, diluted, and the absorbance measured at 540 nm. The control reaction representing 100% enzyme activity did not contain any plant extract. To eliminate the absorbance produced by plant extract, appropriate extract controls were also included. One unit of enzyme activity is defined as the amount of enzyme required to release one micromole of maltose from starch per min under the assay conditions.
For the determination of the inhibitor concentration at which 50% inhibition of enzyme activity occurs (IC 50 ), the assay was performed as above except that the inhibitor/plant extract concentrations were varied from 0.1-150 μg. Acarbose was used as a positive control at a concentration range of 6.5 μg-32.8 μg. The IC 50 values were determined from plots of percent inhibition versus log inhibitor concentration and calculated by logarithmic regression analysis from the mean inhibitory values. The IC 50 values were defined as the concentration of the extract, containing the α-amylase inhibitor that inhibited 50% of the PPA or HPA activity. Other quantifiers were defined as follows: %Relative enzyme activity = (enzyme activity of test/enzyme activity of control) * 100; %inhibition in the α-amylase activity = (100 −% relative enzyme activity).
For kinetic experiments involving concentration independent inhibition, the inhibitor/extracts were taken at their Hypoglycemic effect by inhibition of glucose-6-phosphatase and fructose-1-6-biphosphatase in liver and stimulation of hepatic glucose-6-phosphate dehydrogenase. [25] Murraya koenigii L. Spreng (Curry Leaves) SP-2 Rutaceae Leaves Increases glucogenesis and decreases glycogenolysis and gluconeogenesis. [26] Syzygium cumini L. Skeels (Jamun) SPSC2 Myrtaceae Seeds Reduces blood glucose level, increase in serum insulin level, exhibits insulinase activity. Hypoglycemic activity mediated through insulin release mechanism, glycogen content and hepatic glucokinase, hexokinase, glucose-6-phosphate, and phosphofructokinase levels in diabetic mice. [4,17] Tribulus terrestris L. (Gokshura) SUTT1 Zygophyllaceae Seeds Hypoglycemic activity by inhibiting oxidative stress. [16] a The common Indian name for the plant is given in brackets. b Voucher numbers as given by Botanical Survey of India, Pune.
IC 50 values and incubated with HPA while the substrate (starch) concentration was varied from 0.5-5 mg mL −1 and reaction allowed to proceed as mentioned above. Kinetic analysis was performed based on Lineweaver-Burk double reciprocal plots and the kinetic parameters calculated using Origin 6.0 software from Originlab, USA.

Screening of Plant Extracts on α-Amylase Inhibition.
The plants from the Indian sub-continent exhibiting potential hypoglycemic properties were sequentially extracted with polar to non-polar solvents and a total of 91 extracts obtained. It should be noted that while generally coldand hot-water extracts are most commonly used in the traditional method of preparing medicines in Ayurveda, chances of missing out on bioactive principles with better amylase inhibitory potential from less-polar solvents are high. Hence the rationale for performing extractions from polar to non-polar solvents is not only to confirm and validate the inhibitory activity, if found, in any of the aqueous extractions performed in the traditional manner but also to search for newer and higher specific affinity inhibitors in less-polar solvents. PPA was used as a target enzyme for screening of inhibitory activity from the above mentioned ninety-one plant extracts. The control reaction representing 100% enzyme activity was 0.21 U/mL for PPA and 0.25 U/mL for HPA. Extract samples dissolved in DMSO contained as a final yield of 1.5 mg mL −1 of the dried extract and the enzyme activity of both PPA and HPA was not affected by DMSO at the concentration used. Appropriately diluted plant extract was used for enzyme inhibition assay and the activity obtained with each extract was normalized to percent relative activity (Figure 1) (Figure 2). Of these, B. orellana methanol extract and F. bengalensis hotwater extract showed similar IC 50 values for HPA as seen for PPA, while M. koenigii isopropanol, S. cumini coldwater, and T. terrestris acetone extracts required higher concentrations for 50% HPA inhibition as compared to PPA. However, five extracts, namely, isopropanol extracts of C. longa l (0.16 μgmL −1 ) and C. verum (1.0 μgmL −1 ), coldwater extract of F. bengalensis (4.4 μgmL −1 ), and hot-water extracts of S. cumini (4.1 μgmL −1 ) could inhibit HPA at much lower concentrations than PPA. Acarbose inhibited HPA with an IC 50 value of 10.5 μgmL −1 . Thus, these crude extracts could inhibit HPA at much lower concentrations than even acarbose and would therefore be good candidates to test for high-affinity inhibitors. For this, inhibitor kinetic studies were performed to determine the type of inhibition.

Kinetic Studies.
The inhibitory assays were performed using extracts at different dilutions in order to check for concentration dependent or independent inhibition. Of the 10 plant extracts studied, 6 extracts, namely, C. longa   independent inhibition for HPA. A preliminary study for the probable mode of inhibition was performed using Michaelis-Menten kinetics at their IC 50 value with varying substrate concentrations. The double reciprocal Lineweaver-Burk plot showed a decrease in both V max and K m values for all the 4 extracts studied, thereby suggesting an uncompetitive type of inhibition ( Figure 4) Figure 2: IC 50 values of the extracts exhibiting ≥50% inhibition on Porcine pancreatic α-amylase and Human pancreatic α-amylase enzyme activity. The data is calculated as the mean ± SEM; (n = 3). The students F-test was used and the bars with different asterisks ( * * * , * * , * ) show significant difference with respect to control (P < .05). Acarbose is taken as the standard α-amylase inhibitor.   M. koenigii, and 0.14 μMmin −1 and 1.71 mg for T. terrestris, respectively.

Phytochemical and GC-MS Analysis.
Preliminary qualitative phytochemical analysis was performed to determine the probable type of compounds present in the extracts causing HPA inhibition. The results showed the presence of different types of active constituents such as proteins, cardiac glycosides, saponins, alkaloids, flavonoids, and so forth. (Table 2).
Using the Wiley-NIST library, GC-MS identification of the compounds based on the retention time, peak area, molecular mass, and molecular formula are shown in Table 3.

Discussion
The aim of the present study was to investigate the HPA inhibitory activity from medicinal plants known in the Indian Ayurvedic system for their anti-diabetic properties. To date no reports of compounds responsible for HPA inhibition from these plants exist in the literature and we report here for the first time their inhibitory on HPA. Many of the plants chosen in this study are used by the Indian population not only for food purposes but also form a part of the local pharmacopoeia for the treatment of diabetes. Our results indicate that retardation of starch hydrolysis by inhibition of HPA activity by some of these extracts leads to a reduction in glucose concentrations ( Figure 5). Of the 11 plants and their 91 extracts tested, 10 exhibited significant inhibition of HPA, suggesting that they contain compounds capable of HPA inhibition. Of these, 4 are aqueous extracts while 6    are extracts of less-polar solvents. Preliminary phytochemical analysis to indicate the kind of compounds present in these extracts suggests the occurrence of proteins/peptides and polyphenols in cold-and hot-water extracts while tannins, alkaloids, flavonoids, and saponins are found in non-polar extracts. Flavonoids and polyphenolics have been reported to contribute to hypoglycemic activity [30]. B. orellana methanol extract exhibited concentration independent inhibition with an IC 50 value of 49 μgmL −1 . The major components identified in this extract were found to be β-tocopherols and vitamin E which are known to be present in B. ornella leaves [31].
C. longa is known to contain curcuminoids, glycosides, terpenoids, and flavonoids [34]. Maximal inhibition of the enzyme (HPA) was obtained with C. longa isopropanol extract at a low concentration of 0.16 μgmL −1 in a competitive manner and with acetone extract at 7.4 μgmL −1 in a concentration independent manner. The probable compounds in C. longa isopropanol extract were podocarpic acid, curlone, and cinnamic acid while those in acetone extract were curlone, 3-cyano-7-hydroxy-4-methylcoumarin, and 5amino-2-hydroxybenzoic acid.
F. bengalensis bark extract showed HPA inhibitory activity in the aqueous fraction with an IC 50 value of 4.4 μgmL −1 in a concentration-dependent manner. The major phytochemical components in the aqueous extract were found to be alkaloids, proteins, tannins, flavonoids, and saponins while cardiac glycosides and steroids were found to be absent. A dimethoxy derivative of leucocyanidin 3-Obeta-D-galactosyl cellobioside isolated from the bark of F. bengalensis Linn. has been shown to decrease blood sugar levels significantly both in normal and moderately diabetic rats [35].
The compounds identified in M. koenigii isopropanol extract were cyclohexanone 2-methyl-5-(1-methylethenyl), Evidence-Based Complementary and Alternative Medicine 9 2, 3, 5, 6 tetrachlorohydroquinone, and Vitamin E which have been reported in leaves by other investigators [36,37]. The extract was able to inhibit HPA in a concentration independent manner with an IC 50 value of 127 μgmL −1 . Kinetic analysis also indicated that this extract resulted in the best decrease in both K m and V max values to 1.02 mg and 0.07 μMmin −1 as compared to control, namely, 2.12 mg and 0.17 μMmin −1 .
S. cumini is known to contain proteins, tannins, an alkaloid jambosine, a glycoside antimellin, and fatty acids [38]. The extracts exhibiting concentration-dependent HPA inhibitory activity were cold-and hot-water extracts with an IC 50 value of 42.1 and 4.1 μgmL −1 , respectively. It is most likely that the bioactive component could probably be a protein, a glycoside, or a polyphenol.
Furostanol and spirostanol saponins, flavonoids, alkaloids, steroids, carbohydrates, and amides have been isolated from T. terrestris [39]. In our study, the acetone extractinhibited HPA activity in a concentration independent manner with a high IC 50 value of 511 μgmL −1 . The compounds identified in this extract were ethyl crotonate and sorbinose.
These findings suggest that inhibition of HPA activity, leading to retardation of starch hydrolysis is one of the mechanisms through which the plants tested in this study could be exhibiting their hypoglycemic effect. Modulation of HPA activity by compounds in these extracts would thus eventually lead to a lowering of post-prandial blood glucose levels ( Figure 5). Presence of HPA inhibitor blocks the normal pathway of conversion of dietary starch to maltose, maltotriose, and oligosaccharides and then to glucose in the gut which gets absorbed in the blood. Thus, glucose formation can be blocked in hyperglycemic conditions with pancreatic α-amylase inhibitors. This HPA inhibition could occur in a concentration dependent or independent manner depending upon the bioactive compound. The compounds/bioactive principle from these plants responsible for the HPA inhibition need to be isolated and characterized. The efficacy of HPA inhibition can be increased with the combination of these isolated components as shown by Said et al., where in herbal combination of four plants is more effective as antidiabetic traditional medicine [40].

Conclusions
This study highlights the HPA inhibitory activity of 10 plant extracts exhibiting promising leads as HPA inhibitory molecules and lends scientific support to their use in traditional Ayurvedic medicine. B. orellana, F. bengalensis, S. cumini, C. verum, M. koenigii, C. longa, and T. terrestris plants, which are known for their hypoglycemic property, were found to exhibit strong inhibitory action on HPA, even better than acarbose. The phytochemicals in the extracts causing this inhibition were identified as alkaloids, proteins, tannins, cardiac glycosides, flavonoids, saponins, and steroids. Studies to isolate the bioactive principle(s) from these extracts are required. In the recent years, the search for new molecules as potential α-amylase inhibitors with a high specific affinity has intensified. Thus, modulation of HPA activity through the therapeutic use of high-affinity plant derived inhibitors would be of considerable medical relevance in the treatment of diabetes.