Development and application of a high throughput assay for discovery of starch hydrolase inhibitors

To search for the active ingredients in low GI foods, a 96-well based high-throughput method HTS for rapidly measuring the inhibition of starch hydrolase was developed, by monitoring the

DEVELOPMENT AND APPLICATION OF A THROUGHPUT SCREENING ASSAY FOR DISCOVERY

HIGH-OF STARCH HYDROLASE INHIBITORS

LIU TING TING

NATIONAL UNIVERSITY OF SINGAPORE

2013

DEVELOPMENT AND APPLICATION OF A THROUGHPUT SCREENING ASSAY FOR DISCOVERY

HIGH-OF STARCH HYDROLASE INHIBITORS

LIU TING TING

(BSc, MSc Wageningen University)

DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE

2013

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR

OF PHILOSOPHY

Acknowledgments

First and foremost, I would like to express my sincere gratitude to my

supervisor of my doctoral study, associate professor Huang Dejian for the

valuable guidance and advice With his encouragement, I would be able to

carry on my study and complete this thesis I also would like to thank Dr

Zhang Dawei from Division of Chemistry & Biological Chemistry, Nanyang

Technological University, for his kindly help on my research

Besides, I would like to thank the staff of Food Science and

Technology Program of NUS, Ms Lee Chooi Lan, Ms Lew Huey Lee and Ms

Jiang Xiaohui for their kindly help during my study I also would like to thank

Ms Soh Yee Lyn, Ms Chen Mei Juan and Ms Yew Mun Yip for their

contributions in various experiments

I wish to acknowledge Jing Brand Co Ltd for the financial support of

my research project and the Singapore International Graduate Award (SINGA)

for PhD scholarship

Last but not least, I would like to extent my gratitude to my entire

family for their moral support and courage throughout the years

Table of contents

Acknowledgments ii

Table of contents iii

Summary vii

List of tables ix

List of figures x

List of abbreviations xiii

Chapter 1 Introduction 1

Chapter 2 Literature review 6

2.1 Starch hydrolase 7

2.1.1 Starch digestion by starch hydrolase 7

2.1.2 Starch hydrolase used in diabetes research 8

2.2 Methods of determining starch hydrolase activity 9

2.2.1 Conventional methods 9

2.2.2 Turbidity measurement applications 13

2.3 Starch hydrolase inhibitors 14

2.3.1 Acarbose 14

2.3.2 Phytochemicals as starch hydrolase inhibitors 16

2.3.3 Proanthocyanidins 16

2.3.4 Phenylpropanoid sucrose esters 19

2.4 Functional food for diabetes patients 21

-Chapter 3 A high-throughput assay for quantification of starch hydrolase inhibition based on turbidity measurement 23

3.1 Introductions 24

3.2 Materials and methods 27

3.2.2 Determination of αglucosidase activity 29

3.2.3 Determination of dynamic range of the turbidity measurement 30

3.2.4 Highthroughput assay of starch hydrolase activity 31

3.2.5 Determination of inhibitor activity 33

3.2.6 Determination of precision and accuracy 33

3.2.7 Statistical Analysis 34

3.3 Results and discussions 35

3.3.1 Dose response between turbidity and starch concentration 35

3.3.2 Quantification of the enzyme activity 36

3.3.3 Inhibitor activity 38

-3.3.4 Precision and accuracy………- 40-

3.3.5 Acarbose equivalence 42

3.4 Conclusion 49

-Chapter 4 Amylase and sucrase inhibition activity of diboside A isolated from wild buckwheat (Fagopyrum dibotrys) 50

4.1 Introduction 51

4.2 Materials and methods 53

-4.2.1 Assay guided fractionation of α-amylase inhibitors diboside A from Fagopyrum dibotrys 53

4.2.2 αAmylase inhibition assay 54

4.2.3 αGlucosidase inhibition assay 55

4.2.4 Inhibition kinetics of Diboside A 56

4.2.5 Molecular docking study 57

4.2.6 Statistical analysis 57

4.3 Results and discussions 58

4.3.1 Fractionation of αAmylase inhibitors from F dibotrys extracts 58

4.3.2 Diboside A 63

4.3.3 Inhibition kinetics of diboside A on αamylase and sucrase 64

4.4 Conclusion 75

Smilax glabra rhizomes 76

5.1 Introduction 77

5.2 Material and Methods 79

5.2.1 Plant material and reagents 79

5.2.2 Extractions 79

5.2.3 αAmylase and αglucosidase inhibitory study on raw extracts 80

-5.2.4 Fractionation and isolation of bioactive components from S glabra rhizome 80

-5.2.5 α-Amylase and α-glucosidase inhibitory study on isolated compounds- 81 5.3 Result and Discussion 83

5.3.1 αAmylase and αglucosidase inhibition study on raw extracts 83

-5.3.2 Structure identification of phenylpropanoid sucrose esters compounds from Smilax.glabra rhizome 86

-5.3.3 α-Amylase and α-glucosidase inhibition activity of isolated compounds 91

5.4 Conclusion 96

-Chapter 6 Characterization of proanthocyanidins in wild buckwheat root extracts 97

6.1 Introduction 98

6.2 Material and methods 99

6.2.1 Extraction and characterization of proanthocyanidins 99

6.2.2 Thiolysis of the proanthocyanidins for HPLC analysis 99

6.2.3 αAmylase inhibition assay 100

6.2.4 αGlucosidase inhibition assay 101

-6.3 Results and discussions……… -102-

6.3.1 Characterization and quantification of proanthocyanidins isolated from Fagopyrum dibotrys rhizome by HPLCMS 102

6.3.2 Thiolyzed products of the proanthocyanidin 107

and αglucosidase 109

6.4 Conclusions 111

-Chapter 7 Effects of grape seeds extracts on sensory properties and in vitro digestibility of wheat noodles 112

7.1 Introduction 113

7.2 Materials and methods 115

7.2.1 Reagents and instruments 115

-7.2.2 Determination of IC50 value of grape seed extracts 115

7.2.3 Determination of grape seed extracts total phenolic content 116

-7.2.4 Determination of proanthocyanidin profile of grape seed extracts1167.2.5 Formulation of GSE incorporated wheat noodles 117

7.2.6 In vitro digestion model of GSE incorporated noodle samples 117

7.2.7 Determination of grape seeds proanthocyanidin cooking loss 118

7.2.8 Sensory evaluation and consumer acceptance 119

7.3 Results and discussions 120

7.3.1 Determination of total phenolic content of grape seeds extracts 120 -7.3.2 Determination of the proanthocyanidin profiles of grape seeds extracts 120

-7.3.3 Determination of IC50 values of the grape seed extracts 123

7.3.4 Formulation of grape seed wheat noodles 126

7.3.5 In vitro digestion of GSEincorporated wheat noodles 128

7.3.6 Determination of the loss of proanthocyanidin during cooking 133

7.3.7 Sensory evaluation and consumer acceptance 135

7.4 Conclusion 138

Chapter 8.Conclusions and future work……… 139

List of publications 145

Bibliography 146

Appendices 165

-Several novel approaches to treat Type 2 diabetes have been tested, including preventing hyperglycemia by the consumption of low glycemic index (GI) foods To search for the active ingredients in low GI foods, a 96-well based high-throughput method (HTS) for rapidly measuring the inhibition

of starch hydrolase was developed, by monitoring the decrease in turbidity over time during the enzymatic digestion of starch The area under the curve

of turbidity measured over time was used to quantify the inhibitory effects of polyphenolic compounds on starch hydrolase Acarbose equivalence was introduced for the first time, which has the advantage of eliminating run-to-run variation, and allowing easier comparison of inhibitor activity Using this assay, grape seed and cinnamon bark-extracts were found to be the most active polyphenol extracts in inhibiting starch hydrolase, and proanthocyanidins were identified as the active constituents

By applying HTS, extracts of wild buckwheat root (Fagopyrum dibotrys)

exhibited strong starch hydrolase inhibitory activity Using HTS as a guide, diboside A was isolated as a selective inhibitor of pancreatic α-amylase and sucrase, but not maltase Analysis of the inhibition kinetics revealed that diboside A is a non-competitive inhibitor of sucrase (Ki = 72.4 µM) and uncompetitive inhibitor of α-amylase (Ki = 5.1 µM) Molecular docking studies then showed that the allosteric inhibition of α-amylase by diboside A is attributed to the number of hydrogen bonds formed and electrostatic interactions between the enzyme active site and inhibitor

To continue investigating the anti-diabetic effects of phenylpropanoid sucrose ester compounds (PSEs), smilaside A and smilaside D were isolated

from ethyl acetate extracts of the Smilax glabra rhizome Both compounds

displayed negligible inhibitory effects towards porcine pancreas α-amylase, but weak and moderate inhibitory activity towards rat intestinal α-glucosidase,

report demonstrating inhibition of α-glucosidase by PSE compounds in vitro

Proanthocyanidins in the extracts of wild buckwheat root were also found to be major contributors to the inhibition of starch hydrolase The components of proanthocyanidins were separated by a diol HPLC column, and their molecular weights were determined by ESI-MS The results showed that B-type proanthocyanidin oligomers are the predominant form, with a mean degree of polymerization of 7.2 A dose-response of the proanthocyanidin fraction was used to calculate the IC50, which was 10.7 µg/mL, 35.2 µg/mL and 18.0 µg/mL for amylase, maltase, and sucrase, respectively

The use of grape seeds extracts (GSE) as α-amylase inhibitors using wheat noodles as the food matrix was investigated Wheat noodles were

formulated with the addition of 0.5, 1.0, or 2.0% GSE A 3 hour in vitro

digestion kinetic study showed a good dose response for inhibiting the rate of starch hydrolysis Five minutes of cooking at 100 ºC led to a 15.7% loss of proanthocyanidins and the degradation of large proanthocyanidin oligomers into smaller oligomers Finally, a sensory test showed that the color of the noodles could significantly influence the acceptance score, and 2.0% GSE incorporation gave the highest consumer acceptance

Inhibitors concentrations tested for starch hydrolases

Determination of the linear range of gelatinized starch

Accuracy and precision of the high-throughput method

Table 4 IC50 and acarbose equivalence of inhibitors obtained by starch

hydrolases

Table 5 Estimation of the proanthocyanidins contents expressed as

milligram epicatechin equivalent per 100 mg of selected

extracts

Table 6 IC50 and acarbose equivalence (AE) of different inhibitors on

α-amylase and α-glucosidase (maltase and sucrase)

Table 7 Docking scores of diboside A in the 2 allosteric binding sites

The cells colored in green indicate favourable interactions

while cells colored in red indicate unfavorable interaction All

calculated values are in kcal mol-1

Table 8 The ESI-MS/MS2 peaks of proanthocyanidins in F dibotrys

Table 9 IC50 value of wild buckwheat proanthocyanidin extracts on

α-amylase and α-glucosidase inhibition

Table 10 Estimation of grape seed proanthocyanidins content

Table 11 In vitro kinetics of starch hydrolysis (% total starch hydrolysed

at different time intervals) of wheat noodles and noodles with

addition of different dosage of grape seed extract (GSE)

Table 12 Cooking loss of proanthocyanidins (100 ºC, 5 mins)

Table 13 Sensory evaluation results of grape seed extracts incorporated

noodles

Figure 1 The hydrolysis reaction of pNPG by α-glucosidase

Figure 2 The reaction of the GOD/POD method for determining

α-glucosidase activity

Figure 3 The DNSA assay for determining reducing sugar content

Figure 4 Chemical structure of acarbose

Figure 5 Skeletal structure of proanthocyanidins

(epicatechin-(4ß→8)-epicatechin), R1 = R2 = OH

Figure 6 Phenylpropanoid sucrose esters (PSE) with sucrose as the

core structure, where at least one -OH group is substituted

by a phenylpropanoid moiety

Figure 7 Phenylpropanoid moiety substituents

Figure 8 Experimental set-up and data collection of high-throughput

turbidity measurement

Figure 9 Kinetic curves of starch hydrolysis in the presence of

different hydrolases (A) amyloglucosidase (Aspergilllus niger), (B) α-amylase (porcine pancreatic), and (C) α-

glucosidase (rat intestine acetone powder) Assay conditions are 37 ºC and starch concentration, 5.0 g/mL Insets are dose-response relationships of the AUC vs enzyme concentrations

Figure 10 Representative kinetic curves of starch hydrolysis in the

presence of inhibitors

Figure 11 Comparison of acarbose equivalence among different starch

hydrolases and total phenolic contents of different samples

Figure 12 Structure of diboside A (1,3,6'-tri-p-coumaroyl-6-feruloyl

sucrose)

Figure 13 ESI-MS and MS2 spectra of diboside A

Figure 14 Diboside A 1H-NMR, 300MHZ

at initial sucrose concentration of 18 mg/mL and sucrase concentration of 0.15 U/mL; porcine pancreas α-amylase (B)

at starch concentration of 10 mg/mL, and amylase concentration is 3 U/mL

Figure 16 Eadie-Hofstee plot of acarbose on sucrase

Figure 17 Two binding sites, BS1 and BS2, were found using SiteMap

The binding sites are represented using molecular surfaces Their respective SiteScores are labelled under the structures

Figure 18 (A) Binding conformation of diboside A in BS1 shown as

molecular surface Regions in red represent electronegative areas, regions in blue represent electropositive areas, and white areas are neutral (B) 2-D ligand interaction diagram

of diboside A in BS1

Figure 19 (A) Binding conformation of diboside A in BS2 shown as

molecular surface (B) 2-D ligand interaction diagram in

BS2

Figure 20 Extraction, fractionation and isolation of active compounds

from Smilax glabra rhizome

Figure 21 S glabra ethyl acetate and choloroform extract on

α-glucosidase inhibition (A) sucrase inhibition (B) maltase inhibition

Figure 22 ESI-MS and MS2 spectra of smilaside A

(3,6-diferuloyl-4’,6’-diacetyl sucrose)

Figure 23 ESI-MS and MS2 spectra of smilaside D

(1-p-coumaroyl-3,6-diferuloyl-4-acetyl sucrose)

Figure 24 Chemical structures of smilaside A, smilaside D, diboside A,

astilbin and 5-O-caffeoylshikimic acid

Figure 25 Amylase inhibitory effect of isolated compounds from

rhizome of Smilax glabra

Figure 26 Smilax glabra rhizome isolated compounds on α-glucosidase

inhibition

Figure 27 HPLC chromatograms of proanthocyanidins separated using

diol column

Figure 28 Thiolyzed products of wild buckwheat proanthocyanidins

Figure 29 Kinetic curves of α-amylase inhibition by grape seed

extracts (Tianding)

Figure 30 Sample preparation of noodles, control, 0.5%, 1.0%, 2.0%

GSE (left to right), and cooking medium, cooked noodle samples, uncooked fresh noodles (top to bottom)

Figure 31 In vitro digestion models of GSE incorporated wheat

noodles

CBP - Cranberry Pomace Extracts

CNM - Cinnamon Bark Extracts

DM

DNSA

- Diabetes Mellitus

- 3,5-Dinitrosalicylic Acid ESI-MS - Electrospray Ionization- Mass Spectrometry

GAE - Gallic Acid Equivalents

- Grape Seed Extracts

- Grape Seed Proanthocyandins HPLC

HSA

- High Performance Liquid Chromatography

- Human Salivary Amylase LC-MS - Liquid Chromatography- Mass Spectrometry

LOD - Limit of Detection

LOQ - Limit of Quantification

HTS - High Throughput Screening

MUR

NMR

- Mulberry Root Extracts

- Nuclear Magnetic Resonance

OD - Optical Density

OPCs - Proanthocyanidins

pNPG5 - p-nitrophenyl-α-D-maltopentaoside

PPA

PSE

- Porcine Pancreatic α-Amylase

- Phenylpropanoid Sucrose Esters

SD - Standard Deviations

TCM - Traditional Chinese Medicine

TPC - Total Phenolic Content

Chapter 1 Introduction

Diabetes mellitus (DM) is a group of metabolic diseases characterized

by hyperglycemia, resulting from defects in insulin secretion, insulin action, or both [1] There are two types of diabetes: type 1 and type 2 Type 1 diabetes results from the patients’ failure to produce insulin, and as such they require insulin injections Patients with type 2 diabetes, also called non-insulin dependent diabetes mellitus (NIDDM), either do not produce enough insulin,

or are insulin insensitive [2] Type 2 diabetes is the most common form of the disease, accounting for 90-95% of cases [3] The worldwide increase in the incidence of Type 2 DM is becoming a major health concern

Several strategies can be used to regulate the elevated post-prandial blood glucose levels in type 2 diabetes These include stimulating insulin secretion from the ß-cells of pancreatic islets, inhibiting the insulin degradation process, repairing or regenerating pancreatic beta cells, and inhibiting the starch hydrolases, α-amylase and α-glucosidase [4] Among these, the inhibition of enzymes is a major treatment strategy that has been intensively studied by many research groups [5]

Acarbose and voglibose are anti-diabetic drugs that function as enzyme inhibitors, and are widely used in many countries [3] For example, acarbose prevents the degradation of complex carbohydrates into glucose, so that some carbohydrates remain undigested in the intestine and are delivered to the colon

In the colon, bacteria ferment those undigested carbohydrates, causing gastrointestinal side-effects such as flatulence, and abdominal pain [6] The limitations of currently available oral agents encouraged a concerted effort to identify novel drugs that can manage type 2 diabetes more efficiently Plant-based products have been popular all over the world for the centuries, and many of them have been studied and recommended as alternative treatments for diabetes [7] Babu and coworkers have created a database of 389 medicinal plants that have been traditionally used as diabetes treatments [8] Some of these medicinal plants have been recorded in one of the most famous ancient

medicinal books in China, Ben Cao Gang Mu (Compendium of Materia Medica) However, the effectiveness of these herbs is yet to be convincingly

tested in rigorous clinical trials In addition, the purported bioactive compounds in these herbs have not yet been fully elucidated

Discovering the active compounds in medicinal plants that function as starch hydrolase inhibitors has traditionally been accomplished using low throughput methods, such as the 3,5-dinitrosalicyclic acid (DNSA) reduction

assay [9], glucose oxidase/-peroxidase (GOD/POD) assays [10], and the nitrophenyl-α-D-glucopyranoside (pNPG; [11]) or p-nitrophenyl-α-D- maltopentaoside (pNPG5) method [12] However, these methods share

p-common problems, being labor intensive, time consuming and costly when the number of the samples is large, as is the case with herbal plants In addition,

the pNPG method uses synthetic substrates, which cannot fully mimic the

behavior of enzymes towards natural substrates We therefore developed a high-throughput, inexpensive laboratory method that can screen a large number of samples in a short period of time, importantly using natural substrates to ensure the fidelity of the data Additionally, the high throughput method could guide fractionation of the bioactive compounds in medicinal plants, and could in turn speed up the process of identifying novel bioactive constituents In this study, the anti-diabetic activity of the Traditional Chinese

Medicine (TCM) wild buckwheat (Fagopyrum dibotrys), and rhizomes of Smilax glabra were investigated It was shown that daily consumption of F dibotrys root extracts could significantly reduce postprandial blood glucose levels [13] Hypoglycemic effects of the rhizomes of Smilax glabra was

demonstrated in normal and diabetic mice, where isolated methanol extracts caused enhanced insulin sensitivity [14] The bioactive compounds from these studies were isolated and characterized

Apart from anti-diabetic compounds isolation and development of a high throughput method for efficiently screening those compounds from medicinal

plants Reducing the digestibility of starch based foods by incorporating medicinal plants extracts could also be a meaningful approach to produce foods suitable for consumption by diabetes patients without worrying about hyperglycemic effects Foods, or food products, can be categorized into high (>70), low (<55), and intermediate GI groups, based on their GI value [15] Common starchy staple foods fall into the category of high Glycemic Index (GI) High GI food is positively associated with the risk of developing type 2 diabetes [16, 17] Diabetes patients are recommended to consume low GI foods, since they are digested slowly, and have been demonstrated to improve long-term glycemic control [15] The majority of low-GI food products are designed by modifying the intrinsic properties of the starch However, in the market it is rare to find starch-based foods that were made by incorporating

plant extracts or compounds isolated from botanical sources Many in vitro

studies have shown that grape seed proanthocyanidins may have therapeutic potential for the prevention and treatment of diabetes [18] Part of this research project therefore attempted to add grape seed extracts into wheat noodles to create a slowly digested product

As described above, there is an urgent need to develop a high throughput screening (HTS) method for efficiently discovering botanicals with potential anti-diabetic activities Ideally, the HTS method would facilitate the isolation

of overlooked bioactive compounds present in traditional herbs The objectives of this thesis are therefore to:

 Develop a high throughput method to rapidly screen the starch

hydrolase inhibitory activity of botanicals from medicinal and

traditional food sources (Chapter 3)

 Use the HTS assay as a guide to isolate and characterize the starch hydrolase inhibitors from the promising herbal extracts (Chapter 4, 5, and 6)

 Incorporate starch hydrolase inhibitors into functional food with a low starch digestibility (Chapter 7)

To develop the HTS assay, a microplate reader was used to monitor the changes in turbidity in a 96 well microplate It is expected that this method

will be applicable to the high-throughput screening of many different plant

species Additionally, this research supported the importance of wild

buckwheat and Smilax glabra as promising anti-diabetic plants It is the first time that starch hydrolase inhibitors were isolated, and anti-diabetic activity studied in detail The isolation process was facilitated by the development of the novel HTS method In addition, the proanthocyanidin profiles of the grape seeds were determined With the aim of developing starch based foods for diabetes patients to both promote their health and reduce postprandial blood glucose levels, we used grape seeds as the key beneficial ingredient in our designed product The functional food we created could satisfy the needs of customers who are currently consuming nutritionally inadequate foods

The following sections will review our research on the development of the high-throughput method, the anti-diabetic medicinal plant compounds, and starch-based functional foods

Chapter 2 Literature Review

2.1 Starch hydrolase

2.1.1 Starch digestion by starch hydrolase

Starch is a long polymer of glucose molecules, linked by α-(14) glucosidic bonds in amylose, as well as in amylopectin; although amylopectin also contains branch points formed by α-(16) glucosidic bonds between glucose residues [19] In addition, amylose is digested far more efficiently than the branched amylopectin because of its simple structure [19] In humans, the digestion of starch involves several stages It initiates from salivary α-amylase during chewing, which is followed by extensive hydrolysis of the starch into smaller oligosaccharides by pancreatic α-amylase The resultant mixture of oligosaccharides is then broken down to glucose by α-glucosidases α-Amylase is an endo-acting enzyme that catalyzes the hydrolysis of α-1,4-glucan bonds in starch It liberates glucose molecules in units of two or three, and the combined action of salivary and pancreatic amylase leads to the production of large amounts of maltose and maltotriose, and relatively small amounts of limit dextrin [20] However, negligible amounts of glucose are produced by α-amylase

The resultant mixture formed by amylase digestion is broken down to glucose by α-glucosidase in the brush border This enzyme can be further sub-classified into maltase-glucoamylase and sucrase-isomaltase [19] Generally, these enzymes can remove an α-linked glucose, usually from the non-reducing end of a chain [21] In humans, sucrase-isomaltase is far more abundant than maltase-glucoamylase, and is responsible for 80% of the maltase (and maltotriase) activity in the small intestine [22] The α-(16) glucosidic bonds

in starch are almost exclusively hydrolyzed by the isomaltase subunit of sucrase-isomaltase

2.1.2 Starch hydrolase used in diabetes research

α-Amylase is widely present in fungus, plants, and mammals However, there are significant structural differences between the amylases from different

origins In in vitro experiments, mammalian α-amylase is extensively used,

since it is highly conserved with human amylase, while in contrast, fungal amylase is not a good model enzyme [23] Human pancreatic α-amylase (HPA)

is an important pharmacological target for the treatment of type-2 diabetes However, the cost of the HPA is relatively high for research purposes, and so

porcine pancreas α-amylase (PPA) is commonly used for in vitro digestion

measurements PPA is composed of 496 amino acids, and shows 83% similarity to HPA [24] PPA is an endo-enzyme that functions by catalyzing the hydrolysis of α-1,4-glucosidic bonds in amylose and amylopectin randomly from the non-reducing end [25] In addition, PPA has higher turnover number than human salivary amylase (HSA) [20, 23] The role of HSA in starch digestion is relatively minor, because food is only chewed for a short time, and so salivary amylase is not the main focus of anti-hyperglycemia research Instead, pancreas α-amylase and α-glucosidase are the two main targets Rat intestinal acetone powder is commonly used as α-glucosidase by researchers, although the extracts contain many components

2.2 Methods of determining starch hydrolase activity

2.2.1 Conventional methods

The pNPG method has traditionally been used to determine glucosidase activity In the presence of α-glucosidase, p-nitrophenyl-α-D- glucopyranoside (pNPG) is a synthetic substrate that gets hydrolyzed into α-

α-D-glucose and p-nitrophenol (Figure 1) Importantly, sodium bicarbonate

(Na2CO3) needs to be added to the reaction aliquots before UV detection, to bring the pH of the solution from neutral to basic At alkaline pH, the phenolic

proton dissociates (the pKa for p-nitrophenol is approximately 9), giving a

phenolate anion with an intense yellow color that can be easily measured with

a spectrophotometer At the same time, the enzymes are deactivated at alkaline

pH Since samples need to be periodically withdrawn from the enzymatic hydrolysis reaction, this method is not convenient To create a high throughput quantitative measuring method, Kim et al., (2000) modified this process by removing the alkalization step, and tested twenty-one naturally occurring flavonoid inhibitors on α-glucosidase activity [26] However, without alkalization, the method is much less sensitive Moreover, the short detection wavelength (405 nm) used in this method undergoes background interference from natural yellow pigments such as flavonoids and carotenoids

OH OH OH HO

O HO

N + O

O

-HO

Figure 1 The hydrolysis reaction of pNPG by α-glucosidase

Unlike the pNPG method, the GOD/POD method can determine

α-glucosidase activity using natural substrates In principle, glucose is oxidized

by glucose oxidase (GOD) to produce gluconate and hydrogen peroxide

Hydrogen peroxide then reacts with o-dianisidine in the presence of peroxidase to form a colored product, before the oxidized o-dianisidine reacts

with sulfuric acid to form more stable colored products, which can be

measured at 540 nm (Figure 2) The advantages of this method are that a large

quantity of the samples can be assayed at one time, and it also uses natural substrates such as maltose and sucrose However, some limitations have been reported This assay is prone to interference, especially from reducing substances such as polyphenolics Inhibitors such as those based on flavonoids (for example luteolin) with strong antioxidant activity therefore cannot be measured using this method Generally, flavonoids with dihydroxy groups in

the ortho position of the B-ring will cause strong interference [27, 28],

suggesting that this method can only be used for certain inhibitors

Glucose oxidase

The DNSA method has become the most commonly used protocol for measuring α-amylase activity Here, 3,5-dinitrosalicyic acid is an aromatic synthetic compound that reacts with the reducing sugars that are generated from starch hydrolysis to form 3-amino-5-nitrosalicylic acid, which absorbs

light strongly at 540 nm (Figure 3) [9]

Figure 3 The DNSA assay for determining reducing sugar content

The activity of α-amylase can be also determined using D-maltopentaoside (pNPG5) as the substrate This substrate is similar to pNPG, except that a longer chain of five glucose units connects the

p-nitrophenyl-α-chromogenic group 4-nitrophenyl Once the substrate is cleaved by α-amylase,

the released p-nitrophenol can be continuously monitored at 405 nm [29]

Funke and Melzig (2006) have adapted this method to test several plant extracts for their inhibitory effects on porcine pancreatic α-amylase [29]

As an alternative to the DNSA and pNPG5 methods, the starch-iodine

test can also be used to determine α-amylase activity This procedure measures amylase activity by determining starch hydrolysis, as determined by decreased starch iodine formation One of the principal limitations of such an iodine-based method is interference caused by iodine-reducing substances in the samples [30] Another limitation is that an excess of starch can give negative

data, so that serial dilutions of starch are necessary in order to accurately measure enzyme activity In addition, this method is not suitable for screening

a large number of samples

In practice, each method has certain limitations, as described above In addition, there are large differences between determining enzyme activity using natural, such as starch, and synthetic substrates Starch is composed of macromolecules, and has significant micro-structural diversity due to the different ratios of amylose/amylopectin depending on the plant origin, and different ratios of the amorphous/crystalline phases, depending on the conditions used for processing In addition, starch can form supra-molecular complexes with other biomolecules by hydrogen bonding and host-guest interactions The limited water solubility of starch further complicates its enzymatic hydrolysis dynamics These features make it problematic for

synthetic substrates, such as those used in the pNPG assay Using natural

starch to determine the inhibitor activity of starch hydrolase will eliminate the potential problems caused by synthetic substrates At the same time, it is necessary to pursue a more simple and convenient method to monitor the reaction progress in real time, and in particular one where large quantities of samples could be measured at one time

2.2.2 Turbidity measurement applications

The turbidity method is extensively used for microbiological studies For example, yeast cell numbers can be estimated by changes in turbidity, and measured using a spectrophotometer In essence, samples contain insoluble solids that interfere with the transmittance of light, and so the relationship

between absorbance and transmittance can be expressed as shown in Equation

1

A= -Log10 [T] and T= I / I0 Equation 1

Where I: Light passing through the sample

I0: Light entering the sample

A: Absorbance

T: Transmission

Measuring turbidity changes is simple, and can be applied to determining enzyme activity, as well as microbiological studies For studying starch hydrolase, Satoyama et al (1999) reported a method based on the changes in turbidity caused by the hydrolysis of gelatinized starch by amylase

to quantify α-amylase activity [31] The amylase activity was calculated from initial changes in absorbance caused by starch digestion Similarly, Ichiki et al., (2007) adopted this method for their amylase induction inhibition study using 96-cell microplates, a spectrophotometer, and starch-agar gel as the substrate [32] Both studies measured the initial rate as a key kinetic parameter for the quantification However, it is important to consider that starch and gels are macromolecules, and the movement and structural change of macromolecules significantly affects the accuracy of measuring the transmittance at the initial stage of the reaction

2.3 Starch hydrolase inhibitors

Medicines for controlling diabetes and its complications can be mainly divided into insulin, insulin-secretagogues, insulin sensitizers, insulin-like growth factors, aldose reductase inhibitors, α-glucosidase / α-amylase inhibitors, and protein glycation inhibitors [33] The inhibition of α-glucosidase and α-amylase can significantly reduce the post-prandial increase

in blood glucose after a meal, with starch as the major calorie intake The apparent advantage of the starch hydrolase inhibitors is that they function in the small intestine, and as such avoid the issues faced by oral medicines that require absorption and, in some cases, metabolism in the liver It is therefore

an important prevention strategy in the management of type 2 diabetic patients Among them, acarbose has received considerable attention over the past decades In addition, plants extracts have been extensively studied as drugs complementary to acarbose

2.3.1 Acarbose

Acarbose is a pseudo-tetrasaccharide, because of its structural similarity

to typical tetrasaccharides It is bio-synthesized by Actinoplanes, a type of bacteria Acarbose strongly and broadly inhibits the brush-border enzymes glucoamylase, dextrinase, maltase, and sucrase, as well as pancreatic α-

amylase in vitro and in vivo [34] Due to the presence of intramolecular

nitrogen (Figure 4), acarbose attaches to the carbohydrate binding site of the

α-glucosidase enzyme (such as sucrase) with an affinity exceeding that of the normal substrate (for example sucrose), by a factor of 104-105 [34] The enzymatic reaction stops because the C-N linkage in the acarviosine unit of acarbose cannot be cleaved As long as acarbose remains bound to the α-glucosidase enzyme, ingested carbohydrates cannot be digested, and glucose cannot be released for absorption Acarbose reversibly binds to the active site

of glucosidase, and acts as a competitive inhibitor The side effects of acarbose

are significant, due to the large amount of undigested oligosaccharides, which enter the colon, generate gaseous metabolites by bacterial fermentation, and as such give rise to gastrointestinal complaints such as flatulence, meteorism, and diarrhea [35]

NH HO

OH

O HO

2.3.2 Phytochemicals as starch hydrolase inhibitors

A considerable number of hypoglycemic plants have undergone clinical trials, and were found to effectively prevent diabetes In particular, traditional Chinese herbs have shown great potential in fighting diabetes and its complications The structural diversity of plants is a great resource, but is largely unexplored in terms of starch hydrolase inhibitors The bioactive constituents of plants broadly range from polysaccharides, peptides, alkaloids, glycopeptides, triterpenoids, steroids, xanthonone, flavonoids, phenolics, coumarins, iridoids, alkyl disulphides, and guanidine [3, 36-38]

2.3.3 Proanthocyanidins

Plant tannins, one of the major groups of antioxidant polyphenols found

in food and beverages, have attracted a lot of attention in recent years because

of their beneficial health properties Tannin can be broadly classified into two groups: hydrolysable tannins, and condensed tannins, or proanthocyanidins [39] Proanthocyanidins are oligomers and polymers of flavan-3-ols The

skeletal structure of proanthocyanidin is shown in Figure 5 The flavan-3-ol

units are linked mainly through C4C8 bonds, although C4C6 bonds can also exist, and both are called B-type proanthocyanidins [40, 41] A-type proanthocyanidins are defined by the feature of two interflavanoid linkages: a combination of C4C8, and an additional ether bond between C2C7 Proanthcoyanidins that consist exclusively of (epi) catechins are called procyanidins In general, polymeric procyanidins are very polar compounds,

as the hydrophobicity decreases with increasing degree of polymerization [42] Proanthocyanidns are widely present in the plant kingdom, for example

in fruits, berries, seeds, and the bark of pine trees [43] Proanthocyanidins may have potential therapeutic use for the prevention of diabetes and its complications through several mechanisms For example, cinnamon bark

proanthocyanidins can work as the reactive carbonyl scavengers to prevent the formation of advanced glycation end products (AGE), a toxic group of compounds created during glucose metabolism [44] Anderson et al (2004) also found that cinnamon bark A-type proanthocyanidin polymers are potential insulin sensitizers, and stimulate the an reduced production of reactive oxygen species in collagen-stimulated platelets from rats [45] Cinnamon bark proanthocyanidin extracts may also prevent hyperglycemia, by inhibiting α-amylase and α-glucosidase [46] The enzyme inhibitory effects of proanthocynidin have also been described in the studies of sorghum [47], grape pomace [48], and pine bark [49] Procyanidins from French maritime pine bark have been reported to exert clinical anti-diabetic effects after period intake, by inhibiting α-glucosidase As such, pine bark extracts have been commercialized as health supplements [49] Although proanthocyanidin extracts exhibit the greatest health benefits, their rather astringent taste may limit their application in food

2.3.4 Phenylpropanoid sucrose esters

Phenylpropanoid sucrose esters (PSE) are naturally occurring compounds isolated from various plants, and are structurally characterized by

a sucrose core where at least one –OH group is substituted for a

phenylpropanoid Ph-CH=CH-CO- moiety through an ester linkage (Figure 6)

[50] It has been reported that the phenyl ring (-R) can be substituted or unsubstituted Possible substituents include cinnamic, coumaric, ferulic, caffeic, sinapic acid, and glucopyranosylferulic, 3,4,5-trimethoxycinnamic

acid (Figure 7)

The major sources of PSE are plant species of the Polygalaceae, Polygonaceae, and Liliaceae families These groups of compounds are demonstrated to possess many biological activities such as antitumor, antibacterial, antioxidant, antiviral, anti-inflammatory, neuro-protective, and glycosidase inhibitory activities [50]

The anti-diabetic properties of PSE are overlooked, although the plant is commonly used in folk medicine Many PSE isolated possess phenolic moieties, and could act as potential antioxidants Lapathoside D was found to act as an α-glucosidase inhibitor, and exhibited stronger inhibitory activity than acarbose [51] However, the results obtained using α-glucosidase were derived from fungal instead of mammalian sources

Figure 6 Phenylpropanoid sucrose esters (PSEs) with sucrose as the core

structure, where at least one -OH group is substituted by a phenylpropanoid moiety

Figure 7 Phenylpropanoid moiety substituents

2.4 Functional food for diabetes patients

Although plant based therapies have been studied in vitro and in human

clinical trials, only a small number of these were focused on functional foods formulated for diabetic patients The application of polyphenols for use in functional foods is rather challenging, and many aspects should be carefully considered For example, the exact dose of plant extracts added to the functional foods is critical The concentrations of polyphenols that appear to

be effective in vitro are often of an order of magnitude higher than the levels used in vivo In addition, preserving the bioavailability of the active

ingredients during processing, storage, and following oral administration is challenging [52] Other technical issues such as low stability, conditional solubility, and the potential unpleasant taste of some polyphenols have hindered their application

Glycemic index (GI) is a useful concept applied in functional foods for diabetes GI is defined by the level of rise in postprandial blood glucose, as compared with a reference food such as white bread or glucose [53] Food or food products can be categorized into high ( > 70), low ( < 55), or intermediate

GI foods based on their GI value [15] Common starchy staple foods fall into the category of high GI and are positively associated with the risk of developing type 2 diabetes [16, 17] Diabetes patients are recommended to consume low GI foods, since they are digested slowly, and can improve long-term glycemic control [54] Numerous edible plants, and the bioactive

ingredients derived from plants, have been characterized in vitro or in vivo for

their role in preventing the type 2 diabetes Edible plants are therefore good candidates for incorporation into low GI functional products

Low GI starch can occur naturally, or be fortified by encapsulation Natural low GI starch could exist due to the presence of considerable amounts

of resistant starch (RS), for example in legumes and cereals [55] In addition,

some foods inherently contain anti-diabetic compounds Only little research has attempted to add edible plants extracts to starch based food products However, much food-related research has focused on antioxidant activities, for example, turmeric [56], buckwheat [57], green tea [58, 59] and grape seed extracts [60] have all been added as active ingredients to foods In addition, raspberries and blueberries contain high concentrations of anthocyanidin, and have been added to pancakes, but no reduction in glycemic response was observed [61] The high proanthocyanidin content in cinnamon makes it an ideal candidate to reduce postprandial blood glucose, as demonstrated by Hlebowicz et al., (2007 and 2009), and different doses of cinnamon have been added into rice pudding [62, 63] Losso (2009) incorporated 9% fenugreek seed flour into standard wheat flour to make bread Although no significant change in postprandial blood glucose concentration was seen after consumption of the bread, insulin sensitivity was improved [64] In addition, different plants have been shown to have synergistic effects, for example Shakib (2010) tested different ratios of barley, fenugreek, and ginger in bread Barley bread significantly lowered postprandial blood glucose levels, while fenugreek and ginger produced a synergistic effect, with the lowest GI [65] The development of functional foods provides an important opportunity for diabetic patients who want to improve their health

Chapter 3

A high-throughput assay for quantification of starch hydrolase

inhibition based on turbidity measurement

3.1 Introductions

Diabetes mellitus is a metabolic disorder in the endocrine system; diabetic patients do not produce enough insulin or are insulin insensitive and thus have a high postprandial blood glucose level after consuming high sugar

or easily digestible starchy foods [2] Worldwide, more than 220 million

people have diabetes, and this number is increasing rapidly as the aging

population expands Recent warnings on the side effects of anti-diabetic drugs, Rosiglitazone and Pioglitazone [66], highlight the urgent need of alternative

and safer means of blood glucose control, ideally through functional foods that contain bioactive ingredients that may regulate blood glucose concentration toward the normal range Therefore, opportunities and challenges for the food industry are emerging in the area of evidence-based functional foods with a low glycemic index that may decrease starch digestion rates

The digestive process of starch involves saliva amylase, pancreatic amylase, and the small intestinal brush border α-glucosidase, that is, maltase-amyloglucosidase and sucrase-isomaltase [19, 67] Determination of the starch hydrolase activity and inhibitor effectiveness is conventionally done by chromogenic assays such as 3,5-dinitrosalicyclic acid (DNSA) assay for reducing sugar content [9, 68], glucose oxidase/peroxidase (GOD/POD) assay

α-[10, 69], and nitrophenyl-α-D-glucopyranoside (pNPG;[11]) or nitrophenyl-α-D-maltopentaoside (pNPG5) assay [12] The DNSA assay and

p-GOD/POD method are indirect methods and tedious operationally and are thus

not suitable for high-throughput format The pNPG/pNPG5 assay has been

done in 96-well microplate, but they are only synthetic substrates of starch hydrolase The short detection wavelength (405 nm) used may suffer from food sample background interference from natural pigments such as flavonoids and carotenoids Moreover, the use of synthetic substrate may give false positive results For example, resveratrol was initially found to be a sirtuins activator when the synthetic fluorogenic peptide substrate, Fluor de