Optimisation of an in vitro model for anti-diabetic screening by Gayle Pamela Wilson

Submitted in partial fulfillment of the requirements for the degree of MAGISTER SCIENTIAE in the Faculty of Science at the Nelson Mandela Metropolitan University 2006 Supervisor: Dr S Roux Co-Supervisor: Dr M van de Venter

Contents
Summary..........................................................................................................................IV Acknowledgements .........................................................................................................VI List of Abbreviations .................................................................................................... VII List of Figures................................................................................................................... X List of Tables ................................................................................................................XIII Chapter 1 Literature Review ........................................................................................... 1 1.1 Introduction..........................................................................................................1 1.2 Insulin ..................................................................................................................3 1.2.1 Insulin Secretion ...........................................................................................4 1.2.2 Insulin Receptor ............................................................................................6 1.3 Glucose Absorption By Cells...............................................................................9 1.4 Diabetes Mellitus ...............................................................................................13 1.4.1 Pharmacological Treatment ........................................................................15 1.4.1.1) Biguanides ..........................................................................................17 1.4.1.2) Sulfonylurea Drugs.............................................................................18 1.4.1.3) Thiazolidinediones..............................................................................19 1.4.1.4) Insulin .................................................................................................24 1.5 Plants and the treatment of Diabetes Mellitus ...................................................24 1.5.1 Sutherlandia frutescens (Fabaceae) ...........................................................25 1.5.2 Toxicity of Hypoglycaemic Plants..............................................................26 Chapter 2 Introduction to the present study................................................................ 28 Chapter 3 Models for glucose utilization ...................................................................... 30 3.1 General methods ................................................................................................33 3.1.1 Maintenance of cell lines ............................................................................33 3.1.2 Cytotoxicity Assay......................................................................................34 3.1.3 Sulforhodamine B Assay ............................................................................34 3.1.4 Glucose oxidase assay.................................................................................35 3.1.5 Cell viability assay......................................................................................35 3.2 3T3-L1 cell line..................................................................................................35 3.2.1 Induction of 3T3-L1 preadipocyte differentiation ......................................35 3.2.2 Adipocyte differentiation assays.................................................................36 3.2.2.1 Glycerol-3-phosphate dehydrogenase assay ........................................36 3.2.3 Protein concentration determination ...........................................................37 3.2.3.1 Adapted Folin method for microtiter plates.........................................37 3.2.4 Glucose Uptake...........................................................................................37 3.3 C2C12 cell line ..................................................................................................37 3.3.1 Glucose uptake............................................................................................37 3.4 Chang cell line ...................................................................................................38 3.4.1 Glucose uptake............................................................................................38

II

3.5 Statistical analysis..............................................................................................39 3.6 Results and Discussion ......................................................................................39 3.6.1 3T3-L1 cell line...........................................................................................39 3.6.1.1 Cytotoxicity..........................................................................................39 3.6.1.2 Glycerol-3-phosphate dehydrogenase activity.....................................41 3.6.1.3 Glucose uptake.....................................................................................43 3.6.2 C2C12 skeletal muscle cell line..................................................................51 3.6.2.1 Cytotoxicity..........................................................................................51 3.6.2.2 Glucose uptake.....................................................................................52 3.6.3 Chang liver cell line ....................................................................................56 3.6.3.1 Cytotoxicity..........................................................................................56 3.6.3.2 Glucose uptake.....................................................................................57 Chapter 4 Models for insulin secretion......................................................................... 64 4.1 Materials and methods .......................................................................................65 4.1.1 INS cells maintenance.................................................................................65 4.1.1.1 Mercaptoethanol ..................................................................................65 4.1.1.2 Complete Medium ...............................................................................65 4.1.1.3 Poly-L-Lysine (PLL) ...........................................................................66 4.1.1.4 Pretreatment of culture dishes..............................................................66 4.1.2 Insulin Secretion by INS-1 cells .................................................................66 4.1.3 Plasma insulin determination from rat model.............................................67 4.1.3.1 Preparation of medication ....................................................................67 1) Sutherlandia frutescens (kankerbos) .......................................................67 2) Amitriptyline and Metformin ..................................................................67 4.1.3.2 Experimental Procedure.......................................................................68 4.1.4 Binding studies with INS-1 rat pancreatic cells..........................................69 4.1.4.1 Binding time study...............................................................................69 4.1.4.2 Saturation binding study ......................................................................70 4.1.4.3 Displacement study..............................................................................70 4.1.5 Statistical analysis.......................................................................................70 4.2 Results and Discussion ......................................................................................71 4.2.1 Cytotoxicity.................................................................................................71 4.2.2 Insulin Secretion by INS-1 cells .................................................................72 4.2.3 Serum insulin determination in an animal model .......................................82 4.2.4 Binding studies using INS-1 rat pancreatic cells ........................................83 4.2.4.1 Choice of incubation time ....................................................................83 4.2.4.2 Saturation binding................................................................................84 4.2.4.3 Displacement studies ...........................................................................85 Chapter 5 Conclusion ..................................................................................................... 90 References........................................................................................................................ 99 Appendix A .................................................................................................................... 105

III

Summary
The need for alternative strategies for the prevention and treatment of diabetes is growing rapidly as type II diabetes is reaching epidemic status in our society. This need was the basis for the creation of this study, as it was necessary to start looking towards medicinal plants as potential antidiabetic treatment and no comprehensive in vitro model existed. In creating a model for determining the effects of alternative traditional medicines as antidiabetic potentiates, it was necessary that two metabolic pathways, namely glucose uptake and insulin secretion, which play a significant role in glucose homeostasis, be at the centre of our investigations. The objective of this project was to optimize the methodology required to screen and ultimately determine the effectiveness of the plant extracts Kankerbos and MRC2003, as antidiabetic potentiates, through observing their effects on glucose utilisation and insulin secretion. If these medicinal plants are going to make a positive contribution to the health of type II diabetic South Africans, then the determination of their efficacy is essential. The cell lines used in this study included 3T3-L1 preadipocytes, Chang liver, C2C12 muscle and INS-1 rat pancreatic cells. Each cell line represents a different in vivo organ that is known to have an influence on glucose homeostasis in our bodies, each with its own unique metabolic pathways and mechanisms of activity, thereby making each one a vital component in the study. The positive controls for the two models were insulin and metformin (glucose utilisation) and glibenclamide (insulin secretion). Insulin was shown to provide a significant increase in the amount of glucose taken up in C2C12 muscle and Chang liver cells for acute conditions. Chronic treatments with metformin provided a significant increase in glucose utilised by Chang liver cells. Glibenclamide was an effective positive control for stimulating insulin secretion by INS-1 cells under acute conditions as there was a significant increase in the amount of insulin secreted. MRC2003 did not show any significant antidiabetic activity. Sutherlandia frutescens (Kankerbos) showed biological activities comparable to some of the more recognized

IV

antidiabetic compounds throughout the study. With regards to the glucose utilisation model, Kankerbos was seen to have both acute and chronic effects in different cell lines. In the C2C12 muscle cell line, Kankerbos significantly increased glucose uptake when they were exposed to acute conditions. Kankerbos also had a significant effect on the Chang liver cells as it was observed that under both acute and chronic conditions, this plant extract induced the uptake of glucose into these cells. With respect to the insulin secretion model involving INS-1 cells, no significant effect was seen during acute exposure with Kankerbos treatment. However during chronic exposure, an increase in insulin secretion was initiated by this plant extract. Overall, the results of this study suggest that Kankerbos has a twofold mechanism of action for its glucose-lowering effects. Given that Kankerbos is widely available in South Africa, this study was valuable as it provided an indication that Kankerbos has antidiabetic activities and could possibly be used as an alternative antidiabetic medication.

Acknowledgements

I would like to extend my sincere thanks and appreciation to everyone who made this study possible, especially to: God, for His strength and faithfulness Dr Roux and Dr van de Venter for their unwavering guidance and support as well as their willing assistance in the preparation of this manuscript All staff and postgraduate students in the Department of Biochemistry and Microbiology, Nelson Mandela Metropolitan University, for their continual support and encouragement The National Research Foundation and University of Port Elizabeth for their financial support My Mom and Dad, Greg, family and friends for their unconditional love and support

VI

List of Abbreviations

ADD1/SREBP1 ADP AMP Arg ATP BSA Ca2+ CaCl2 C/EBP DEX DHAP DMEM DMSO EDTA FCS FFA GABA GLUT G3PDH g HDL HCl HEPES IRS IBMX KD kg

Adipocyte determination and differentiation factor 1/ sterol regulatory element binding protein-1 Adenosine diphosphate Alpha Adenosine monophosphate Arginine Adenosine triphosphate Beta Bovine serum albumin Calcium ion Calcium chloride CCAAT/enhancer-binding proteins Delta Dexamethazone Dihydroxyacetone phosphate Dulbeccos Modification of Eagles Medium Dimethyl sulfoxide Ethylenediaminetetraacetic acid tetrasodium salt Fetal Calf Serum Free fatty acid Gamma Gamma-aminobutyric acid Glucose transporter Glycerol-3-phosphate dehydrogenase Gram High density lipoprotein Hydrochloric acid N-[2-Hydroxyethyl]piperazine-N- [2- ethanesulfonic acid] Insulin receptor substrate Isobutylmethylxanthine Dissociation constant Kilogram

VII

Kir 6.x K+ K+ATP KRBH l LXR Lys m MAPK MgSO4 2 - ME M n NaCl NaHCO3 PBS PBSA PC-1 PI3-kinase PLL PPAR P r
2

Potassium channel subunit Potassium ion ATP- sensitive potassium channel Krebs-Ringer-bicarbonate-HEPES buffer Litre Lipid-modulated nuclear receptor Lysine Milli Ras/mitogen-activated protein kinase Magnesium sulphate 2 - Mercaptoethanol Micro Molar Nano Sodium chloride Sodium bicarbonate Complete phosphate buffered saline Phosphate buffered saline (without Ca2+ and Mg2+) Plasma cell antigen-1 Phosphatidylinositol 3-kinase Poly-L-lysine Peroxisome proliferator-activated receptor Probability Correlation coefficient Radioimmunoassay Revolutions per minute Roswell Park Memorial Institute culture medium Retinoid X receptor Standard deviation Standard error of mean Sulforhodamine B Sulfonylurea receptor Trichloroacetic acid Triethylamine

RIA rpm RPMI 1640 RXR SD SEM SRB SUR TCA TEA

VIII

Tris TNF TZD WHO

Tris(hydroxymethyl)aminomethane Tumor necrosis factor Thiazolidinediones World Health Organisation

IX

List of Figures
Figure 1.1: Insulin is the principal regulator of energy metabolism..3 Figure 1.2: Schematic diagram of the probable structure of the insulin receptor tetramer in the activated state..7 Figure 1.3: Pathways of insulin control 7 Figure 1.4: Central role of Akt in the actions of insulin8 Figure 1.5: Hypothetical model of the action of insulin on glucose transport specifically for GLUT 412 Figure 1.6: Diabetes mellitus type II: the tip of the iceberg.16 Figure 1.7: Common mechanism of insulin release in response to hyperglycaemia or sulfonylurea administration...18 Figure 1.8: The transcriptional control of adipogenesis involves the activation of several families of transcription factors20 Figure 1.9: Sutherlandia frutescens..26 Figure 3.1: Cytotoxicity results of compounds (BDM1, BDH1 BDP1 and metformin) and plant extracts (Kankerbos and MRC2003) on 3T3-L1 adipocyte cells after a 48 hour exposure period.40 Figure 3.2: The effect of plant extracts (Kankerbos and MRC; 12.5g/well) (with and without inducers) on glycerol-3-phosphate dehydrogenase (G3PDH) specific activity.42 Figure 3.3: The percentage glucose taken up by undifferentiated (a) and differentiated (b) 3T3-L1 cells when exposed to treatments under acute conditions..45 Figure 3.4: Acute effect of treatments on glucose uptake in 3T3-L1 cells after 10 minute incubation time..45 Figure 3.5: Acute effect of treatments on glucose uptake in 3T3-L1 cells after 120 minute incubation time46 Figure 3.6: Comparison of glucose standard curve with and without 10% serum in 8mM RPMI 1640..47 Figure 3.7: Glucose uptake as a percent of control in 3T3-L1 cells with wells lined with PLL (a) and FCS (b)..48 Figure 3.8: Acute effects of treatments (1M BDM1, BDH1 and BDP1) on glucose uptake in 3T3-L1 cells after 60 minute incubation...49

Figure 3.9: Cytotoxicity results of compounds (BDM1, BDH1, BDP1 and metformin) and plant extracts (Kankerbos and MRC2003) on C2C12 muscle cells after a 48 hour exposure period.51 Figure 3.10: The acute effect of positive controls, metformin and insulin (1M) on glucose uptake in C2C12 cells after 90 minute exposure..53 Figure 3.11: Acute effects of treatments on glucose uptake by C2C12 cells after 90 minutes...54 Figure 3.12: Effect of insulin (1M) on glucose uptake in C2C12 muscle cells compared to control after chronic exposure55 Figure 3.13: Cytotoxicity results of compounds (BDM1, BDH1 and BDP1) and plant extract (MRC2003) on Chang liver cells after a 48 hour exposure period...56 Figure 3.14: Acute effect of two positive controls (1M metformin and insulin) on glucose uptake in Chang liver cells57 Figure 3.15: Acute effect of treatments on glucose uptake in Chang liver cells..58 Figure 3.16: Acute effect of treatments on glucose uptake in Chang liver cells..59 Figure 3.17: Chronic effects of treatments exposed to Chang liver cells for 48 hours prior to glucose uptake experimentation 60 Figure 3.18: Effect of different treatments (BDM1, BDH1, BDP1; 1M and Kankerbos and MRC2003; 12.5g/well) on glucose uptake in Chang liver cells under chronic conditions61 Figure 3.19: CellTiter-BlueTM results of Chang liver cells exposed to different treatments for chronic experimentation62 Figure 4.1: Cytotoxicity results of compounds (Glibenclamide and Amitriptyline) and plant extracts (Kankerbos and MRC2003) on INS-1 cell line72 Figure 4.2: Effect of 1M glibenclamide (30 minute exposure) on insulin secretion in INS-1 including 2mM glucose in the medium..74 Figure 4.3: Effect of extracellular glucose concentration (glucose-free, 2mM and 10mM glucose) on insulin secretion in INS-1 cells...74 Figure 4.4: Insulin secretion after 30 minute treatment without glucose and with effect of extracellular glucose concentration (glucose-free, 2mM and 10mM glucose) (10 and 30 minute exposure) on insulin secretion in INS-1 cells including glibenclamide in the test medium..75

XI

Figure 4.5: Effect of extracellular glucose concentration (glucose-free, 2mM and 10mM glucose) (10 and 30 minute exposure) on insulin secretion in INS-1 cells including glibenclamide in the test medium..76 Figure 4.6: Effect of different glucose concentrations (basal, 2mM and 10mM) on insulin stimulation in INS-1 cells under acute conditions.77 Figure 4.7: Effect of compounds (glibenclamide and amitriptyline, 1M) and plant extracts (Kankerbos and MRC2003, 0.5ng/well) on insulin secretion in INS-1 cells under acute conditions (10 minute exposure)78 Figure 4.8: Effect of extracellular glucose as well as the test compounds (Glibenclamide and Amitriptyline, 1M) and plant extracts (Kankerbos and MRC2003, 0.5ng/well) on insulin secretion in INS-1 cells under chronic conditions80 Figure 4.9: Fasting serum insulin concentrations as a result of a 19 week treatment with the medication...82 Figure 4.10: Specific binding of [3H]glibenclamide by INS-1 cells as a function of incubation time84 Figure 4.11: Saturation binding curve..84 Figure 4.12: Saturation binding curve..85 Figure 4.13: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M) and amitriptyline (0.67M and 6.67M)87 Figure 4.14: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M), amitriptyline and trimipramine (0.67M and 6.67M)..88 Figure 4.15: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M) and varying concentrations of unlabelled amitriptyline (0.67nM 6.67M)...88 Figure 5.1: Summary of carbohydrate metabolism .92 Figure 5.2: Mechanisms of insulin action93 Figure 5.3: Mechanisms of metformin action on hepatic glucose production and muscle consumption...94

XII

List of Tables

Table 4.1: Experimental plan for medicinal exposure .................................................68 Table 4.2: Probability (P) of binding to SUR-1 receptors ...........................................86 Table A1: Conversion table for plant extract concentrations ....................................105

XIII

Chapter 1 Literature Review

1.1 Introduction The identification of substances that mediate or mimic the action of insulin could lead to the development of novel structures which may be of clinical use in the treatment of persons having disorders of glucose metabolism, such as impaired glucose tolerance, elevated blood glucose associated with type II diabetes and insulin resistance (Larner et al., 1997). Modern drug discovery requires a systematic approach to optimize time and resource use in order to test the maximum number of samples in systems which hopefully predict therapeutic efficacy (Wagner and Farnsworth, 1994). Techniques for the study of hypoglycaemic activity in vivo employ animals with normoglycaemia or induced hyperglycaemia, as well as diabetic humans. In vivo bioassays are essential to prove the value of new hypoglycaemic agents, however, animal tests reveal relatively little about the specific mechanisms of action of the compound, and it is evident that there are a great many mechanisms by which blood glucose levels may be reduced. Some of these mechanisms, such as those producing hypoglycaemia as a side-effect of their toxicity, are obviously not useful in diabetes treatment. The lack of perfect models for type II diabetes, coupled with financial restrictions on obtaining and maintaining animals, and social restrictions on extensive use of animals in experimentation, indicate that a more practical approach would involve a series of in vitro prescreens before testing a potential new hypoglycaemic agent in animals. Many in vitro techniques have been developed to elucidate the varied mechanisms of action of hypoglycaemic agents discovered by in vivo bioassays. Three aspects of the hypoglycaemic response are commonly studied in vitro: insulin release from the pancreatic islets, peripheral insulin binding and glucose uptake, and effects on hepatic enzymes (Wagner and Farnsworth, 1994).

During the last 20 years, the considerable and significant advances in tissue culture methodology, the use of chemically-defined cell and tissue culture media, and the availability of mammalian cells have transformed in vitro methods from a new technology to a valuable research tool. In the past, in vitro methodology was used as the last approach in product development. Today, the use of in vitro tests in product development, drug discovery and safety evaluation has become commonplace, resulting almost exclusively from the evolution of science rather than any fundamental change in philosophy. Yet all in vitro methods are alternatives to animal testing. It is not known how well any in vitro system would recapitulate the in vivo system. Thus, it would be difficult to design an in vitro test battery to replace in vivo test systems. In vitro systems are well suited to the study of biological processes in a more isolated context, therefore, in vitro tests have their greatest potential in providing information on basic mechanistic processes in order to refine specific experimental questions to be addressed in the animal. Advantages of in vitro models include the ability to directly manipulate cultivation conditions and isolate the target tissue from the physiological effects of other organs and tissues. Additional unique competitive advantages of in vitro models include: 1) Rapid large-scale screening of drug candidates allows promising substances to be identified at an early stage of the drug development process; 2) The time and costs involved in developing active agents are significantly reduced and 3) Cost-intensive and ethically controversial animal experiments can be reduced to a minimum. In order to determine the best model for antidiabetic screening, it is essential to elucidate and understand the molecular machinery involved in the regulation of blood glucose levels.

1.2 Insulin Insulin is the most potent anabolic hormone known and it is essential for appropriate tissue development, growth, and maintenance of whole-body glucose homeostasis. It regulates glucose homeostasis at many sites, reducing hepatic glucose output (via decreased gluconeogenesis and glycogenolysis) and increasing the rate of glucose uptake, primarily into striated muscle and adipose tissue (Figure 1.1). It also profoundly affects lipid metabolism, increasing lipid synthesis in liver and fat cells, and attenuating fatty acid release from triglycerides in fat and muscle (Pessin and Saltiel, 2000).

Figure 1.1: Insulin is the principal regulator of energy metabolism. When glucose or other nutrients are absorbed from the gastrointestinal tract, this elicits insulin secretion. Insulin regulates the metabolism of multiple fuels. Selected actions of insulin are indicated (+, activation; -, inhibition). Insulin activates transport of glucose into muscle and adipose tissue, and promotes synthesis of glycogen and triglycerides. Insulin also inhibits hepatic glucose production by inhibiting both glycogenolysis and glucogenesis. Insulin does not directly regulate the metabolism of red blood cells, which uses glycolysis to provide energy. Although the brain uses glucose in the fed state, it can also use ketone bodies when levels rise high enough (eg. during fasting) (Taylor, 1999)

Despite significant advances in past years on the chemistry and biology of insulin and its receptor, the signaling mechanisms involved in the various biologic responses to insulin remain somewhat elusive (Pessin and Saltiel, 2000). Progress in this area has been complicated by the pleiotropic nature of the actions of insulin. The relative activation and coordination of these distinct cellular processes by insulin varies with cell type, state of differentiation of the cell, presence of other hormones, and insulin dose response and time course, suggesting that insulin action involves a network of inter-related and independent pathways with differing levels of divergence regarding mechanisms of regulation (Saltiel, 1990). 1.2.1 Insulin Secretion Insulin is released from the pancreatic -cell directly from the granules by exocytosis and the movement of the granules to the cell membrane in response to stimulation probably involves microfilaments and microtubules. Insulin is released from pancreatic -cells at a low basal rate and at a much higher stimulated rate in response to a stimulus. The most important stimulus for insulin secretion is an increase in the extracellular concentration of glucose. Within one minute of the addition of glucose to the tissue an increased rate of secretion occurs. The response of insulin secretion to the change in glucose concentration is sigmoid so that there is little response below 5mM, and a 50% response at about 8mM (Newsholme and Leech, 1992). Pancreatic cells secrete insulin in a pulsatile fashion and, in response to a square-wave increase in interstitial glucose concentration, release insulin in a biphasic manner, characterized by a spike lasting approximately 10 minutes (first-phase release) and followed by gradually increasing release (second-phase release). It has been suggested that these different phases of insulin released represent two different intra-islet pools: one a rapidly releasable pool accounting for about 5% of islet insulin represents granules close to the cell membrane and is thought to be responsible for first-phase insulin release. The second is a reserve pool, the release of which requires adenosine triphosphatedependent mobilization of insulin-containing granules into the rapidly releasable pool for

subsequent exocytosis. Both phases of insulin release are important for maintaining normal glucose homeostasis. However, considerably more emphasis has been placed on the importance of first-phase insulin, assuming that this is the major determinant of early insulin release, that is, the increase in plasma insulin levels observed during the initial 30 minutes following glucose or meal ingestion (Gerich, 2000). The proposed hypothesis of insulin secretion: An increase in the extracellular glucose concentration above 5mM increases proportionally the rate of glycolysis (through operation of the glucose/glucose 6-phosphate cycle) and this raises the concentration of phosphoenol pyruvate which, increases the rate of uptake of calcium ions and probably increases the rate of release from intracellular calcium stores. An increased cytosolic concentration of calcium ions, via calmodulin, causes contraction of the microfilaments or microtubules and hence results in an increased rate of exocytosis and insulin secretion. The hypothesis attempts to link the rate of glycolysis, calcium ions and rate of insulin secretion (Newsholme and Leech, 1992). Potassium channels are sensitive to ATP and function in coupling cell metabolism to membrane potential in many tissues (Hernandez-Sanchez et al., 1999). K+ATP channels have been found in a variety of tissues including heart, pancreatic -cells, skeletal muscle, smooth muscle and the central nervous system (Isomoto et al., 1996). K+ATP channels comprise an octameric complex of pore-forming Kir6.x subunits and regulatory sulfonylurea receptors (SURs) (Gribble and Ashcroft, 2000). There are three isoforms of the sulfonylurea receptor, SUR1 and two spliced variants of SUR 2, SUR 2A and SUR 2B. The SUR1-Kir 6.2 and SUR2B-Kir 6.2 or Kir 6.1, constitute the cardiac and vascular smooth muscle-type channels, respectively (Hernandez-Sanchez et al., 1999). SUR is a member of the family of ATP-binding cassette (ABC) transporter proteins and appears to be the major determinant of the pharmacological properties of K+ATP channels (Gribble and Ashcroft, 2000; Isomoto et al., 1996). K+ATP channels which consist of SUR1 and Kir 6.2 do not only occur in cells but are also present in the alpha, delta and pancreatic polypeptide cells of the pancreatic islets (Fujikura et al., 1999). The K+ATP (SUR1-Kir 6.2) channels mediate glucose-induced insulin secretion in pancreatic cells. K+ATP

channels are modulated by intracellular ATP/ADP ratios: ATP closes the K+ATP dependent channels, while ADP opens them (Raab-Graham et al., 1999). When the level of blood glucose increases, glucose enters -cells via GLUT 1, 2 and 3 transporters, which are insulin independent (Walker, 2000). Following glucose metabolism, the intracellular ATP/ADP ratio increases thereby inactivating the K+ATP channels. This increase in the ATP concentration causes closure of the K+ATP sensitive channels and the efflux of K+ through the channel is decreased causing depolarization of the membrane. This in turn results in the opening of the voltage-dependent calcium channels with a subsequent increase in the intracellular Ca2+, which initiates insulin secretion (Hernandez-Sanchez et al., 1999). 1.2.2 Insulin Receptor Once insulin has entered the circulation it reacts with target cells that have insulin receptors on their plasma membranes (Katzung, 1995). The most important target cells, which have insulin receptors are liver, muscle and fat. The number of insulin receptors on individual cells and the affinity of the receptor to insulin varies. The receptors bind insulin with high specificity and affinity in the picomole range. Insulin action is initiated through the binding to and activation of its cell-surface receptor, which consists of two subunits and two subunits that are disulfide linked into an 22 heterotetrameric complex, as illustrated in figure 1.2. Insulin binds to the extracellular subunits, transmitting a signal across the plasma membrane that activates the intracellular tyrosine kinase domain of the subunit. Although PI 3-kinase activity is clearly necessary for insulin-stimulated glucose uptake, additional signals are also required for the stimulation of GLUT4 translocation (Pessin and Saltiel, 2000).

Figure 1.2: Schematic diagram of the probable structure of the insulin receptor tetramer in the activated state (Katzung, 1995)

Upon binding insulin, the protein tyrosine kinase phosphorylates itself as well as target substances (Figure 1.3), such as the insulin receptor proteins (IRS-1 and IRS-2), Cbl and p52Sho (Galic et al., 2005). IRS-1 plays a more prominent role in stimulating glucose uptake by muscle and fat, whereas IRS-2 functions mainly in the liver. It has been discovered that IRS-2 boosts insulin production by the pancreas (Alper, 2000).

Figure 1.3: Pathways of insulin control (Alper, 2000)

These phosphorylation events allow for recruitment and activation of signaling pathways, including Ras/mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3kinase (PI3-kinase)/Akt pathways that mediate the metabolic, transcriptional and mitogenic actions of insulin (Figure 1.4) (Galic et al., 2005).

Figure 1.4: Central role of Akt in the actions of insulin (Galic et al., 2005)

1.3 Glucose Absorption By Cells The entry of glucose into cells is a crucial step in life-supporting processes since glucose is the main monosaccharide in nature that provides carbon and energy for almost all cells. The passage of glucose into cells depends on different parameters, including expression of the appropriate glucose transporters in the target tissues and hormonal regulation of their function (Gorovits and Charron, 2003). Glucose is a hydrophilic compound; it cannot pass through the lipid bilayer by simple diffusion, and therefore requires specific carrier proteins to mediate its specific transport into the cytosol. Cells take up glucose by facilitated diffusion, via glucose transporters (GLUTs) associated with the plasma membrane (Medina and Owen, 2002). Until now the search for the mammalian facilitative glucose transporters has yielded 12 carriers including GLUT15 and the recently discovered GLUT612. The kinetic properties and substrate specificities of the different isoforms are specifically suited to the energy requirements of the particular cell types (Medina and Owen, 2002). The transporters form a family whose members are broadly alike in structure and function. Each transporter consists of a polypeptide chain consisting of approximately 500 amino acids (Lienhard et al., 1992). All GLUTs have been predicted to have 12 membranespanning domains (helices) connected by hydrophilic loops, the first of which is exofacial and contains an N-glycosylation site in GLUT15. Both the amino and carboxyl termini of GLUTs reside on the cytoplasmic side of the cell membrane. Models of GLUTs suggest that five of the transmembrane helices form an aqueous pore providing a channel for substrate passage. Lack of a crystal structure leaves the precise structure of GLUTs hypothetical (Gorovits and Charron, 2003). Glucose passage into the cell is an intricate one. In the process, the transporter takes on two shapes, one which binds glucose on the extracellular side and the other which binds it on the intracellular side. It is suggested that a glucose molecule enters the cell through a four step process. First, it occupies the outward-facing binding site. Second, the complex of transporter and glucose changes conformation. The glucose occupies the binding site

facing into the cell. Third, the transporter releases the glucose into the cytoplasm. Fourth, the unoccupied transporter changes back into the conformation in which the binding site is on the outward side, thereby enabling the binding of another glucose molecule. The transporter is envisioned as a conformational oscillator which shifts the binding pocket for glucose between opposite sides of the cell membrane (Lienhard et al., 1992). GLUTs expression is cell-specific and subject to hormonal and environmental control. (Medina and Owen, 2002). GLUT1 is a high affinity glucose transporter that is ubiquitously expressed in most mammalian tissues. It provides basal glucose transport and, most importantly, transport of glucose through the blood-brain barrier, erythrocytes, and neuronal cell membranes. GLUT2 is a low affinity glucose transporter with a high turnover rate expressed in adult liver, kidney, intestinal epithelium, and pancreatic -cells (Gorovits and Charron, 2003). These kinetic properties allow GLUT2 to function in the liver where glucose transport must not be rate limiting for influx or efflux. When circulating glucose levels are high there needs to be net hepatic uptake as the intracellular glucose is metabolized or converted into glycogen. Conversely, when glucose levels are low, the liver needs to export glucose to the plasma. This is achieved by GLUT2 coupled with the regulated phosphorylating activity of hexokinase IV. GLUT2 therefore functions in the regulation of insulin levels because changes in the blood glucose level are effectively transmitted to the liver and the pancreatic cells by GLUT2. In order to regulate insulin secretion, pancreatic cells need to be highly sensitive to changes in plasma glucose concentrations. Therefore, a low-affinity transporter, such as GLUT2 will not be saturated at physiological levels and glucose flux will be proportional to plasma glucose concentration. As in the liver, hexokinase regulates the entry of glucose into the glycolytic pathway and, along with GLUT2, plays a role in glucose sensing by -cells (Medina and Owen, 2002). GLUT3 is found in the neuronal cells of the brain and functions to ensure a constant movement of the glucose into these cells. This transporter takes up glucose and converts it into energy-yielding compounds in an insulin-mediated step (Katzung, 1995; Lienhard et al., 1992). The insulin-stimulated, high affinity glucose transporter protein GLUT4 is expressed predominantly, but not exclusively, in all insulinsensitive tissues: skeletal muscle, adipose tissue, and heart. GLUT4 is unique in that it is

10

the only insulin-responsive GLUT that has been characterized so far. Stimuli leading to the increase in translocation of GLUT4 to the plasma membrane include insulin, contraction, and hypoxia (Gorovits and Charron, 2003). One of the primary functions of insulin is to facilitate the disposal of blood glucose into the peripheral tissues during the post-absorptive state. Insulin increases cellular uptake and metabolism of glucose by accelerating its transmembrane transport (Johnson, 1998). Undoubtedly GLUT4 translocation to the plasma membrane from the intracellular pool leads to an increase in glucose uptake. Insulin-stimulated translocation of GLUT4 to the plasma membranes requires phosphatidylinositol 3-kinase (PI3K) activity. However, the activity of PI3K alone is not sufficient to cause a GLUT4-mediated increase in glucose transport, suggesting that additional stimuli are important for GLUT4 translocation (Gorovits and Charron, 2003). GLUT5 is found in the small intestine and kidney, where it functions in the absorption of fructose (Katzung, 1995; Lienhard et al., 1992). GLUT1 and GLUT4 are the main GLUTs functioning in adipose and muscle tissue. GLUT1 is thought to play a constitutive role, and is responsible for basal glucose uptake while GLUT4 is the inducible transporter. One of the most important, and well established, models of GLUT regulation is the stimulation of GLUT expression and translocation in adipose and muscle tissue by insulin. It is this process that provides the regulation of whole-body glucose homeostasis and, when dysfunctional, plays a vital role in diabetes mellitus. GLUT4 is almost completely responsible for insulin-stimulated glucose transport. In rat adipocytes, the most studied cell system for insulin action on glucose transport, more than 95% of GLUT4 and 30-40% of GLUT1 is associated with intracellular membranes, and are therefore non-functional. These GLUTs are translocated to the plasma membrane in response to insulin, where they are then able to facilitate the transport of substrate (Medina and Owen, 2002). It is well established that the mechanism by which glucose transport (Figure 1.5, step 6) is upregulated in fat and skeletal muscle, in response to insulin, involves recruitment of the insulin-responsive GLUT4 transporter from an intracellular compartment to the plasma membrane (steps 3-5). Insulin induces intracellular vesicles containing GLUT4 to

11

move to the inner surface of the cell membrane and fuse with it. This occurs after intracellular signaling apparatus turns on a multistep signaling pathway that activates the glucose transporter, due to binding of and activation by insulin to its receptor (steps 1-2). The transporters are retrieved to the interior when small vesicles formed through membrane invagination and fission fuse with larger endosomes, where the transporter segregates into the tubular extentions that pinch off to form new vesicles (step 8). As long as insulin remains, the vesicles continue to fuse with the cell membrane, but a lowered level of insulin breaks the cycle (step 7) and the glucose transporter accumulates in intracellular vesicles (Lienhard et al., 1992).

Figure 1.5: Hypothetical model of the action of insulin on glucose transport specifically for GLUT 4 (Johnson, 1995)

12

1.4 Diabetes Mellitus Type II diabetes mellitus is a heterogeneous disorder due to a combination of inherited and acquired factors that adversely affect glucose metabolism. It is thought that these factors lead to diabetes mainly by affecting -cell function and tissue insulin sensitivity. If the amount of insulin produced is too little to allow for glucose to be used or stored, or if the insulin being produced does not work effectively, glucose accumulates in the blood. Hyperglycaemia develops when rates of glucose release into the circulation exceed rates of tissue glucose uptake. This may occur because release is increased, because uptake is reduced, or due to a combination of factors such as increased release with a lesser increase in uptake (Gerich, 2000). In the normal individual, the concentration of glucose in blood is maintained at about 90 mg/dL of plasma. However, fasting blood glucose in diabetics may be 300-400 mg/dL and may even reach 1000 mg/dL (Johnson, 1998). Type II diabetes is associated with insulin resistance initially and later, as the function of the -cell decreases, insulin deficiency (Cerasi, 2000). Type II diabetes is characterized both by abnormalities of insulin secretion progressively leading to secretion failure as well as insulin resistance of all major target tissues (Haring, 1999). Although insulin resistance is important in the early stages of type II diabetes, the failure in adequate -cell compensation leads to the progression to the diabetic state. Compensation for insulin resistance is through increased secretion per -cell or by an increase in -cell mass through neogenesis or replication of the existing -cells (Withers et al., 1998). Beta-cell mass is normally tightly maintained through a balance of -cell birth (-cell replication and islet neogenesis) and -cell death through apoptosis. Most of the increase in -cell mass with insulin resistance is probably due to increased -cell number, but -cell hypertrophy may also contribute (Weir and Bonner-Weir, 2004). It is suggested that the disease is triggered when the delicate balance between insulin production and insulin responsiveness goes awry. First, cells in muscle, fat and liver lose some of their ability to respond fully to insulin. In response to growing insulin resistance, pancreatic cells temporarily over produce insulin causing hyperinsulinemia (Alper, 2000). Much of the increase in insulin secretion undoubtedly results from the increase in 13

-cell mass. At some point, -cells are no longer able to keep glucose levels in the prediabetic range. This failure presumably occurring because of a critical decline of -cell mass and/or increase in insulin resistance (Weir and Bonner-Weir, 2004). But the insulinproducing cells eventually die, leading to full-blown diabetes. Therefore type II diabetes results when the body loses the fine-tuned balance between insulin action and insulin secretion. For years up until recently, it was believed that a malfunction in the insulin receptor leads to insulin resistance, however researchers are converging on a new hypothesis to explain this metabolic disorder. The shift in thinking occurred as a result of failure in linking insulin receptor malfunction and the disease. It is now believed by some researchers that two related pathways that normally respond to insulin by signaling cells in the tissues to remove glucose, lie at the heart of insulin resistance. They are however unsure as to why the biochemical pathways believed to be involved are not functioning properly. It is believed that defects in the insulin signaling pathway leading to the disease are subtle, as no mutations in the insulin receptor substrate (IRS) genes in diabetics have been uncovered. Due to insulins regulation of glucose metabolism being so finely tuned, it is believed that one or two subtle mutations will upset the entire system when combined with the proper environmental insults. Another new insight into type II diabetes is that in order for this disease to develop, insulin resistance must occur in both muscle and liver, leading researchers to conclude that the insulin regulating system must fail at multiple points (Alper, 2000). The new insights into type II diabetes are mainly from studies in knockout mice, however, human genetics does not necessarily correlate with animal data, therefore not everyone is convinced and the old dogma is still being followed. The major causes of insulin resistance of the skeletal muscle in the prediabetic state may be discussed as genetic background, obesity related insulin resistance and of physical inactivity. Type II diabetes is known to have both genetic and environmental determinants and is strongly associated with age and obesity (Sheard and Clark, 2000). Among the environmental factors causing insulin resistance, obesity is of predominant importance (Haring, 1999). Obesity is defined as an excess of body weight that is mainly attributable to an increased body fat accumulation (Tremblay and Doucet, 2000). Obesity is the most important

14

modifiable risk factor for type II diabetes (Sakurai et al., 1999). About 90 percent of people with this form of the disease are overweight (Sheard and Clark, 2000). An excess of body fat is associated with a deterioration of glucose utilisation and promotes development of type II diabetes, particularly in those with a genetic predisposition for the disease (www.geocities.com/jqjacobs). There are a number of different hypotheses explaining the mechanism of insulin resistance in obesity. It was proposed that tumor necrosis factor alpha (TNF-) is released by adipose tissue and is able to impair insulin signaling through serine kinase and tyrosine phosphatase dependent modulation of the insulin signaling chain at the level of the insulin receptor and substrates (IRS) (Haring, 1999). Another proposed mechanism of insulin resistance in obesity is related to a protein molecule which interferes with insulin action. Membrane glycoprotein PC-1 (plasma cell antigen-1) has been shown to reduce the uptake of glucose by cells through inhibiting insulin receptor tyrosine kinase activity. Obesitys contribution to the onset of diabetes may be by increasing the levels of PC-1 (www.pslgroup.com). The increase in obesity prevalence has led the World Health Organization (WHO) to refer to a global epidemic to describe the obesity issue. Body weight and fat losses are essential if the health burden which obesity imposes on a great proportion of individuals throughout the world is to be alleviated (Tremblay and Doucet, 2000). 1.4.1 Pharmacological Treatment In figure 1.6, the simplified schematic presentation illustrates the evolution of type II diabetes mellitus. Type II diabetes is a progressive disease. In these patients the blood glucose level often does not rise high enough to produce many symptoms and therefore goes undiagnosed. Because insulin is being produced they can often be treated without insulin and may respond to dietary measures with or without addition of medication (Vardaxis, 1994). Diabetes mellitus type II represents the end stage of long lasting metabolic disturbances caused by insulin resistance associated with hyperinsulinemia, obesity, dyslipoproteinemia, arterial hypertension and consequently premature atherosclerosis. Since this detrimental metabolic milieu is present for many years before 15

plasma glucose levels are elevated, it is not surprising that type II diabetic patients have already micro- and/or macrovascular complications at the time of the initial diagnosis. Subjects in stage I have normal glucose tolerance due to the ability of their -cells to compensate for the insulin-resistant state. At this stage elevated triglyceride levels and reduced HDL levels as well as an increased waist to hip ratio may indicate insulin resistance and should lead to therapeutic action. In stage II, glucose tolerance after an oral glucose load (75g) is impaired due to developing insulin-secretory deficiency. To avoid progression to clinically overt type II diabetes (stage III), these IGT subjects must receive treatment options to reduce insulin resistance, such as dietary advice and increase of physical activity (Matthaei et al., 2000).

Figure 1.6: Diabetes mellitus type II: the tip of the iceberg (Matthaei et al., 2000)

Until the importance of screening for, as well as treating, the early stage of the metabolic syndrome is appreciated, the optimal treatment of patients with type II diabetes mellitus to avoid diabetic complications and to preserve quality of life is a major focus in todays medical world. Although nonpharmacological treatment modalities such as reduced caloric intake and increased physical activity represent the basis of the treatment of insulin resistance and their efficacy have been demonstrated in numerous studies, the

16

actual number of patients sufficiently treated without pharmacological agents is comparatively low. Therefore, pharmacological treatment is required in the vast majority of type II diabetic patients (Matthaei et al., 2000). The choice of an oral antidiabetic agent may be influenced by a large number of factors, but often comes down as much to personal preference or experience as to detailed knowledge of the differential actions of each molecule. The following are the available antihyperglycaemic agents which are known to ameliorate insulin resistance, namely, biguanides, sulfonylureas, thiazolidinediones and insulin. 1.4.1.1) Biguanides Metformin is a hypoglycaemic drug effective in the treatment of type II diabetes mellitus (Klip and Leiter, 1990). Despite almost 40 years of research, the precise mechanism of metformin action is still not entirely understood. Several cellular mechanisms have been described but a single unifying site of action has yet to be identified (Matthaei et al., 2000). Biguanides have no direct effect on insulin release and in fact actually reduce the serum insulin level in both the basal and stimulated states. This decrease is believed to be secondary to the biguanide-induced decrease in blood glucose concentration and is usually attributed to one or more influences, working individually or in combination, that include: decreased lumen-to-plasma glucose transport, suppression of hepatic gluconeogenesis coupled with a net decrease in hepatic glucose output, and increased anaerobic glucose utilization in tissues. At the subcellular level, metformin has been shown to increase insulin binding to its receptor both in vitro and in vivo. Therefore, metformin promotes glucose uptake at tissue level and insulin secretion is not enhanced by its action (Foye et al., 1995; Klip and Leiter, 1990). Metformin appears to be the drug of choice to start pharmacological treatment in insulin resistant and overweight/obese diabetic subjects. Unlike other pharmacological therapies for type II diabetes, metformin is not associated with weight gain (Matthaei et al., 2000).

17

1.4.1.2) Sulfonylurea Drugs Sulfonylureas are a class of compounds available for treating hyperglycemia in noninsulin-dependant diabetics. Sulfonylurea drugs are oral hypoglycaemic agents, which increase the release of endogenous insulin as well as improve its peripheral effectiveness (Katzung, 1995). Sulfonylurea drugs have three mechanisms of action including the release of insulin from pancreatic cells, the reduction of serum glucagon levels and directly increasing tissue responsiveness to insulin (Clark et al., 1997). Adenosine triphosphate-sensitive potassium (K+ATP) channels are the targets for the sulfonylurea drugs used in the treatment of type II diabetes mellitus (Figure 1.7). They are found in a wide range of tissues, however, in cells, K+ATP channels provide a link between the glucose concentration and the rate of insulin secretion (Gribble and Ashcroft, 2000).

Figure 1.7: Common mechanism of insulin release in response to hyperglycaemia or sulfonylurea administration (Harrower, 2000).

18

In type II diabetes, once insulin secretion becomes defective, the drug, sulfonylurea, is administered to increase insulin release. Sulfonylureas promote insulin secretion by favouring the closure of K+ATP channels. Sulfonylureas block K+ATP (SUR1-Kir 6.2) channels by binding to the receptor (SUR1). Binding of the sulfonylurea inhibits the efflux of K+ through the channel and results in depolarization. Depolarization of the membrane provokes calcium influx by opening a voltage-gated Ca2+ channel. The increased intracellular Ca2+ concentration triggers the secretion of insulin (Katzung, 1995; Quesada et al., 1999). The use of sulfonylureas in the treatment of diabetes mellitus is controversial. Hypoglycaemia is a danger with sulfonylureas and may be caused by drug overdose, drug interactions, altered drug metabolism or the patient failing to eat (Clark et al., 1997) 1.4.1.3) Thiazolidinediones Adipose tissue has been the subject of intense scrutiny, and one important reason for this is that this tissue provides a critical link in maintaining systemic energy balance. The ongoing explosion in the incidence of obesity and its ugly stepsister, type II diabetes, has focused attention on all aspects of adipocyte biology, including adipogenesis (Rosen et al., 2000). Functionally, cellular differentiation can be thought of as a shift in gene expression patterns, such that transcripts that determine the primitive, multipotent state give way to those that define the final phenotype. Morphological changes result from the actions of the genes that are induced as the cells differentiate, including alterations in the cell shape and the accumulation of lipid that accompany adipogenesis. A transcriptional cascade has been found to drive adipogenesis. Three classes of transcription factors have been identified that directly influence fat cell development. These include peroxisome proliferator-activated receptor (PPAR), CCAAT/enhancer-binding proteins (C/EBPs) and the basic helix-loop-helix family, adipocyte determination and differentiation factor 1/sterol regulatory element binding protein-1 (ADD1/SREBP1c) (Rosen et al., 2000)

19

After the onset of differentiation, a cascade of gene expression begins with the rapid induction of C/EBP and (Figure 1.8).

Figure 1.8: The transcriptional control of adipogenesis involves the activation of several families of transcription factors (Rosen et al., 2000)

Concomitantly, synchronous re-entry into the cell cycle occurs and cells proceed through a mitotic clonal expansion phase that consists of approximately two rounds of mitosis. Near to or upon completion of mitotic clonal expansion, expression of C/EBP and PPAR is induced, and expression of C/EBP and begins to decline (Dowell et al., 2000). This induction is likely to be a direct transcriptional effect through C/EBP binding sites in the PPAR promoter. PPAR, in the current model for a transcriptional network in adipogenesis, is then responsible for inducing C/EBP. C/EBP exerts positive feedback on PPAR to maintain the differentiated state (Rosen et al., 2000). Both C/EBP and PPAR, the latter in combination with the obligate heterodimeric partner and nuclear hormone receptor family member, retenoid X receptor (RXR), have been shown to bind regulatory elements within the promoter of the 422/aP2 gene.

20

In addition ADD1/SREBP1c, a member of the basic helix-loop-helix leucine zipper transcription factor class, promotes lipogenic gene expression and stimulates production of an unidentified PPAR ligand (Dowell et al., 2000). ADD1/SREBP1 can activate PPAR by inducing its expression as well as by promoting the production of an endogenous PPAR ligand. Thus, C/EBP, PPAR and ADD1/SREBP1 cooperatively promote adipogenesis and subsequent maintenance of the adipocyte phenotype (Rosen et al., 2000). Thiazolidinediones (TZDs) are structurally related agents used for the treatment of type II diabetes. In addition to improving glucose metabolic insulin sensitivity and reducing the requirement for insulin, these agents impressively reduce hypertriglyceridemia and have been reported to lower plasma free fatty acid (FFA) levels in various animal models of insulin resistance. Despite the fact that these compounds have been studied for almost 20 years, the in vivo mechanisms of insulin-sensitization remain unclear. Improvements in glucoregulation may occur via their action on lipid metabolism; a body of evidence suggests that an oversupply of fatty acids to insulin-sensitive target tissues, particularly liver and skeletal muscle, contributes to the development of insulin resistance. Treatmentinduced reductions in fatty acid availability may thus contribute to the insulin-sensitizing effects of these compounds (Oakes et al., 2001). TZDs may exert their effects via ligand activation of PPAR, which is primarily expressed in adipose tissue. PPAR activation contributes to the triggering of differentiation of preadipocytes and induces expression of genes involved in the transport and sequestration of FFA. These cellular and molecular effects of PPAR agonism could influence FFA exchange between adipose and other tissues, but direct evidence based on measurements of in vivo fluxes and the metabolic fate of FFA are lacking (Oakes et al., 2001). Compounds that bind to and activate the PPAR subunit of the PPAR - RXR nuclear receptor heterodimer alter transcription of genes involved in glucose and lipid metabolism. Included in these target genes are lipid transporters (CD36, aquaporin), key 21

metabolic enzymes (lipoprotein lipase, phosphoenolpyruvate carboxykinase, uncoupling protein-1), adipocyte-enriched signaling molecules (leptin, resistin, ACRP30, FIAF/PGAR), lipid-modulated nuclear receptors (LXR), and an intermediate in the insulin signaling pathway (c-Cbl-associating protein) (Forman, 2002). TZD treatment therefore results in expression of a number of adipocyte-specific genes so that PPAR activation and/or overexpression is essentially associated with adipose-cell differentiation and adipogenesis. Therefore, the clinical observations that treatment with TZDs improves insulin-stimulated glucose uptake (muscle) and endogenous (essentially hepatic) glucose production while PPAR is mainly expressed in fat cells makes it difficult to link cellular and metabolic mechanisms of action. In addition, considering the well known connection between obesity and insulin resistance, it seems paradoxical that an agent that promotes adipogenesis should improve insulin sensitivity (Matthaei et al., 2000). A number of hypothetical schemas to reconcile these apparent quandaries and explain the overall mode of action of TZDs have been put forward. First, the minute quantities of PPAR expressed in muscle may be sufficient or alternatively might be induced during treatment with TZDs, leading to a direct PPAR-mediated response. Second, the effect of TZDs may also be mediated by FFA, which have been shown to interfere with muscle glucose metabolism and contribute to the impaired insulin-stimulated glucose. Since TZDs have been shown to selectively stimulate lipogenic activities in fat cells, a TZD/PPAR-mediated fatty-acid-steal phenomenon has been proposed leaving less FFAs available for muscle. Third, TZDs have been shown to reduce expression levels in fat cells of TNF and leptin, both of which have been implicated in obesity-related insulin resistance. Although the definite role of the cytokine TNF for human insulin resistance remains to be determined, TNF has been shown to interfere with proximal insulin signaling events. In addition, leptin has been shown to impair insulin signaling in isolated rat adipocytes. Since TZDs have been shown to reduce expression levels of both TNF and leptin, in fat cells, they could contribute to alleviating obesity-related insulin resistance. Which of these mechanisms plays the most important role in vivo is unclear at

22

present, but since they are not mutually exclusive all of them may be involved (Matthaei et al., 2000) It is evident that there are a number of critical gaps in our understanding of the type IIPPAR connection. For example, PPAR is required for adipogenesis and its synthetic agonists increase adipose mass in vivo. This is unexpected, since insulin resistance worsens in most patients as fat mass increases. Therefore the question raised is how an adipogenic agent can also act as an antidiabetic agent. The PPAR ligand deficiency hypothesis claims to provide the answer to this particular unresolved matter. LG754 defines a new class of nuclear receptor agonist that has minimal coactivator recruitment activity and therefore minimal inherent transcriptional activity. Instead, this compound activates transcription by allosterically enhancing the ligand binding activity of its partner receptor, PPAR. LG754 therefore represents the first example of a nuclear receptorsensitizing agent. As well as being a PPAR sensitizer, it has been found that LG754 relieves insulin resistance in vivo. These findings have important implications, since the molecular events that result in insulin resistance remain obscure. PPAR agonists have the interesting property of lowering blood glucose in diabetic animals but not in nondiabetic animals. This implies that PPAR ligands reverse or replace a deficiency that is unique to the diabetic state. It is speculated that insulin resistance arises from a relative deficiency in endogenous PPAR and that PPAR agonists are antidiabetic agents, because they correct this deficiency (Forman, 2002). Although the identity of the endogenous PPAR ligand is unknown, it is known that the transcription factor ADD1/SREBP1c is required to produce an endogenous ligand in adipocytes. It has been shown that SREBP-1c levels are lower in obesity, suggesting that the obese state may be associated with a corresponding decrease in endogenous PPAR ligands. While this response may provide short term benefits, a chronic deficiency in PPAR ligands eventually lead to the development of insulin resistance (Forman, 2002).

23

1.4.1.4) Insulin The primary goal in the treatment of diabetes mellitus is to prevent the development of the metabolic abnormalities experienced by diabetics. For many diabetics, the goal can be realized only by supplemental or replacement insulin therapy. Insulin is required by type II diabetics who cannot be maintained adequately on oral hypoglycaemic agents or dietary regulation. No single insulin preparation or combination of preparations can successfully meet the demands of such a diverse group. Consequently, a large number of insulin preparations have been developed, each of which has certain advantages and disadvantages. Although all of these preparations consist primarily of insulin and exhibit the biologic effects of insulin, they do differ in their onset and duration of action. Accordingly, they are classified on the basis of their duration of action into short (usual onset 0.5-2 hours; usual duration 3-6 hours), intermediate (usual onset 3-6 hours; usual duration 12-20 hours) and long-acting categories (usual onset 6-12 hours; usual duration 18-36 hours). The time course of action of any insulin may vary considerably in different individuals, or at different times of day in the same individual. Consequently, these time courses should be considered only as general guidelines (Foye et al., 1995).

1.5 Plants and the treatment of Diabetes Mellitus Medicinal plants have been part of the great healing traditions around the world going back thousands of years. The World Health Organisation (WHO) defines traditional medicine as health practices, approaches, knowledge and beliefs incorporating plant, animal and mineral based medicines, spiritual therapies, manual techniques applied singularly or in combination to treat, diagnose and prevent illness or maintain well-being. In 2002 WHO launched its first comprehensive traditional medicine strategy to assist efforts to promote affordable, effective and safe use of traditional medicine and complimentary alternative medicine. In Africa, traditional medicine is used by up to 80% of the population to meet primary healthcare needs and is crucial in the fight against diseases. The ratio of a conventional, or western-trained general practitioner to patients is 1:20 000, whereas the availability of traditional medicine practitioners is 1:200 to 1:400.

24

This highlights the need for reliable and affordable herbal medicines that are locally available (www.i-sis.org.uk/). More than 1123 species of plants have been used enthnopharmacologically or experimentally to treat symptoms of diabetes mellitus. These are very large and widely distributed families and the phylogenetic distance between even this select group is a good indication of the varied nature of the active constituents. It is therefore necessary to learn more about particular groups of hypoglycaemic natural products and their mechanism of action before this method of drug discovery can be successfully employed (Wagner and Farnsworth, 1994). The high percentage of active plants found probably reflects, at least in part, the great variety of possible active constituents and mechanism of action. 1.5.1 Sutherlandia frutescens (Fabaceae) There is no distinction made between the use of S. frutescens and its subspecies microphylla for medicinal purposes. These plants are among the most multi-purpose and useful of the medicinal plants in southern Africa. Conditions that have been treated with these plants include fever, poor appetite, indigestion, gastritis, oesophagitis, peptic ulcer, dysentery, cancer (prevention and treatment), diabetes, colds and flu, cough, asthma, chronic bronchitis, kidney and liver conditions, rheumatism, heart failure, urinary tract infections and stress and anxiety (Wagner and Farnsworth, 1994). Sutherlandia frutescens (subspecies microphylla) genus Fabaceae (pea and bean/leguminosae) is a perennial shrub that grows wild in the arid regions of Botswana, Namibia, Zululand, Western and Eastern Cape regions of Africa. Sutherlandia can grow up to 1.5 metres in height in optimum conditions of stony grasslands exposed to constant sunshine in daylight hours. A display of blood red flowers bloom from June to December and its seeds are carried in greenish- red papery pods, which are almost transparent, as can be seen in figure 1.9. The pinnate and compound shaped leaves have a green-grey colour giving the bush a silvery appearance (www.i-sis.org.uk/). Although this plant has been renamed to Lessertia, it is still referred by its more favoured name of Sutherlandia (Muller, 2002). In this study, Sutherlandia frutescens subspecies microphylla was referred to by its more common name, Kankerbos.

25

Figure 1.9: Sutherlandia frutescens (Seier et al., 2002)

A number of highly active compounds, including canavanine, pinitol and the amino acid GABA (gamma-aminobutyric acid), occur in high quantities in Sutherlandia species, suggesting that there is indeed a scientific basis for some of the folk uses for serious medical conditions. L-canavanine is a potent L-arginine antagonist with documented antiviral, anti-bacterial, antifungal and anticancer activities. Pinitol (2-deoxy-2-aminogalactopyranosyl) is a known anti-diabetic agent (www.sutherlandia.org). Narayanan et al, (1987) showed that pinitol isolated from Bougainvillea spectabilis had hypoglycaemic and anti-diabetic action. GABA is an inhibitory neurotransmitter that could account for the plant being used for stress and anxiety, and for the improvement in mood and wellbeing experienced by many patients. In addition a novel triterpenoid glucoside, SU 1, has been isolated and characterized, and has promising biological activities (www.sutherlandia.org). 1.5.2 Toxicity of Hypoglycaemic Plants An in-depth literature review and properly controlled experimental bioassays should be carried out in order to confirm the non-toxicity of a specific antidiabetic plant. Toxicity is influenced by the plant part, method of preparation, route of administration and test organism. Diabetes mellitus is a chronic condition with no known cure and antidiabetic drugs must be administered for a patients entire lifetime. Therefore it is important that

26

chronic toxicity studies be performed before recommending a plant-derived drug for antidiabetic therapy (Wagner and Farnsworth, 1994). An investigation into the possible toxicity of consumption of Sutherlandia leaf powder (Sutherlandia frutescens subspecies microphylla) in vervet monkeys (Chlorocebus aethiops), was carried out by determining a variety of biochemical, haematological, physiological and physical variables. These variables reflected liver, kidney, muscle, respiratory, intestinal, bone and general biological function. The conclusions referred to Sutherlandia leaf powder consumption in adult male vervet monkeys for three months. At the recommended dose, 3x the recommended dose and 9x the recommended dose, Sutherlandia leaf powder consumption was found not to be associated with toxic or other side-effects (Seier et al., 2002).

27

Chapter 2 Introduction to the present study

Insulin resistance and non-insulin-dependent diabetes mellitus (type II) have reached epidemic status in industrialized societies. Over 125 million people worldwide suffer from type II diabetes, and these individuals face a dramatically increased risk for developing atherosclerotic heart disease, stroke, renal disease, blindness and limb amputations. It is thus alarming that the number of type II cases have increased 5-fold in the past decade, a trend that is predicted to continue. Equally worrisome is that type II, initially defined as a disease of adult onset, is now appearing in adolescents (Forman, 2002). Type II diabetes is more common than type I and accounts for 90% of diabetic cases in South Africa. It has been estimated that of the four million diabetics in South Africa, half remain undiagnosed. The enormous costs of modern treatment and the evidence that current methods of treatment fail to achieve the ideals of normoglycaemia and the prevention of diabetic complications, indicate that alternative strategies for the prevention and treatment of diabetes must be developed. Due to the fact that almost 90% of people in developing countries still rely on traditional medicines for their primary health care and the fact that scientific investigations of traditional medicines have led to the discovery of drugs now in professional use worldwide, a synthesis of local traditional and modern knowledge and techniques for the management of diabetes should be feasible (Wagner and Farnsworth, 1994). Therefore indigenous, renewable, medicinal plant resources could prove to be the practical and cost-efficient alternative that is clearly and desperately needed. The present study was planned keeping the above issues in mind. The objective of this project was to optimize the methodology required to screen and determine the effectiveness of the specific plants as antidiabetic potentiates, through observing if increases in glucose utilisation and insulin secretion occurred under treatment. Once optimal methodology is achieved through this specific project, the parameters used to

28

validate the models will in the future be able to be applied to scientifically establish the antidiabetic effect(s) and mechanism(s) of the plants under investigation. The cell lines used in this study, namely 3T3-L1 preadipocytes, Chang liver cells, C2C12 muscle cells and INS-1 rat pancreatic cells, were investigated as potential in vitro models for type II diabetes. Indigenous antidiabetic plants were used as test samples in order to determine optimal conditions for screening and clearing mechanisms.

29

Chapter 3 Models for glucose utilization

The raison detre of the adipocyte is to store energy (in the form of triacylglycerol) for use during periods of caloric insufficiency. Adipocytes first appear late in fetal development preparatory to postnatal life when substantial energy reserve is needed to survive periods of fasting (Mandrup and Lane, 1997). Adipogenesis has been one of the most intensely studied models of cellular differentiation. In part this has been because of the availability of in vitro models that faithfully recapitulate most of the critical aspects of fat cell formation in vivo. This includes morphological changes, cessation of cell growth, expression of many lipogenic enzymes, extensive lipid accumulation, and the establishment of sensitivity to most or all of the key hormones that impact on this cell type, including insulin (Rosen and Spiegelman, 2000).

3T3-L1 is a continuous mouse strain of 3T3 developed through clonal isolation, which has a fibroblast-like morphology. Cells undergo a pre-adipose to adipose-like conversion as they progress from rapidly dividing to a confluent and contact inhibited state (www.biotech.ist.unige.it/cldb/c173.html). Considerable progress has been made in understanding the molecular mechanisms of adipocyte biology using the 3T3-L1 cell line as a model. A large body of evidence shows that differentiation of 3T3 preadipocytes faithfully mimics the in vivo process giving rise to cells that possess virtually all the biochemical and morphological characteristics of adipocytes. 3T3-L1 cells propagated under normal conditions have a fibroblastic phenotype. However, when treated with a combination of dexamethasone (DEX), isobutylmethylxanthine (IBMX) and insulin, 3T3-L1 cells adopt a rounded phenotype and within 5 days begin to accumulate lipids intracellularly in the form of lipid droplets. Treatment of cells with DEX activates the transcription factor CCAAT/enhancer-binding protein b (C/EBPb). IBMX inhibits soluble cyclic nucleotide phosphodiesterases and results in increased intracellular cAMP levels (Elks and Manganiello, 1985). At the nuclear level, treatment with IBMX results in activation of the related transcription factor C/EBPd. C/EBPb and d in turn induce

30

transcription of C/EBP and PPAR. Within 3 days of exposure to inducers, the cells undergo two rounds of mitosis, termed mitotic clonal expansion, which are required for differentiation. Insulin or insulin-like growth factor-1 promote adipocyte differentiation by activating PI3-kinase and Akt activity. Modulation of the activity of the forkhead transcription factor Foxo1 appears to be necessary for insulin to promote adipocyte differentiation. C/EBP and PPAR direct the final phase of adipogenesis by activating expression of adipocyte-specific genes, such as fatty acid synthetase, fatty acid binding protein, leptin and adiponectin. Endogenous negative regulators of adipocyte differentiation, such as Pref-1 and Wnt-10b, are highly expressed on undifferentiated 3T3-L1 cells, and are down-regulated upon addition of adipogenesis inducers (www.chemicon.com).

The evolution of glycerol-3-phosphate dehydrogenase (G3PDH) activity as well as of G3PDH protein and mRNA is used as an indicator of adipocyte differentiation. The enzyme accumulates to a low extent during culture in the absence of insulin. When insulin is present, the enzyme level is dramatically increased. DEX accelerates the insulin-dependent adipose conversion but alone does not ensure the complete differentiation process (Gaben-Cogneville et al.,1990). IBMX also functions in accelerating the differentiation of 3T3-L1 preadipocytes. This effect seems to be mediated by increased cyclic AMP levels (Wiederer and Loffler, 1987). Following hormonal induction, confluent preadipocytes undergo mitotic clonal expansion, become growth arrested, and then coordinately express adipocyte gene products (Mandrup and Lane, 1997).

The C2C12 muscle cell line is a subclone from a myoblast line established from normal adult C3H mouse leg muscle. The C2C12 cell line differentiates rapidly, forming contractile myotubes and producing characteristic muscle proteins. Cultures must not be allowed to become confluent, as this will deplete the myoblastic population in the culture (Tortorella and Pilch, 2002).

31

Insulin promotes the postprandial clearance of glucose from the blood primarily into skeletal muscle in both humans and rodents. Glucose transport into skeletal muscle is regulated by translocation of the glucose transporter GLUT4 from intracellular vesicles to the plasma membrane (Aslesen et al., 2001). Several lines of independent experimentation support the notion that insulin-dependent regulation of GLUT4 movement is similar or identical in skeletal muscle and adipocytes. However, GLUT4 translocates in response to exercise and hypoxia in the former and not in the latter tissue. Insulin-dependent GLUT4 translocation is PI3-kinase dependent in both tissues. Overall, it is the inability of insulin to promote glucose uptake in skeletal muscle that causes insulin resistance, and, eventually type II diabetes (Tortorella and Pilch, 2002). The development of muscle and fat cell lines has facilitated our understanding of the translocation process, specifically C2C12 and 3T3-L1 cell lines, which are unique in their expression of the GLUT4 protein. Both of these cell lines undergo differentiation in culture from myoblasts into myotubes in the case of C2C12 and from fibroblasts into adipocytes in the case of 3T3-L1. In both models, GLUT4 expression occurs on and after differentiation into myotubes or adipocytes (Ueyama et al., 1999). The liver plays a central role in glucose homeostasis. During the postabsorptive and fasting period, >90% of glucose production is derived from the liver. When glucose enters the blood following feeding, the liver switches from net glucose production to net glucose uptake to lessen the rise in blood glucose. Such switching is regulated by raised blood glucose per se and plasma insulin. One third of an oral glucose load is taken up by the liver. Glucose uptake by the liver is dependent on the amount of glucose reaching it, the insulin level within the hepatic sinusoids and a signal generated by portal glucose delivery. Impaired suppression of hepatic glucose production and a defect in hepatic glucose uptake in response to the raise in plasma glucose and insulin are major pathogenesis in fasting and excessive postprandial hyperglycemia, that are common features in obesity and diabetes (Rutter et al., 2003). Liver cell membranes have noninsulin-dependent glucose transporters and therefore do not require insulin for glucose entry. However insulin, acting through intracellular signaling systems, stimulates the

32

liver cell enzymes that promote utilization of glucose for the synthesis of glycogen, amino acids, proteins and fats, particularly fatty acids. The glucose transporter, GLUT2, is primarily expressed in hepatocytes. GLUT2 is a low-affinity receptor with a high turnover rate. These kinetic properties allow GLUT2 to function in the liver where glucose transport must not be rate limiting for influx or efflux (Medina and Owen, 2002). The Chang liver cell line is derived from non-malignant human tissue. These cells are epithelial in morphology and tend to pile up in cultures with high population density (www.biotech.ist.unige.it). Plant extracts were tested for potential antidiabetic activity by exposing these three cell lines (3T3-L1 adipocyte cells, C2C12 muscle cells and Chang liver cells) to the samples and determining the amount of glucose taken up. Each cell line represents a different in vivo organ that is known to have an involvement in glucose homeostasis, each with its own unique metabolic pathways and mechanism of glucose uptake. Therefore in order to develop a model for determining glucose utilization, all three cell lines were required as each represents a small part of the complete entity. 3.1 General methods The methods described in the following sections are the final optimised ones used in the glucose utilization models. 3.1.1 Maintenance of cell lines The cells were grown in a humidified atmosphere containing 5% CO2 at 37C (Asfari et al., 1992). The cells were routinely maintained in antibiotic free growth medium which consisted of RPMI 1640 (Sigma) supplemented with 10% heat-inactivated fetal calf serum (FCS) (Highveld Biological, SA). Proliferating preadipocytes were refed fresh growth medium every 2-3 days. For routine maintainance, trypsin in phosphate-buffered saline without Ca2+ and Mg2+ (PBSA) was used as the treatment to detach the cells from the culture plate and after 5-10 minute exposure at 37C the cells were seeded in growth

33

medium using a split ratio of 1:5. Routine cell counting was carried out using an improved Neubauer haemocytometer (Superior). 3.1.2 Cytotoxicity Assay A cell suspension was prepared in growth medium to give 4 000 cells per well using a 96-well plate (Nunc). Cells were inoculated in a volume of 200l per well and 200l aliquots of growth medium was added to cell-free wells. The cells were preincubated for 24 hours after which they were fed fresh culture medium containing test compounds and exposed for 48 hours after which cell survival percentage was determined using the sulforhodamine B (SRB) assay. 3.1.3 Sulforhodamine B Assay Determination of cell growth and cell viability was performed by in situ fixation of cells, followed by staining with a protein-binding dye, sulforhodamine B (SRB) (Sigma). The SRB binds to the basic amino acids of cellular macromolecules; the solubilized stain is measured spectrophotometrically to determine relative cell growth (Monks et al., 1991). After 48 hours incubation with plant extract/medicine, 50l cold 50% TCA (4C) was added on top of the growth medium in each well (final concentration 10%) and an incubation period of 1 hour at 4C followed. After removal of the contents of all the wells, the plate was washed five times with tap water and air-dried. 100l of 0.4% SRB (w/v in 1% acetic acid) was added to all the wells and incubated for 10 minutes at room temperature. The plate was then washed four times with 1% acetic acid to remove unbound dye and air-dried until no moisture was visible. The plates could then be stored for a period of up to 7 days or processed immediately as follows: 100l 10mM Tris base (pH 10.5) was added to solubilise the bound stain and the 96-well plate was placed on the shaker for 5 minutes. Optical densities were read using the Multiscan MS microtiter plate reader (Labsystem Multiscan MS, 665 Dosimat) at 540nm.

34

3.1.4 Glucose oxidase assay The amount of glucose taken up by the cells in each experiment was indirectly measured by assaying the amount of glucose remaining in the incubation medium using a glucose oxidase kit (Glu-cinet, Bayer), after a limited incubation time. 3.1.5 Cell viability assay Due to the variability of cell numbers in different wells, the viability of the cells in each well in each experiment should be determined so that any significant difference between the cells exposed to the treatments compared to the control wells under chronic conditions can be determined. The CellTiter-BlueTM viability assay kit (Promega, USA) was used in accordance with the instruction manual protocol. This assay uses resazurin (indicator dye) to measure the metabolic capacity of cells, thereby indicating cell viability. 3.2 3T3-L1 cell line The 3T3-L1 preadipocyte cell line, used in the present study, was obtained from the American Type Culture Collection (ATCC) through Highveld Biological (Johannesburg, South Africa). 3.2.1 Induction of 3T3-L1 preadipocyte differentiation Cells were plated at a density of 12 000 cells/well in 24-well plates, which were coated with poly-L-lysine (PLL) (procedure described in section 4.1.1.3) and incubated for 24 hours in the above growth medium (section 3.1.1). On day 1, the growth medium was replaced by supplemented medium, which consisted of RPMI 1640, supplemented with 10% FCS, 10 g insulin/ml, 10-8 M DEX, 0.1mM IBMX and the appropriate plant extract (Kankerbos or MRC2003). Cells were refed 48 hours later with the same supplemented medium and after another 24 hours (day 4), this medium was removed and replaced with growth medium including the plant extracts. After a further 48 hour incubation (day 6), the cells were assayed in their appropriate experiments. 35

3.2.2 Adipocyte differentiation assays This assay was used as the method of measuring the degree of fat cell differentiation. 3.2.2.1 Glycerol-3-phosphate dehydrogenase assay In this assay glycerol-3-phosphate dehydrogenase (G3PDH) acts by reducing dihydroxyacetone phosphate (DHAP) to glycerol-3-phosphate and at the same time there is oxidation of the cofactor NADH to NAD+, which can be followed at 340nm (Wise and Green, 1979). 3T3-L1 cells were grown and subcultured in 24-well plates, after which they were harvested at 4C by the addition of 250l Tris buffer (pH 7.5, containing 1mM EDTA) and scraping the bottom of each well. Once the cells were suspended and homogenized using a Dounce homogeniser with a Teflon pestle, the homogenate of each separate well was transferred to individual eppendorfs and microfuged at 4C for 5min. The pellet formed contained cell debris and the supernatant was collected for further assaying. The substrate was prepared by adding 0.5g wet weight of a Dowex-50 (H+) cation exchange resin to 25mg cyclohexylammonium dihydroxyacetone phosphate dimethyl ketal dissolved in 2ml water. This mixture was swirled for 30s and then the resin was allowed to settle. The supernatant was removed and incubated at 40C for 4 hours at which time the ketal was hydrolysed. The pH was then increased to 4.5 with a saturated sodium bicarbonate solution. To assay the supernatant for G3PDH activity, the following substances were sequentially introduced into a microcuvette: 325l water (37C); 50l TEA buffer (pH 7.5, containing 6.66ml triethylamine, 3.6l 2-mercaptoethanol and 420mg EDTA in 50ml water) (37C); 100l sample supernatant (4C). The reaction was initiated by addition of 20l DHAP (4C). After rapid but gentle mixing, the cuvette was immediately placed in the 37C heating mantle of a spectrophotometer and the decrease in absorption at 340nm followed against air.

36

3.2.3 Protein concentration determination Protein concentration was determined for calculation of the specific activity of the enzyme, G3PDH. The protein concentration of the samples was determined using an adapted Folin-Lowry method for microtiter plates. 3.2.3.1 Adapted Folin method for microtiter plates A 2mg/ml BSA stock solution was prepared and diluted 10x on the day of use to prepare the dilutions for the protein standard curve. The 0.2mg/ml sample was serially diluted to give 6 points on the standard curve. The procedure involved addition of 100l serially diluted standard or sample (controls contained 100l of 25mM Tris buffer, pH 7.4, containing 1mM EDTA) and 25l coppertartrate-alkaline solution. This was mixed and incubated for 10min at room temperature. 10l diluted Folin-Ciocalteau phenol reagent (1:1 with 0.1 N NaOH) was added, mixed and a further incubation period of 20min at room temperature followed. The absorbance was read at 600nm on a microplate reader. 3.2.4 Glucose Uptake 3T3-L1 cells were grown and differentiated into adipocytes in medium which consisted of RPMI 1640, supplemented with 10% FCS, 10 g insulin/ml, 10-8 M DEX, 0.1mM IBMX and the appropriate test compound or plant extract. The final method was not achieved, however the reasons and explanations are discussed in section 3.6.1.3. 3.3 C2C12 cell line 3.3.1 Glucose uptake The C2C12 skeletal muscle cell line was purchased from Highveld Biological (Johannesburg, South Africa) and passage 8 to 11 was used in this study. Prior to this, a previous study had been carried out and the growth and differentiation conditions of this cell line had been optimised (Askew, 2003). It was determined that after being cultured in 37

RPMI 1640 (including 10% FCS) for three days, the cells were differentiated and ready to be used for experimentation. After this time period, the cells could not be left indefinitely otherwise they died suddenly. For the glucose uptake experiment, cells were seeded at 4 500 cells/well into 96-well microtiter plates and cultured for 3 days in RPMI 1640 (including 10% FCS). For the glucose uptake experiment, all procedures were completed at 37C. The incubation medium (8mM glucose RPMI 1640 + 0.1% BSA + specific treatment) was added to the appropriate wells (50l/well) for an incubation time of 1.5 hours. Only 8mM glucose RPMI 1640 (0.1% BSA) was added to the control wells. Once incubation was complete, 10l of the medium in the well was removed and placed into a new 96-well microtiter plate after which 200l glucose oxidase reagent (SERAPAK Plus, Bayer) was added to the sample. This was incubated at 37C for 15 minutes and absorbance read at 492nm using a Multiscan MS microtiter plate reader (Labsystems). Treatments tested for their efficacy on glucose uptake by C2C12 muscle cells included the compounds BDM1, BDH1, pinitol (BDPl), metformin and insulin, all at the chosen concentration of 1M. Kankerbos and MRC2003 were tested using a concentration of 2.5g/well. The acute effect of these treatments was observed by exposing the cells to them for 90 minutes on the day of experimentation, whereas chronic effect was seen by pre-exposing the cells to treatment for 48 hours prior to the glucose uptake experiment and then again for 90 minutes during the experiment. 3.4 Chang cell line 3.4.1 Glucose uptake Cells of the Chang liver cell line (Passage 22 to 24) were seeded at 6 000 cells/well into 96-well microtiter plates and cultured for 5 days in growth medium. On the day of the experiment, all procedures were carried out at 37C. 25l of incubation medium (8mM glucose RPMI 1640 + 0.1% BSA + the specific treatment) was added to the appropriate wells for 3 hours. Control wells contained only 8mM glucose RPMI 1640 (0.1% BSA). After incubation, 20l was removed and placed into a new 96-well plate into which 80l double distilled water was added to give a 5x dilution. 50l of this dilution was removed 38

and 200l glucose oxidase reagent (Glu-cinet, Bayer) was added to the 50l medium left in the well. After an incubation of 15 minutes at 37C, the absorbance was measured at 492nm using a Multiscan MS microtiter plate reader (Labsystems). The treatments tested for their effect on glucose uptake in Chang liver cells included the following: BDM1, BDH1, pinitol (BDP1), metformin and insulin, all used at 1M concentration. The plant extracts, Kankerbos and MRC2003, were used at 12.5g/well. Acute and chronic effects were both investigated. Acute effect was observed by exposing the cells to the treatments only on the day of the experiment for the 3 hour incubation time. However, chronic effect included adding the treatments to the cells for 48 hours prior to the day of experiment and then again for the 3 hours. 3.5 Statistical analysis All results were analyzed using the data analysis program in Microsoft Excel. The statistical tests were carried out on original data obtained from the experiments performed. The relationship between two variables was determined by the use of the t-test, two-sample assuming equal variances. 3.6 Results and Discussion 3.6.1 3T3-L1 cell line 3.6.1.1 Cytotoxicity In order to determine the effect of treatments on the viability of 3T3-L1 cells over time, cytotoxicity tests were carried out whereby cells where exposed to treatments for 48 hours. These tests allow one to determine the concentrations of each treatment that are not cytotoxic to the specific cell line and thereby use one that is non-cytotoxic in further experimentation. Basically, any concentration that shows less than 80% of control is too cytotoxic for the specific cell line it is being tested on and is not eligible for further use. Ideally the concentration of the treatment chosen should not be cytotoxic to the cell line however it should elicit an effect. Concentrations tested for the compounds (BDM1,

39

BDH1, pinitol (BDP1) and metformin) included 0.1, 1, 10, 50 and 250M. The concentrations tested for the plant extracts, namely Kankerbos and MRC(2003), ranged from (0.0025 25g/well) (Figure 3.1).

120 100 80 60 40 20 0 -2 -1 0 1 2 3

120 100

% of control

% of control

80 60 40 20 0 -2 -1 0 1 2 3

log [BDM1]M

log [BDH1]M

100

100

% of control

60 40 20 0 -2 -1 0 1 2 3

% of control

80

80 60 40 20 0 -2 -1 0 1 2 3

log[BDP1]M

log[Metformin]M

120

120

% of control

100 80 60 40 20 0 0 1 2 3 4 5

% of control

100 80 60 40 20 0 0 1 2 3 4 5

log[Kankerbos]ng/well

log[MRC2003]ng/well

Figure 3.1: Cytotoxicity results of compounds (BDM1, BDH1 BDP1 and metformin) and plant extracts (Kankerbos and MRC2003) on 3T3-L1 adipocyte cells after a 48 hour exposure period. Data points represent the mean SD (n=6 wells) from a single experiment.

40

In figure 3.1, the compounds (BDM1 and BDH1) showed no cytotoxic effect on the 3T3L1 cell line. The concentrations of BDP1 showed varied results for percentage control and metformin was only cytotoxic to the cells at 250M. For these compounds, no significant cytotoxicity was seen and it was decided that all further experimentation carried out with 3T3-L1 cells would be done so at a concentration of 1M. For the plant extracts (Kankerbos and MRC2003) only the cells exposed to the highest concentration of 25g/well showed signs of cytotoxicity. All results shown for the plant extracts are g/well, however a conversion table showing the g/ml values is shown in Appendix 1 3.6.1.2 Glycerol-3-phosphate dehydrogenase activity As the 3T3-L1 cells differentiated, observations were an alteration in cell shape and the accumulation of lipid droplets in the cytoplasm. Samples tested included Kankerbos (12.5g/well/1ml) and MRC (12.5g/well/1ml) plant extracts and the control included cells not exposed to any form of treatment other than inducers where applicable. Once the cells had reached confluency, the inducers of differentiation (insulin, DEX and IBMX) were added for 3 days to the appropriate cells. After this they were removed and the cells fed without inducers for another 7 days. The glycerol-3-phosphate dehydrogenase assay functions in measuring fat cell differentiation. The reason for using this enzymatic assay is because G3PDH is an intermediatory enzyme in the pathway by which triacylglycerols are formed, therefore before fat droplets form the presence of G3PDH is detectable. G3PDH reduces DHAP to G-3-P and corresponding to this reduction, NADH is oxidized to NAD+. With the oxidation is a decrease in absorbance at 340nm and by using the slope of this decrease, one can calculate the enzyme activity. Determination of protein concentration then enables one to finally calculate the specific activity of G3PDH.

41

0.35 0.30

G3PDH specific activity (mU/mg protein)

0.25 0.20 0.15 0.10 0.05 0.00 Control Kankerbos MRC

Without Inducers With Inducers

Figure 3.2: The effect of plant extracts (Kankerbos and MRC; 12.5g/well) (with and without inducers) on glycerol-3-phosphate dehydrogenase (G3PDH) specific activity. Values shown are the mean SEM (n=4 wells) from an individual experiment. *P < 0.001 when compared to the control.

Figure 3.2 shows the G3PDH specific activity in the 3T3-L1 cells after exposure to the plant extracts with and without induction of differentiation by inducers. The control sample gave results that were expected, that is an increase in G3PDH specific activity when cells were exposed to the inducers of differentiation compared to without inducers added. The plant extracts showed increased G3PDH specific activity with and without inducers compared to the control, although not all samples were significantly different to the control. This indicates that both Kankerbos and MRC extracts stimulate fat cell differentiation. These findings would lead us to believe that the plant extracts act on PPAR, which is a transcription factor that directly influences fat cell development (Rosen et al, 2000). The reason why the plant extract samples with inducers gave lower values compared to without, was due to a large variation in sample results and therefore greater error bars. This experiment was planned to be repeated; however problems with the 3T3-L1 cell line pursued, this will be discussed in the following section. However it was felt based on these results, that a positive control for PPAR involvement should be included in further investigations. An example of a positive control would have been the

42

use of a thiazolidinedione, such as pioglitazone, which is a PPAR ligand/activator, which has been found to increase insulin sensitivity and decrease blood glucose. 3.6.1.3 Glucose uptake In the introduction to this project, it was mentioned that one of the objectives would be to optimize the methods required to screen and determine the effectiveness of specific plant extracts as antidiabetic potentiates. One of the ways to do this was through creating a model for observing glucose uptake in cells. The idea of stream-lining a method entails increasing the amount of samples to be tested but at the same time using materials sparingly. Based on this it was initially decided to seed the 3T3-L1 cells (4 000 cells/well) into 96-well plates and test the amount of glucose taken up by undifferentiated and differentiated cells. However this method was found not to be sensitive enough due to an inadequate amount of cells and it was decided following experiments would be carried out in 24-well plates (24 000 cells/well). The positive controls chosen for glucose uptake in 3T3-L1 cells due to their antidiabetic activity were metformin and insulin. Insulin specifically for its positive effect on the translocation of GLUT4 to the cell surface thereby promoting glucose uptake. In addition to these, BDP1 was included as a test substance as it is suspected to have antidiabetic effects (Narayanan et al., 1987). The first experiment making use of the 24-well plates involved testing the positive controls (1M), as well as BDP1 (1M) and the plant extract, Kankerbos (12.5g/well) on differentiated 3T3-L1 cells. The effects of the treatments were tested by exposing the cells to the specific treatment only during the glucose uptake experiment (acute conditions). Glucose transport in 3T3-L1 cells is facilitated due to the presence of the glucose transporters. The intracellular GLUTs are both GLUT1 and GLUT4, and both are recruited to the cell membrane (Zierler, 1999). Under basal conditions, GLUT4 is efficiently sequestered in intracellular membrane compartments. When cells are treated

43

with insulin, the sequestered pool of GLUT4 is mobilized and translocates to the plasma membrane (Perrini et al., 2004). Insulin increases the rate of GLUT4 exocytosis, with little or no decrease in its rate of endocytosis, so that in adipocytes the proportion of GLUT4 at the cell surface increases from less than 10% in the absence of insulin to 35 to 50% in its presence (Bogan et al., 2001). Therefore one expects to see a marked increase in glucose uptake in these cells in response to insulin. Although GLUT4 expression only occurs on and after differentiation into adipocytes (Ueyama et al., 1999), it was decided that a comparison of glucose uptake in undifferentiated and differentiated 3T3-L1 cells would be beneficial in the first phases of method optimization. In figure 3.3, the results of the comparison of glucose uptake in undifferentiated (a) and differentiated (b) 3T3-L1 cells are shown. It can be seen that over time there is an increase in the amount of glucose taken up in the basal samples as well as the treated samples in both scenarios. However in the differentiated 3T3-L1 cells there was greater amount taken up overall, thus confirming the presence of GLUT4 transporters due to their expression in these cells. Due to this confirmation, it was obvious for streamlining purposes, to only use differentiated 3T3-L1 cells for further glucose uptake experiments. For reproducibility purposes, the experiments were repeated and the following graphs (10 minute and 120 minute incubation times) show the combined results of two individual glucose uptake experiments under acute conditions. Figure 3.4 shows the results of glucose uptake as a percent of the control cells (basal-state) after 10 minute incubation time. The positive controls, metformin and insulin both at a 1M concentration, caused a significant increase in glucose utilization compared to the control within 10 minutes. This immediate effect by insulin was expected as it is known to initiate the translocation of GLUT4 to the cell surface plasma membrane. Also seen within 10 minutes of treatment exposure was that the plant extracts, Kankerbos (12.5g/well) and MRC2003 (2.5g/well), significantly enhanced glucose uptake in 3T3-L1 adipocytes.

44

a)
% Glucose uptake

45 40 35 30 25 20 15 10 5 0 Control Metformin Insulin

10 minutes 120 minutes

b)
% Glucose uptake

45 40 35 30 25 20 15 10 5 0 Control BDP1 Kankerbos Metformin Insulin 10 minutes 120 minutes

Figure 3.3: The percentage glucose taken up by undifferentiated (a) and differentiated (b) 3T3-L1 cells when exposed to treatments under acute conditions. Positive controls: 1M metformin and 1M insulin. Treatments tested included: 1M BDP1 and 12.5g/well Kankerbos. Values shown are the averages from two experiments tested in quadruplicate.

400 350 300

**

***

**

% of control

250 200 150 100 50 0 Control Kankerbos MRC2003 Metformin Insulin

Figure 3.4: Acute effect of treatments on glucose uptake in 3T3-L1 cells after 10 minute incubation time. Values shown are the mean SEM (n=6 or 8) from two experiments combined. *P < 0.05; **P < 0.01; ***P<0.001 when compared to the control.

45

200 150 100 50 0 Control Kankerbos MRC2003

***

% of control

Metformin

Insulin

Figure 3.5: Acute effect of treatments on glucose uptake in 3T3-L1 cells after 120 minute incubation time. Values shown are the mean SEM (n=6 or 8) from two experiments combined. ***P<0.001 when compared to the control.

The results of glucose utilization as a percent of control after 120 minutes (Figure 3.5), showed once again an increase in glucose uptake in the cells exposed to the treatments. In other words, the treatments stimulated glucose utilization compared to basal glucose uptake. Although there was an increase, it can be seen that the difference between the control and treatments is not as large at 120 minutes compared to 10 minute incubation. This is expected for a number of reasons, including saturation of the cells with glucose after 120 minutes of uptake, a decrease in the amount of glucose available in the media as well as possible decrease in the activity of the treatments over such a long exposure time. Due to these results as well as the need to streamline the method, it was decided to choose an incubation time between 10 and 120 minutes, where a significant difference between control and treated cells would still be seen. For all further glucose uptake experiments with 3T3-L1 cells it was decided to sample at 60 minutes. During further experimentation, it was observed that the viability of the cells during the glucose uptake experiments were questionable, so it was thought that the addition of 10% serum might be beneficial and glucose standard curves with and without serum were carried out.

46

Glucose standard curve without 10% serum

2.0 1.5 1.0 0.5 0.0 0.00

A492nm

0.02

0.04

0.06

0.08

0.10

0.12 y = 15.866x R2 = 0.9957

[Glucose] g/well

Glucose standard curve with 10% serum

2.0

A492nm

1.5 1.0 0.5 0.0 0 0.02 0.04 0.06 0.08 0.1 0.12 y = 15.802x R2 = 0.9816

[Glucose] g/well

Figure 3.6: Comparison of glucose standard curve with and without 10% serum in 8mM RPMI 1640. Data points are the mean SD (n=4) from an individual experiment.

Figure 3.6 shows that adding 10% serum to the incubation media (8mM glucose RPMI 1640) during experimentation would not affect the amount of glucose taken up by the cells as the glucose standard curves (with or without 10% serum) gave similar results. Based on these findings, it was decided to include 10% serum in the media in future glucose uptake experiments involving 3T3-L1. This would hopefully increase their viability during the experiment. The results obtained up until this stage in the project allowed some decisions to be made regarding the optimization of the experiment. As mentioned 10% serum would be included in the incubation media during the glucose uptake experiment. Also, as mentioned, it was decided to take samples at 60 minutes only in order to streamline the

47

procedure. However due to the cells beginning to peel during the experiment, it was thought that lining the wells with PLL when seeding, would possibly be beneficial and aid in cell attachment. Originally during the G3PDH experiments, the wells used were treated with PLL and no problems were seen. Following these changes in the methodology, seeding and differentiation induction of 3T3-L1 cells was continued on numerous occasions with glucose uptake experiments being the final step. However despite the above attempts to overcome the problems, the peeling of the cells from the edges into the centers of the wells persisted. This hampered obtaining accurate results, as it was not known what the effect of the peeling was on the ability of the cells to take up glucose. Although the wells were being plated with PLL before seeding this did not seem to be improving the situation, so another possibility was tested, that of plating the wells with FCS (Figure 3.7).

a)

450 400 350

% of control

300 250 200 150 100 50 0 Control BDP1 MRC(methanol) MRC(2002) MRC(2003)

b)
450 400 350 300 250 200 150 100 50 0 Control BDP1 Kankerbos MRC(2003)

Figure 3.7: Glucose uptake as a percent of control in 3T3-L1 cells with wells lined with PLL (a) and FCS (b). Values shown are the averages from a single experiment tested in quadruplicate.

% of control

48

Changing certain criteria continued, with the aim of finding the right method that allowed glucose uptake experiments to proceed without the cells lifting and peeling. Pretreating the wells with PLL and FCS was one change that was abandoned as they were not improving the situation. Another change was that of cell number. Up until this stage, 24 000 cells/well were used, however it was thought that by decreasing this, peeling might be prevented. Seeding at 12 000, 8 000 and 6 000 cells/well was tried, however no improvement was observed and with low seeding numbers it prolonged induction of differentiation due to waiting for the cells to get to confluency. The samples, BDM1 and BDH1, were obtained from the Medical Research Council (MRC). These compounds were expected to be within the plant extracts being tested on 3T3-L1 cells, therefore they were included in the following experiment.

2000 1800 1600 1400 1200 1000 800 600 400 200 0 Control BDM1 BDH1 BDP1 Metformin Insulin

The graph in figure 3.8 indicates glucose uptake in 3T3-L1 cells, however the results confirmed the problems with the cells. When the incubation time for a glucose uptake experiment is completed, the amount of glucose left in each well is determined with a glucose oxidase assay and absorption values are determined (492nm). In this specific

% of control

Figure 3.8: Acute effects of treatments (1M BDM1, BDH1 and BDP1) on glucose uptake in 3T3-L1 cells after 60 minute incubation. Positive controls: 1M metformin and insulin. Values represent the averages from a single experiment tested in quadruplicate.

49

experiment, the differences in absorption between the 0 time and 60 minute samples for the control cells were minimal. Due to these nominal values, the subsequent increases in glucose uptake calculated for treated cells compared to control cells were invalid. Contamination also plagued the 3T3-L1 cell line and this was also thought to be adding to their inability to grow and behave normally. Micoplasmas were suspected and so on two separate occasions the cells were run through treatments with ciprobay. However this did not seem to affect their well-being. Different types of preincubation media were also tested, including Krebs Ringer bicarbonate buffer and RPMI 1640 instead of DMEM, however neither of these made any significant difference in preventing the cells from peeling into the center of the wells and forming clumps. Also on numerous occasions, the number of days the cells were being cultured was decreased, however once again this made no positive impact on the cells. Over time, the 3T3-L1 behaviour had changed and their morphology had changed to more spindly cells, so overall they were not behaving as expected. Finally it was confirmed by the European Collection of Cell Culture (ECACC) that the 3T3-L1 cell line had transformed and was not exhibiting contact inhibition, so this aided in explaining all the problems discussed.

50

3.6.2 C2C12 skeletal muscle cell line 3.6.2.1 Cytotoxicity Cytotoxicity studies on C2C12 skeletal muscle cells were conducted to ascertain which concentrations of the compounds and plant extracts would be noncytotoxic to these cells and thereby be used for further glucose uptake experiments (Figure 3.9).

120

120

% of control

% of control

100 80 60 40 20 0 -1.5 -1 -0.5 0 0.5 1 1.5

100 80 60 40 20 0 -1.5 -1 -0.5 0 0.5 1 1.5

log[BDM1]M

log[BDH1]M

% of control

60 40 20 0 -1.5 -1 -0.5 0 0.5 1 1.5

% of control

120 100 80

120 100 80 60 40 20 0 -1.5 -1 -0.5 0 0.5 1 1.5

log[BDP1]M
120

log[Metformin]M
100

% of control

80 60 40 20 0 0 1 2 3 4 5

% of control

100

80 60 40 20 0 0 1 2 3 4 5

log[Kankerbos]ng/well

log[MRC2003]ng/well

Figure 3.9: Cytotoxicity results of compounds (BDM1, BDH1, BDP1 and metformin) and plant extracts (Kankerbos and MRC2003) on C2C12 muscle cells after a 48 hour exposure period. Data points represent the mean SD (n=4 wells) from a single experiment.

51

The results of the range of concentrations (0.1 to 10M) tested for the compounds, BDM1, BDH1, BDP1 and metformin on C2C12 cells can be seen in figure 3.9. No cytotoxic effect was observed at either of these concentrations of the compounds on this cell line. For streamlining purposes as well as the benefit of being able to compare results of these compounds between cell lines, it was decided to continue using 1M as the concentration for further use. Figure 3.9 reveals that Kankerbos was not cytotoxic at any of the concentrations tested (0.25, 2.5 and 25g/well). The MRC2003 sample however gave less than 80% of control for the highest concentration of 25g/well. In order to maintain comparability between the plant extracts within the C2C12 cell line, a concentration of 2.5g/well for Kankerbos and MRC2003 was chosen to be used for further glucose uptake experiments. All results shown for the plant extracts are g/well, however a conversion table showing the g/ml values is shown in Appendix 1. 3.6.2.2 Glucose uptake Muscle is the major site of glucose disposal, in which insulin increases glucose uptake by promoting the translocation of GLUT4 from an intracellular compartment to the plasma membrane via a PI 3-kinase pathway. Skeletal muscle accounts for more than 80% of the total insulin mediated glucose uptake (Kumar and Dey, 2003). Although the C2C12 muscle cell line expresses far less GLUT4 than muscle tissue and its insulin responsiveness is minimal, it has been used extensively for in vitro cell culture studies of glucose transport and signaling mechanisms. Therefore it would seem as if insulin stimulation of glucose transport is not solely dependent on the presence of the insulin receptor and the GLUT4 protein (Purintrapiban and Ratanachaiyavang, 2003; Tortorella and Pilch, 2002). As we were trying to create a model for glucose uptake using C2C12 muscle cells as one of the cell lines, it had to be deciphered which positive control would best suit this specific cell line.

52

180 160 140 120 100 80 60 40 20 0 Control Metformin Insulin

Figure 3.10: The acute effect of positive controls, metformin and insulin (1M) on glucose uptake in C2C12 cells after 90 minute exposure. Values shown are the mean SEM (n=8 wells) of two experiments.

Figure 3.10 reveals the results of the two positive controls tested for their effect on the amount of glucose taken up by C2C12 skeletal muscle cells under acute conditions. One can see that the obvious choice of positive control for further experiments of this nature is insulin (1M). Metformin is known to increase glucose uptake in skeletal muscle, however its effect is somewhat more chronic due to its slow permeation across the inner membrane (Kirpichnikov et al., 2002). Therefore with regards to the C2C12 cell line, it would seem not to have an effect within acute conditions. The acute effects of treatments on glucose uptake by C2C12 cells can be seen in figure 3.11. The C2C12 cells were exposed to the specific treatments for 90 minutes and this graph shows the typical glucose uptake experiment results found in this cell line. Insulin, our positive control, significantly increased the amount of glucose taken up by these cells compared to the control cells (*P<0.05). Under acute conditions, BDM1 had no effect on glucose uptake by C2C12 cells, however cells exposed to BDH1 took up glucose significantly more than the control cells (*P<0.05). BDP1 is a compound that has been reported to possess insulin-like properties (Davis et al., 2000). It has also been said to have a direct effect on insulin sensitization, which thereby facilitates glucose uptake. Greenwood et al. (2001) reported BDP1 to stimulate glucose uptake into L6 skeletal muscle cells and as can be seen in figure 3.11 it would seem to stimulate glucose uptake by C2C12 skeletal muscle cells, as there was an increase in glucose uptake compared to

% of control

53

the control (**P<0.01). The two main plant extracts under investigation, namely Kankerbos and MRC2003, gave differing results in their effect on glucose uptake into C2C12 cells. Kankerbos significantly increased the glucose entering the cells (**P<0.01), whereas the cells exposed to the MRC2003 extract showed no significant glucose uptake.

300 250

** **

% of control

200 150 100 50 0 Control BDM1 BDH1 BDP1

Kankerbos

MRC2003

Insulin

Figure 3.11: Acute effects of treatments on glucose uptake by C2C12 cells after 90 minutes. Treatments included: 1M BDM1, BDH1 and BDP1 as well as plant extracts, Kankerbos and MRC2003 (2.5g/well). Positive control: 1M insulin. Values shown are the mean SEM (n=8 wells) from two experiments. *P < 0.05; **P < 0.01 when compared to the control.

C2C12 cells were exposed to treatments over a longer time period, by adding them 48 hours prior to experimentation, to investigate the chronic effect of the treatments. Glucose uptake experiments were carried out under chronic conditions a number of times, however the results were unexplainably inconsistent every time. It was therefore decided that for the C2C12 cell line model, acute exposure experiments would only be carried out in future investigations. The one finding that did remain consistent time after time, was for the cells exposed to insulin (1M) under chronic conditions (figure 3.12).

54

160 140 120

% of control

100 80 60 40 20 0 Control Insulin

Figure 3.12: Effect of insulin (1M) on glucose uptake in C2C12 muscle cells compared to control after chronic exposure. Data points are the mean SEM (n=6 wells) from two experiments.

Figure 3.12 revealed that the chronic presence of insulin inhibited the uptake of glucose by C2C12 cells. It would seem that resistance to insulin was induced in the C2C12 cells under these conditions. Insulin resistance in skeletal muscle is characteristic in type II diabetic patients (Dela et al., 1994). The results obtained in this experiment are not unknown, because insulin resistance has been developed in skeletal muscle cell lines such as C2C12 muscle cells (Kumar and Dey, 2003). This part of the glucose utilisation model could possibly play an important role, as it could be used to screen the potential antidiabetic compounds and plant extracts for their effects on insulin resistance in skeletal muscle.

55

3.6.3 Chang liver cell line 3.6.3.1 Cytotoxicity Assays to determine the effect of different concentrations of specific treatments on Chang liver cells were carried out. The compounds tested included BDM1, BDH1 and BDP1, where the cells were exposed to these using a range of concentrations (0.1 to 50M). The effect of different concentrations (0.025, 0.25, 2.5 and 25g/well) of the plant extract, MRC2003, on these cells was also determined.

160 140 120 100 80 60 40 20 0 -2 -1 0 1 2

160 140 120 100 80 60 40 20 0 -2 -1 0 1 2

% of control

% of control

log[BDM1]M

log[BDH1]M

160 140 120 100 80 60 40 20 0 -2 -1 0 1 2

250

% of control

% of control

200 150 100 50 0 0 1 2 3 4 5

log[BDP1]M

log[MRC2003]ng/well

Figure 3.13: Cytotoxicity results of compounds (BDM1, BDH1 and BDP1) and plant extract (MRC2003) on Chang liver cells after a 48 hour exposure period. Data points represent the mean SD (n=8 wells) from a single experiment.

Based on the theory of cytotoxicity testing previously discussed, figure 3.13 indicates that none of the compounds tested (BDM1, BDH1 and BDP1) showed any cytotoxicity toward the Chang liver cell line they were tested on. Therefore to maintain some consistency and

56

thereby be able to compare the effect of these compounds on different cell lines, it was decided to continue with the 1M concentration for BDM1, BDH1 and BDP1. The plant extract sample, MRC2003, appeared to have no cytotoxic effect in the Chang liver cells. It was decided to use the MRC2003 sample at a concentration of 12.5g/well (acute) and 10g/well (chronic) for further experimentation with the Chang liver cell line. Although the plant extract, Kankerbos, was not tested specifically for cytotoxicity in this cell line, it was decided, based on previous findings and on the results obtained for MRC2003, that its concentration for glucose uptake experiments in Chang liver cells under acute conditions would also be 12.5g/well and for chronic conditions, 10g/well. All results shown for the plant extracts are g/well, however a conversion table showing the g/ml values is shown in Appendix 1. 3.6.3.2 Glucose uptake For any experiment carried out one needs a positive control in order to substantiate the results obtained. The two possible candidates for their positive effect on glucose uptake in Chang liver cells were metformin and insulin (Figure 3.14).

160 140 120

% of control

100 80 60 40 20 0 Control Metformin Insulin

Figure 3.14: Acute effect of two positive controls (1M metformin and insulin) on glucose uptake in Chang liver cells. Values shown are the mean SEM (n=8 or 15 wells) from an individual experiment. *P < 0.01 when compared to the control.

57

Shown in figure 3.14 is the glucose uptake results found when Chang liver cells were exposed to metformin and insulin under acute conditions. It has been found for metformin that there is an effect if the exposure to the drug is over a longer time period. This is probably due to the relatively slow permeation of the drug across cell membranes and its subsequent accumulation within the mitochondrial matrix, where it exerts a dosedependent inhibitory effect on complex I of the respiratory chain (Rutter et al., 2003). Therefore these results correlate to this as it is evident that under acute conditions, metformin has no effect and there is no significant uptake of glucose by the cells. Insulin, with regards to exposure time, appears to work oppositely as can be seen that it has acute effect allowing a significant increase in glucose uptake. Therefore based on these findings, it was decided that for the glucose uptake experiment under acute conditions that insulin would act as a positive control and seeing that metformin is known to have a chronic effect on glucose uptake that it would be used as the positive control for chronic conditions. The graph in figure 3.15 indicates the results obtained in a typical glucose uptake experiment in Chang liver cells, where cells were exposed to treatments for the acute exposure time of three hours.

160 140 120

**

% of control

100 80 60 40 20 0 Control BDM1 BDH1 BDP1 Kankerbos MRC2003 Insulin

Figure 3.15: Acute effect of treatments on glucose uptake in Chang liver cells. Treatments included: 1M BDM1, BDH1 and BDP1 as well as plant extracts, Kankerbos and MRC2003 (12.5g/well). Positive control: 1M insulin. Values shown are the mean SEM (n=8 or 15 wells) from an individual experiment. *P < 0.05; **P < 0.01 when compared to the control.

58

Our positive control, namely insulin (1M), showed a significant increase in glucose uptake in Chang liver cells compared to the control (**P<0.01). The two samples supplied by the MRC, namely BDM1 and BDH1, showed no effect on glucose uptake in Chang cells under acute conditions. BDP1, a compound found in Kankerbos, is alleged to have anti-diabetic properties (Davis et al., 2000). BDP1 supposedly enhances the action of insulin and therefore the uptake of glucose. However it was found that under acute conditions, BDP1 (1M) had no significant impact on glucose uptake in Chang liver cells. The plant extracts had differing effects. The MRC2003 sample (12.5g) did not show any effect on glucose uptake however Kankerbos (12.5g) had an effect, as it increased the uptake of glucose by the Chang cells significantly compared to the control (*P<0.05).

160 140 120

% of control

100 80 60 40 20 0 Control BDM1 BDH1 BDP1 Kankerbos MRC2003 Insulin

Figure 3.16: Acute effect of treatments on glucose uptake in Chang liver cells. Treatments included: 1M BDM1, BDH1 and BDP1 as well as plant extracts, Kankerbos and MRC2003 (12.5g/well). Positive control: 1M insulin. Data points represent the mean SEM (n=8 or 15) of three experiments.

Figure 3.16 shows the combined results of three individual glucose uptake experiments under acute conditions, to give an indication of reproducibility. As one can see, all the effects of the different treatments are maintained and are comparable to those shown in figure 3.15.

59

The chronic effect of the treatments on glucose uptake in Chang liver cells are shown in the figure 3.17.

140 120 100 80 60 40 20 0 Control BDM1 BDH1 BDP1 Kankerbos MRC2003

**

**

% of control

Metformin

Figure 3.17: Chronic effects of treatments exposed to Chang liver cells for 48 hours prior to glucose uptake experimentation. Treatments included: 1M BDM1, BDH1 and BDP1 as well as plant extracts, Kankerbos and MRC2003 (10g/well). Positive control: 1M metformin. Data points represent the mean SEM (n=15 wells) from a single experiment. *P < 0.05; **P < 0.001 when compared to the control.

Metformin decreases blood glucose levels by reducing hepatic glucose output (gluconeogenesis) (Moriaka et al., 2005). Therefore a bigger difference between glucose levels before and after incubation was expected. Also prolonged exposure to metformin is more effective, so the effect of metformin seen in figure 3.17 was to be expected. Metformin as our positive control for glucose uptake under chronic conditions gave a significant increase in the amount of glucose taken up by the cells compared to the control (**P<0.001). BDH1 (1M) showed an effect under chronic conditions, whereas it had no effect when it was exposed to the cells for only three hours. A significant increase in glucose uptake was observed compared to the control (*P<0.05), as seen in figure 3.17. However, BDM1 (1M) had no effect again, therefore this sample seems to have no effect on glucose uptake in Chang liver cells under acute or chronic conditions. Once again BDP1 showed no significant effect on glucose uptake and so under chronic conditions no effect was observed. An effect was observed when Kankerbos (10g/well) was exposed to the cells for a long time, a significantly increased amount of glucose was

60

taken up by the Chang cells compared to the control (**P<0.001). Therefore it would seem that the plant extract, Kankerbos, has an impact on glucose uptake under acute and chronic conditions. The other plant extract tested, namely MRC2003 (10g/well), showed no effect on glucose uptake under chronic conditions, as under acute conditions. Therefore the MRC2003 sample seems to have no effect on this specific cell line when it comes to glucose uptake experiments. Through combining three different glucose uptake experiments, one can see the reproducibility of the results in figure 3.18.

160 140 120 % of control 100 80 60 40 20 0 Control BDM1 BDH1 BDP1 Kankerbos MRC2003 Metformin

Figure 3.18: Effect of different treatments (BDM1, BDH1, BDP1; 1M and Kankerbos and MRC2003; 10g/well) on glucose uptake in Chang liver cells under chronic conditions. Positive control: 1M metformin. Data points represent the mean SEM (n=8 or 15 wells) of three experiments

After exposing the Chang liver cells to the different treatments for 48 hours prior to the glucose uptake experiment, it was very important to determine the viability of the cells. This was in order to determine if any specific treatment had a positive or negative impact on cell growth and viability and to make sure that at the time of experimentation the cells in each well were all at a similar density. By accomplishing this, one could cancel out any doubt as to the difference cell numbers play in the role of differing results.

61

90 80 70

Alamar blue

60 50 40 30 20 10 0 BDM1 BDH1 BDP1 Kankerbos MRC2003 Metformin

Figure 3.19: CellTiter-BlueTM results of Chang liver cells exposed to different treatments for chronic experimentation. Treatments included: 1M BDM1, BDH1 and BDP1 as well as plant extracts, Kankerbos and MRC2003 (10g/well). Positive control: 1M metformin. Values represent the mean SEM (n=8 or 9 wells) from a single experiment.

Figure 3.19 shows the CellTiter-BlueTM results and it was found that there was no significant difference between the cells exposed to the treatments compared to the control well under chronic conditions. These results were not used in the calculations of the previous figures (figures 3.17 and 3.18) because there was no significant difference found, however it is advisable that CellTiter-BlueTM treatment be carried out for all experiments to confirm cell viability. In conclusion, the two successful cell lines used for the glucose utilisation model in this study were the C2C12 skeletal muscle and Chang liver cell lines. Unfortunately the problems experienced with the 3T3-L1 cell line prevented any conclusions being made with regards to the best methods for glucose uptake in these cells. The C2C12 skeletal muscle cell line was an important cell line to include in this model, as it is known that skeletal muscle accounts for a large amount of the total insulin mediated glucose uptake and is therefore one of the major insulin target tissues. Although the chronic exposure experiments with C2C12 cells were not achieved, the acute

62

condition (1.5 hours) experiments were successful using insulin as the positive control for glucose uptake. As a result of these findings, it was decided that in the future, all glucose uptake experiments with C2C12 muscle cells would only include acute exposure conditions. For the Chang liver cell line, both the acute (3 hours) and chronic (48 hour pretreatment + 3 hours) condition experiments were successful. When observing the acute effects of the treatments on glucose uptake in these cells, it was found that insulin was the best positive control. However due to metformins mechanisms of action, it showed the required effect when exposed to the Chang liver cells chronically and was therefore chosen as the positive control under these conditions. All together, these two cell lines provided a good basis for the creation of a model whereby the effects on glucose utilisation could be investigated for treatments that were possible antidiabetic potentiates.

63

Chapter 4 Models for insulin secretion

The insulin-secreting cell lines (INS-1 and INS-2) were established from the cells isolated from an x-ray-induced rat transplantable insulinoma. INS-1 cell morphology, insulin biosynthesis and secretion are remarkably similar to those of the parent tumour propagated in vivo (Asfari et al., 1992). The continuous growth of INS cells depends on the presence of 2-mercaptoethanol (2ME) in the culture medium. It was found that the presence of 2-ME (optimal concentration 50M) promoted cell proliferation. In the absence of 2-ME, these cells do not only stop proliferating, but actually do not survive after trypsinization for culture transfer. The high and stable insulin content of INS-1 cells (averaging 8g/106 cells) as compared to normal islets (40g/106 cells), is probably due to the continuous presence of 2-ME. Several mechanisms have been proposed to explain the effect of 2-ME, however whether 2-ME participates in sustaining the differentiated state of INS-1 cells remains hypothetical. 2-ME may exert its effects through participation in cellular metabolism, which leads to the enhancement of growth factor receptor activation, or on the expression of oncogenes (Asfari et al., 1992). Therefore, compared to normal -cells, INS-1 cells retain a differentiated -cell phenotype with respect to glucose transport and metabolism, which are important functions for normal metabolism-secretion coupling (Janjic et al., 1999). Unwanted weight gain has been associated with the use of tricyclic antidepressants and it has been postulated that the mechanism is due to antidepressants increasing blood insulin levels (Garland et al., 1988 ; Hurr, 1996). Obesity, which is a condition caused by an excessive amount of adipose tissue, is strongly associated with the development of type II diabetes (Sheard and Clark, 2000). Due to type II diabetes complications and chronicity, reducing risk factors such as obesity through lifestyle modification is crucial to the longterm health of patients. Therefore weight gain, which is an undesired effect of some 64

antidepressants, is a major problem, which needs to be solved and further investigation is required. One of the main aims of the project was to create an in vitro model using an insulinsecreting cell line in order to determine the effect of different drugs (antidepressants and traditional medicine) on insulin secretion. Through setting up a model, one would be able to determine the positive or negative effects of the drugs on insulin secretion. 4.1 Materials and methods Methods described in the following sections are the final optimised methods used in the model, decided upon when the initial set of experiments were completed. 4.1.1 INS cells maintenance INS1 cells were kindly donated by Professor Guy Rutter from the University of Bristol, England. 4.1.1.1 Mercaptoethanol Stock mercaptoethanol (3.48l) (0.5mM) was made up to 10ml with RPMI 1640 and this solution was filtered through a 0.22m filter. Aliquots (0.6ml) were frozen in cryo vials at 20C. 4.1.1.2 Complete Medium The INSI cells are maintained in complete medium composed of RPMI 1640 supplemented with 10% heat-inactivated FCS and 50M 2-ME. The complete medium was made up by adding 0.5ml mercaptoethanol from frozen aliquots to 5ml heat inactivated serum (10%) and then adding 44.5ml RPMI to make 50ml, which could be kept for 10 days at 4C.

65

4.1.1.3 Poly-L-Lysine (PLL) Twenty five millilitres of sterile water was added to 5mg PLL (Sigma) in the bottle and 1ml aliquots were placed into vials. For use, 19ml sterile water was added to the 1ml PLL in the vial. 4.1.1.4 Pretreatment of culture dishes Prior to use of culture dishes, 2ml PLL was placed into the culture dish, coating the surface. This was left for 10 minutes at room temperature, removed through aspiration and the surface washed once with PBSA. 4.1.2 Insulin Secretion by INS-1 cells Optimal seeding density was determined as 30 000 cells/well. Cells were seeded into 96well microtiter plates and cultured for 2 days in complete medium. For the experiments, each well was washed once with 50l Krebs-Ringer-bicarbonate-HEPES buffer (KRBH) composed of 118.4mM NaCl, 4.75mM KCl, 1.192mM MgSO4, 2.54mM CaCl2, 10mM HEPES, 2mM NaHCO3, 0.1% BSA (Janjic et al., 1999). The cells were then exposed to 50l preincubation medium (KRBH) for 30 minutes and after removal, treatment incubation medium (KRBH including applicable medication) (100l/well) was added for 10 minutes. After this incubation was complete, 50l incubation medium was removed and a 50x dilution was made for insulin determination using the LINCO Research rat insulin radioimmunoassay (RIA) kit (Cat no. RI-13K), which makes use of rat insulin as a standard. All incubations were carried out at 37C. Samples were stored at -20C until required for insulin determination. Assay buffer from the LINCO kit was used as the diluant for all the dilutions required in these experiments. INS-I cells were exposed to an antidepressant, namely amitriptyline (1M), plant extracts of S frutenscens (Kankerbos) (0.5g/well) and MRC2003 (0.5g/well) as well as a positive control: glibenclamide (a sulfonylurea) (1M), at different exposure times.

66

4.1.3 Plasma insulin determination from rat model To determine whether results obtained with cell culture experiments correlate with results in an animal model, a collaborative project between Wayne Chadwick (NMMU Biochemistry PhD Student) and myself was done where I determined the concentration of insulin in plasma samples from rats. The care of the animals as well as the preparation of the medication was the responsibility of Wayne. All procedures were first tested on rats that were not part of the experimental setup. 4.1.3.1 Preparation of medication 1) Sutherlandia frutescens (kankerbos) Dried kankerbos leaves (2.5g) were weighed and placed in 100 ml boiling water to form an infusion, which was then brewed overnight. The extract was filtered and the rats received a dose concentration of 0.01ml/g body weight in 100ml tea extract in water. The drinking water with extract was replaced daily and the volume remaining each day was recorded. 2) Amitriptyline and Metformin Amitriptyline, the best documented antidepressant that is still used by some patients and is known to cause weight gain was used in this study. Metformin, a biguanide drug, was given as a positive control for glucose uptake in Waynes project. To be consistent in the project the pancreases of all the rats were tested for insulin secretion. An amitriptyline tablet (0.11g) containing 25 mg amitriptyline (highest dose available) or a metformin tablet (0.55g) containing 500 mg metformin (highest dose available) was dissolved in 1 ml of a 0.1 M HCl solution, ensuring that the drugs would dissolve. The relevant drug and water was then added to the water bottles (200ml) and these were placed in the respective cages. Due to the fact that the stability of these drugs in solution is unknown, the medication was changed twice daily and the volume remaining was monitored. Rats were weighed weekly. The doses of the different treatments were relevant to the body weights of the rats in the specific cage. Rats were administered a daily dosage of 67

0.0044mg amitriptyline tablet/g body weight/day and 0.012mg metformin tablet/g body weight/day. 4.1.3.2 Experimental Procedure Table 4.1: Experimental plan for medicinal exposure. Each group consisted of 4 rats. Time 0 1.5 weeks 1.5 months 3 months Control Amitriptyline Sutherlandia frutescans Metformin Each group to be sacrificed consisted of 4 rats (Table 4.1) therefore the 48 rats were divided into 12 cages. The cages were cleaned weekly and the rats were fed dog pellets and received their various medications in their drinking water on a daily basis. To give an indication on whether any of the drugs played a role on weight gain and to what degree, the rats were weighed on a weekly basis throughout the experiment. On the day of sacrifice (after a 12 hour fast), blood was taken from the heart so as to determine the blood glucose level and the blood insulin level prior to death. An intramuscular injection (3l ketamine/g body weight) was administered and as soon as there was no response from the rat, blood was taken from the heart and then placed into a tube (2ml) containing sodium fluoride oxalate, which acts as an anticoagulant. The tube was then immediately placed on ice and once the rat sacrifice was complete, the blood was centrifuged for 20 minutes at 3 000 rpm. Two equal volume aliquots of plasma were taken and frozen at -80C. Insulin quantification was carried out with the use of the LINCO Research rat insulin RIA kit. Assay buffer from the LINCO kit was used as the diluant for the dilution of the samples.

68

4.1.4 Binding studies with INS-1 rat pancreatic cells INS-1 cells were seeded into 24-well plates (maintained in complete medium) and once confluency was reached, experimentation was carried out in quadruplicate. For total binding wells, 100l RPMI 1640 was added, whereas 100l RPMI 1640/unlabelled glibenclamide solution was added to each non-specific binding well (see sections 4.1.4.1/2 for concentrations). The incubation, carried out at 37C, was initiated through the addition of 50l of [3H]Glibenclamide (50 Ci/mmol) in RPMI 1640 to each well. Once incubation with [3H]Glibenclamide was complete, the medium was removed and each well rapidly washed three times with 1ml ice cold PBS. 50l of 1mM NaOH was added to each well to aid in detachment of the cells from the well surface. Each aliquot was then added to a scintillation vial containing 3ml scintillation cocktail (Packard BioScience). Each well was rinsed with 50l PBS and that was also transferred to the scintillation vials. The vials were shaken and radioactivity was determined using a Packard Liquid Scintillation Analyzer, Tri-carb 2300TR (Packard Instrument Company). Non-specific binding was determined by incubations in the additional presence of unlabelled glibenclamide in excess (10 -100x more than labeled). Specific binding was determined by subtracting non-specific from total binding. 4.1.4.1 Binding time study Initially, an incubation time for [3H]Glibenclamide binding was determined. Time intervals investigated included: 2, 6, 12, 30, 60 and 90 minutes at an incubation temperature of 37C. The final concentration of [3H]Glibenclamide in the incubation medium was 0.67nM, which was in accordance with the KD value reported for glibenclamide in the Amersham catalogue. Non-specific binding was defined by binding of [3H]Glibenclanide in the presence of 67nM glibenclamide (100x excess).

69

4.1.4.2 Saturation binding study For saturation experiments, the concentration of [3H]Glibenclamide was varied in order to determine the actual KD value in the experimentation. This was performed using concentrations of [3H]Glibenclamide between 0.25M and 67nM. Non-specific binding was determined by incubations in the presence of 1.34M unlabelled glibenclamide. Saturation binding studies were performed using a 30 minute incubation time at 37C. 4.1.4.3 Displacement study Inhibition of [3H]Glibenclamide binding to INS-1 cells was studied in the presence of [3H]Glibenclamide (8nM) and unlabelled glibenclamide (1.34nM - 1.34M) or the two antidepressants, amitriptyline and trimipramine (0.67M - 6.67M). The concentration of the two antidepressants was in accordance with reported plasma levels. Displacement studies were carried out at a temperature of 37C for 30 minutes. 4.1.5 Statistical analysis Statistics were carried out on results using the data analysis program in Microsoft Excel as described in section 3.5. The actual KD value for [3H]Glibenclamide was calculated using GraphPad Prism Version 4.

70

4.2 Results and Discussion 4.2.1 Cytotoxicity The assays to determine the cytotoxicity of treatments to INS-1 cells were carried out in the same way seen in sections 3.1.1 and 3.1.2. Similarily the concentrations tested for the plant extracts, namely Kankerbos and MRC(2003), ranged from (0.0025 25g/well). The antidepressant, amitriptyline, was tested using the same concentrations used for other treatments tested, ranging from 0.1 to 250M. Glibenclamide was used as the positive control and the concentrations tested included: 0.001, 0.01, 0.1, 1 and 10M. The reason that the highest concentration tested was only 10M compared to the usual 250M, was due to solubility problems experienced as well as information obtained from the binding studies performed. The cytotoxicity results shown in figure 4.1 reveal at which concentration each treatment should be tested on the INS-1 cells. The basic rule of cytotoxicity tests is that any concentration that shows less than 80% of control is too cytotoxic for the specific cell line it is being tested on. The graph of amitriptyline indicates that at high concentrations (50 and 250 M) it is very cytotoxic to INS-1 cells. One does not want too low a dosage of a treatment for experimentation because one requires an effect of some sort to be seen. Therefore the 1M concentration was chosen to be used on all the tests on INS-1 cells. All the concentrations of glibenclamide tested on INS-1 showed no cytotoxic effect. 1M glibenclamide was chosen for further studies on the INS-1 cell line, as it was decided to keep the treatment concentrations the same. The plant extract Kankerbos was found to be cytotoxic at the two highest concentrations tested, namely 2.5 and 25 g/well. A concentration of 0.5 g/well was decided upon for further experimentation and the same concentration would be used for the MRC2003 plant extract as well. All results shown for the plant extracts are g/well, however a conversion table showing the g/ml values is shown in Appendix 1.

71

140 120

120 100

% of control

% of control

100 80 60 40 20 0 -4 -3 -2 -1 0 1 2

80 60 40 20 0 -2 -1 0 1 2 3

log[Glibenclamide]M

log[Amitriptyline]M

100

100

% of control

80 60 40 20 0 0 1 2 3 4 5

% of control

80 60 40 20 0 0 1 2 3 4 5

log[Kankerbos]ng/well

log[MRC2003]ng/well

Figure 4.1: Cytotoxicity results of compounds (Glibenclamide and Amitriptyline) and plant extracts (Kankerbos and MRC2003) on INS-1 cell line. Data points represent the mean SD (n=6 or 12 wells) from an individual experiment

4.2.2 Insulin Secretion by INS-1 cells In order to optimize the specific method to determine the amount of insulin secreted, an initial series of experiments was performed to evaluate the incubation times and combinations of treatments required before the final procedure was chosen. Due to the fact that these initial experiments were carried out just to get an indication of the most suitable protocol to use in the final insulin secretion experiments, cell number or viability was not determined. For each individual experiment it was visually observed that the surface area of INS-1 cells of each well tested was similar. However this could not be determined between different experiments and therefore comparison of these initial

72

experiments is not possible. Each specific experiment carried out was done so in order to determine some specific feature for the final model. Glibenclamide, the sulphonylurea used by type II diabetic patients, was used as the positive control. Firstly, the concentration of glucose to be used in the medium had to be tested. Three options were tested: glucose-free (basal), 2mM glucose and 10mM glucose. For the first experiment these options were all carried out at two time intervals that needed investigating, namely 10 and 30 minutes. These concentrations and incubation times were chosen due to results described in literature (Salt et al., 1998). In addition, each glucose concentration was tested with and without glibenclamide (1M). All of the above combinations were tested in triplicate and using a 5x dilution of the 20l sample, however due to the cost of the LINCO kit and budget constraints thereby leading to only limited numbers of samples being able to be tested, it was decided to only carry out the kit on the 30 minute samples collected. This was also in order to get an indication of the kits sensitivity, which according to the results obtained was found to be too sensitive and therefore further dilutions would have to be done in future experiments. As previously mentioned glibenclamide was used as the positive control in the insulin secretion experiments. Sulphonylureas inhibit KATP-channels in pancreatic -cell plasma membranes and thereby initiate insulin release. The released insulin caused by the interaction of sulphonylureas with their receptors is largely preformed insulin because these agents have little immediate effect on insulin synthesis. In general, the secretory pattern obtained with the sulphonylureas is similar to, but distinct from that obtained with glucose stimulation. Because sulphonylureas apparently stimulate insulin secretion by a mechanism that differs from that of glucose, they can be used in the treatment of diabetics who have lost their ability to respond to a glycaemic stimulus, but who have retained residual pancreatic -cell function, for example type II diabetics (Foye et al., 1995). As shown in figure 4.2, the amount of insulin secreted was increased when glibenclamide was added to the test medium. Insulin secretion is promoted synergistically by glucose and sulphonylurea sensitization of the pancreas. It has been seen that if glucose levels are reduced, that despite sulphonylurea concentrations being maintained, insulin release decreases rapidly. Therefore based on these facts and the need for a

73

positive control (glibenclamide) for insulin stimulation, a glucose-free medium would not be feasible.

1800 1600 1400 1200 1000 800 600 400 200 0 2mM glucose (-glibenclamide) 2mM glucose (+glibenclamide)

Figure 4.2: Effect of 1M glibenclamide (30 minute exposure) on insulin secretion in INS-1 including 2mM glucose in the medium. Data points represent the mean (SD) of an individual experiment assayed in triplicate.

An observation made that can be seen in figures 4.3 and 4.5 was that of a decrease in insulin stimulation when cells were exposed for 30 minutes to 10mM glucose with the addition of glibenclamide. This decrease was in comparison to the effect of 2mM glucose. It would seem that the higher glucose concentration inhibits the full effect of glibenclamide.

[Insulin] ng/ml

1800 1600 1400 1200 1000 800 600 400 200 0 Glucose-free (+glibenclamide) 2mM glucose (+glibenclamide) 10mM glucose (+glibenclamide)

Figure 4.3: Effect of extracellular glucose concentration (glucose-free, 2mM and 10mM glucose) on insulin secretion in INS-1 cells. Cells exposed for 30 minutes and included treatment with glibenclamide. Data points represent the mean SD (n = 3 wells) from one experiment.

[Insulin] ng/ml

74

As previously mentioned, dilutions of the 20l sample was decided upon because there was too much insulin secreted by the INS-1 cells under the specific treatments and therefore the concentrations of insulin were off-scale and did not fall on the standard curve. A 5x dilution had already been tested, so further dilutions including 10x, 50x and 100x were decided upon. For this experiment only two of the treatments were tested to get an indication of where the sample concentrations fell on the standard curve. The 30 minute glucose-free and 10mM glucose samples without glibenclamide were tested in this experiment and results are shown in Figure 4.4.

600 500

[Insulin] ng/ml

400 300 200 100 0 10x 50x Glucose-free 100x 10x 50x 10mM glucose 100x

Figure 4.4: Insulin secretion after 30 30 minute treatment without glucosewith with Figure 4.4: Insulin secretion after minute treatment without glucose and and 10mM glucose, including dilutions of the glucose concentrations. All results shown 10mM glucose, including dilutions of the glucose concentrations. All results shown are without glibenclamide. Values shown are the means of one experiment assayed in are without glibenclamide. Values shown are the means of one experiment assayed in duplicate. duplicate.

Based on the results seen in figure 4.4, it was decided that for all 30 minute samples to be tested in the future, a 100x dilution would be made. At this dilution the values obtained fell in the middle of the standard curve, thereby assuring accurate determination of the amount of insulin secreted. Another decision made at this point was that the incubation medium volume would be increased from 20l to 100l. The 10 minute samples also needed to be tested and a dilution factor of 50x was decided upon as less insulin would have been secreted after 10 minutes as compared to 30 minutes. These results can be seen in figure 4.5.

75

450 400 350

[Insulin] ng/ml

300 250 200 150 100 50 0 Glucose-free (+glibenclamide) 2mM glucose (+glibenclamide) 10mM glucose (+glibenclamide) 10 minutes 30 minutes

Figure 4.5: Effect of extracellular glucose concentration (glucose-free, 2mM and 10mM glucose) (10 and 30 minute exposure) on insulin secretion in INS-1 cells including glibenclamide in the test medium. Data points represent the mean SD (n = 4 wells) of an individual experiment.

At this point the initial set of experiments were complete and based on the results seen in figure 4.5 and previous findings, it was decided that the best protocol to use for optimal determination of insulin secretion would be 2mM glucose in the incubation medium and a 10 minute incubation time. Remembering that the effect of the samples to be tested was unknown, the reasoning for choosing 2mM instead of 10mM glucose was that the purpose of this experiment was to determine the effect of the plant extracts or treatments on insulin secretion. Therefore knowing that the lower glucose concentration would not saturate the glucose transporter and/or the expressed isoform of hexokinase, preventing any effect of the treatments, it was chosen. It is known that glucose-provoked release of insulin is biphasic, including early and late phases. The acute phase occurs almost immediately and peaks within 5 to 10 minutes, whereas the second phase of insulin release occurs 15 to 20 minutes later. This second phase only persists if the glucose level remains elevated (Foye et al., 1995). Based on this theory, the choice of the 10 minute samples was decided upon. One of the main purposes of this project was to finalize

76

methods for the screening of plant extracts. To facilitate the screening of many plant extracts in the future for possible stimulation of insulin release, the aim was to streamline the method including practical and financial consideration. Therefore in finalizing this model for insulin secretion, many decisions were made keeping this in mind. At this stage all the initial experiments were completed and the model for determining insulin secretion was complete. Therefore our aim to create an in vitro model using the INS-1 cell line had been achieved and the determination of the effect of different drugs (antidepressants and traditional medicine) on insulin secretion was now possible.

55 50 45 40

[Insulin] ng/ml

35 30 25 20 15 10 5 0 Basal 2mM glucose 10mM glucose

Figure 4.6: Effect of different glucose concentrations (basal, 2mM and 10mM) on insulin stimulation in INS-1 cells under acute conditions. Data points represent the mean SEM (n = 4 wells) from an individual experiment. *P <0.05 when compared to basal.

Although 2mM glucose had been chosen to be included in the incubation medium for the chronic exposure experiments, a control for the acute exposure experiments was to include basal (glucose-free), 2mM and 10mM glucose and see their effect on the secretion of insulin after 10 minutes exposure. Figure 4.6 provides proof that as the concentration of glucose increases, so the amount of insulin secreted into the medium is increased, both the 2mM and 10mM glucose treatments showed significantly increased

77

levels compared to the glucose-free treatment (*P<0.05). This confirmed that the final model and insulin determination method chosen for insulin secretion was functioning correctly. The findings found in figures 4.7 and 4.8 represent an individual experiment, however the experiment was repeated in triplicate and the results were confirmed. For the acute exposure experiments where treatments were only added for the duration of the experiment (10 minutes), the control used was basal wells where glucose-free medium was added.

50 45 40 35

[Insulin] ng/ml

30 25 20 15 10 5 0 Basal Glibenclamide (1uM)

Amitriptyline (1uM)

Kankerbos (0.5g/w ell)

MRC2003 (0.5g/w ell)

Figure 4.7: Effect of compounds (Glibenclamide and Amitriptyline, 1M) and plant extracts (Kankerbos and MRC2003, 0.5g/well) on insulin secretion in INS-1 cells under acute conditions (10 minute exposure). Positive control: 1M glibenclamide. The values shown are the mean values (SEM) from an individual experiment assayed in quadruplicate. The experiment was performed three times with similar results. *P <0.05 when compared to basal.

Figure 4.7 shows the stimulation of insulin secretion by exposure to the different medicines tested. Firstly, it can be seen that the 1M concentration of glibenclamide produced a significant difference compared to the basal control (*P<0.05), under acute exposure as expected.

78

In the experiment, amitriptyline (1M) also stimulated the INS-1 cells to produce significantly more insulin than the basal control (*P<0.05). Unwanted weight has been associated with the use of tricyclic antidepressants. It has been postulated that the mechanism is due to antidepressants increasing blood insulin levels thereby lowering blood glucose levels which leads to carbohydrate craving as described in literature (Flava, 2000). It is interesting that these results correlate with this hypothesis. The graph indicates that the plant extracts, Kankerbos and MRC2003 (0.5g/well), do not seem to have any significant effect on the stimulation of insulin secretion when INS-1 cells are exposed to these under acute conditions. Chronic exposure is defined as such because the cells are exposed to the treatments for 48 hours prior to the experiment and then are re-exposed for the duration of the experiment (10 minutes) to the applicable treatments. As previously mentioned, the wells were preexposed to the specific treatment before the day of experimentation. On the day of the experiment, basal wells were exposed to glucose-free medium for 10 minutes, this was to determine the effect of the treatment under chronic conditions only, as there was no further stimulation. Two millimolar glucose added for 10 minutes gave an indication of results when further insulin stimulation occurred. The final query was to see the effect of the treatment when the cells were re-exposed for an additional acute period. Therefore this was the effect of pre-exposure (chronic) and re-exposure (acute). This test included 2mM glucose, so when comparing to the 2mM glucose only treatment, one could get an indication of insulin stimulation from the treatment only.

79

60 50

[Insulin] ng/ml

40 Basal 30 20 10 0 Glibenclamide (1M) Amitriptyline (1M) Kankerbos (0.5g/w ell) MRC2003 (0.5g/w ell) 2mM glucose 2mM glucose + treatment

Figure 4.8: Effect of extracellular glucose as well as the test compounds (Glibenclamide and Amitriptyline, 1M) and plant extracts (Kankerbos and MRC2003, 0.5g/well) on insulin secretion in INS-1 cells under chronic conditions. The values shown are the mean values (SEM) from an individual experiment assayed in quadruplicate. The experiment was performed three times with similar results. *P <0.05 when compared to 2mM glucose.

The hyperinsulinemia realized with sulphonylureas (glibenclamide) is transient. When the sulphonylureas are administered chronically, the plasma insulin levels usually, but not always, decline to pretreatment levels or below. The decline has been shown to be associated with decreases in proinsulin synthesis, pancreatic insulin content and insulin secretion (Foye et al., 1995). Based on the theory discussed, glibenclamide (1M) was not expected to have a chronic effect, as its primary effect is seen with acute exposure conditions. This was confirmed with the results obtained (figure 4.8), as no significant difference was noted. As discussed previously, amitriptyline increased the amount of insulin secreted by the INS-1 cells under acute conditions (figure 4.7), however one can see in figure 4.8 that there was no significant difference between the basal, 2mM glucose, 2mM glucose including amitriptyline, thereby indicating that amitriptyline does not affect insulin secretion on the chronic exposure level. There was no significant difference seen between basal and 2mM for the plant extract, Kankerbos (0.5g/well), indicating that

80

chronic exposure alone does not seem to have an impact on insulin secretion. However, of important significance to this project were the results found when Kankerbos (0.5g/well) was added in the experiment for an additional 10 minutes, so representing the accumulative effect of chronic and acute exposure. A significant increase in the amount of insulin secreted was seen for 2mM glucose including Kankerbos (*P<0.05) compared to the 2mM glucose only results. MRC2003 (0.5g/well), the other plant extract of interest in this project, was shown to have no significant effect on insulin secretion.

81

4.2.3 Serum insulin determination in an animal model The hypoglycaemic effect of certain treatments was investigated using male Wistar rats, the findings of which can be seen in figure 4.9. The rats were starved for approximately 12 hours prior to sacrifice. Fasting blood glucose concentration of rats receiving amitriptyline was significantly lower than the control group, indicating an increase in the transfer of glucose from the plasma into the cytoplasm of the cells. Also, an increase in the stimulation of insulin secretion by amitriptyline treated INS-1 cells was seen (figure 4.7). Experiments done by Wayne Chadwick showed that the percentage insulin degraded in the liver, kidney and muscle for the amitriptyline treated group was significantly higher than the control group. Due to these results, increased serum insulin levels were expected for the amitriptyline treated rats. However, it was found that there was no significant difference between the control and amitriptyline for the concentration of insulin in the serum (Figure 4.9).

0.8 0.7

[Insulin] ng/ml

0.6 0.5 0.4 0.3 0.2 0.1 0.0 Control Amitriptyline Metformin Kankerbos

Figure 4.9: Fasting serum insulin concentrations as a result of a 19 week treatment with the medication. Data are the mean SD (n=6 or 7 wells) from one experiment.

82

A possible explanation could be that the concentration of amitriptyline was a threshold dose as the lowest dose (75 mg/kg/day) was given to the rats. There could have been an increase in insulin secreted into the blood under amitriptyline treatment, but it was being degraded before detection could be made. Metformin enhances glucose uptake at tissue level and therefore insulin secretion is not enhanced. So the results found were to be expected and metformin acted more as a negative control for the experiment of insulin secretion into the blood. The rats receiving the traditional anti-diabetic medicine, Sutherlandia frutescens (Kankerbos), did not have any significant difference in serum insulin levels compared to the control group. 4.2.4 Binding studies using INS-1 rat pancreatic cells Binding experiments were performed in order to measure the binding of [3H]Glibenclamide (50 Ci/mmol, Amersham) to SUR-1 receptors present in INS-1 rat pancreatic cells. Glibenclamide is a sulphonylurea drug that is used by diabetics and has been reported to bind with high affinity to SUR-1. To enhance the model being created, it was thought that it would be beneficial to have experiments in place, whereby one could determine if a test compound or plant extract interacted with SUR-1 specifically. This would then enable us to determine if its mechanism of action with regards to it stimulating insulin secretion was through binding to these receptors or not. 4.2.4.1 Choice of incubation time The radioligand associates with the SUR-1 receptor in a time dependant manner. In order to determine the time needed for maximum saturation, [3H]Glibenclamide was incubated for different time intervals with the INS-1 cells at a constant concentration (0.67nM). The results are shown in figure 4.10. Non-specific binding was determined in the presence of 67nM unlabelled glibenclamide. For all further binding experiments, the duration of incubation required to reach binding equilibrium was chosen as 30 minutes.

83

Specific binding (CPM/well)

160 140 120 100 80 60 40 20 0 0 20 40 60 80 100

Time (min)

Figure 4.10: Specific binding of [3H]glibenclamide by INS-1 cells as a function of incubation time. Data points represent the mean SEM (n = 3 or 4 wells) from a single experiment.

4.2.4.2 Saturation binding Non-specific binding was measured at different radioligand concentrations in combination with an excess (saturating) of unlabelled competing ligand. Through the subtraction of these values from the total binding at each radioligand concentration, the specific binding was calculated (figure 4.11). Non-specific binding increased linearly as expected, whereas specific binding was saturable indicating that all the receptors were saturated with glibenclamide.

1200

Bound ligand (cpm/well)

1000 800 600 400 200 0 0 10 20 30 40 50 60 70 80

Total Non-specific Specific

[Glibenclamide] (nM)

Figure 4.11: Saturation binding curve. Data points for total, non-specific and specific binding are mean SD (n=3 or 4 wells) from a single experiment.

84

The actual dissociation constant (KD) for [3H]Glibenclamide was then calculated at halfmaximal bound and as in figure 4.12, was shown to be 14.23 2.04nM (R2=0.9682; 99% confidence values: 8.5 to 19.97nM). The actual KD value calculated was found to be higher than what was theoretically stated (0.4nM according to Amersham catalogue) and this was probably due to variations between cell types and experimental conditions. Based on the calculated KD value in the range of 8.5 to 19.97nM, displacement studies were performed using a [3H] Glibenclamide concentration of 8nM.

650 600 550

Bound ligand (cpm /well)

500 450 400 350 300 250 200 150 100 50 0 0 14.23 20 40 60 80

Kd

Glibe nclamide (nM )

Figure 4.12: Saturation binding curve. Data points for specific binding are mean SD (n=3 or 4 wells) from a single experiment.

4.2.4.3 Displacement studies The purpose of carrying out displacement studies is to determine if there is a compound in a plant extract that binds to SUR-1 and thereby potentially stimulating insulin secretion. As there were no plant extracts in this study that were expected to do this, the antidepressants amitiptyline and trimipramine were used to test the above hypothesis.

85

The hypothesis to be tested followed a computational overlay study (Wilson, 2002). By determining the similarity of the sulphonylurea and antidepressants 3-D structures, the possibility of the antidepressant binding to SUR-1 receptors, blocking the K+ATP channel and resulting in increased insulin secretion was explored. Through the overlay study the probability (P) of a specific antidepressant agent binding to SUR-1 and thereby initiating insulin secretion was determined. For the purpose of the current investigation, the two tricyclic antidepressants amitriptyline and trimipramine were focused on. Table 4.2 shows the P values for these two antidepressants as found in the previous study. Table 4.2: Probability (P) of binding to SUR-1 receptors Tricyclic Antidepressant P Amitriptyline Trimipramine 0.023711 0.48985

P is the probabilty of a specific antidepressant agent binding to SUR1 and thereby initiating insulin secretion. If P<0.05 this indicates that the Root Mean Square (RMS) of a particular antidepressant is significantly different to the training sets selected RMS range and there is therefore no possibility of binding. However, if P>0.05 there is a 40% chance that binding will occur. The P value calculated for amitriptyline revealed that there would be no possibility of binding to SUR-1, whereas trimipramine has a 40% chance of binding. For these studies, a constant amount of [3H]Glibenclamide was added to each of the wells together with varying concentrations of unlabelled competitors of interest. In all the displacement studies carried out, unlabelled glibenclamide was shown to be a significant competitor to [3H]Glibenclamide. Figure 4.13 shows as the concentration of unlabelled glibenclamide increases, so the percent bound [3H]Glibenclamide decreases indicating displacement of the radioligand.

86

% Bound [3H]Glibenclamide

180 160 140 120 100 80 60 40 20 0 1.34 13.4 134 1340 670 6670 Amitriptyline Unlabelled Glibenclamide

[Competitor]nM

Figure 4.13: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M) and amitriptyline (0.67M and 6.67M). The values shown are the mean values (SEM) from one experiment assayed in triplicate or quadruplicate. The experiment was repeated with similar results. *P <0.05 compared to the total bound in absence of competitor.

The initial displacement assay (figure 4.13) revealed that amitriptyline did not significantly displace [3H]Glibenclamide and therefore did not significantly interact with the SUR-1 receptors of INS-1 cells. This finding correlates with results found through computational structural overlay studies done by Wilson, 2002 as seen in Table 4.2. Although the binding study confirmed the computational overlay results with regard to amitriptyline, it was decided to carry out another displacement study with trimipramine in order to confirm the previous findings. It is evident in figure 4.14 that there is a large difference between amitriptyline and trimipramine, which could explain the results found in the computational overlaying experiments which identified a 40% chance of binding for trimipramine compared to a 0% chance for amitriptyline. However figure 4.14 reveals that there was no significant displacement of [3H]Glibenclamide by trimipramine.

87

175

% Bound [ H]Glibenclamide

150 125 100 75 50 25 0 1.34 13.4 134 1340 670

*
Amitriptyline Trimipramine Unlabelled Glibenclamide

6670

[Competitor]nM

Figure 4.14: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M), amitriptyline and trimipramine (0.67M and 6.67M). The values shown are the mean values (SEM) from one experiment assayed in triplicate or quadruplicate. *P <0.05 compared to the total bound in absence of competitor.

Figure 4.14 shows a continual finding during the binding studies was that amitriptyline competition produced an increase in bound [3H]Glibenclamide greater than 100%. This was investigated by implementing the use of a range of lower concentrations of amitriptyline in the displacement studies, as seen in figure 4.15.

% Bound [3H]Glibenclamide

140 120 100 80 60 40 20 0 1.34 13.4 134 1340 0.67 67 670 6670

140 120 100 80 60 40 20 0 Amitriptyline Unlabelled Glibenclamide

[Competitor]nM

Figure 4.15: Displacement of [3H]glibenclamide by unlabelled glibenclamide (1.34nM 1.34M) and varying concentrations of unlabelled amitriptyline (0.67nM 6.67M). Mean SEM (n=3 or 4 wells) from an individual experiment.

88

Figure 4.15 confirms that amitriptyline does not bind to the receptor, as there is no displacement of [3H]Glibenclamide. Even at the lower concentrations of amitriptyline tested, results produced binding of [3H]Glibenclamide that was greater than 100%. Previous experiments (see figure 4.7), however show that amitriptyline does stimulate insulin secretion by INS-1 cells under acute conditions. The mechanism is therefore not clear at this stage, though we know it is not via the SUR-1 receptors, like sulphonylureas. This confirms the results obtained previously ie. theoretical binding possibilities of tricyclic antidepressants to SUR-1 (Wilson, 2002) in the computational overlaying studies. This series of experiments successfully determined the conditions needed to screen the effects of drugs (compounds) or plant extracts on pancreatic cells.

89

Chapter 5 Conclusion
The present study investigated the in vitro antidiabetic properties of two indigenous plants, with the aim to optimise the processes required for their screening and ultimately the determination of their effectiveness. The study was based on the fact that these plant extracts are being widely used as medicinal plants, however there is a lack of scientific evidence for their efficacy. Determination of their efficacy is vital as these medicinal plants may serve as important participants in the management of type II diabetes in the future and thereby make a positive contribution to South Africas health care services. Type II diabetes is characterised by abnormal glucose homeostasis leading to hyperglycaemia. The mechanism of action of many hypoglycaemic drugs used for type II diabetes is based on their ability to positively affect glucose utilisation and insulin secretion. Therefore in creating a model for determining the effects of alternative traditional medicines as antidiabetic potentiates, it was necessary that these two metabolic pathways (glucose uptake and insulin secretion) play the role of centre pivot in our investigations. In chapter 1 the pharmacological treatments available for type II diabetic patients were discussed. The hypoglycaemic drugs which base their mechanisms on glucose utilisation are namely the biguanides (metformin), thiazolidinediones and insulin. Sulfonylurea drugs such as glibenclamide in this study, employ insulin secretion as their mechanism of action. This study encompassed these two vital processes in creating a model for antidiabetic screening, therefore a model for glucose utilisation (Chapter 3) and a model for insulin secretion (Chapter 4) was developed. The biguanides function as antidiabetic agents by stimulating hepatic glucose metabolism, reducing hepatic glucose output and reducing intestinal glucose absorption (Moriaka et al., 2005). The thiazolidinediones increase peripheral glucose uptake and metabolism as well as increase insulin sensitivity (Oakes et al., 2001). Insulin functions by stimulating hepatic glucose metabolism as well as increasing peripheral glucose uptake and metabolism (Levinthal and Tavill, 1999). The 90

sulfonylureas stimulate insulin secretion from pancreatic -cells (Foye et al., 1995). So overall one can see that drugs used for the treatment of diabetes have a variety of tissues that they act on, therefore it was necessary to use the relevant cell lines to represent all the possible scenarios. Due to this the 3T3-L1 adipocyte, Chang liver, C2C12 skeletal muscle and INS-1 rat pancreatic cell lines were involved in the glucose utilisation and insulin secretion models of this study. The entry of glucose into cells is a crucial step in life-supporting processes since glucose is the main monosaccharide in nature that provides carbon and energy for almost all cells. The passage of glucose into cells depends on different parameters, including expression of the appropriate glucose transporters in the target tissues and hormonal regulation of their function. In mammalian cells a tight regulation of blood glucose levels is needed to meet the energetic demands of the brain, a tissue that uses glucose as its primary energy source. Adequate glucose flux into tissues provides maintenance of glucose homeostasis that is critical in well being (Gorovits and Charron, 2003). Glucose that is taken up by a cell may be oxidized to form energy (glycolysis). In type II diabetes, excessive hepatic glucose output contributes to the fasting hyperglycaemia. Increased gluconeogenesis is the predominant mechanism responsible for this increased glucose output, while glycogenolysis has not been shown to be increased in patients with type II diabetes. Hyperglucagonemia has been shown to augment increased rates of hepatic glucose output, probably through enhanced gluconeogenesis (Figure 5.1) (Levinthal and Tavill, 1999).

91

Figure 5.1: Summary of carbohydrate metabolism (www.np.edu.sg)

In this study, the two contenders for use as positive controls in the glucose utilisation model were insulin and metformin. Insulin's actions are summarised in figure 5.2. Insulin promotes glycogen synthesis (glycogenesis) in the liver and inhibits its breakdown (glycogenolysis). It promotes protein, cholesterol, and triglyceride synthesis and stimulates formation of very-lowdensity lipoprotein cholesterol. It also inhibits hepatic gluconeogenesis, stimulates glycolysis, and inhibits ketogenesis (Levinthal and Tavill, 1999).

92

Figure 5.2: Mechanisms of insulin action. Key: insulin. (Saltiel and Kahn, 2001)

stimulated by insulin;

inhibited by

Exposure to insulin, which dominates in the prandial and immediate postprandial period, encourages storage of carbohydrate and fat and reduces or prevents hepatic glucose production. There is a general agreement that the major factor in insulin's indirect suppression of endogenous glucose production is decreased plasma FFA concentration, due to the antilipolytic effect of insulin on adipose tissue. When hyperinsulinaemia becomes sufficiently high, the insulin-receptor complex in adipocytes and in skeletal muscle initiates a transduction chain that results in translocation of intracellular vesicles to fuse with the plasma membrane. Laced in the walls of these vesicles are GLUTs, GLUT1 and GLUT4 in adipocytes, but almost exclusively GLUT4 in skeletal muscle (Zierler, 1999). Insulin does not directly stimulate hepatic glucose uptake through translocation of glucose transporters to the plasma membrane as it does in skeletal muscle. Instead, insulin promotes glucose uptake in the liver cell, at least in part, through increased expression of hepatic glucokinase (Iozza et al., 2003).

93

Metformin is an insulin-sensitising agent with potent antihyperglycaemic properties (Tankova, 2003). Its glucose-lowering effects are mainly a consequence of reduced hepatic glucose output (primarily through inhibition of gluconeogenesis and, to a lesser extent, glycogenolysis) and increased insulin-stimulated glucose-uptake in skeletal muscle and adipocytes. Its major mode of action is to reduce hepatic glucose production, which is increased at least twofold in patients with type II diabetes. Metformin facilitates insulin-induced suppression of gluconeogenesis from several substances (lactate, pyruvate, glycerol and amino acids) and opposes the gluconeogenic actions of glucagon. In addition, it increases intramitochondrial levels of calcium, a modulator of mitochondrial respiration (Figure 5.3) (Kirpichnikov et al., 2002).

Figure 5.3: Mechanisms of metformin action on hepatic glucose production and muscle consumption (Kirpichnikov et al., 2002)

The primary site of metformins action on reduction of hepatic glucose production appears to be hepatocyte mitochondria, where it disrupts respiratory chain oxidation of complex I substrates. Inhibition of cellular respiration decreases gluconeogenesis and may induce expression of glucose transporters, thereby facilitating glucose utilisation. In several tissues, including skeletal muscle and adipocytes, metformin facilitates glucose

94

transport by increasing tyrosine kinase activity in insulin receptors and enhancing the trafficking of glucose transporters 1 and 4 to the cell membrane (Kirpichnikov et al., 2002). Insulin and metformin were chosen as positive controls for glucose uptake in 3T3-L1 cells. In the C2C12 cell line, the results for the determination of the ideal positive control correlated to theory, as it is known that insulin increases peripheral glucose uptake and metabolism in skeletal muscle. Insulin was shown to provide a significant increase in the amount of glucose taken up by these cells (Figure 3.11), however this only qualified for acute conditions as no viable results were obtained with these cells when they were exposed to chronic circumstances. Considering the role of muscle in glucose clearance in the body, and the mechanism of glucose uptake in this tissue, it would be more appropriate for an antidiabetic agent to act immediately upon administration rather than over a long period of time. Knowing that both insulin and metformin stimulate hepatic glucose metabolism, the results obtained for Chang liver cells in this study were expected. For acute exposure, insulin showed a significant effect on glucose utilisation by Chang liver cells (Figures 3.14 and 3.15) and because of this significant effect was chosen as the positive control for this condition. Due to the results found in figure 3.14 as well as the fact that it is known that metformin has an effect with chronic exposure, it was decided to use it as the positive control for glucose utilisation in the cells under chronic conditions. Chronic treatments with metformin provided a significant increase in glucose utilised by Chang liver cells, as was seen in figure 3.17. The positive control used for the insulin secretion model in this study was the sulfonylurea, glibenclamide. Glibenclamide is used by type II diabetics and it functions through its ability to bind to SUR-1 receptors of pancreatic -cells, thereby initiating the release of insulin. Glibenclamide was an effective positive control for stimulating insulin secretion by INS-1 cells under acute conditions as there was a significant increase in the amount of insulin secreted when these cells were exposed to this drug (Figure 4.7).

95

However no significant effect was seen with regards to the chronic conditions tested when glibenclamide was included. This was not unexpected as it is known that glibenclamides activity does not favour the chronic case but rather the acute one. Therefore the results from this study correlated with previous information of glibenclamides mechanism of action. BDM1 and BDH1 are compounds expected to be present in the plant extracts tested in this study. Although these compounds were included as test substances for their effect on glucose utilisation in 3T3-L1 cells, results were inconclusive due to the unforeseen problems experienced with the cell line. BDM1 was shown to have no effect on glucose uptake in C2C12 muscle cells for either acute or chronic conditions. However, BDH1 significantly increased the amount of glucose taken up by the C2C12 cells being exposed to acute conditions (Figure 3.11) but not in chronic exposure conditions. When tested on the Chang liver cell line, BDM1 once again seemed to have no significant effect on glucose utilisation in these cells. When the Chang cells were exposed to BDH1 under acute conditions, no significant effect was seen either. On the other hand, when these cells were treated to BDH1 chronically, a significant increase in glucose uptake was seen (Figure 3.17). BDP1, as mentioned, is a compound that in previous studies has been found to display similar characteristics to insulin and has been said to be an antidiabetic compound. In this study, the only positive results that correlate to these previous findings were observed in C2C12 muscle cells. It was found that when these cells were exposed to BDP1 acutely, that a significant increase in glucose utilisation was seen (Figure 3.11). Other than BDP1s significant effect on C2C12 cell (acute exposure), no other significant results were found for this compound on any of the other cell lines. Throughout the course of this study it was seen that Kankerbos has biological activities which are comparable to some of the more well-known antidiabetic compounds. The three successful cell lines that Kankerbos was tested on included: C2C12 muscle, Chang liver and INS-1 pancreatic cells. With regards to the glucose utilisation model, Kankerbos was seen to have both acute and chronic effects in different cell lines. In the C2C12 muscle cells, Kankerbos significantly increased glucose uptake when they were

96

exposed to acute conditions (Figure 3.11). Kankerbos also had a significant effect on the Chang liver cell line as it was observed that under both acute (Figure 3.15) and chronic (Figures 3.17) conditions, this plant extract induced the uptake of glucose into these cells. With respect to the insulin secretion model involving INS-1 cells, no significant effect was seen during acute exposure when they were treated with Kankerbos. However during chronic exposure, this plant extract initiated an increased insulin secretion from INS-1 cells (Figure 4.8). Taken together, the results of this study suggest that Kankerbos has a dual mechanism of action for its glucose-lowering effects: It stimulates glucose usage by C2C12 muscle and Chang liver cells as well as enhances insulin release from INS-1 cells. Since Kankerbos is widely distributed in South Africa and therefore accessible to many communities, this study was useful as it provides an indication that Kankerbos has antidiabetic activities and could therefore be used as a possible alternative once further analysis has been completed. The plant extract, MRC2003, proved not to be as effective as Kankerbos in this study. Although it was seen in two experiments that the MRC2003 sample significantly enhanced glucose utilisation in 3T3-L1 adipocytes (Figure 3.4), no conclusions could be drawn due to the fact that no feasible results were obtained from the experiments carried out on the 3T3-L1 cell line. Further experimentation involving the MRC2003 sample included it being tested on the C2C12 skeletal muscle, Chang liver and INS-1 pancreatic cell lines. For all these cell lines within both models, no significant effect was seen when the cells were exposed to MRC2003 under acute and chronic conditions. Therefore no conclusions could be made about this plant extracts potential as an antidiabetic compound. Seeing that glibenclamide is known to bind with high affinity to SUR-1 and thereby initiate insulin secretion, it was decided to use this theory for further investigations. This was used as the basis and reasoning for carrying out binding studies. It was thought that by carrying out these studies it would allow one to determine if the mechanism of a compound or plant extract in stimulating insulin secretion was subsequent to its binding to the SUR-1 receptor or whether it was accomplished by another mechanism. Owing to

97

the fact that there was no plant extract in this specific study that was expected to do this, amitriptyline and trimipramine were used. Due to amitriptyline being included as a test substance for the binding studies in this study as well as the computational overlay study previously done, it was included in the insulin secretion model of this study. The amitriptyline-treated INS-1 pancreatic cells showed a significant increase in insulin stimulation and secretion when exposure was acute (Figure 4.7). These results correlated with a previous hypothesis in literature: Unwanted weight gain associated with a few antidepressants could be due to their increasing blood insulin levels causing blood glucose to drop and thereby leading to the food cravings experienced by the patients. This significant effect was however only observed when the INS-1 cells were exposed to amitriptyline under acute conditions and not chronic. Although amitriptyline had this effect on insulin secretion by INS-1 cells, the binding studies revealed that amitriptyline does not bind to the SUR-1 receptor of pancreatic cells (Figures 4.13, 4.14, 4.15). Amitriptylines mechanism of action on insulin secretion is still elusive at this stage, however it seems that amitriptyline does not exert its effect on insulin secretion in the same manner as the sulphonylureas, that is via the SUR -1 receptors. In conclusion, the objective of this study, to optimize the techniques required for the screening and determination of the effectiveness of specific compounds and plants as antidiabetic potentiates, through observing if they had effects on glucose utilisation and insulin secretion, was accomplished. Since optimal methodology was achieved through this specific project, the criteria used to validate the models will now be able to be applied to scientifically ascertain the antidiabetic effect(s) and mechanism(s) of plants being researched in the future.

98

References
Alper J (2000) New Insights into type 2 diabetes, Science 289: 37-39 Asfari M, Janjic D, Meda P, Li G, Halban PA, Wollheim CB (1992) Establishment of 2 Mercaptoethanol-Dependent Differentiated InsulinSecreting Cell Lines, Endocrinology 130(1): 167-178 Askew SL (2003) Evaluation of methods to culture C2C12 skeletal muscle cells and detect differentiation markers, BSc. Honours, University of Port Elizabeth Aslesen R, Engebretsen EML, Franch J, Jensen J (2001) Glucose uptake and metabolic stress in rat muscles stimulated electrically with different protocols, J. Appl. Physiol. 91(3): 1237-1244 Bogan JS, McKee AE, Lodish HF (2001) Insulin-Responsive Compartments Containing GLUT4 in 3T3-L1 and CHO Cells: Regulation by Amino Acid Concentrations, Molecular and Cellular Biology 21(14): 4785-4806 Cerasi E (2000) Type 2 Diabetes: To Stimulate or Not to Stimulate the Cell, Metabolism: Clinical and Experimental 49(10) Suppl 2: 1-2 Clark JBF, Queener SF, Karlb VB (1997) Pharmacologic Basis of Nursing Practice, 5th Ed Mosby-Year Book Inc.: 755-763 Davis A, Christiansen M, Horowitz JF, Klein S, Hellerstein MK, Ostlund RE (2000) Effect of Pinitol Treatment on Insulin Action in Subjects with Insulin Resistance, Diabetes Care 23: 1000-1005 Dela F, Ploug T, Handberg A, Peterson LN, Larsen JJ, Mikines KJ, Galbo H (1994) Physical Training Increases Muscle GLUT4 Protein and mRNA in Patients With NIDDM, Diabetes 43: 862-865 Dowell P, Flexner C, Kwiterovich PO, Lane MD (2000) Suppression of Preadipocyte Differentiation and Promotion of Adipocyte Death by HIV Protease Inhibitors, J. Biol. Chem. 275(52): 41325-41332 Elks ML, Manganiello VC (1985) Antilipolytic action of insulin: role of cAMP phosphodiesterase activation, Endocrinology 116: 2119-2121 Flava M (2000) Weight gain and antidepressants, J.Clin. Psychiatry 61(11): 37-41 Forman BM (2002) The Antidiabetic Agent LG100754 Sensitizes Cells to Low Concentrations of Peroxisome Proliferator-activated Receptor Ligands, J. Biol. Chem. 277(15): 12503-12506 Foye WO, Lemke TL, Williams DA (1995) Principles of Medicinal Chemistry, 4th Ed. Williams and Wilkins: 37-599 99

Fujikura K, Suzuki M, Kotake K, Inagaki N, Seino S, Takata K (1999) Immunolocalization of Sulfonylurea receptor 1 in rat pancreas, Diabetologia 42(10): 1204-1211 Gaben-Cogneville AM, Breant B, Coudray AM, Hoa DH, Mester J (1990) Differentiation of newborn rat preadipocytes in culture: effects of insulin and dexamethasone, Exp. Cell. Res. 191(1):133-140 Galic S, Hauser C, Kahn BB, Haj FG, Neel BG, Tonks NK, Tiganis T (2005) Coordinated Regulation of Insulin Signaling by the Protein Tyrosine Phosphatases PTP1B and TCPTP, Molecular and Cellular Biology 25(2): 819829 Garland EJ, Remick RA, Zis AP (1988) Weight gain with antidepressants and lithium, Journal of Clinical Psychopharmacology 8(5): 323-330 Gerich JE (2000) Addressing the Insulin Secretion Defect: A Logical First-Line Approach, Metabolism: Clinical and Experimental 49(10) Suppl. 2: 12-16 Gorovits N, Charron MJ (2003) What we know about facilitative glucose transporters, Biochemistry and Molecular Education 31(3) : 163-172 Greenwood M, Kreider RB, Rasmussen C, Almada AL, Earnest CP (2001) DPinitol Augments Whole Body Creatine Retention in Man, Journal of Exercise Physiology, 4(4) : 41-47 Gribble FM, Ashcroft FM (2000) Sulphonylurea Sensitivity of Adenosine Triphosphate-Sensitive Potassium Channels From Cells and Extrapancreatic Tissues, Metabolism: Clinical and Experimental 49(10) Suppl. 2: 3-6 Haring HU (1999) Pathogenesis of type 2 diabetes : are there common causes for insulin resistance and secretion failure ?, Exp Clin Endocrinal Diabetes 107(2) : 17-23 Harrower A (2000) Gliclazide Modified Release : From Once-Daily Administration to 24-Hour Blood Glucose Control, Metabolism : Clinical and Experimental 49(10) Suppl. 2 : 7-11 Hernandez-Sanchez C, Ito Y, Ferrer J, Reitman M, Le Roith D (1999) Characterization of the mouse sulfonylurea receptor 1 promoter and its regulation, J Biol Chem 274 (26): 1-5 Hurr NA (1996) The effect of aerobic exercise on the adipose tissue of subjects using antidepressant medication, MSc. Dissertation, University of Port Elizabeth Iozzo P, Hallsten K, Oikonen V, Virtanen KA, Kemppainen J, Solin O, Ferrannini E, Knuuti J, Nuutila P (2003) Insulin-mediated hepatic glucose uptake is impaired in type 2 diabetes: evidence for a relationship with glycemic control, J Clin Endocrinol Metab 88: 205560.

100

Isomoto S, Kando C, Yamada M, Matsumoto S, Higashiguchi O, Hario Y Matsuzawa Y, Kurachi Y (1996) Communication : A novel sulfonylurea receptor forms with Bir (Kir 6.2) a smooth muscle type ATP-sensitive K+ channel, J Biol Chem 271 (40): 1-14 Janjic D, Maechler P, Sekine N, Bartley C, Annen AS, Wollheim CB (1999) Free Radical Modulation of Insulin Release in INS 1 Cells Exposed to Alloxan, Biochemical Pharmacology 57: 639-648 Johnson LR (1998) Essential Medical Physiology, 2nd Ed Lippincott-Raven Publishers: 565-581 Katzung BG (1995) Basic and Clinical Pharmacology, 6th Ed Appleton and Lange :448-457, 637-650 Kirpichnikov D, McFarlane S, Sowers J (2002) Metformin: An Update, Ann. Intern. Med. 137: 25-33 Klip A, Leiter LA (1990) Cellular mechanism of action of metformin, Diabetes Care 13(6): 696-704 Kumar N, Dey CS (2003) Development of insulin resistance and reversal by thiazolidinediones in C2C12 skeletal muscle cells, Biochemical Pharmacology 65: 249-257 Larner J, Price J, Picariello T, Huang L (1997) Method of treating defective glucose metabolism using synthetic insulin substances, United States Patent Levinthal GN, Tavill AS (1999) Liver Disease and Diabetes Mellitus, Clinical Diabetes 17(2) Lienhard GE, Slot JW, James DE, Muecklar MM (1992) How cells absorb glucose, Scientific American 206: 34-39 Mandrup S, Lane MD (1997) Regulating Adipogenesis, JBC 272(9): 5367-5370 Matthaei S, Stumvoll M, Kellerer M, Haring H (2000) Pathophysiology and Pharmacological Treatment of insulin Resistance, Endocrine Reviews 21(6): 585-618 Medina RA, Owen GI (2002) Glucose transporters: expression, regulation and cancer, Biol. Res. 35(1) Monks A, Scudiero D, Skehan P, Shoemaker R, Paull K, Vistica D, Hose C, Langley J, Cronise P, Vaigro-Wolff A, Gray-Goodrich M, Campbell H, Mayo J, Boyd M (1991) Feasibility of a High-flux Anticancer Drug Screen Using a Diverse Panel of Cultured Human Tumor Cell Lines, Journal of the National Cancer Institute 83(11): 757-766

101

Morioka K, Nakatani K, Matsumoto K, Urakawa H, Kitagawa N, Katsuki A, Hori Y, Gabazza EC, Yano Y, Nishioka J, Noboni T, Sumida Y, Adachi Y (2005) Metformin-induced suppression of glucose-6-phosphatase expression is independent of insulin signaling in rat hepatoma cells, Int J Mol Med 15(3): 449-452 Muller F (2002) Sutherlandia: a useful African herb in medical practice, The South African Journal of Natural Medicine 8: 69-94 Narayanan CR, Joshi DD, Mujumdar AM, Dhekne VV (1987) Pinitol-A new antidiabetic compound from the leaves of Bougainvillea Spectabilis, Current Science 56(3): 139-141 Newsholme EA, Leech AR (1992) Biochemistry for the medical sciences, John Wiley & Sons Ltd: 562-582, 694-709, 828-838 Oakes ND, Thalen PG, Jacinto SM, Ljung B (2001) Thiazolidinediones Increase Plasma-Adipose Tissue FFA Exchange Capacity and Enhance Insulin-Mediated Control of Systemic FFA Availability, Diabetes 50: 1158-1165 Perrini S, Natalicchio A, Laviola L, Belsanti G, Montrone C, Cigarelli A, Minielli V, Grano M, De Pergola G, Giorgino R, Giorgina F (2004) Dehydroepiandrosterone Stimulates Glucose Uptake in Human and Murine Adipocytes by Inducing GLUT1 and GLUT4 Translocation to the Plasma Membrane, Diabetes 53: 41-52 Pessin JE, Saltiel AR (2000) Signaling pathways in insulin action: molecular targets of insulin resistance, J. Clin. Invest. 106(2): 165-169 Purintrapiban J, Ratanachaiyavong S (2003) The Effects of Insulin and Metformin on Glucose Uptake in L8 Myotubes, ScienceAsia 29: 341-346 Quesada I, Nadal A, Soria B (1999) Different effects of tolbutamide and diazoxide in alpha-, beta-, and delta- cells within intact islets of Langerhans, Diabetes 48(12): 2390-2397 Raab-Graham KF, Cirilo LJ, Boettcher AA, Radeke CM, Vandenberg CA (1999) Membrane Topology of the Amino-terminal Region of the Sulfonylurea Receptor, J Biol Chem 274(41): 1-19 Rosen ED, Spiegelman BM (2000) Molecular Regulation Of Adipogenesis, Annu. Rev. Cell Dev. Biol 16: 145-171 Rosen ED, Walkey CJ, Puigserver P, Spiegelman BM (2000) Transcriptional regulation of adipogenesis, Genes and Development 14: 1293-1307 Rutter GA, da Silva Xavier GDS, Leclerc I (2003) Roles of 5AMP-activated protein kinase in mammalian glucose homeostasis, Biochemical Journal 375: 116

102

Sakurai Y, Teruya K, Shimoda N, Umeda T, Tanaka H, Muto T, Kondo T, Nakamura K, Yoshizawa N (1999) Association between duration of obesity and the risk of non-insulin-dependant Diabetes Mellitus, American Journal of Epidemiology 149(3): 256-260 Salt IP, Johnson G, Ashcroft SJH, Hardie DG (1998) AMP-activated protein kinase is activated by low glucose in cell lines derived from pancreatic cells, and may regulate insulin release, Biochem. J. 335: 533-539 Saltiel AR (1990) Second Messengers of Insulin Action, Diabetes Care 13(3): 244-256 Saltiel AR, Kahn CR (2001) Insulin signaling and the regulation of glucose and lipid metabolism Nature 414: 799-806 Seier JV, Mdhluli M, Dhansay MA, Loza J, Laubscher R (2002) A Toxicity Study Of Sutherlandia Leaf Powder (Sutherlandia microphylla) Consumption, Medical Research Council: 1-35 Sheard NF, Clark NG (2000) The Role of Nutrition Therapy in the Management of Diabetes Mellitus, Nutrition in Clinical Care 3(6): 334-348 Tankova T (2003) Current indications for metformin therapy, Rom. J. Intern. Med. 41(3): 215-225 Taylor SI (1999) Deconstructing Type 2 Diabetes, Cell 97: 9-12 Tortorella LL, Pilch PF (2002) C2C12 myocytes lack an insulin-responsive vesicular compartment despite dexamethasone-induced GLUT 4 expression, Am. J. Physiol. Endocrinol. Metab. 283(3): 514-524 Trembley A, Doucet E (2000) Obesity: a disease or a biological adaptation?, Obesity Reviews 1:27-35 Ueyama A, Yaworsky KL, Wang Q, Ebina Y, Klip A (1999) GLUT-4myc ectopic expression in L6 myoblasts generates a GLUT-4-specific pool conferring insulin sensitivity, Am. J. Physiol. Endocrinol. Metab. 277(3): E572-E578 Vardaxis NJ (1994) Pathology for the health sciences, Macmillan Education Wagner H, Farnsworth NR (1994) Economic and Medicinal Plant Research, Volume 6, Academic Press Ltd: 149-187 Walker P (2000) Studies on the influence of antidepressants on insulin secretion by the RIN-m5F cells, BSc. Honours, University of Port Elizabeth Weir GC, Bonner-Weir S (2004) Five Stages of Evolving Beta-Cell Dysfunction During Progression to Diabetes, Diabetes 53:S16-S21 Wiederer O, Loffler G (1987) Hormonal regulation of the differentiation of rat adipocyte precursor cells in primary culture, J. Lipid. Res. 28(6): 649-658

103

Wilson GP (2002) Experimental models to determine the effect of different drugs on insulin secretion, BSc. Honours, University of Port Elizabeth Wise LS, Green H (1979) Participation of one isoenzyme cytosolic glycerophosphate dehydrogenase in the adipose precursor cells, J. Biol. Chem. 254: 273-275 Withers DJ, Sanchez Eutierrez J, Towery H, Burks DJ, Ren J, Previs S, Zhang Y, Bernal D, Pans S, Shulman GI, Banner-Weier S, White MF (1998) Disruption of IRS-2 causes type 2 diabetes in mice, Nature 391: 900-904 Zierler K (1999) Whole body glucose metabolism, Am J Physiol Endocrinol Metab 276 (3): E409-E426 www.biotech.ist.unige.it www.chemicon.com/adipogenesis assay www.geocities.com www.i-sis.org.uk/ www.lifeclinic.com www.np.edu.sg www.pslgroup.com www.sutherlandia.org

104

Appendix A
Table A1: Conversion table for plant extract concentrations Cell line C2C12 Chang Plant extract Kankerbos MRC2003 Kankerbos Condition Acute Acute Acute Chronic (48hr exposure) Acute Chronic (48hr exposure) Acute Chronic (48hr exposure) Acute Chronic (48hr exposure) g/well 2.5 2.5 12.5 10 12.5 10 0.5 0.5 0.5 0.5 l/well 50 50 25 200 25 200 100 200 100 200 g/ml 50 50 500 50 500 50 5 2.5 5 2.5

MRC2003

INS-1

Kankerbos

MRC2003

105