Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effect of metformin treatment on isolated cardiac function and blood pressure in diabetic and hypertensive… Verma, Subodh 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_fall_verma_subodh.pdf [ 3.98MB ]
Metadata
JSON: 831-1.0086319.json
JSON-LD: 831-1.0086319-ld.json
RDF/XML (Pretty): 831-1.0086319-rdf.xml
RDF/JSON: 831-1.0086319-rdf.json
Turtle: 831-1.0086319-turtle.txt
N-Triples: 831-1.0086319-rdf-ntriples.txt
Original Record: 831-1.0086319-source.json
Full Text
831-1.0086319-fulltext.txt
Citation
831-1.0086319.ris

Full Text

We accept this thesis as conformingto the required standardEFFECT OF METFORMIN TREATMENTON ISOLATED CARDIAC FUNCTION AND BLOOD PRESSUREIN DIABETIC AND HYPERTENSIVE RATSbySUBODH VERMAB. Pharm., Gujarat University, India, 1991A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCESinTHE FACULTY OF GRADUATE STUDIESFaculty of Pharmaceutical SciencesDivision of Pharmacology and ToxicologyTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1993© Subodh Verma, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of1 i17ercor(c /IL Ce.g.The University of British ColumbiaVancouver, CanadaDate^ (”3DE-6 (2/88)ABSTRACTThe general purpose of this thesis was to examine the effect of chronicmetformin treatment on isolated cardiac function in rats made diabetic with STZ andsecondly to explore the effect of metformin treatment on the hyperinsulinemic statusof two rodent models of hypertension, the spontaneously hypertensive rat and thefructose-induced hypertensive rat.In the first set of experiments, the effects of oral metformin administrationwere studied in isolated perfused working hearts from control and diabetic rats.Control (C) and streptozotocin (STZ) diabetic (D) rats were treated for 8 weeks withmetformin hydrochloride. Treatment was initiated at 350 mg/kg/day and wasgradually increased to a dose of 650 mg/kg/day which was maintained over a six weekperiod. Isolated heart performance was assessed under conditions of increasingpreload in order to evaluate the performance of each heart to "stress". Hearts fromuntreated D rats exhibited a depressed response to increases in left atrial fillingpressure from 17.5 to 22.5 cm H2O in terms of left ventricular developed pressure(LVDP), ventricular contractility (+dP/dt) and ventricular relaxation (-dP/dt) whencompared to age matched controls. The diabetic hearts also exhibited a delayed halftime to relaxation (T1/2R) at filling pressures from 15 to 22.5 cm H2O. The functioncurves were performed at a constant heart rate of 300 beats/minute. These responseswere restored to control values in D rats treated with metformin. Metformin treatmentdid not affect these ventricular indices in the C rats. Metformin reduced plasmaglucose levels in the diabetic rats from 24.3 mM to 14.4 mM without any increase inthe plasma insulin level. The D group had higher triglyceride levels than age matcheduntreated C rats and metformin administration in the D rat reduced triglyceride levelsto control values but had no effect in C rats. In conclusion, metformin administrationimproved cardiac performance in STZ-diabetic rats under conditions of increasingpreload.The second series of experiments aimed at exploring the relationship betweenelevated insulin levels and experimental hypertension using metformin as theexperimental intervention. In particular, the effect of chronic metformin treatmentwas studied in a genetic and acquired model of hyperinsulinemia and hypertension,;the Spontaneously hypertensive rat (SHR) and the fructose-induced hypertensive rat.SHR and their genetic normotensive controls, the Wistar-Kyoto (WKY) rat,were procured at 5 weeks (when hypertension is not manifest in the SHR). The ratswere divided into four groups: SHR. untreated, SHR metformin-treated, WKYuntreated and WKY metformin-treated. Baseline measurements of plasma insulin,plasma glucose and systolic blood pressure were performed from week 9 to 13. Atweek 16 insulin implants were surgically implanted in the treated animals to evaluatethe effect of artificially raising plasma insulin and systolic blood pressure weremeasured.The SHR exhibited full blown systolic hypertension by 9 weeks of age whichpersisted throughout the experiment. Paralleling the elevation in systolic bloodpressure, the SHR exhibited sustained hyperinsulinemia by 9 weeks of age.Metformin treatment of the SHR prevented the rise in plasma insulin when comparedto untreated SHR. The treatment also attenuated the systolic blood pressure byapproximately 30-35 mm Hg. Metformin administration did not change the plasmaglucose level in any group. Raising the plasma insulin level in the treated groups viainsulin implants caused an elevation in the plasma insulin level in the SHR-treatedgroup which was accompanied by an increase in systolic blood pressure.The next step was to explore if the insulin-blood pressure relationship in thefructose-induced hypertensive rat, where hypertension is not genetically determinedbut is induced by feeding rats a fructose-enriched diet. Sprague Dawley rats weredivided into four groups: control untreated, control metformin-treated, fructose treatedand fructose-fed metformin-treated. A baseline reading of plasma insulin, plasmaglucose and systolic blood pressure was performed at week 6 following which chronicmetformin treatment was initiated in the treated groups at the same dose as discussedin the previous experiment. At week 7, the animals in the fructose groups were startedon a fructose-enriched diet. Weekly measurements of plasma insulin, plasma glucoseand systolic blood pressure were performed from week seven to eleven.The fructose-fed untreated rats exhibited an elevated plasma insulin level byweek 9 as compared to control rats. This hyperinsulinemia was sustained over theexperimental period. the fructose-fed untreated rats also exhibited an elevation insystolic blood pressure of about 20-25 mm Hg. As seen in the previous study,metformin treatment did not change the plasma glucose levels in any group.Treatment of Sprague Dawley rats with metformin prior to starting fructose, preventedthe rise in plasma insulin when compared to fructose-fed untreated rats. This was alsoaccompanied by a complete prevention in the rise of systolic blood pressure seen inthe fructose fed untreated rats.The fact that a specific drug intervention (metformin) caused a decrease in theplasma insulin level and systolic blood pressure, provides strong support to the notionthat elevated insulin levels contribute at least in part to the development ofhypertension in these two rodent models of hypertension. Further studies addressingthis issue need to be carried out to clearly establish a cause-effect relationship ofinsulin and hypertension.ivTABLE OF CONTENTSABSTRACT^ iiTABLE OF CONTENTSLIST OF TABLES^ viiLIST OF FIGURES viiiLIST OF ABBREVIATIONS^ ixACKNOWLEDGEMENTSDEDICATION^ xiINTRODUCTIONA. Overview^ 1B. Complications of chronic diabetes^ 3C. Chronic diabetes and heart disease 3D. Left ventricular systolic function in diabetes^ 4E. Left ventricular diastolic function^ 4F. Pathogenesis of diabetic heart disease 5G. Models of cardiac dysfunction in diabetic rats^9H. Prevention of diabetes-induced cardiac changes 10I. Metformin^ 10J. Glucose lowering effect of metformin^ 12K. Effect of metformin on glucose utilization 13L. Metformin and lipid metabolism^ 17M. Hyperinsulinemia and hypertension 18N. Hypertension and carbohydrate metabolism^ 190.^Insulin-hypertension:possible mechanisms 20P^Rodent models used to evaluate the interplay betweenhyperinsulinemia/insulin resistance and hypertension^21Q.^Rationale of the proposed experiments^ 24SPECIFIC GOALS OF THE PRESENT INVESTIGATION^26MATERIALS AND METHODS^ 27STUDY 1: EFFECT OF METFORMIN ON ISOLATEDCARDIAC FUNCTION^ 271. Animals and Methods 272. Heart Perfusion^ 283. Plasma analysis 294. Statistical analysis^ 30STUDY 2: EFFECT OF METFORMIN TREATMENT ONTHE HYPERINSULINEMIC STATUS OFRODENT MODELS OF HYPERTENSION^30(A) Spontaneously hypertensive rat study^30Animals and Experimental design 30(B) Fructose-induced hypertension stud^31Animals and Research design 31(C) Methods^321. Blood Pressure Measurement^322. Biochemical Measurement 333. Statistical Analysis^ 33RESULTS^ 34DISCUSSION 65CONCLUSIONS^ 79REFERENCES 81viLIST OF TABLESTables Pages1 General features of IDDM and NIDDM diabetes 22 General characteristics of the rats at terminationin the isolated working heart study 383 Time to half relaxation from 15 to 22.5 cmH2O in the isolated working heart study 454 Food and fluid intake in the SHR study 465 Plasma glucose in the SHR study 476 Plasma insulin levels pre- and post-implant, in the SHR study 547 Plasma glucose levels pre- and post-implant, in the SHR study 558 Systolic blood pressure pre- and post-implant, in the SHR study 569 Food and Fluid intake in fructose study 5710 Plasma glucose in the fructose study 58viiLIST OF FIGURESFigures^ Pages^1^Factors important in the development ofdiabetic cardiac dysfunction^ 7^2^Chemical Structure of Metformin 11^3^Possible links between hyperinsulinemia/insulin resistance and hypertension^ 224^Effect of metformin treatment on +dP/dT inthe isolated working heart study^ 395^Effect of metformin treatment on -dP/dT inthe isolated working heart study^ 416^Effect of metformin treatment on LVDP inthe isolated working heart study^ 437^Body weight in the four experimental groups in the SHR study^488^Plasma insulin in the four experimental groups in the SHR study^509^Systolic blood pressure in the four experimental groups in the SHR study 5210^Body weight in the experimental groups in the fructose study^5911^Plasma insulin in the different groups in the fructose study^6112^Systolic blood pressure in the different groups in the fructose study^63VIIILIST OF ABBREVIATIONSCoA^Coenzyme A-dP/dT^Rate of Left Ventricular Pressure Decline+dP/dT^Rate of Left Ventricular Pressure DevelopmentFFA^Free Fatty AcidsHGO^Hepatic Glucose OutputIDDM^Insulin Dependent Diabetes MellitusLVDP^Left Ventricular Developed Pressuremg^milligramsmin^minutesms^millisecondsSHR^Spontaneously Hypertensive RatsSNS^Sympathetic nervous systemSTZ^StreptozotocinT 1 /2R^Time to half relaxationu^microU^unitsWKY^Wistar-Kyoto RatsACKNOWLEDGEMENTSI am extremely grateful to Dr. John McNeill for not only providing me with anopportunity, but for being an excellent supervisor and an exceptional human being.I would like to thank the members of my supervisory committee, Dr. FrankAbbott, Dr. Jack Diamond, Dr. Kath MacLeod and Dr. David Godin for theirscientific input and advice.My special thanks to Dr. Sanjay Bhanot for his knowledgeable guidancethroughout my project and for being a sincere, trustworthy and truthful friend.My thanks to our laboratory manager, Ms. Mary Battell for her cooperationthroughout my masters. I would also like to thank Dr. Soter Dai for helping me withthe working heart preparation.My special thanks to Ms. Sylvia Chan for helping me with my thesis and forbeing a constant support.I would also like to thank all my laboratory colleagues for their constantsupport and encouragement.DEDICATIONI dedicate this thesisin the fond memory of my fatherwho would have been the happiest man todaymy motherfor leading her children into intellectual pursuitsmy sisterwho makes everything worthwhilemy friendsBaljeet and Sanjaywho were always thereand to timefor allowing me to get this far.xiINTRODUCTIONA. OVERVIEWDiabetes mellitus is a group of syndromes characterized by hyperglycemia,relative insulin deficiency, altered lipid, carbohydrate and protein metabolism and bylong term complications involving the eyes, kidney, cardiovascular system and nerves(Kahn and Shechter 1990). Virtually all forms of diabetes mellitus are due to either adecrease in the circulating concentration of insulin (insulin deficiency) or a decreasein the response of the peripheral tissues to insulin (insulin resistance) in associationwith an excess of hormones with actions opposite to those of insulin (for example,glucagon, growth hormone, cortisol and catecholamines). These hormonal alterationslead to abnormal carbohydrate, lipid, fat and ketone metabolism with the centralfeature of the syndrome being hyperglycemia. Insulin lowers the concentration ofglucose in the blood by inhibiting hepatic glucose production (gluconeogenesis andglycogenolysis) and by stimulating the uptake of glucose into the peripheral tissueslike muscle and fat (Kahn and Shechter 1990; Foster 1991). Most patients can beclassified as having either insulin-dependent diabetes mellitus (IDDM) or non insulin-dependent diabetes mellitus (NIDDM). There are genetic and environmentalcomponents involved in the pathogenisis of both IDDM and NIDDM. Some generalcharacteristics of IDDM and NIDDM diabetes are listed in Table 1.TABLE 1GENERAL FEATURES OF IDDM AND NIDDM DIABETESIDDM^NIDDMGenetic locusAge of onsetBody habitusPlasma insulinPlasma glucagonAcute complicationInsulin therapySulfonylurea therapyChromosome 6<40Normal to wastedLow to absent .High, suppressibleKetoacidosisResponsiveUnresponsiveChromosome I 1 (?)>40ObeseNormal to high .High. resistantHyperosmolar comaResponsive toresistant.ResponsiveFROM FOSTER D.W. 19912B. COMPLICATIONS OF CHRONIC DIABETESDiabetes mellitus is a multifaceted syndrome which affects various systems andorgans. The two major acute complications of insulin deficiency include ketoacidosisand hyperosmolar non-ketotic coma (Carroll and Matz, 1983; Foster and McGarry,1983). The former is a complication of IDDM while the later is generally seen inNIDDM diabetes. More importantly, the diabetic patient has an increased risk ofdeveloping serious long-term pathophysiology that leads to morbidity and prematuremortality. These include cardiovascular alterations, increased risk of coronary arterydisease (CAD), increased frequency of silent myocardial infarction (Brownlee et al.,1988), diabetic retinopathy (Ramsay et al., 1988), nephropathy (Sequist et al., 1989),neuropathy (Mogenesen and Christensen, 1984) and diabetic foot disease (LoGerfoand Coffmann, 1984).C. CHRONIC DIABETES AND HEART DISEASESeveral clinical and experimental studies have shown that the incidence ofcardiac disease is much greater in the diabetic population compared to the non-diabetic population and accounts for almost 80% of all diabetic deaths (Pierce et al.,1988).The causal role of diabetes mellitus in the development of congestive heartfailure was most conclusively demonstrated in the Framingham Study, where therelative risk of developing cardiac failure during the 18-year follow-up was 2.4 timeshigher in male and 5.1 times higher in female diabetic subjects than in the respectivenon-diabetic subjects (McGee et al., 1971; Kannel et al., 1974). More importantly,this greater risk persisted after taking into account the age, blood pressure, weight andserum cholesterol concentrations of the subjects. The Framingham Study also showed3that the association of diabetes and congestive heart failure was confined to IDDMsubjects. However, other studies, such as that of Hamby et al. (1974) showed thatcongestive heart failure is also more prevalent in NIDDM.D. LEFT VENTRICULAR SYSTOLIC FUNCTION IN DIABETESEarly studies using systolic time intervals to evaluate left ventricular systolicfunction at rest revealed a prolonged pre-ejection period (PEP) and a shorter leftventricular ejection time (LVET) and a higher PEP/LVET ratio in diabetic subjects(Ahmed et al., 1975). Similar findings have been reported in several other studies ondiabetic patients who had not been assigned to a specific diagnostic classification ofdiabetes and in patients on insulin treatment or with verified IDDM (Shapiro et al.,1981; Jermendy et al., 1983).As left ventricular function disturbances are more evident upon exercise than atrest, several studies focussed on the performance of diabetic hearts upon exercise.Despite normal left ventricular ejection fraction (LVEF) at rest, Vered et al. (1984)demonstrated an abnormally low LVEF response to exercise in 43% of diabeticpatients but in none of the control subjects. Similar findings have been reported byMildenberger et al. (1984). It is important to note that in most of the studies theabnormal LVEF response to exercise could not be related to the duration of diabetes,metabolic control, autonomic nervous function or diabetic microvascularcomplications.E. LEFT VENTRICULAR DIASTOLIC FUNCTIONDiastolic abnormalities have been extensively studied in diabetes. This ismainly because diastolic filling alterations are one of the earliest signs of left4ventricular dysfunction (Inouye et al., 1984). The diastolic abnormalities observedinclude: an increased atrial contribution to left ventricular filling, prolongedisovolumic relaxation time and slowing of the thinning of the left ventricular wall (forreview see Matti et al., 1990).The abnormal diastolic function of the diabetic heart is indicative of diminishedleft ventricular compliance and prolonged left ventricular relaxation. Support for theview that chamber stiffness (decreased ventricular compliance) contributes to diastolicdysfunction in the diabetic heart is provided by a study in which 12 NIDDM patientswith chest pain but no significant coronary artery narrowing were catheterized andstudied. An elevated left ventricular end diastolic pressure and a reduced strokevolume were noted in these patients supporting the view of a decreased ventricularcompliance in diabetes (Regan et al., 1977).F. PATHOGENESIS OF DIABETIC HEART DISEASEThe pathogenesis of diabetic heart disease is believed to be due to severalfactors and multiple mechanisms (Giles and Sander, 1989; Rodrigues and McNeill,1992). A summary of the established and potential factors important in thedevelopment of cardiac dysfunction in diabetes are illustrated in Figure 1. Until 20years ago, it was assumed that the increased risk of congestive heart failure in thediabetic population reflected the increased incidence of coronary atherosclerosis andmyocardial infarctions in these patients. However, beginning with the study of Rubleret al. (1972) clinical, epidemiological and pathological data have mounted to supportthe existence of a specific cardiomyopathy in the absence of coronary artery disease,macroangiopathy, microangiopathy, autonomic neuropathy or hypertension in thediabetic population (for review see Zarich and Nesto, 1989). Clinical evidence for theexistence of diabetic cardiomyopathy was provided by Hamby et al.(1974), who noted5an increased incidence of diabetes in patients with idiopathic cardiomyopathy.Sixteen of 73 patients with idiopathic cardiomyopathy were diabetic, compared withonly 11% in an age and sex-matched cohort without cardiomyopathy. More importantwas the finding that only 1 out 16 diabetic patients with cardiomyopathy was insulin-dependent. This study coupled with the Framingham study clearly demonstrated thatdiabetes per se, either IDDM or NIDDM, is associated with cardiac abnormalities.Regan et al. (1983) described angiographic and hemodynamic findings in a group of17 patients with NIDDM in the absence of hypertension or valvular disease. D'Elia etal. (1979) found diastolic and systolic dysfunction in 59% of a diabetic cohort withrenal failure in the absence of coronary artery disease. The fact that diabetic heartdisease can occur both in IDDM and NIDDM despite their divergence in respect to thepathogenisis of hyperglycemia, suggests that raised blood glucose and/or metabolicchanges associated with hyperglycemia probably have a key role in the pathogenesisof diabetic heart muscle disease.Not only has cardiac disease in diabetes been described in clinical settings, it isnow well characterized in animal models of chemically induced diabetes. Isolatedworking hearts from streptozotocin (STZ)- or alloxan-induced diabetic rats show adecreased ability to respond to increases in left-atrial filling pressure in terms of leftventricular developed pressure (LVDP), rate of contraction (+dP/dt), rate of relaxation(-dP/dt), cardiac and aortic output (Vadlamudi et al., 1982). The lack of ability of thehearts to respond to increases in filling pressure suggests that the diabetic heart,although capable of functioning like controls under normal conditions, does notadequately respond to "stress" simulated by increases in preload. A diminished abilityof diabetic hearts to respond to increases in afterload has also been demonstrated(Ingebreston et al., 1980, Nichol et al., 1992). Moreover, isolated papillary musclesfrom STZ-induced diabetic rat hearts exhibit a depressed velocity of shortening and adelayed onset of relaxation (Fein et al., 1980). Regan et al. (1974) using one-year old6FIGURE 1. FACTORS IMPORTANT IN THE DEVELOPMENT OFDIABETIC CARDIAC DYSFUNCTIONGENETIC^ENVIRONMENTALDIABETES MELLITUSHyperglycemiaGlycosylation Polyol r- HyperlipidemiaDecreased T, of Proteins PathwayAltered1Myosin Isozyme Shift Altered^Increased Cardio- Increased^Phospholipid/From V, to V, Membrane^CollagenComponents Cross-linkingNeuropathy LCAC^CholesterolRatioDepressed MyosinATPase I^I Increased MVO,DIABETIC CARDIOMYOPATHYTAKEN FROM: Giles T.D. et al., 19897alloxan-diabetic dogs, showed ventricular stiffness associated with shortening of leftventricular ejection time. In isolated papillary muscle from rabbits made diabetic withalloxan, Fein et al. (1985) found a prolonged duration of isotonic and isometriccontraction.In addition to the factors depicted in Figure 1, in the last few years considerableattention has been directed towards the role of altered myocardial energetics in thedevelopment of diabetic cardiomyopathy (For review see Rodrigues and McNeill,1992). Under normal conditions, an estimated 60-70% of myocardial energy isderived from the metabolism of lipids; the remainder is derived from non-lipid sourcesincluding carbohydrates, ketone bodies and amino acids. In diabetes, the circulatinglevels of free fatty acids (FFA) are increased as determined by their mobilization andsynthesis by the adipose tissue and liver. As a result, there is an increased uptake,oxidation and storage of fatty acids by the myocardial cell. Thus, there is an almostexclusive dependence on FFA oxidation as a source of myocardial energy, whichincreases from about 60-70% to about 90% in diabetes. Elevated FFA oxidation leadsto a build up of intermediates such as long chain acyl carnitines which havedeleterious effects on the myocardial cell through a variety of mechanisms.Moreover, the inability of the diabetic heart to utilize glucose leads to an elevatedoxygen demand per molecule of ATP produced. Support for this hypothesis comesfrom studies in which perfusion of diabetic rat hearts with dichloroacetate (a glucoseoxidation stimulator) was shown to acutely reverse the depression of cardiac function.Thus, it is now believed that the inability of the diabetic heart to utilize glucose as asubstrate for energy production and the total reliance on FFA oxidation as a source ofATP may be a potential factor important in the development of diabeticcardiomyopathy (for review see Rodrigues and McNeill, 1992).Several studies from our laboratory point towards alterations in lipidmetabolism as being an important determinant of cardiac dysfunction in diabetes. For8example, in a study in which Wistar and Wistar-Kyoto (WKY) rats were injected withidentical doses of STZ, although both groups of rats exhibited elevated plasmaglucose levels, only the Wistar rats exhibited elevated levels of circulating lipids.Interestingly, depression of the myocardial function was seen in the group of Wistarrats while the cardiac function in the STZ-injected WKY rats remained unaffected(Rodrigues and McNeill, 1986). In another study, STZ-diabetic rats treated withhydralazine showed elevated blood glucose but normal circulating lipids. Thefunction of the hearts from diabetic rats treated with hydralazine was similar to that ofnon-diabetic controls, supporting the notion that hyperlipidemia may be an importantdeterminant of cardiac disease in the diabetic rat (Rodrigues et al. 1986).G. MODELS OF CARDIAC DYSFUCTION IN DIABETIC RATSTwo chemicals have been extensively used to produce diabetes in laboratory .animals: alloxan and streptozotocin (STZ). Both these agents produce beta cellnecrosis after a single dose in laboratory animals, thereby causing markedhyperglycemia and hypoinsulinemia, the severity of which can be varied by alteringthe dose of the agent. The methylnitrosourea analog STZ has now largely replacedalloxan as a primary agent for the induction of diabetes in animals because of itshigher selectivity for the B-cells of the pancreas and longer half life (Rakienten et al.,1963).Animals injected with STZ exhibit typical signs of diabetes, includingpolydypsia, polyphagia, polyuria, decreased body weight gain, hyperglycemia,hypoinsulinemia, hyperlipidemia and depressed cardiac performance. The depressionof the cardiac function is noticed as early as 6-8 weeks after the induction of diabeteswith STZ or alloxan (McNeill and Tahiliani, 1986)9H. PREVENTION OF DIABETES-INDUCED CARDIC CHANGESInsulin treatment is successful in preventing or retarding the cardiacabnormalities in diabetes. The effectiveness of in vivo insulin treatment of STZdiabetic rats on various functional and biochemical cardiac parameters has beendemonstrated (Tahiliani and McNeill, 1986; Dillmann 1980; Lopaschuck et al., 1983).Insulin treatment of diabetic rats seems to be less effective in preventing and reversingcardiac alterations in the more chronic stages (five months) versus the early stages (6-8 weeks). Treatment of 5 month old diabetic animals with insulin was successful incontrolling plasma glucose however, only a partial reversal of cardiac performancewas noticed (Tahiliani et al., 1983). In a chronic canine model of diabetes, cardiacfunction was unaffected by insulin control of plasma glucose (Regan et al., 1981). Inconjunction with these experimental studies, myocardial abnormalities have also beenreported in clinical settings despite insulin treatment and tight glucose control (TheUniversity Group Diabetes Program, 1975). It would, therefore, be desirable to havedrug treatments, in addition to insulin, that have (a) an insulin-like/insulin-enhancingeffect and (b) a lipid-lowering effect, to prevent the diabetes-induced myocardialalterations.I. METFORMINThe biguanides, metformin and phenformin, were introduced in 1957 asglucose lowering agents. Phenformin initially received greater use but was withdrawnin many countries during the 1970's because of its association with lactic acidosis.Metformin is now accepted as the biguanide of choice. It is used extensivelyworldwide, except in the U.S., where it is undergoing clinical trials. Metformin isreadily distinguishable from sulfonylureas because it ameliorates hyperglycemia1 0FIGURE 2. CHEMICAL STRUCTURE OF METFORMINNH NHII^II(CH3)2 N-C-NH-C-NH21 1without stimulating insulin, promoting weight gain, or causing clinical hypoglycemia(Bailey, 1985). The chemical structure of metformin is given in Figure 2.Absorption of metformin occurs mainly from the small intestine. Theabsorption half life of the drug is 0.9-2.6 h and the bioavailability is about 50-60%.Concentrations of metformin in peripheral plasma reach a maximum of about 2 ug/mlabout 2 hours after an oral dose. Metformin is stable and does not bind to the plasmaproteins. The drug is excreted in the urine apparently unchanged. The elimination israpid, with about 90% being cleared in 12 hours (Tucker et al., 1981).J. GLUCOSE LOWERING EFFECT OF METFORMINMetformin is better described as an antihyperglycemic rather than anhypoglycemic, as it rarely causes hypoglycemia (Bailey 1985, 1988). Early studiesshowing the antihyperglycemic effect of metformin have been thoroughly reviewed byHermann (1979). Metformin appears to be as effective in non-obese as in obeseNIDDM subjects (Clarke and Duncan 1968; Clark and Campbell, 1977). The drug isgiven in combination with sulfonylureas to patients in which sulfonylureas alone donot achieve an acceptable level of glycemic control or in IDDM patients inconjunction with insulin.The antihyperglycemic effect of metformin cannot be attributed to an increasein insulin concentration. Basal insulin concentrations and concentrations after an oralglucose load are typically unchanged or slightly reduced during metformin therapy(Lord et al., 1983; Prager and Schernthaner, 1983; Fantus and Brosseau, 1986;Barzilai and Simonson, 1988; Wu et al., 1990; Prager et al., 1986; Rizkalla et al.,1986; Campbell et al., 1987; Nosadini et al., 1987; Pederson et al., 1989; Benzi et al.,1990). Similar observations have been made in animal models of obese and non-obese hyperinsulinemic and hypoinsulinemic diabetes (Lord et al., 1983; Penicaud et12al., 1989; Rossetti et al., 1990). The drug has little or no effect on the secretion ofglucagon, somatostatin, growth hormone or cortisol (Penicaud et al., 1989).The main mechanism of the antihyperglycemic effect of metformin is thoughtto be via an enhancement of insulin-mediated glucose utilization in the peripheraltissues such as the muscle and fat. The effect of metformin on basal hepatic glucoseproduction (HGP) has also been extensively evaluated, and it is observed that amodest decrease in basal hepatic glucose production (HGP) is seen when metforminproduces a substantial decrease in basal glycemia, which may involve a modestreduction in gluconeogenesis (for review see Bailey, 1992).K. EFFECT OF METFORMIN ON GLUCOSE UTILIZATIONIn the basal state, glucose utilization is largely independent of insulin and themain tissues utilizing glucose are the brain, blood cells, kidney medulla, intestine andskin (Felig et al., 1990). Metformin does not affect aerobic or anaerobic glucoseutilization by the brain, kidney medulla or skin of normal mice (Wilcock and Bailey,1990). Also basal glucose utilization by hepatocytes of normal and STZ-treated ratsand mice was not affected by therapeutic concentrations of metformin. The intestine,which accumulates much higher concentrations of metformin than other tissues, hasrecently been identified as an important site of metformin-stimulated glucoseoxidation. In obese fa/fa rats, metformin (350 mg/kg/day) increased basal glucoseutilization in the jejunum (Penicaud et al., 1989). Administration of metformin (250mg/kg/day) to fasted rats also increased anaerobic glycolysis by the intestine (Baileyet al., 1989).The antihyperglycemic effect of metformin is evident after a glucose challenge(Lord et al., 1983; Prager and Schernthaner, 1983; Fantus and Brosseau, 1986;Barzilai and Simonson, 1988; Frayn et al., 1971; Wu et al., 1990). Glucose disposal is13shared by the liver and peripheral tissues, particularly muscle. Glucose uptake by theliver is an insulin-independent process and studies have shown that metformin haslittle effect on hepatic glucose uptake after an oral glucose challenge (Bailey, 1992).Diabetes is characterized by an increased hepatic glucose output (HGO) which isaccounted for mainly by an increased gluconeogenesis (Consoli et al., 1989). Severalstudies have investigated whether metformin can supress gluconeogenesis. Metforminreduced gluconeogenesis in normal guinea pigs (Meyer et al., 1967). Highconcentrations of the drug reduced basal and glucagon-stimulated gluconeogenesis byanimal liver and kidney in vitro in the absence of added insulin (Meyer et al., 1967;Alengrin et al., 1987). A synergistic reduction in gluconeogenesis by isolated rathepatocytes for a range of substrates (lactate, pyruvate, glutamine, alanine, glycerol)was observed when physiological concentrations of insulin were added to therapeuticconcentrations of metformin (Wollen and Bailey, 1988). In vivo studies are notconsistent with the antigluconeogenic effect of metformin. For example, in a study inwhich 14C-lactate was administered to rats via the hepatic portal vein, the appearanceof 14C_ glucose in the plasma was not affected by metformin (Bailey 1992). It isspeculated that as metformin increases the supply of gluconeogenic substrate to theliver in the form of lactate produced by the intestine, it is probable that even a smallincrease of gluconeogenic substrate can override the inherent antigluconeogenicaction of the drug. This has been proposed as an important mechanism that preventsthe drug from causing clinical hypoglycemia because the supply of lactate from theintestine to the liver ensures that gluconeogenesis is not critically impaired (Bailey1992)The fact that metformin improves glucose tolerance without increasing insulinsecretion suggests that the drug may enhance insulin-mediated glucose disposal.Several lines of evidence now lend support to this concept. Prager et al (1986)showed a 23% increase in glucose disposal in obese and non-obese NIDDM subjects14receiving metformin. A similar increase (43%) has been reported in obese NIDDMsubjects receiving metformin (Nosadini et al., 1987). Data from experimental studiesare consistent with clinical findings. During a steady-state hyperglycemic-hyperinsulinemic clamp in normal rats, 250 mg/kg/day metformin increased glucosedisposal by 20% (Bailey, 1992). Similar findings have been reported in mildlydiabetic STZ rats given metformin (Rossetti et al., 1990). Studies documenting areduction in insulin requirement by metformin in overweight insulin-treated NIDDMand IDDM patients lends further support to the idea that metformin'santihyperglycemic effect is mediated via an improvement in insulin action (Leblanc etal., 1987). Of importance is the study by Gin et al. (1982) in which addition ofmetformin (1.7g/day for 2 days) reduced postprandial insulin requirement of IDDMpatients by 26%. Gin et al (1985) also showed that IDDM patients receivingmetformin supplement (1.7 g/day for 7 days) had an 18% improvement in glucoseuptake assessed during a euglycemic hyperinsulinemic clamp.Two major insulin responsive peripheral tissues include the muscle and theadipose tissue, the skeletal muscle being quantitatively the major site of insulinmediated glucose uptake. Several studies have consistently demonstrated thatmetformin improves insulin-stimulated glucose uptake in the muscle of both humansand animals (for review see Bailey, 1992). Soleus muscle isolated from STZ-induceddiabetic mice treated with metformin at a dose of 250 mg/kg/day showed a 20%increase in insulin-stimulated glucose uptake and oxidation (Bailey and Pauh, 1986).When metformin was incubated with hemidiaphrams of alloxan-induced diabetic rats,insulin-mediated glucose uptake was increased by 27%, suggesting that metformincan directly act on diabetic muscle (Frayn and Adnitt, 1972). In muscle strips isolatedfrom insulin-resistant humans, metformin increased insulin-mediated glucose uptake(Galuska et al., 1991).The effect of metformin on insulin-mediated glucose utilization in adipose15tissue remains unclear. In normal rat adipocytes, metformin increased basal andinsulin-stimulated glucose uptake by 19-43% (Jacobs et al., 1986; Matthaei et al.,1991). This increase was associated with an increase in insulin-induced translocationof glucose transporters from microsomes into the plasma membrane. In a study inwhich adipose tissue biopsies from normal human subjects were utilized, metformin(2 ug/ml) increased insulin-stimulated glucose oxidation and incorporation intotriglyceride (Cigolini et al., 1984); however, a similar study failed to detect any effect(Pedersen et al., 1989). Also, in a study using adipocytes from obese NIDDM subjectstreated with metformin, no alterations in glucose transport, oxidation or lipogenesiscould be seen in spite of clear evidence that the drug decreased glycemia (Pedersen etal., 1989). Thus, it appears that the chief mechanism through which metforminenhances insulin action is via an increase in insulin mediated glucose utilization inmuscle.The cellular mechanism of action of metformin could potentially involveinsulin receptor and/or post-receptor effects. There appears to be a poor correlationbetween the effects of metformin on insulin receptor binding and glucose metabolism.There are many reports that metformin can improve glucose homeostasis in diabeticstates without any alteration in insulin binding (Prager and Schernthaner, 1983; Fantusand Brosseau, 1986). Moreover, in view of the large number of spare receptors,increased binding does not appear to have a significant effect on glucose homeostasisat normal circulating insulin concentrations. It thus appears that effects of metformindistal from the receptor may play a significant role in defining metformin's insulin-enhancing ability.16L. METFORMIN AND LIPID METABOLISMMetformin therapy is associated with a decrease in circulating triglycerideconcentrations in non-diabetic and NIDDM patients. Reductions of 10-20% are oftenfound in non-hypertriglyceridemic subjects and up to 50% in patients with elevatedtriglycerides (Gustafson et al., 1971; Fedele et al., 1976; Sirtori et al., 1977; Sirtori etal., 1984; Montaguti et al., 1979; Janka, 1985). The decrease in the triglycerides isbelieved to be via a decrease in very low density lipoproteins (VLDL-triglyceride)(Zavaroni et al., 1984).In animals with hypertriglyceridemia, a significant hypocholesterolemic effectcan be noted (Sirtori et al., 1977). This effect on cholesterol is not evident in animalswith lipid parameters within the normal range (Billingham et al., 1980). The decreasein cholesterol is believed to be via a decrease in low density lipoprotein (LDL)cholesterol or VLDL cholesterol. In patients, metformin treatment has beenassociated with a small decrease in cholesterol (10%) in non-diabetic and NIDDMsubjects (Wu et al., 1990). The effect of metformin treatment on high densitylipoprotein cholesterol concentrations is unclear. In one study, metformin modestlyincreased (10%) HDL concentrations in non-diabetic and NIDDM subjects (Sirtori,1988). Other clinical and animal studies do not support this finding (Rains et al.,1988; Billingham et al., 1980). In alloxan-diabetic rats, metformin administration ledto a decrease in hydroxymethyl-glutaryl-CoA (HMG CoA) reductase activity,suggesting that the drug may decrease cholesterol biosynthesis in these rats.In the past several years, there has been an increased awareness thatabnormalities in lipid and lipoprotein abnormalities contribute to the prematuredevelopment of arterial vascular disease. In conjugation with this, a substantial bodyof evidence now implicates peripheral hyperinsulinemia as an important risk factor forarterial vascular disease. Bearing in mind the ability of metformin to improve17glycemia without increasing insulin concentrations and the beneficialantihypertriglyceridemic effect, one can observe a potentially favorable profileemerging that might help to delay or ameliorate some chronic complications ofdiabetes.M. HYPERINSULINEMIA AND HYPERTENSIONEssential hypertension in man is a multifactorial condition in whichenvironmental influences act upon an unknown number of genetic factors. Althoughhypertension remains the major risk factor for the development of coronary arterydisease (CAD), it is difficult to demonstrate an improvement in morbidity andmortality rates from CAD despite successful programs to pharmacologically controlhypertension (Reaven et al., 1991). In the past several years, a growing wealth of datasuggest that a subset of hypertension is associated with metabolic abnormalitiesinvolving hyperinsulinemia and/or insulin insensitivity, obesity and lipidabnormalities (for review see Reaven, 1988; Zavaroni et al., 1987). Of particularconcern is the finding that these metabolic abnormalities persist with conventionalantihypertensive therapy like thiazides and 13-blockers (Skarfors et al., 1989; Pollare etal., 1989; Weinberger 1986; Frishman, 1988). Central to this concept, evidencerecently has accumulated showing that hyperinsulinemia, suggestive of reducedinsulin sensitivity or insulin resistance, is frequently found in essential hypertensivepatients even when they have normal body weight and glucose tolerance. Evidence,both from clinical and experimental studies, now suggests that hyperinsulinemia andinsulin resistance may be central in the development of hypertension, dyslipidemiaand atherosclerosis (Reaven, 1991; DeFronzo and Ferrannini, 1991; Ferrannini et al.,1991; Haffnner et al., 1992).18N. HYPERTENSION AND CARBOHYDRATE METABOLISMParillo et al. (1988) investigated differences in carbohydrate metabolismbetween normotensive and hypertensive subjects using an oral glucose tolerance test.They found that hypertensive subjects had a higher serum glucose level and seruminsulin levels 60-180 minutes after ingestion of oral glucose than did normotensivesubjects. Perhaps the earliest publication demonstrating the presence of higher thannormal plasma insulin concentrations in patients with high blood pressure was that ofWelborn et al. (1966). They studied 19 patients diagnosed as having essentialhypertension, 10 of whom were not being treated, and found that the group with highblood pressure had significantly higher plasma insulin concentrations. Moreimportantly, the hyperinsulinemia was noted before, and at every time point measuredafter an oral glucose load, and was found in both the treated and untreated patientswith high blood pressure. Modan et al. (1985) in a survey of about 2500 patientsfound that hypertensive patients exhibited fasting and postprandial hyperinsulinemiaindependent of obesity, age, or magnitude of glucose tolerance. Several clinical andepidemiological studies confirm these observations (for review see Kannel et al.,1991; Ferrannini and Natali, 1991). More direct evidence for the involvement ofhyperinsulinemia and hypertension comes from a study in which physical exercisetraining of obese hyperinsulinemic subjects led to a decrease in insulin level and bloodpressure without a change in weight (Rocchini et al., 1989).Not only have elevated insulin levels been documented in human essentialhypertensives, they have also been shown in rodent models of hypertension. Theinterplay between elevated insulin levels and hypertension in rats is best elucidated inthe fructose-fed hypertensive rat model. In this model insulin resistance,hyperinsulinemia and hypertension are induced in normal Sprague Dawley rats byfructose feeding (Hwang et al., 1987). In these rats, exercise training resulted in an19improvement in insulin sensitivity and a resultant decrease in blood pressure (Reavenet al., 1988). Furthermore, somatostatin administration led to a decrease inhyperinsulinemia and an attenuation in blood pressure in these rats (Reaven et al.,1989). Further support for the involvement of insulin in the development ofhypertension comes from data obtained in spontaneously hypertensive rat (SHR) inwhich hyperinsulinemia precedes the development of hypertension (Reaven andChang, 1991).In summary, data from clinical, epidemiological and animal studies areconsistent with the hypothesis that changes in insulin metabolism are related tohypertension. The next necessary question to ask is how do insulin levels lead tohypertension ?0. INSULIN - HYPERTENSION: POSSIBLE MECHANISMSIn an effort to maintain normal glycemia in the presence of insulininsensitivity, the pancreatic beta cells try to offset the insulin resistance by secretingmore insulin. In other words, the trade off in maintaining euglycemia in the presenceof insulin insensitivity is hyperinsulinemia. High insulin levels can lead tohypertension through several mechanisms. Elevated insulin levels could lead tosympathetic stimulation and hypertension (Rowe et al., 1981; Liang et al., 1982;Young, 1988). Another possible link between hyperinsulinemia and hypertensioninvolves the effect of insulin on the handling of sodium and water by the kidney.There is evidence that insulin can act to promote renal tubular sodium resorption inman (DeFronzo et al., 1985). More recently it has been demonstrated that insulin actsat the level of the proximal tubule to increase volume reabsorption (Mondon andReaven, 1988). Thus, elevated insulin levels have the potential to cause volumeoverload and high blood pressure. In addition, elevated insulin levels can increase20intracellular sodium and calcium concentrations (Mahnensmith and Aroson, 1985;Canessa et al., 1987; Wiedmann et al., 1985), which in turn may lead to increasedvascular reactivity to pressor amines and finally may cause an increase in vascularsmooth muscle proliferation (Stout, 1990). Another suggested mechanism involvesresistance of the vascular smooth muscle cells to the vasodilatory actions of insulin,i.e. in the face of insulin insensitivity/ insulin resistance the vascular smooth muscle(VSM) is more responsive to the actions of vasoconstrictor amines. This would inturn lead to an increase in peripheral vascular resistance and an increase in bloodpressure (Sowers et al., 1991; Zemel et al., 1991). Figure 3 illustrates the possiblelinks between hyperinsulinemia/insulin resistance and high blood pressure. All ofthese events may potentially act independently or in an additive fashion to lead tohypertension.P. RODENT MODELS USED TO EVALUATE THE INTERPLAYBETWEEN HYPERINSULINEMIA /INSULIN RESISTANCEAND HYPERTENSIONThe Spontaneously Hypertensive Rat (SHR) The SHR has a genetic propensity to develop hypertension. These rats havealso been shown to be insulin-resistant and hyperinsulinemic when compared to theirgenetic controls, the Wistar Kyoto rats (WKY) (Mondon and Reaven, 1988).Studies on adipocytes isolated from SHR indicate that insulin-stimulated glucoseuptake is lower when compared to adipocytes isolated from WKY rats (Reaven et al.,1989). Reaven and Chang (1991) found a positive correlation (r=0.6) between thedegree of insulin resistance in isolated adipocytes and the degree of blood pressure. Asignificant relationship (P<0.001) was seen between plasma insulin concentration andblood pressure in the same rats. Although two recent reports do not support the21FIGURE 3. POSSIBLE LINKS BETWEEN HYPERINSULINEMIA /INSULIN RESISTANCE AND HYPERTENSIONEnvironmental Factors Genetic FactorsSelective Insulin ResistanceHyperinsulinemiaAltered VascularFunctionAltered IonFluxSodiumRetentionSNST ActivityHypertensionFROM: ROCCHINI, 199222presence of insulin resistance in the SHR, the presence of hyperinsulinemia wasconfirmed even in those studies (Buchanan et al., 1992; Buchanan et al., 1992). It hasbeen proposed that nutrient-stimulated hyperinsulinemia may play a role in thedevelopment and regulation of blood pressure in the SHR (Buchanan et al., 1992).These studies support the notion that hyperinsulinemia may play a role in theregulation of blood pressure in this animal model.The Fructose-Fed Hypertensive RatZavoroni et al. (1980) showed that substituting fructose for the carbohydrateconventionally present in rat chow led to insulin resistance and hyperinsulinemia inSprague Dawley rats independent of obesity. More recently Hwang et al. (1987)demonstrated that this dietary manipulation also led to an increase in blood pressureby about 20 mm Hg in these animals. Furthermore, administration of clonidine tothese rats inhibited the hypertension but did not improve the associated metabolicdefects, suggesting that the defects in glucose- metabolism may not be secondary to anincrease in sympathetic outflow. Fructose-induced hypertension is not accompaniedby any changes in plasma renin or angiotensin levels (Hwang 1989); however, it ischaracterized by elevated atrial natriuretic peptide levels and decreased plasmaaldosterone levels suggestive of volume overload (Hwang, 1989). Ashyperinsulinemia can potentially lead to sodium and water retention, it is possible thatfructose hypertension is caused by volume overload secondary to hyperinsulinemia.In contrast to the SHR, the fructose-induced hyperinsulinemic, insulin resistant andhypertensive rat represents an acquired form of hypertension.23Q. RATIONALE OF THE PROPOSED EXPERIMENTSThe general purpose of this thesis was to examine the effect of metformintreatment on isolated cardiac responses in STZ-diabetic rats and secondly to explorethe effect of metformin treatment on the hyperinsulinemic status of two rodent modelsof hypertension, the SHR and the fructose-induced hypertensive rat.In the first set of experiments, we hypothesized that if metformin improvesglucose homeostasis and has beneficial antihypertriglyceridemic actions, it shouldprevent the development of cardiac dysfunction in the diabetic rat. The experimentaldesign involved treatment of STZ-induced diabetic rats with metformin hydrochlorideand measurement of isolated cardiac function under conditions of increasing preload.Four indices of ventricular function were evaluated, i.e. rate of contraction (+dP/dt),rate of relaxation (-dP/dt), left ventricular developed pressure (LVDP) and time to halfrelaxation (T1/2R). Recent data indicate that insulin itself may be incapable ofpreventing cardiac disease in the chronic diabetic. In light of these findings, ifmetformin treatment could prevent the development of functional cardiacabnormalities, it may serve as a novel metabolic approach for the treatment of seriousand frequently fatal cardiovascular complications of chronic diabetes.Our next series of experiments were aimed at exploring the relationshipbetween elevated insulin levels (hyperinsulinemia) and experimental hypertension.We hypothesized that if defects in carbohydrate metabolism manifested ashyperinsulinemia/insulin insensitivity were responsible for the development andregulation of hypertension, drugs which improve insulin sensitivity (e.g. metformin)and decrease insulin levels may potentially ameliorate hypertension. Particularlyrelevant was the study by Landin et al. (1991) in which metformin treatment of non-obese, non-diabetic untreated hypertensives led to improved glucose tolerance andlower insulin levels. Surprisingly, the treatment also lowered their blood pressure. In24light of this recent and interesting finding, we decided to examine the effect ofmetformin treatment on (a) a genetic model of hyperinsulinemic hypertension, theSHR and (b) an acquired model of hyperinsulinemia and hypertension, the fructose-fed hypertensive rat. In an effort to establish causality between insulin levels andhypertension, we proposed to examine the effect of artificially raising the plasmainsulin (via insulin implants) and studying the resultant effect on blood pressure inSHR animals treated with metformin. If metformin lowers plasma insulin and bloodpressure in SHR animals, then artificially increasing plasma insulin levels shouldincrease blood pressure. As mentioned earlier, the aim of these studies was directedtowards providing insight into the role of hyperinsulinemia in the development ofexperimental hypertension and secondly to examine the potential antihypertensiveeffect of metformin.25SPECIFIC GOALS OF THE PRESENT INVESTIGATION1. To study the effect of chronic treatment of STZ-induced diabetic rats withmetformin on various indices of ventricular function, i.e., +dP/dt, -dP/dt,LVDP and Ti/2R.2. To examine the effect of metformin treatment on the hyperinsulinemic status ofthe SHR and the fructose-fed hypertensive rat.3. To study the effect of metformin treatment on the hypertensive status of theSHR and the fructose-fed hypertensive rat.4. To evaluate the inter-relationship between hyperinsulinemia and hypertensionby attempting to correlate changes in plasma insulin with changes in bloodpressure in these two models of experimental hypertension andhyperinsulinemia.5.^To evaluate the effect on blood pressure of artificially raising plasma insulinlevels in the SHR.26MATERIALS AND METHODSSTUDY 1: EFFECT OF METFORMIN TREATMENTON ISOLATED CARDIAC FUNCTION1.^Animals and MethodsForty four male Wistar rats (175-200g) obtained locally were used in thisstudy. The rats were divided into two groups. One group received a single tail veininjection of streptozotocin (STZ, obtained from Sigma, St. Louis, MO) at a dose of50mg/kg and served as the diabetic group. Previous studies from our laboratory haveindicated that animals so treated develop hyperglycemia, hypoinsulinemia anddepressed cardiac function. The other group was injected with 0.9% sodium chloridesolution and served as age-matched controls. The rats injected with STZ were checkedfor hyperglycemia at 48 hours and 72 hours: Only those rats with plasma glucosevalues greater than 15 mM were included in the study. The control and diabeticgroups were then further divided into two groups: untreated control (n=10),metformin-treated control (n=10), untreated diabetic (n=12) and metformin-treateddiabetic(n=12). The rats were housed two to three per cage and received food andwater ad libitum. The treated groups received metformin hydrochloride (a gift fromNordic Pharmaceuticals Inc.) dissolved in the drinking water. Metformin treatmentwas initiated at a dose of 350 mg/kg/day and was gradually increased over a two weekperiod to a final dose of 650 mg/kg/day. The treated groups received this final dosefor a 6 week period. At the end of 8 weeks, hearts were isolated and perfused asdescribed in the next section. At termination, whole blood (arterial and venous) wascollected in heparinized tubes and the plasma separated by centrifugation (3000Xg)27for 10 minutes. The plasma samples were stored at -80C until analyzed for glucose,insulin and triglyceride.2.^Heart PerfusionHearts were removed from control and diabetic rats anesthetized with sodiumpentobarbital and perfused by the Langendorff procedure for approximately 5 minutesand subsequently perfused in the working heart mode as previously described byRodrigues et al. (1988). The hearts were perfused with oxygenated (95% Oxygen, 5%CO2) Chenoweth-Koelle (CK) buffer maintained at 37±1 C. The CK buffercomposition was : NaC1 120 mM, KCl 5.6 mM, CaC12 2.18 mM, MgC12 2.1 mM,NaHCO3 19 mM, glucose 10 mM. The left ventricular developed pressure (LVDP)was measured using a Staham P23 AA transducer (Statham Gould) attached to a 3 cmpiece of polyethylene glycol (PE-90) tubing which in turn was attached to a 20 gaugeneedle which was inserted into the left ventricle through the apex of the heart.Cardiac work was initiated by switching the perfusionfrom the Langendorff to theworking heart mode. Briefly, in the working heart mode, the CK buffer entered theleft ventricle through the left atrium and was pumped out through the aorta. Theaortic outflow was subjected to an afterload of a 75-mm column of water. The leftventricular pressure and first derivative of the left ventricular pressure was recordedon a Grass model 79D polygraph. A platinum electrode from a Grass Model SD9Dstimulator was connected to the left atrium of each heart and stimulated at twice thethreshold voltage with square wave pulses of 5 ms to give a rate of 300 beats/minute.The cardiac function data [left ventricular developed pressure (LVDP), rate of forcedevelopment (+dP/dt) and rate of relaxation (-dP/dt)] were collected and analyzedusing a computer program. The pressure transducer signal from the polygraph wassampled at 667 Hz over 1.5 sec at each function curve. This resulted in data being28collected for six complete cardiac pulses. Three of the six were analyzed with curvefitting techniques in order to determine pulse height, area, start and finish. The valuesof these three pulses were averaged to give the data at that point. This was allcalculated by the computer program. Each heart was equilibrated at 15 cm H2O for10-15 minutes before the function curves were performed. The function curves wereperformed by estimating the left ventricular function against varying left atrial fillingpressures. The filling pressures were altered by changing the height of the atrialfilling reservoirs from 7.5 to 22.5 cm in 2.5 cm increments. The filling pressure wasfirst reduced stepwise from 15 to 7.5 cm H2O after which it was increased stepwise to22.5 cm and finally decreased to 15 cm H2O. At each point, the pressure developedwas allowed to stabilize before it was recorded. In general, stable pressuredevelopment was achieved within 2 minutes after the left atrial filling pressure wasaltered. A complete function curve required about 20-30 minutes. Time to halfrelaxation (Tu2R) was calculated manually from each hard copy of the tracing. Aperpendicular line was drawn from the peak of the function curve intersecting thebaseline of the tracing. The horizontal line from the midpoint of this perpendicularline was drawn to intersect the pressure decline portion of the curve and the distancewas noted in millimeters (mm). With the help of the chart speed the distance in mmwas converted into milliseconds (ms), and expressed as T112R.3. Plasma AnalysisThe plasma samples were analyzed for glucose and triglycerides usingdiagnostic kits from Boehringer Mannheim Diagnostic (Dorval, Quebec, Canada).Plasma insulin was assayed using a double antibody radioimmunoassay kit from ICNBiomedicals, Costa Messa ,CA, USA..294.^Statistical Analysis:Values are indicated as mean±S.E. of the mean. "n" indicates the number ofanimals in each group. Statistical analysis was performed with a two way analysis ofvariance followed by a Newman-Keuls test. A probability of P<0.05 was taken as thelevel of statistical significance.STUDY 2: EFFECT OF METFORMIN TREATMENTON HYPERINSULINEMIC STATUSIN TWO RODENT MODELS OF HYPERTENSION(A) Spontaneous Hypertensive Rat StudyAnimals and Experimental Design:As discussed in the introduction, the SHR has been characterized as a model ofinsulin resistant and hyperinsulinemic hypertension, similar to human essentialhypertension. The purpose of this study was to examine if metformin treatment ofSHR can prevent the development of hyperinsulinemia and hypertension as comparedto untreated SHR.Male SHR and WKY rats were procured at 4 weeks of age from Charles River,Montreal Canada. The rats were allowed one week to adapt and were then dividedinto the following experimental groups: SHR-untreated (n=9), SHR metformin-treated(n=9), WKY-untreated (n=7) and WKY metformin-treated (n=7). Basal values ofsystolic blood pressure, plasma glucose and insulin were recorded at week 5 (weeksdenote age of the animals). Following the initial measurements, chronic metformintreatment was initiated at a dose of 350mg/kg/day and gradually titrated to a final doseof 500mg/kg/day over a two week period. The treated groups received this final dose30throughout the experiment. This dose was chosen based on our previous study.Starting week 8, weekly measurements of systolic blood pressure, plasma glucose andplasma insulin (all samples collected from 5 hour fasted rats) were performed untilweek 13. Food intake, fluid intake and body weight of all the animals was recordedonce a week. In order to examine the effect of increasing exogenous plasma insulinconcentration on blood pressure, insulin implants were surgically placed in the treatedanimals at week 16. The implants were placed subcutaneously on the dorsal side ofeach rat. The implant delivered insulin at a dose of 1 unit/rat/day. At week 18 (post-implant) systolic blood pressure, plasma insulin and glucose was measured.(B) Fructose-induced Hypertension StudyAnimals and Research Design:As previously discussed, in normal Sprague Dawley rats substituting fructosefor the starch usually present in the rat chow results in hyperinsulinemia, insulinresistance and hypertension. The model thus represents an acquired form of elevatedsystolic blood pressure. This study was designed to examine the effect of metformintreatment on elevated insulin levels and blood pressure in the fructose-inducedhypertensive Sprague Dawley rat.Male Sprague Dawley rats were procured locally (180-200g body weight) at 6weeks of age. The animals were divided into four experimental groups: controluntreated (n=8), control treated (n=8), fructose untreated (n=9) and fructosemetformin treated (n=10) and basal values of plasma insulin, glucose and systolicblood pressure were recorded. Starting at week 7 (weeks signify the age of theanimal) chronic metformin treatment was initiated in the control treated and thefructose treated groups. Treatment was initiated at a dose of 350 mg/kg/day and31500mg/kg/day over a two week period. One week after initiating metformintreatment, the animals in the fructose and fructose treated group were started on a 66%fructose diet. The fructose diet (66% fructose, 12% fat and 22% protein, TekladLabs, Madison, WI, U.S.A.) has an electrolyte, protein and fat content verycomparable to the standard rat chow. The only difference is that the 60% vegetablestarch present in normal rat chow is replaced by 66% fructose in the fructose diet.Starting at week 8, blood pressure, plasma insulin and plasma glucose were measuredeach week for the next four weeks. In addition, the food intake, fluid intake and bodyweights of the animals were recorded each week. Results of this study would indicateif starting metformin prior to the administration of the fructose diet prevents fructose-induced metabolic changes and hypertension in this rodent model of hypertension.(C) METHODS1.^Blood Pressure MeasurementIndirect systolic BP was measured in conscious rats using the indirecttail cuff method without external preheating (Bunag 1973). The animals werepreconditioned to the experimental procedure before conducting the actualmeasurements. The apparatus used includes a BP sensor cuff, a BP amplifier and ananalog/digital recorder and printer (Model 179 semi-automatic BP analyzer, IITCINC., Woodland Hills, CA, USA). The various parameters, such as cycling intervaland inflation/deflation rate, are kept constant by the semi-automatic equipment. Inthis method, the reappearance of pulsations on gradual deflation of the BP cuff aredetected by a photoelectric sensor and are amplified and recorded digitally as thesystolic BP. An average of 5 such readings was taken to obtain the individual systolicBP. The major advantage of this method is that the recordings are carried out at a low32ambient temperature, thus eliminating the heat stress typical of other BP measuringdevices.2. Biochemical MeasurementsPlasma glucose was measured with the glucose oxidase method using adiagnostic kit from Boehringer Mannheim. Plasma insulin was assayed using thedouble antibody radioimmunoassay using a kit from ICN biomedicals, Costa Messa,CA, USA.3. Statistical AnalysisIn the series of experiments outlined above, the independent variablewas the drug intervention (treated vs. control). Since there were several dependentvariables e.g. glucose, blood pressure, insulin, the differences among groups wereevaluated using multivariate analysis of variance (MANOVA), using a NumberCruncher statistical Program (NCSS). MANOVA is the most powerful statisticalprocedure available for this type of analysis. In MANOVA, the mean vector (made upof the individual variate means) is examined for any difference. In the current study, aprobability of P<0.05 was taken to indicate a significant difference between themeans. When the MANOVA detected a significant difference in the mean vector, theindividual variables were analyzed employing the Newman-Keuls test for multiplecomparisons.33RESULTSSTUDY 1: EFFECT OF METFORMIN TREATMENTON ISOLATED HEART FUNCTION IN STZ-DIABETIC RATSGeneral Features of Experimental AnimalsThe general features of the experimental animals are presented in Table 2.Induction of diabetes with STZ resulted in symptoms characteristic of the diabeticstate. The untreated diabetic animals had a lower body weight, higher food and fluidintake, higher plasma glucose and trigylceride levels and lower plasma insulin valueswhen compared to age-matched control rats.Treatment of the diabetic rats with metformin resulted in a significant reductionin plasma glucose levels to near control values. The insulin levels in the diabetictreated and untreated groups were similar indicating that the glucose lowering effectoccurred without any increase in the insulin levels. Metformin treatment also loweredthe elevated plasma triglyceride levels in the diabetic group to control level.Treatment of the control group with metformin resulted in no changes in the glucoseand triglyceride values when compared to age matched untreated control rats.However, metformin significantly decreased the plasma insulin values in control-treated rats. Metformin's antihyperglycemic action is believed to be mediated via anenhancement of insulin action (Bailey 1985, 1992). In the diabetic group, this wasmanifested as an decrease in glucose levels without any increase in the insulin level,whereas in the control-treated animals metformin allowed similar glucose levels in theface of lower plasma insulin levels. This effect of a decrease in the plasma insulinlevels in control animals has not only been seen with metformin but has also beennoticed with other insulin enhancing agents like vanadyl sulphate (Cam et al., 1993).34Effect of metformin on heart function of experimental animalsCardiac performance of control and diabetic hearts was assessed by measuringleft ventricular responses to changing left atrial filling pressures in terms of leftventricular developed pressure (LVDP), rate of contraction (+dP/dt) and rate ofrelaxation (-dP/dt). Figures 4-6 illustrate the left ventricular responses measured inthe various experimental groups. The time to half relaxation (T1/2R), an index ofbeginning heart failure, was calculated at left atrial filling pressures of 15 to 22.5 cmH2O (Table 3). The myocardial performance in control, metformin-treated control,diabetic and metformin-treated diabetic groups was similar at atrial filling pressuresfrom 7.5-15 cm H2O. However, hearts from untreated diabetic animals exhibited adepressed response to increases in atrial filling pressure from 17.5 to 22.5 cm H2O interms of +dP/dT, -dP/dT and LVDP. More importantly, the diabetic hearts exhibited adelayed time to half relaxation at filling pressures of 15 to 22.5 cm H2O. Metformintreatment completely restored the ability of the diabetic hearts to respond to increasesin atrial filling pressure from 17.5 to 22.5 cm H2O. The treatment had no effect onthe control animals.STUDY 2: EFFECT OF METFORMIN ON THE HYPERINSULINEMICSTATUS IN EXPERIMENTAL HYPERTENSIONMetformin treatment of the SHRTable 4 illustrates the food and fluid intake of the rats in the four experimentalgroups. Treatment of the SHR and WKY rats with metformin led to a 20-25%decrease in the food and fluid intake when compared to untreated animals. Parallelingthe decrease in food and fluid intake the SHRM and WKYM rats exhibited a35decreased weight gain when compared to untreated animals (Figure 7). Anothernoteworthy observation was that as the rats grow older, the WKY appear to gainsignificantly more than age matched SHR. At week 16 the WKY weighapproximately 40 g more than the SHR.Figure 8 illustrates the plasma insulin level (uU/ml) in the various experimentalgroups. The SHR group exhibited sustained hyperinsulinemia by week 9 whencompared to the WKY group. Treatment of the SHR rats with metformin (SHRM)prevented the development of hyperinsulinemia. The treatment had no effect on theinsulin levels in the WKYM group. Table 5 illustrates the plasma glucose level in thefour experimental groups. Metformin treatment did not affect the plasma glucoseprofile in either the SHR or WKY groups.Figure 9 illustrates the systolic blood pressure (mm Hg) in the variousexperimental groups. Paralleling the elevation in plasma insulin, the SHR exhibitedsystolic hypertension by week 9 when compared to the normotensive WKY.Treatment of the SHR with metformin (SHRM) led to a 30-35 mm Hg decrease insystolic blood pressure. This effect was maintained during the entire treatment timeframe. Treatment with metformin had no effect on the systolic blood pressure in theWKYM rats.Table 6 indicates the plasma insulin levels in the four experimental groupsprior (week 16) to and two weeks post (week 18) the insulin implants. The insulinimplants were placed only in the treated animals i.e. WKYM and SHRM. The SHRMgroup exhibited lower plasma insulin levels than the SHR prior to the implant (week16). The insulin implants caused a marked increase in the plasma insulin level in theSHRM group but had no effect in the WKYM group. However, artificially raising theplasma insulin also caused a decrease in plasma glucose in the SHRM and WKYMrats (Table 7). Table 8 illustrates the blood pressure in the animals before and twoweek after the implants. Paralleling the elevation in plasma insulin in the SHRM36group, an increase in systolic blood pressure was also observed which was similar tothe SHR group. The implants did not affect the blood pressure in the WKYM group.Metformin treatment of the fructose-fed Sprague Dawley ratTable 9 illustrates the food and fluid intake in the various experimental groups.As noticed in the SHR experiment, treatment of the control and fructose animals withmetformin (con+m and fruc+m) led to a 20-25% decrease in food and fluid intake.Figure 10 illustrates the body weight in all the groups. The treated groups exhibited alower body weight gain, which was significant at weeks 10 and 11. Table 10illustrates the plasma glucose levels in the four experimental groups. Metformintreatment did not affect the plasma glucose level in either the control or the fructosegroups.Figure 11 illustrates the plasma insulin level (uU/ml) in the various groups.Fructose feeding of control animals led to elevated insulin levels when compared tocontrol animals. Metformin treatment of the fructose group (fruc+m) prevented theincrease in plasma insulin when compared to the fructose group. The treatment alsoled to a significant decrease in the plasma insulin level in control treated animals(con+m).The systolic blood pressure in the four groups is shown in figure 12. Thefructose group exhibited a higher systolic blood pressure when compared to thecontrol group. This elevation in systolic blood pressure was sustained over the entireexperiment. This increase in blood pressure induced by fructose feeding wascompletely prevented by pre-treatment with metformin (fruc+m). Metformintreatment had no effect on the blood pressure of the control treated group.37TABLE 2GENERAL CHARACTERISTICS OF THE RATS IN THE FOUREXPERIMENTAL GROUPS AT TERMINATION (8 WEEKS) IN THEISOLATED WORKING HEART STUDYParameter CExperimental GroupsCT DTWeight (g) 468±12 402±16 342±11* 348±13*Fluid Intake (ml/day) 47±1 30±2 /.` 159±6* 58±9#Food Intake (g/day) 32-1- 1 25±0.5 * 53±1 * 29±- 1 #Plasma Glucose (mM) 9.9±0.3 8.9±0.3 24.3±4.0* 14.4±3.0*#Plasma Insulin (uU/m1) 57.2±5.0 32±1.9*+ 9.1±1.0* 16±4.5*Plasma Triglyceride (mM) 2.5±0.5 1.3±0.8 5.9±1.9* 2.9±1.3#P < 0.05 different from control#^P < 0.05 different from diabeticP < 0.05 different from diabetic and diabetic—treatedStatistical differences were determined by ANOVA followed by a Newman—Keulstest.[Control (C, n=10), Control-treated (CT, n=10), Diabetic (D, n=11), Diabetic Treated(DT, n=12)]. Rats were injected with streptozotocin (50mg/kg/day) to inducediabetes. The treated groups received metformin hydrochloride in the drinking waterfor 8 weeks. At the end of 8 weeks of treatment, isolated working heart performancewas assessed under conditions of increasing preload.38FIGURE 4Effect of metformin treatment on rate of contraction (+dP/dT) in the isolated workingheart preparation at left atrial filling pressures from 7.5 to 22.5 cm H2O. Hearts fromthe four experimental groups [Control V (n=7), Control metformin-treated • (n=6),Diabetic 0 (n=8) and Diabetic metformin-treated • (n=8)] were isolated andperfused following 8 weeks of metformin treatment. Values are expressed asmean±SE. "n" indicates the number of rats in each group. * P<0.05, ANOVA,different from diabetic.39900080007000-‹.z 6000-oE. 5000a_<400010^•5^20FELLING PRESSURE (cm H20)300020005 2540FIGURE 5Effect of metformin treatment on rate of relaxation (-dP/dt) in the isolated workingheart preparation at left atrial filling pressures from 7.5 to 22.5 cm H2O. Hearts fromthe four experimental groups [Control V (n=7), Control metformin-treatedv(n=6),Diabetic Q(n=8) and Diabetic treated .(n=8)] were isolated and perfused following 8weeks of metformin treatment. Values are expressed as mean±SE. "n" represents thenumber of rats per group. * P<0.05, ANOVA, different from diabetic.41FILLING PRESSURE (cm H20)42FIGURE 6Effect of metformin treatment on left ventricular developed pressure (LVDP) in theisolated working heart preparation at left atrial filling pressures from 7.5 to 22.5 cmH2O. Hearts from the four experimental groups [Control V(n=7), Control metformin-treated v (n=6), Diabetic 0 (n=8) and Diabetic metformin-treated • (n=8)] wereisolated and perfused following 8 weeks of treatment. Values are expressed asmean±SE. "n" indicates the number of rats in each group. * P<0.05 different fromdiabetic.435^10^15^20^25FILLING PRESSURE (cm H20)44TABLE 3TIME TO HALF RELAXATION (T112R mseconds) IN THE FOUREXPERIMENTAL GROUPS AT PRELOADS OF 15 TO 22.5 CM H2O IN THEISOLATED WORKING HEART EXPERIMENTFillingPressure^ C^CT^D^DT(cm H20)15 34.0±4 28.8±4 64.0±8* 32.0±417.5 29.6±4 28.0±4 56.0±8* 25.6±820 20.8±4 27.2±4 56.0±4* 29.6±422.5 25.6±8 22.4±4 56.0±8* 31.2±8[Control (C, n=7), control metformin-treated (CT, n=6), diabetic (D, n=8) and diabeticmetformin-treated (DT, n=8)]. Time to half relaxation expressed in msec from fillingpressures of 15 to 22.5 cm H2O. * P<0.05, ANOVA, different from C, CT, and DT.45TABLE 4FOOD AND FLUID INTAKE IN THE SHR STUDYFood intake (g/day) in the four experimental groupsGroup Week 5(baseline)Week 9 Week 11 Week 13WKY 17.0±0.4 26.711.1 21.110.8 25.010.9WKY-M 18.010_7 26.710.7 15.610.4* 17.3±0.6*SHR 21.310.2 22.611.0 20.610.4 23.010.4SHR-M 21.610.4 26.011.4 19.710.3 18.3±0.5Fluid intake in the four experimental groups (ml/day)Group Week 5(baseline)Week 9 '-Week 11 Week 13WKY 34±1.3 41107 46±1.7 . 4911.2WKY-M 3210.5 30±0.1* 3912.9*  32±2.3*SHR 3710.5 39±1.0 45±2.1 48±2.0SHR-M 35±0.5 30±1.411 33±1.011 31±1.00WKY (n=7), WKY metformin-treated (WKYM, n=7), SHR (n=9) and SHRmetformin-treated (SHRM, n=9). "n" indicates the number of rats in each group.Values expressed as mean±SE.* P<0.05 MANOVA different from WKY■1)<0.05 MANOVA different from SHR46TABLE 5PLASMA GLUCOSE (mM) IN THE FOUR EXPERIMENTAL GROUPSIN THE SHR STUDYGroups Week 5 Week 9 Week 11 Week 13SHR(n=7) 6.1±0.3 6.7±0.2 6.1±0.3 6.2±0.5SHRM(n=7) 6.5±0.1 6.4±0.1 5.8±0.2 5.7±0.2WKY(N=7) 6.9±0.2 6.6±0.4 6.6±0.3 6.5±0.3WKYM(N=8) 6.6±0.3 5.9±0.2 5.7±0.1 6.0±0.2SHRM=SHR metformin-treated, WKYM=WKY metformin-treated. "n" indicates thenumber of rats in each group.47FIGURE 7Body weight (g) in the four experimental groups in the SHR study. [WKY Q (n=7),WKYM (n=7), SHR 0 (n=9) and SHRM • (n=9)]. Metformin treatment wasinitiated at week 6 (week denotes age of animal).* P<0.05, MANOVA different from WKYM and SHRM11P<0.05, MANOVA different from SHR484504003503007: 1)I - - -= 2500ED200150100504^6^8^10^12^14^16^18^20WEEKS49FIGURE 8Plasma insulin (uU/ml) in the four experimental groups in the SHR study. [WKY(n=7), WKYM • (n=7), SHR v (n=9), SHRM (n=9)]. Metformin treatmentinitiated at week 6 (week denotes age of animal). "n" indicates the number of rats ineach group. Values expressed as mean±SE.*P<0.001, MANOVA, different from WKY, WKYM and SHRM.50WEEKSPLASMAINSULINuU/m11614124^6^8^10706050403020100251FIGURE 9Systolic blood pressure (mm Hg) in the four experimental groups in the SHR study.[WKY 0 (n=7), WKYM • (n=7), SHR v (n=9), SHRM • (n=9)].Metformin treatment was initiated at week 6 (week denotes age of animal). "n"indicates the number of rats in each group. Values are expressed as mean±SE.@ P<0.001, MANOVA different from WKY, WKYM and SHRM* P<0.001, MANOVA different from WKY, WKYM and SHR52SYSTOLIC BLOOD PRESSURE (mm Hg)-o,a)TABLE 6PLASMA INSULIN LEVELS (µU/ml) DURING THE IMPLANT EXPERIMENTGROUP PRE-]MPLANT(16 WEEKS OLD) POST-IMPLANT(18 WEEKS OLD)WKY(n=7)52.613.5 59.211.2WKYM 58.1+5.0 63.912.9(Ja=7)SHR(n=7)57.513.6 63.711.3SHRM 36.5+1.2 66.112.6 *(n=8)* P<0.001 ANOVA, DIFFERENT FROM SHRMAt week 16 implants were surgically implanted in the dorsal neck region of theWKYM and SHRM groups. The untreated groups served as the controls (withoutimplant) to evaluate the effect of artificially raising plasma insulin in the treatedgroups on blood pressure.54TABLE 7PLASMA GLUCOSE LEVELS (mM) DURING THE IMPLANT EXPERIMENTGROUP PRE-IMPLANT(16 WEEKS OLD) POST-EtvfPLANT(18 WEEKS OLD)WKY(n=7)6.610.5 6.81-0.7WKYM(n=7)5.91-0.4 4.4±0.7 *SHR(n=7)6.11-0.4 6.31-0.3SHRM 6.4± 0.7 5.1±0.4*(n=8)*'P<0.05, ANOVA, different from pre-implant WKYM and SHRMAt week 16 implants were surgically implanted in the dorsal neck region of theWKYM AND SHRM GROUPS. The untreated groups served as the controls(without implant) to evaluate the effect of artificially raising plasma insulin in thetreated groups on blood pressure.55TABLE 8BLOOD PRESSURE IN THE FOUR EXPERIMENTAL GROUPSIN THE INSULIN IMPLANT STUDYGROUP/B.P^WKY^WKYM SHR^SHRMPre-implant 143+-3.6 148±1.8 212+6.3 168+2.3Post-implant 150±3.1 156+2.4 208±4.6 189+3.1 #WKY (n=7), WKYM (n=7), SHR(n=9), SHRM(n=9). Values are expressed asmean±SE. "n" indicates the number of rats in each group.# P<0.05, ANOVA different from SHRM (pre-implant)56TABLE 9Food (g/day) and fluid (ml/day) intake in the various experimental groups. Theexperimental groups consisted of control (n=8), control metformin-treated (con+M,n=8), fructose (n=9) and fructose metformin-treated (n=10). Metformin was initiatedat week 6 and fructose was started at week 7. Weeks denote the age of the rats.Values are expressed as mean±SE. "n" indicates the number of rats in each group.FOOD INTAKE IN THE FOUR EXPERIMENTAL GROUPS (g,/day)Group/Week 6(baseline)9 10 11CONTROL 18.0-1-0.3 26.90.3 29.11-0.3 29.8±0.2CON+M 16.4±0.2 24.51-0.2 24.510.4$ 24.9±0.2@FRUCTOSE 18.60.3 24.6±0.6 25.2±0.3 30.1±0.3FRUC+M 17.6±0.3 24.1±0. 24.2±0.2 25.410.44FLUID INTAKE IN Ink. FOUR EXPERIMENTAL GROUPS (ml/day)Group/Week 6(baseline)9 10 11CONTROL 34.51-0.7 39.610.4 43.61-0.4 45.31-0.4CON+M 31.410.3 36.91-0.3 36.01-0.3$ 37.910.3@FRUCTOSE 33.81-0.3 38.0±0.8 39.710.8 40.30.1FRUC+M 33.410.6 35.610.3 35.310.3 34.210.411@ P<0.05 different from Control(Week 11), MANOVA/I P<0.05 different from Fructosc(Week 11), MANOVA$ P<0.05 different from Control(Week 10), MANOVA57TABLE 10PLASMA GLUCOSE (mM) IN THE VARIOUS EXPERIMENTAL GROUPSIN THE FRUCTOSE STUDYGroup/Week 6 9 10 11Con (n=8) 6.61-0.3 6.41-0.2 6.01-0.4 6.310.1Con+m(n=8) 5.91-0.5 6.110.3 6.71-0.5 5.80.5Fruc (n=9) 6.010.1 6.81-0.5 6.61-0.2 6.11-0.3Fruc+m(n=10) 5.81-0.5 6.11-0.5 6.21-0.2 6.31-0.4Con = control; Con+m = control metformin-treated; fruc = fr)actose; frjic+m = fructosemetformin-treated. Metformin initiated at week 6 and fructose started at week 7 in thetreated and fructose groups respectively. Values expressed as mean±SE. "n" =number of rats in each group.58FIGURE 10Body weight (g) in the four experimental groups in the fructose study. [Control (con•n=8), control metformin-treated (con+m v^n=8), fructose (fruc •^n=9) andfructose metformin treated (fruc+m ^ n=10)] Metformin treatment was started atweek 6 and fructose was initiated at week 7 in the treated and fructose groupsrespectively. Weeks denote the age of the rats. Values are expressed as mean±SE."n" indicates the number of rats in each group.# P<0.001, MANOVA different from CON+M and FRUC+M.594^6^8^1 0^12WEEKS60FIGURE 11Plasma insulin (uU/ml) in the four experimental groups in the fructose study. [Control(con, 0 n=8), control metformin-treated (con+m • n=8), fructose (fruc v (n=9),fructose metformin-treated (fruc+m v (n=10)]. Metformin treatment was initiatedat week 6 and fructose was started at week 7 in the treated and fructose groupsrespectively. Weeks denote the age of the rats. "n" indicates the number of rats ineach group. Values are expressed as mean ± SE.* P<0.001, MANOVA, different from CON, CON+M, FRUC+M# P<0.001, MANOVA, different from CON6110090807060504030201004 6^8^10WEEKS12^1462FIGURE 12Systolic blood pressure (mm Hg) in the four experimental groups in the fructosestudy. [Control (con, 0 n=8), control metformin-treated (con+m,^n=8), fructose(fruc, v n=9), fructose metformin-treated (fruc+m, •^n=10)]. Metformintreatment was initiated at week 6 and fructose was started at week 7 in the treated andfructose groups resectively. Weeks denote the age of the rats. Values are expressedas mean±SE. "n" denotes the number of rats in each group.* P<0.001, MANOVA, different from con, con+m and fruc+m.63220an200N 180crlix160cnrnNIxa,^140Q0o 120.1aio:=1^1000E--1U)>-,^8cn0604^6^8^10^12^14WEEKS64DISCUSSIONThe work outlined in this thesis examined the effect of an oralantihyperglycemic agent, metformin, on isolated cardiac function and blood pressurein diabetic and hypertensive rats, respectively. In the first series of experiments, weexamined the effect of metformin treatment on various functional cardiac parametersin rats made diabetic with STZ. The second part of the project explored therelationship between elevated insulin levels and hypertension in two rodent models ofhyperinsulinemia and hypertension, the SHR and the fructose-fed rat using metforminas the experimental intervention.The association between chronic diabetes and cardiac dysfunction has beendemonstrated in both human and animal studies (Ahmed et al., 1975; Fein et al., 1985;Penpargul et al., 1980; Regan et al., 1977; Vadlamudli et al., 1982). The cardiacproblems in diabetes include a decreased ventricular contractility manifested as a lowstroke volume, cardiac index and ejection fraction and a diminished ventricularcompliance manifested as an elevated left ventricular end diastolic pressure (D'Elia etal., 1979; Hamby et al., 1974; Ledet et al., 1979). The presence of these functionalabnormalities, independent of atherosclerosis of the coronary vessels, suggests theexistence of a specific cardiac muscle disease (cardiomyopathy) in diabetes.The first study describes the effects of metformin treatment on isolated cardiacperformance in the STZ-diabetic rat. Induction of diabetes with STZ resulted intypical symptoms of diabetes; the rats exhibited hyperglycemia, hypoinsulinemia,elevated food and fluid intake and a decreased weight gain when compared to non-diabetic rats. Treatment of the diabetic rats with metformin on a chronic basisdecreased plasma glucose to near control values (Table 2). The improvement inglucose levels was seen without any increase in the plasma insulin profile, supportingprevious work that showed that the antihyperglycemic effect of metformin is not65mediated by an increase in insulin secretion but via enhancement of insulin-mediatedglucose uptake in peripheral tissues (Bailey and Puah, 1986; Galuska et al., 1991).Similar observations have been made in animal models of obese and non-obesehyperinsulinemic and hypoinsulinemic diabetes (Lord et al., 1983; Penicaud et al.,1985; Rossetti et al., 1990).STZ-induced diabetic rats exhibited elevated triglyceride levels (Table 2) whencompared to non-diabetic rats and these were restored to control values in thediabetic-treated group, again demonstrating the insulin enhancing effects ofmetformin.Hearts from control and diabetic rats were isolated and cardiac performancewas evaluated under physiological and super-physiological filling pressures, simulatedby increases in left atrial filling pressure from 7.5 to 22.5 cm H2O. This allowedassessment of cardiac function under normal and stressful conditions.Paralleling the elevation of plasma glucose and triglycerides, the diabetic groupalso exhibited a decreased cardiac performance as assessed by their inability torespond to increases in left atrial filling pressure from 17.5 to 22.5 cm H2O (Figures4-6). The diabetic rat hearts exhibited an elevated T1i2R at filling pressures of 15 to22.5 cm H2O. Although the diabetic heart performance differed significantly fromcontrols only at filling pressures of 17.5 to 22.5 cm H2O when measured in terms of±dP/dt and LVDP, calculation of the T 1/2R revealed that the diabetic hearts had adelayed onset of relaxation even at filling pressures of 15 cm H2O (Table 3). This isconsistent with the fact that T 1/2R is a highly sensitive index of beginning heartfailure. Metformin treatment restored the ability of the diabetic animals to respond toincreases in filling pressure for all the functional indices studied.The mechanisms by which metformin improved diabetic cardiac performanceare not known. Several studies have attempted to correlate plasma and myocardiallipid levels to the cardiac dysfunction seen in the diabetic rat (Heyliger et al., 1986;66Shipp et al., 1973). The involvement of lipids is suggested by a study from ourlaboratory, in which Wistar and Wistar-Kyoto (WKY) rats were injected withidentical doses of STZ (Rodrigues and McNeill, 1986). Although both group of ratsexhibited elevated plasma glucose levels, only the Wistar rats exhibited elevatedlevels of circulating lipids. Interestingly, depression of myocardial function was seenin the group of Wistar rats while the cardiac function in the STZ-injected WKY ratsremained unaffected. In another study, treatment of STZ-diabetic rats withhydralazine normalized lipid levels and restored cardiac function to non-diabeticcontrol values (Rodrigues et al., 1986). In addition, other studies have shown thatdiabetic rats pretreated with L-carnitine and choline, and methionine did not showelevated plasma lipid levels and exhibited an improved cardiac performance whencompared with untreated Wistar rats (Heyliger et al., 1986; Rodrigues et al., 1988). Itthus appears that the improvement in the glucose homeostasis and normalization oftriglycerides may be contributing, at least in part, to the improvement in cardiacfunction in the diabetic rats pretreated with metformin.Several subcellular organelles and myocardial enzyme systems have beenshown to be defective in diabetes (Dilmann 1980; Heyliger et al., 1987; Lopaschuck etal., 1983; Malhotra et al., 1981; Penpargul et al., 1981). One of the more prominentsystems affected, which is known to regulate contractility, is the sarcoplasmicreticulum. In diabetic rat hearts, the ability of the sarcoplasmic reticulum to take upcalcium is impaired, which explains the impairment of relaxation in the myocardialmuscles (Lopaschuck et al., 1983). Also, as a consequence of lowered uptake, theamount of calcium available for release during the following beats may be lower thannormal and thus lead to impaired contraction. The mechanism by which diabetesalters sarcoplasmic reticulum calcium uptake is believed to be via an elevation oflong-chain acyl carnitine, a metabolic intermediate which is responsible for thetransport of fatty acids into the mitochondria. This intermediate is a strong inhibitor67of sarcoplasmic reticulum calcium uptake in-vitro and also inhibits other enzymessuch as Na+/K+ATPase. Along with the elevation of long-chain acyl carnitines, thelevels of long-chain acyl CoA are also elevated, which may have detrimental effectson the myocardium (Dilman 1980; Heyliger et al., 1987; Lopaschuck et al., 1983;Malhotra et al., 1981; Penpargul et al., 1981). It would be interesting to examine ifpre treatment of diabetic rats with metformin could restore such subcellular defects.As outlined earlier in the introduction, in the last several years considerableattention has been directed towards investigating the role of altered myocardialenergetics in the development of diabetic heart disease (Rodrigues and McNeill 1992).A view which is gaining acceptance is that diabetes causes metabolic changes and thatthese metabolic changes precede the development of overt cardiac failure in a chronicsetting (Matti et al., 1990). Under normal conditions, an estimated 60-70% ofmyocardial energy is derived from the metabolism of lipids; the remainder is derivedfrom non-lipid sources. Diabetes is associated with elevated triglyceride and FFAlevels and as a result there is an increased uptake, oxidation and storage of fatty acidsby the myocardium. Thus under conditions of diabetes the heart exhibits an exclusivedependence on FFA oxidation as a source of energy, which increases from 60-70% toalmost 90% in diabetes. Elevated FFA oxidation has deleterious metabolic effects onthe myocardium. Firstly, the inability of the diabetic heart to utilize glucose leads toan elevated oxygen demand per molecule of ATP produced. Secondly, increased FFAoxidation leads to an accumulation of intermediates of FFA oxidation, such as longchain acyl carnitines, which have deleterious effects on the myocardial cell viadifferent mechanisms (Rodrigues and McNeill, 1992). One may speculate that theimprovement in cardiac function seen in the present study may be secondary to adecrease in the trigylceride level and a switch in myocardial metabolism frompredominantly FFA to glucose.68Other studies from our laboratory have shown that vanadium treatment ofSTZ-diabetic rats also prevents the development of cardiac dysfunction (Heyliger etal., 1986; Yuen et al., 1993). There appears to be some similarity between the effectsof inorganic and organic vanadium compounds and metformin with respect toexperimental diabetes. Both vanadium compounds and metformin enhance the actionof insulin (Lord et al., 1983; Penicaud et al., 1985; Rossetti et al., 1990; Yuen et al.,1993) and prevent the development of secondary cardiovascular complications ofdiabetes. The exact mechanism(s) of action of biguanides and vanadium is not knownand some overlap may exist. Although the effect of vanadium and metformin inpreventing cardiac disease in humans remains to be determined, these agents mayprovide a unique metabolic approach to the treatment of cardiovascular complicationsof chronic diabetes.The second series of experiments was aimed at exploring the inter-relationshipbetween hyperinsulinemia, insulin resistance and essential hypertension in a geneticmodel of hypertension, the SHR, and an acquired model of hypertension, the fructose-induced hypertensive rat. We hypothesized that if elevated insulin levels and/orinsulin resistance play a role in the development of hypertension, an agent such asmetformin, which enhances the action of endogenous insulin and causes a decrease inthe plasma insulin level, may alleviate or prevent the development of high bloodpressure.Treatment of the SHR and the WKY rats with metformin led to a 20-25%decrease in their food and fluid intake. This was accompanied by a decreased bodyweight gain in the treated groups which was significant at week 11, 13, 16 and 18. Thedecreased body weight gain in the SHRM and WKYM group could be attributed tothe decrease in food and fluid intake seen in these rats. Moreover, metformintreatment often promotes weight loss (Hermann 1979), notably in obese and energy-restricted NIDDM patients (Wales, 1980) and animals (Bailey et al., 1986), possibly69via an increased thermogenic activity of brown adipose tissue and increased "futile"cycling of substrates (Leslie et al., 1986) . Therefore, the decreased body weight gainin the treated groups may also be due to a direct action of the drug.As discussed earlier, the SHR has a genetic propensity for hypertension andhave also been shown to be insulin resistant and hyperinsulinemic when compared totheir genetic controls, the WKY (Mondon and Reaven, 1988; Finch et al., 1990). Asillustrated in figure 8, by week 9 the SHR group exhibited hyperinsulinemia whencompared to the WKY. Paralleling the elevation in plasma insulin, the SHR exhibitedfull-blown systolic hypertension by week 9 (Figure 9). Treatment of SHR withmetformin at a dose of 500 mg/kg/day prevented the development of hyperinsulinemiain these rats. The treatment also decreased the blood pressure in the SHRM group tonear control values. Metformin did not completely restore the blood pressure tocontrol levels; however, it was effective in alleviating the rise in blood pressure whencompared to untreated SHR rats. This is not unexpected as plasma insulin is not theonly determinant of hypertension in the SHR; it may be one contributing factor. It canbe speculated that if elevated plasma insulin is one of the determinants ofhypertension in the SHR, then improving this cause (preventing hyperinsulinemia)alleviated hypertension in the SHRM animals. In an attempt to further explore therelationship between insulin and blood pressure, we artificially raised plasma insulinin the WKYM and SHRM groups via insulin implants and studied the resultant effecton blood pressure. Table 6 illustrates the pre- and post-implant plasma insulin levelsof the four experimental groups. It is clear on observation that in response to theimplants the SHRM group exhibited marked elevations in the plasma insulin level.This was accompanied by an increase in systolic blood pressure (Table 8) in theSHRM group but not in the WKYM. In other words pretreatment of the SHR withmetformin before hypertension is manifested, prevents the development ofhyperinsulinemia and alleviates hypertension. Artificially raising the plasma insulin70level in these rats (SHRM) leads to restoration of hyperinsulinemia and an elevation insystolic blood pressure. Both these observations cannot confirm a causal relationshipof insulin and hypertension yet indicate that these parameters may be related.Moreover, as we did not measure insulin resistance (using a hyperinsulinemic clamp),it is difficult to comment on the association of insulin resistance and hypertension inthis rodent model of hypertension. The question that arises is why do changes inplasma insulin correlate with changes in blood pressure in the SHR but not in theWKY rats? It has been suggested that the SHR may have an increased geneticsusceptibility towards certain insulin effects such as sodium retention, vascularhypertrophy and increased sympathetic nervous activity (Buchanan et al., 1992; alsodiscussed later) when compared to the WKY. This could result in the development ofhigh blood pressure selectively in the SHR. However, artificially raising plasmainsulin in the WKYM and SHRM groups did cause a decrease in the plasma glucoselevel. A noteworthy observation from our data was the fact that from weeks 9 to 13the SHR demonstrated hyperinsulinemia when compared to the WKY rats; however,the pre-implant plasma insulin values taken at week 16 were identical between theSHR and the WKY. A plausible explanation for this lies in the difference in the bodyweights between the WKY and the SHR at week 16 (pre-implant). At week 16 theWKY rats weighed about 40 g more than the SHR which may explain their higherplasma insulin levies. This observation highlights the importance of weight-matchedanimals in accurate interpretation of data on carbohydrate metabolism.The next step was to examine if the insulin-blood pressure hypothesis was validin an another model of hyperinsulinemic hypertension, the fructose-induced rat. Thefructose rat model represents an acquired form of systolic hypertension, wherein therise in blood pressure is not genetically determined but is diet-induced (Hwang et al.,1987; Reaven et al., 1988). Given the fact that we still do not know the relativecontribution of genetic and acquired factors towards the insulin resistance observed in71hypertension, we thought it would be very pertinent to examine the validity of theinsulin-blood pressure hypothesis in this model.Fructose feeding of normotensive Sprague Dawley rats led to sustainedhyperinsulinemia and elevations in blood pressure when compared to control rats(Figure 11 and 12) This effect was apparent about 2 weeks after beginning thefructose-diet. Treatment with metformin prior to the administration of the fructosediet (fruc+m) prevented fructose-induced hyperinsulinemia and hypertension. Thetreatment did not affect the blood pressure in the control rats; however, metformintreatment of the control group caused a reduction in the plasma insulin level whichwas significant at weeks 9 and 11. This was not unexpected as in our first study onisolated working heart, a decrease in plasma insulin was observed in control treatedanimals also. This is consistent with the idea that metformin's action is not mediatedvia an increase in insulin secretion but via an enhancement of insulin mediatedglucose disposal in the peripheral tissues. The question which comes to mind is whydecreases in plasma insulin correlate with decreases in blood pressure in the fructose-treated group but not in the control treated group ? The first point to consider here isthat it is difficult to compare the action of an insulin-enhancing agent in an insulin-sensitive (control) and insulin resistant (fructose) tissue. As outlined in theintroduction, one potential mechanism through which insulin resistance can lead tohypertension is thought to be via resistance to the vasodilator effects of insulin. It canbe postulated that the insulin-resistant rats (in our case the fructose group) haveattenuated vasodilator responses to hyperinsulinemia. An insulin enhancing agent likemetformin may potentially improve the vasodilator responses to hyperinsulinemia andthereby decrease blood pressure.The fact that an experimental intervention which decreased insulin levelscaused a decrease in blood pressure in two models of hyperinsulinemia andhypertension certainly does not establish a causal role for insulin in the development7261,2^Co—/of hypertension, but strongly suggesta-stronwessociation. Further support for theinterplay between insulin and hypertension is provided in the first part of this studywhere artificially raising plasma insulin levels led to an increase in blood pressure inthe SHRM group. Other agents which improve insulin sensitivity and decreaseplasma insulin levels have been shown to both prevent and reverse the development ofhypertension in both the SHR and fructose rat model. These agents include vanadylsulphate, an organic vanadium compound, bismaltolato (oxo) vanadium IV (Bhanotand McNeill 1993), and lastly pioglitazone (Bhanot and McNeill unpublishedobservations). The fact that multiple and specific drug interventions which enhancethe action of insulin lead to a decrease in the plasma insulin and are accompanied by afall in systolic blood pressure provide strong support to the notion that elevated insulinlevels contribute at least in part to the development of hypertension in these twomodels of rodent hypertension. As mentioned earlier, two interventions in fructose-induced hypertensive rats provide indirect evidence for the hypothesis under study.Exercise training of fructose fed rats improved insulin sensitivity, decreased plasmainsulin levels and decreased blood pressure- (Mondon et al., 1980). Somatostatinadministration to these rats prevented hyperinsulinemia and decreased blood pressure(Reaven et al., 1989). The observation that hyperinsulinemia and/or insulin resistanceoccur not only in untreated human hypertensives, but also in rodent models ofhypertension and that preventing hyperinsulinemia via interventions like metformin,vanadyl sulphate and pioglitazone alleviates hypertension, strengthens our contentionthat these abnormalities are intrinsically linked with hypertension and are not merecoincidental findings.As outlined earlier, hyperinsulinemia can lead to hypertension through threewell documented mechanisms. Firstly, insulin and insulin like growth factors aremitogens capable of stimulating smooth muscle proliferation (Stout et al., 1975; Nakoet al., 1985; King et al., 1985; Pfeifle and Ditschuneit, 1981; Sinha et al., 1989).73Therefore, hyperinsulinemia could result in vascular smooth muscle hypertrophy,narrowing of the lumen of resistance vessels and ultimately in the development ofhypertension. With respect to the relationship between changes in vascular structureand hyperinsulinemia, Rocchinni et al. (1988) demonstrated in obese adolescents thatnot only is obesity associated with the presence of structural changes in the forearmresistant vessels but also that these changes directly correlate with the degree ofinsulin resistance. In a group of obese adolescents they determined whether or notstructural changes in the forearm resistance vessels were present by measuringforearm vascular resistance after 10 minutes of ischemic exercise. Obese individualshad significantly elevated minimum forearm vascular resistance and reduced maximalforearm blood flows.The second method by which hyperinsulinemia/insulin resistance can lead tohypertension is with regard to insulin's ability to stimulate the sympathetic nervoussystem. This is an area which has long been emphasized by Landsberg and Young(Young et al., 1984; Landsberg and Young, 1978; Young et al., 1985; Young et al.,1982; Landsberg and Kriger, 1989; Rowe -et al., 1981). They and others havedocumented that under conditions of euglycemic hyperinsulinemia, an activation ofthe sympathetic nervous system (SNS) can be observed in obese and normal humansand animals (Rocchini et al., 1990; Liang et al., 1982). The increase in the SNSactivity is manifested as an increase in heart rate, blood pressure and plasmanorepinephrine. More recently, Anderson et al. (1991) demonstrated thathyperinsulinemia in humans is not only associated with an increase in plasmacatecholamines but also with an increase in sympathetic nerve activity. In addition,overfeeding with both carbohydrates and fat are associated with stimulation of thesympathetic nervous activity. Fisher rats fed a high fat diet develop obesity,hypertension and insulin resistance. When euglycemic insulin infusions areperformed, increases in systolic blood pressure are seen in the fat-fed animals but not74in the control animals. As this blood pressure rise was reversible by combined alphaand beta-blockade, a role for increased sympathetic nervous activity was suggested(Assy et al., 1991). A recent cross-sectional epidemiological study demonstrated thatabdominal obesity correlated with plasma insulin level and urinary epinephrineexcretion (an index of sympathetic activity) (Landsberg et al., 1991).The third proposed mechanism through which elevated insulin levels can leadto high blood pressure is via sodium and water retention. There is human and animaldata which suggest that insulin resistance and/or hyperinsulinemia can result inchronic sodium retention (DeFronzo et al., 1975; Baum 1987). Insulin can enhancechronic renal sodium retention both directly, through its effects on renal tubules, andindirectly through stimulation of the sympathetic nervous system, augmentingangiotensin II mediated aldosterone production and by altering the secretion of atrialnatriuretic peptide (Vierhapper et al., 1983). Rocchinni et al. (1989) also showed inobese adolescents that insulin resistance and sodium sensitivity of blood pressure aredirectly related. They demonstrated that the blood pressure of obese adolescents ismore dependent on dietary sodium intake than the blood pressure of non-obeseindividuals and that hyperinsulinemia and increased sympathetic nervous systemactivity appear to be responsible for the observed sodium sensitivity and hypertension.In addition, Antishin et al. (1990) demonstrated that the endogenous hyperinsulinemiathat occurs in obese subjects following a glucose meal can result in urinary sodiumretention. In that study, the investigators also showed that the obese adolescents whowere the most sodium sensitive had significantly higher fasting insulin concentrations,higher glucose-stimulated insulin levels and greater urine sodium retention in responseto an oral glucose load. There are animal data that suggest that insulin resistance maybe in part responsible for the sodium retention associated with obesity inducedhypertension. In a dog model of obesity-induced hypertension, Rocchini et al. (1987)demonstrated that during the first week of the high fat diet, the increase in sodium75retention appeared to best related to an increase in the plasma norepinephrine activity;whereas, during the latter weeks of the high fat diet, an increase in plasma insulinappeared to be the best predictor of sodium retention. Finally, Rocchini et al. (1990)recently demonstrated that the hypertension associated with weight gain in the dogoccurs only if adequate salt is present in the diet. Thus, in both obese man and dog,insulin appears to play an important role in sodium retention.Another suggested mechanism through which elevated insulin levels/insulinresistance can lead to hypertension is through alterations in cation transport. Insulinhas been shown to affect both sodium and calcium transport, although controversystill exists regarding the molecular mechanism of this effect. A direct effect of insulinon sodium/hydrogen exchange has been demonstrated in vitro (Moore 1985; Lagadie-Grossman et al., 1988). Insulin has been reported to increase and decrease Na-K-ATPase activity (Tedde et al., 1988; Khadouri et al., 1987). Insulin has been linkedwith Na-Li countertransport and Na-K co-transport (Hunt et al., 1986).The question that has been raised by several investigators is that if the abovementioned mechanisms are physiological actions of insulin, then in the face of insulininsensitivity/resistance shouldn't these actions be down-regulated ? A view that isgaining acceptance is that of "selective insulin resistance". Selective insulin resistanceimplies that although an individual or animal may have an impaired ability of insulinto cause whole body glucose uptake, some of the other physiological actions of insulinmay be preserved. With respect to hypertension, one of the potentially importantactions of insulin is its ability to induce renal sodium retention (Rocchini et al., 1989).Rocchini et al. (Rocchini et al., 1989) recently demonstrated that obese adolescentshave selective insulin resistance in that they are resistant with respect to glucoseuptake yet are still sensitive to the renal sodium retaining effects of insulin.It may be speculated that in the present study metformin caused a decrease inthe plasma insulin level which via one or more of the above mechanisms led to a76decrease in blood pressure in both the SHR and fructose rat models. As this studyaimed at establishing a role for insulin in the regulation of blood pressure, nomechanistic studies were performed to identify which pathway was affected bylowering insulin levels with metformin.The association of hyperinsulinemia/insulin resistance and hypertension is wellestablished but the causal relationship remains unclear. Basically there are threepossibilities: firstly, hypertension leads to insulin resistance and hyperinsulinemia,secondly, insulin resistance and hyperinsulinemia lead to hypertension and lastly, bothinsulin resistance/hyperinsulinemia co-exist in the same individual but are not relatedto each other (Ulf 1991).The first relationship seems unlikely since there is evidence that hypertensioncaused by renovascular disorders is not associated with insulin resistance (Mariglianoet al., 1990) Furthermore, treating hypertension with most commonly usedantihypertensive agents does not improve insulin sensitivity, but if anything leads to afurther deterioration (Review Lithell, 1991).In the context of the second relationship, many studies have shown acorrelation between elevated insulin levels and blood pressure. In most studies, thisremains even when concomitant obesity or other confounding factors are accountedfor. However it also clear that not all individuals with hypertension arehyperinsulinemic or that all individuals with hyperinsulinemia are hypertensive. Forinstance there is evidence that patients with an insulinoma do not develophypertension to an excessive extent (Tsutsu et al., 1990). A possibility is that onlycertain individuals are genetically susceptible to the effects of hyperinsulinemia. Thiscould account for the increased prevalence of hypertension in other insulin-resistantstates such as obesity and impaired glucose tolerance. Thus, any factor leading toinsulin resistance and hyperinsulinemia would then lead to hypertension in susceptible77individuals. Such a concept could also account for the correlation seen between bloodpressure and ambient insulin levels (Ulf, 1991).The third possibility that insulin resistance and hypertension co-exist but areunrelated to each other cannot be excluded. However, the consistent relationship andthe improvement of both factors by weight reduction, exercise and agents thatimprove insulin sensitivity (metformin) make it less likely (Landin et al., 1991).Further studies addressing these three possibilities need to be carried out toestablish a cause-effect relationship of insulin and hypertension and to define themechanisms and pathways which are involved. It is likely that as our knowledge ofthe importance of insulin resistance/hyperinsulinemia in hypertension increases therewill be new possibilities to treat both the blood pressure elevation and the manyimportant risk factors linked to hypertension and coronary artery disease.78CONCLUSIONS1. Induction of diabetes with STZ resulted in hyperglycemia, hypoinsulinemia,elevated triglycerides and depressed cardiac function when compared to agematched non-diabetic rats.2. Treatment of the diabetic rats with metformin hydrochloride at a dose of 650mg/kg/day improved glucose homeostasis without an increase in the plasmainsulin level, decreased the elevated triglycerides and improved isolated cardiacfunction assessed under conditions of increasing preload in terms of +dP/dt, -dP/dt, LVDP and T1/2R. Metformin significantly decreased the plasma insulinlevel in the control animals. The treatment had no effect on the heart functionof control animals.3. SHR rats exhibited hyperinsulinemia when compared to their genetic controls,the WKY. The hyperinsulinemia was persisted throughout the experiment.4. Paralleling the elevation in plasma insulin, the SHR exhibited full blownsystolic hypertension by nine weeks of age when compared to the WKY. Thiseffect was also sustained throughout the experiment.5.^Metformin treatment of the SHR prevented the development ofhyperinsulinemia and alleviated hypertension when compared to untreatedSHR.796. Artificially raising plasma insulin by insulin-implants caused a markedelevation in plasma insulin level in the SHRM group. This was accompaniedby an increase in systolic blood pressure.7. Fructose feeding of Sprague Dawley rats caused hyperinsulinemia by weeknine when compared to control rats. As with the previous experiment, thehyperinsulinemia was sustained and maintained over the experimental period.8. This dietary manipulation also caused an increase in systolic blood pressure byapproximately 20-25 mm Hg. This effect was seen throughout the experiment.9. Pre-treatment of Sprague Dawley rats with metformin prior to theadministration of fructose completely prevented the hyperinsulinemia and risein blood pressure when compared to untreated fructose fed rats.80REFERENCESAhmed SS, Jaferi GA, Narang RM and Regan TJ (1975) Preclinical abnormality ofleft ventricular function in diabetes mellitus. Am Heart J 89:153-158.Alengrin F, Grossi G, Canivet B and Dolais-Kitabgi J (1987) Inhibitory effects ofmetformin on insulin and glucagon action in rat hepatocytes involves post-receptor alterations. Diabete & Metab 13:591-597.Anderson EA, Hoffman RP, Balon TW, Sinkey CA and Mark AL (1991)Hyperinsulinemia produces both sympathetic neural activation and vasodilationin normal humans. J Clin Invest 87:2246-2252.Antishin K, Rocchini AP, Moorhead C and Brown C (1990) Circulation 82(SupplIII):III-20 (abstract).Assy W, Wan JM, Coyle S, Blackburn GL, Vistrian BR and Istfan NW (1991) ClinRes A 34.Bailey CJ, Wilcock C, Bull H and Day C (1989) Metformin increases intestinalglucose utilization (abstract). Diabetic Med 6(Suppl 2):32A.Bailey CJ, Wilcock C and Day C (1992) Effect of metformin on glucose metabolismin the splanchnic bed. Br J Pharmacol 105:1009-1015.81Bailey CJ, Flatt PR and Ewan C (1986) Anorectic effect of metformin in lean andgenetically obese hyperglycemic (ob/ob) mice. Arch Int Pharmacodyn Ther282:233-239.Bailey CJ (1985) The antihyperglycemic action of metformin. R Soc Med Int CongrSymp Ser 79:17-26.Bailey CJ (1988) Metformin revisited:its actions and indications for use. Diabetic Med5:315-320.Bailey CJ (1992) Biguanides and NIDDM. Diabetes Care 15(6):755-772.Bailey CJ and Puah JA. (1986) Effect of metformin on glucose metabolism in mousesoleus muscle. Diabete & Metab 12:212-218.Barzilai N and Simonson D (1988) Mechanism of metformin action in NIDDM(abstract). Diabetes 37(Suppl 1):61A.Baum M (1987) Insulin stimulates volume absorption in the rabbit proximalconvolvulated tubule J Clin Invest 79:1104-1109.Benzi L, Trischitta V, Ciccarone A, Cecchetti P, Brunetti A, Squatrito S, Marchetti P,Vigneri R and Navalesi R (1990) Improvement with metformin in insulininternalization and processing in monocytes from NIDDM patients. Diabetes39:844-849.82Bhanot S and McNeill JH (1993) Bis (maltolato) oxo vanadium (IV) lowers plasmainsulin and blood pressure in spontaneously hypertensive rats (abstract).Proceedings Endocrine Society Annual Meeting p 96 (104)Billingham MS, Hall RA, Simpson S and Bailey CJ (1980) Lack of effect ofglibencamide, chlorpropamide and metformin of plasma cholesterol and HDL-cholesterol in non diabetic rats. Horm Metab Res 12:340-341Brownlee M et al. (1988) Advanced products of non enzymatic glycosylation and thepathogenesis of diabetic vascular disease. Diabet Metab Rev 5:437.Buchanan TA, Youn JH, Campese VM and Sipos GF (1992) Enhanced glucosetolerance in spontaneously hypertensive rats. Diabetes 41:872-878.Bunag RD (1973) Validation in awake rats of a tail-cuff method for measuringsystolic pressure. J Appl Physiol 34:279-282.Cam MC, Pederson RA, Brownsey RW and McNeill JH (1993) Long-termeffectiveness of oral vanadyl sulfate in streptozotocin-diabetic rats.Diabetologia 36:218-224.Campbell IW, Duncan CD, Patton NW, Broadhead T, Tucker GT and Woods HF(1987) The effect of metformin on glycaemic control, intermediary metabolismand blood pressure in non-insulin-dependent diabetes mellitus. Diabetic Med4:337-341.83Canessa M, Brugnara C and Escobales N (1987) The Li+-Na+ exchange and Na- K-Cl cotransport systems in essential hypertension. Hypertension 10 (Suppl 1):4-10.Carrol P and Matz R (1983) Uncontrolled diabetes mellitus in adults:experience intreating diabetic ketoacidosis and hyperosmolar nonketotic coma with low-doseinsulin and a uniform treatment regimen. Diabetes Care 6:579-583.Cigolini M, Bossello 0, Zancanaro C, Orlandi PG, Fezzi 0 and Smith U (1984)Influence of metformin on metabolic effect of insulin in human adipose tissuein vitro. Diabete & Metab 10:311-315.Clarke BF and Campbell IW (1977) Comparison of metformin and chlorpropamide innon obese maturity-onset diabetics uncontrolled by diet. Br Med J 275:1576-1578.Clarke BF and Duncan UP (1968) Comparison of chlorpropamide and metformintreatment on weight and blood glucose response of uncontrolled obesediabetics. Lancet 1:123-126.Consoli A, Nurjham N, Capani F and Gerich P (1989) Predominant role ofgluconeogenesis in increased hepatic glucose production in NIDDM. Diabetes38:550-557.D'Elia JA, Weinrauch LA, Healy RW, Libertino RW, Bradley RF and Lelan OS(1979) Myocardial dysfunction without coronary artery disease in diabeticrenal failure. Am J Cardiol 43:193-199.84DeFronzo RA, Cooke C, Andres R, faloona GR and Davis PJ (1975) The effect ofinsulin on renal handling of sodium, pottassium, calcium and phosphate inman. J Clin Invest 55:845-855.DeFronzo RA and Ferrannini E (1991) Insulin resistance: a multifaceted syndromeresponsible for NIDDM, obesity, hypertension, dyslipidemia andatherosclerotic cardiovascular disease. Diabetes Care 14:173-194.Dilmann WH (1980) Diabetes Mellitus induces changes in cardiac myosin of the rat.Diabetes 29:579:582.Draznin B, Kao M, Sussman KE (1987) Insulin and glyburide increase cytosolic freeCa+2 concentration in isolated rat adipocytes. Diabetes 36:174-178.Fantus IG and Brosseau R (1986) Mechanism-of action of metformin: insulin receptorand postreceptor effects in vitro and in vivo. J Clin Endocrinol Metab 63:898-905.Fedele D, Tiengo A, Nosadini R, Marchiori E, Briani G, Garotti MC and Muggeo M(1976) Hypolipidemic effects of metformin in hyperprebetalipoproteinamia.Diabetes Metab Rev 2:127-34.Fein FS, Miller-Green B and Sonnenblick EH (1985) Altered myocardial mechanicsin diabetic rabbits. Am J Physiol 248: H729-H736.85Fein FS, Kornstein IB, Strobeck JE, Capasso JM and Sonnenblick EH (1980) Alteredmyocardial mechanics in the diabetic rat. Circ Res 47:922-933.Fein FS (1990) Diabetic cardiomyopathy. Diabetes Care 13(Suppl 4):1169-1179.Felig P and Bergman M (1990) Integrated physiology of carbohydrate metabolism. InDiabetes Mellitus. Theory and Practice 4th ed. Rifkin H, Porte D Jr, Eds. NewYork, Elsevier p 51-60.Ferrannini and Natali (1991) Essential hypertension, metabolic disorders and insulinresistance. Am H J121(4 pt 2):1274-1282.Ferrannini E, Haffner SM, Mitchell BD and Stern MP (1991) Hyperinsulinemia: thekey feature of a cardiovascular and metabolic syndrome. Diabetologia 34:416-422.Finch D, Davis G, Bower J and Kirchner K (1990) Effect of insulin on renal sodiumhandling in hypertensive rats. Hypertension 15:14-18.Foster DW (1991) Diabetes Mellitus In Harrissons Principles of Internal Medicine,12th edition, Wilson JD, Braunwald E, Isselbacher KJ, Petersdorf RG, MartinJB, Fauci AS and Root RK, Eds. p 1739-1765.Foster DW and McGarry JD (1983) The metabolic derangements and treatment ofdiabetic ketoacidosis. N Engl J Med 309:159-163.86Frayn KN and Adnitt PI (1972) Effects of metformin on glucose uptake by isolateddiaphram from normal and diabetic rats. Biochem Pharmacol 21:3153-3162.Frishman WH (1988) Beta-adrenergic receptor blockers. Adverse effects and druginteractions. Hypertension 11(Suppl 2):21-29.Galuska D, Zierath J, Thorne A, Sonnenfeld T and Wallberg-Henriksson H (1991)Metformin increases insulin-stimulated glucose transport in insulin-resistanthuman skeletal muscle. Diabete & Met 17:159-163.Giles TD and Sander GE (1989) Myocardial disease in hypertensive-diabetic patients.The Am J Med 87(Suppl 6A):6A-23S.Gin H, Slama G, Weissbrodt P, Poynard T, Vexiau P, Klein JC, and Tchobroutsky G(1982) Metformin reduces post-prandial insulin needs in type I (insulindependent) diabetic patients:assesment by the artifical pancreas. Diabetologia24:351-354.Gin H, Messerchmitt C, Brottier E and Aubertin J (1985) Metformin improved insulinresistance in type I insulin-dependent, diabetic patients. Metabolism 34:923-925.Gustafson A, Bjorntorp P and Fahlen M (1971) Metformin administration inhyperlipidemic states. Acta Med Scand 190:491-494.87Haffner SM, Valdez RA, Hazuda HP, Mitchell BD, Morales PA and Stern MP (1992)Prospective analysis of the insulin resistance syndrome (syndrome X). Diabetes41:715-722.Hamby RI, Zoneraich S and Sherman S (1974) Diabetic cardiomyopathy JAMA228:1749-1754.Hermann LS (1979) Metformin:A review of its pharmacological properties andtherapeutic use. Diabetes Metab Rev 5:233-245.Heyliger CB, Rodrigues B and McNeill JH (1986) Effects of choline and methioninetreatment on cardiac dysfunction in the diabetic rat. Diabetes 35:1152-1157.Heyliger CE, Prakash A and McNeill JH (1987) Alterations in cardiac sarcolemmalCa+2 pump activity during diabetes mellitus. Am J Physiol 252:540-544.Hunt SC, Williams RR, Smith JB and Ash KO (1986) Association of threeerythrocyte cation transport system with plasma lipids in the Utah subjectsHypertension 8:30-36.Hwang IS, Ho H, Hoffrnan BB and Reaven GM (1987) Fructose induced insulinresistance and hypertension in rats. Hypertension 10:512-516.Hwang IS, Huang WC, Wu JN, Shim LR and Reaven GM (1989) Effects of fructoseinduced hypertension on the renin angiotensin aldosterone system and atrialnaturetic factor. Am J Hyperten 2:424-427.88Ingebretsen CG, Moreay P, Hawelu-Johnson C and Ingebretsen WR Jr. (1980)Performance of the diabetic rat hearts:effect of anoxia and increased work. AmJ Physiol 239:H614-H620.Inouye I, Massie B and Loge D (1984) Abnormal left ventricular filling: an earlyfinding in mild to moderate systemic hypertension. Am J Cardiol 53:120-126.Jacobs DB, Hayes GR, Truglia JA and Lockwood DH (1986) Effects of metformin oninsulin receptor tyrosine kinase activity in rat adipocytes. Diabetologia 29:798-801Janka HU (1985) Platlet and endothelial function tests during metformin treatment indiabetes mellitus. Horm Metab Res 15(Suppl 1): 120-122.Jermendy G, Kammerer L, Koltai MZ et al. (1983) Preclinical abnormality of leftventricular performance in patients with insulin-dependent diabetes mellitus.Acta Diabetol Lat 20:311-320.Kahn RC and Shechter Y (1990) Insulin, oral hypoglycemic agents, and thepharmacology of the endocrine pancreas. In The Pharmacological Basis ofTherapeutics Volume II, Pergamon Press, p 1463-1495.Kannel WB, Wilson PWF and Zhang TJ (1991) The epidemiology of impairedglucose tolerance and hypertension. Am H J121(4 part 2):1268-1273.Kannel WB, Hjortland M and Castelli WP (1974) Role of diabetes in congestive heartfailure. Am J Cardiol 34:29-34.89Khadouri C, Barlet-Bas C and Doucet A (1987) Mechanism of increased tubular Na-K-ATPase during streptozotocin induced diabetes. Pflugers Arch 409:296-301.King GI, Goodman D, Bunzey S, Moses A and Kahn (1985) J Clin Inves 75:1028-1036.Kleinman KS and Fine LG (1988) Prognostic implication of renal hypertrophy anddiabetes mellitus. Diabetes Metab Rev 4:179-189.Lagadie-Grossman D, Chesnais JM, Fewray D (1988) Intracellular pH regulation inpapillary muscle cells from streptozotocin diabetic rats: an ion sensitivemicroelectrode study. Pflugers Arch 412:613-617.Landin K, Tengborn L and Smith M (1992) Treating insulin resistance in hypertensionwith metformin reduces both blood pressure and metabolic risk factors. J IntMed 229:181-187.Landsberg L and Krieger DR (1989) Obesity, metabolism and the sympatheticnervous system. Am J Hyperten 2:125S-132S.Landsberg L and Young JB (1978) Fasting, feeding and the regulation of thesympathetic nervous system. New Engl J Med 298:1295-1301.Leblanc H, Mane M, Billault B and Passa PH (1987) Combined subcutaneous insulininfusion in 10 overweight insulin requiring diabetic patients. Diabete & Metab13:613-617.90Ledet B, Neubauer B, Christensen NJ and Lundback K (1979) Diabeticcardiomyopathy. Diabetologia 16:207-209.Leslie P, Jung RT and Isles TE (1986) Effect of optimal glycaemic control withcontinous subcutanous insulin infusion on energy expenditure in type I diabetesmellitus. Br Med J293:1121-1126.Liang C, Doherty JU, Faillace R, Maekawa K, Arnold S, Gavras H and Hood WB Jr(1982) Insulin infusion in conscious dogs: effects on systemic and coronaryhemodynamics, regional blood flow and plasma catecholamines. J Clin Inves69:1321-1336Lithell HO (1991) Effects of antihypertensive drugs on insulin glucose and lipidmetabolism. Diabetes Care 14:203-209.LoGefro FW and Coffman JD (1984) Vascular and microvascular disease of the footin diabetes. N Engl J Med 311:1615-1618Lopaschuck GD, Tahiliani A, Vadlamudli RVSV, Katz S and McNeill JH (1983)Cardiac sarcoplasmic reticulum function in insulin or carnitine-treated diabeticrats. Am J Physiol 245:969-976.Lord JM, White SI, Bailey CJ, Atkins TW, Fletcher RF and Taylor KG (1983) Effectof metformin on insulin receptor binding and glycaemic control in type IIdiabetes. Br Med J 286:830-831.91Lord JM, Atkins TW and Bailey CJ (1983) Effect of metformin on hepatocyte insulinreceptor binding in normal, streptozotocin diabetic and genetically obesediabetic (ob/ob) mice. Diabetologia 25:108-113.Mahnensmith RL and Aronson PS (1985) The plasma membrane sodium hydrogenexchanger and its role in physiological and pathophysiological processes. CircRes 56:773-778.Malhotra A, Penpargkul S, Fein FS, Sonnenblick EH and Scheuer J (1981) The effectof streptozotocin-induced diabetes in rats on cardiac contractile proteins. CircRes 491:243-250.Marigliano A, Tedde R and Sechi LA (1990) Insulinemia and blood pressure.Relationships in patients with primary and secondary hypertension, and with orwithout glucose metabolism impairment. Am J Hyperten 3:521-526.Matthaei S, Hammann A, Klein HH, Benecke H, Kreymann G, Flier JS and Greten H(1991) Association of metformin's effect to increase insulin stimulated glucosetransport with potentiation of insulin induced translocation of glucosetransporters from intracellular pool to plasma membrane in rat adipocytes.Diabetes 40:850-859.McGee PA, Castelli WP, McNamara PM and Kannel WB (1971) The natural historyof congestive heart failure. The Framingham Study. N Engl J Med 285:1441-1446.92McNeill JH and Tahiliani AG (1986) Diabetes-induced cardiac changes. Trends inPharmacol Sci 7:364-367.Meyer F, Ipaktchi M and Clauser H (1967) Specific inhibition of gluconeogenesis bybiguanides. Nature (Lond) 213:203-204.Mildenberger RR, Bar-Shlomo B, Druck MN et al. (1984) Clinically unrecongnizedventricular dysfunction in young diabetic patients. J Am Coll Cardiol 4:234-238.Modan M, Halkin H, Almong S, et al (1985) Hyperinsulinemia:a link betweenhypertension, obesity and glucose intolerance. J Clin Invest 75:809-817.Mogensen CE and Christensen CK (1984) Predicting diabetic nephropathy in insulin-dependent patients. N Engl J Med 311:89-94.Mondon CE, Dolkas CB and Reaven GM (1980) Site of enhanced insulin sensitivityin exercise trained rats at rest. Am J Physiol 239:E169-E177.Mondon CE and Reaven GM (1988) Evidence of abnormalities of insulin metabolismin rats with spontaneous hypertension. Metabolism 37:303-305.Montaguti U, Celin D, Ceredi C, Descovich GC (1979) Efficacy of long-termadministration of metformin in hyperlipidemic patients. Res Clin Forums 1:95-103.93Moore RD (1985) In Molecular basis of insulin action, Czech MP, Ed. New York,Plenum Press p 145-170.Nakao J, Ito H, Kanayasu T and Murota SI (1985) Stimulatining effect of insulin onaortic smooth muscle cell migration induced by 12-L-hydroxy-5,8,10,14eicosatetraenoic acid and its modulation by elevated glucose levels. Diabetes34:185-91.Nichol TA, Lopaschuck GD and McNeill JH (1991) Effects of free fatty acids anddichioroacetate on isolated working diabetic rat heart. Am J Physiol261:H1053-H1059.Nosadini R, Avogaro A, Travisan R, Valerio A, Tessari P, Duner E, Tiengo A,Velussi M, Del Prato S, Kreutzenberg S, Muggeo M and Crepaldi G (1987)Effect of metformin on insulin-stimulated glucose turnover and insulin bindingto receptors in type II diabetes. Diabetes Care 10:62-67.Parillo M, Coulston A, Hollenbeck C and Reaven GM (1988) Effect of a low fat dieton carbohydrate metabolism in patients with hypertension. Hypertension11:244-248.Pedersen 0, Nielsen 0, Bak J, Rechelsen B, Beck-Neilsen H and Sorensen N (1989)The effects of metformin on adipocyte insulin action and metabolic control inobese subjects with type II diabetes. Diabetic Med 6:249-56.Peirce N, Beamish RE and Dhalla NS (1988) Heart dysfunction in diabetes. CRCPress, Inc Boca Raton, Florida.94Penicaud L, Hitier Y, Ferre P and Girard J (1989) Hypoglycemic effect of metforminin genetically obese (fa/fa) rats results from an increased utilization of bloodglucose by intestine. Biochem J262:881-85.Penpargul S, Fein F, Sonnenblick EH and Scheuer J (1981) Depressed cardiacsarcoplasmic reticular function from diabetic rats. J Mol Cell Cardiol 13:303-309, 1981.Penpargul S, Scharble T, Vipintsoi T and Scheuer J (1980) The effects of diabetes onthe function and metabolism of rat hearts. Circ Res 47:911-921.Pfeifle B and Ditschuneit H (1981) Effects of insulin on growth of cultured humanarterial smooth muscle cells. Diabetologia 20:155-158.Pollare T, Lithell H, Selinus I and Berne C (1989) Sensitivity to insulin duringtreatment with atenolol and metoprolol: a randomized double blind study ofeffects on carbohydrate, lipoprotein metabolism in hypertensive patients. BrMed J 298:1152-1157.Prager R, Schernthaner G and Graf H (1986) Effect of metformin on peripheral insulinsensitivity in non insulin dependent diabetes mellitus. Diabetes Metab Rev12:346-350.Prager R and Schernthaner G (1983) Insulin receptor binding to monocytes,insulinsecretion and glucose tolerance following metformin treatment. Diabetes32:1083-1086.95Rains SGH, Wilson GA, Richmond W, Elkeles RS (1988) The effect of glibenclamideand metformin on serum lipoproteins in type 2 diabetes. Diabetic Med 5:653-658.Rakienten N, Rakienten ML and Nadkarni MV (1963) Studies on the diabetogeniceffects of streptozotocin (NSC-37917) Can Chem Rep 29: 91-98.Ramsay RC et al. (1988) Progression of diabetic retinopathy after pancreastransplantation in insulin-dependent diabetes mellitus. N Engl J Med 318:208.Reaven GM, Ho H, Hoffman BB (1989) Somatostatin inhibition of frucotse-inducedhypertension. Hypertension 14:117-120.Reaven GM, Ho H, Hoffman BB (1988) Attenuation of fructose-inducedhypertension in rats by exercise training. Hypertension 12:129-132.Reaven GM, Chang H, Hoffman BB, Azhar S (1989) Resistance to insulin stimulatedglucose uptake in adipocytes isolated from spontaneously hypertensive rats.Diabetes 38:1155-1160.Reaven GM (1988) Role of insulin resistance in human disease. Diabetes 37:1495-1507Reaven GM (1991) Insulin resistance and compensatory hyperinsulinemia: role inhypertension, dyslipidemia and coronary heart disease. Am H J121:1283-1288.96Reaven GM (1991) Insulin resistance, hyperinsulinemia, hypertriglyceridemia andhypertension: existence in patients and rodent models. In Hypertension as aninsulin resistant disorder: Genetic Factors and Cellular Mechanisms Smith U,Bruun NE, Hedner T and Hokfelt B, Eds. Excerptica Medica p 55-67.Reaven GM (1991) Insulin resistance and compensatory hyperinsulinemia:Role inhypertension, dyslipidemia, and coronary heart disease. Am H J121(4 part 2):1283-1287.Reaven GM and Chang H (1991) Relationship between blood pressure, plasma insulinand triglyceride concentration, and insulin activity in spontaneouslyhypertensive rats and Wistar kyoto rats. Am J Hypert 4:34-38.Regan TJ, Lyons MM, Ahmed SS, Levinson GE, Oldewurtel HA, Ahmed MR andHaider B (1977) Evidence for cardiomyopathy in familial diabetes mellitus. JClM Invest 60:885-899.Regan TJ, Wu FC, Yey KC, Oldewurtel HA and Haider B (1981) Myocardialcomposition and function in diabetes. Circ Res 49:1268-1277.Regan TJ, Ettinger PO, Khan MI, Jesran MU, Lyons MM, Oldewurtel HA and WeberM (1974) Altered myocardial function and metabolism in chronic diabeticmellitus without ischemia in dogs. Circ Res 35:222-237.Regan TJ (1983) Congestive heart failure in the diabetic. Ann Rev Med 34:161-168.97Rizkalla SW, Elgrably F, Tchobroutsky G and Slama G (1986) Effects of metformintreatment of erythrocyte insulin binding in normal weight subjects, in obesenon diabetic subjects and type I and type II diabetic patients. Diabetes MetabRev 12:219-224.Rocchini AP, Katch V, Kveselis D, Moorehead C, Martin M, Lampman R andGregory M (1989) Insulin and renal sodium retention in obese adolescents.Hypert 14:367-374.Rocchini AP, Katch V, Anderson J, Hinderliter J, Becque D, Martin M and Marks C(1989) Blood pressure in obese adolescents: effects of weight loss. Pediatrics82:16-23.Rocchini AP, Moorehead C, DeRemer S, Goodfriend TL and Ball DL (1990)Hyperinsulinemia and the aldosterone and pressor responses to angiotensin IIHyperten 15:861-866.Rocchini AP, Moorehead C, Wentz E and DeRemer S (1987) Obesity inducedhypertension in the dog. Hypertension 9(Suppl III) III-64-111-68.Rocchini AP (1992) Is insulin resistance in hypertension a generalized phenomena?.In Hypertension as an insulin resistant disorder: Genetic Factors and CellularMechanisms. Smith U, Bruun NE, Hedner T and Hokfelt B, Eds. ExcerptaMedica, p 213-225.98Rodrigues B, Xiang H and McNeil JH (1988) Effect of L-carnitine treatment on lipidmetabolism and cardiac performance in chronically diabetic rats. Diabetes37:1358-1364.Rodrigues B and McNeill JH (1986) Cardiac function in spontaneously hypertensivediabetic rats. Am J Physiol 251:H571-H580.Rodrigues B and McNeill JH (1992) The diabetic heart:metabolic causes for thedevelopment of a cardiomyopathy. Cardiovasc Res 26:913-922.Rossetti L, DeFronzo RA, Gherzi R, Stein P, Andraghetti G, Falzetti G, Shulman GI,Klein-Robbenhaar E and Cordera R (1990) Effect of metformin treatment oninsulin action in diabetic rats: in vivo and in vitro correlations. Metabolism39:425-435.Rowe JW, Young BY, Minaker KL, Stevens-AL, Pallata J and Landsberg L (1981)Effects of insulin and glucose infusion on sympathetic nervous system activityin normal man. Diabetes 30:219-225.Rubler S, Dlugash J, Yuceoglu YZ et al.(1972) New type of cardiomyopathyassociated with diabetic glomerulosclerosis. Am J Cardiol 30:595-602.Sequist ER et al. (1989) Evidence for genetic susceptibility and diabetic nephropathy.N Engl J Med 320:1161.99Shapiro LM, Leatherdale BA, Mackinnon J and Fletcher RF (1981) Left ventricularfunction in diabetes mellitus. II:Relation between clinical features and leftventricular function. Br Heart J 45:129-132.Shipp JC, Menahan LA, Crass MF and Chandhuri SN (1973) Heart triglycerides inhealth and disease In Recent Advances in Studies on cardiac Structure andMetabolism, Dhalla NS, Ed. Baltimore University Park p 179-204.Sinha MK, Buchanan C, Leggett N, Martin L, Khazanie PG, Dimarchi R, Pories WJand Caro JF (1989) Mechanism of IGF stimulated glucose transport in humanadipocytes. Demonstration of specific IGF-I receptors not involved instimulation of glucose transport.Diabetes 38:1217-1225.Sirtori CR, Tremoli E, Sirtori M, Conti F, and Paoletti R (1977) Treatment ofhypertriglycerdemia with metformin. Atherosclerosis 26:853-862.Sirtori CR, Franceshini G, Gianfranschi G, Sirtori M, Montanan G, Bosisio E,Mantero E and Bondioli A (1984) Metformin improves peripheral vascularflow in non hyperlipidemic patients with arterial disease. J CardiovasPharmacol 6:914-923.Sirtori CR (1988) Clinical experience with metformin in patients with peripheralvascular disease. Diabete & Metab 14:601-607.Skarfors ET, Lithell HO, Selinus I and Aber H (1989) Do antihypertensive drugsprecipitate diabetes in predisposed men ? Br Med J 298:1147-1151.100Smith U (1991) Insulin resistance and hypertension-a relationship with severalpossibilities In Hypertension as an insulin-resistant disorder: Genetic factorsand Cellular Mechanisms, Smith U, Bruun NE, Hedner T and Hokfelt B, Eds.Excerptica Medica p 447-450.Sowers JR, Khoury S, Standley P, Zemel P and Zemel M (1991) Mechanism ofhypertension in diabetes. Am J Hypert 4:177-182.Stout RW, Bierman E and Ross R (1975) Effect of insulin on the proliferation ofcultured primate arterial smooth muscle cells.Circ Res 36:319-327.Stout RW (1990) Insulin and atheroma. 20 year perspective. Diabetes Care 13:631-654.Tahiliani AG, Lopaschuck GD and McNeill JH (1983) Effect of insulin treatment onlong term diabetes induced alterations of myocardial function. Gen Pharmacol15:545-547.Tahiliani AG and McNeill JH (1986) Effects of insulin perfusion and altered glucoseconcentration on heart function of 3-day and 6-day diabetic rats. Can J PhysiolPharmacol 64:188-192.Tedde R, Cechi LA, Marigliano A, Scano L, Pala A (1988) In vitro action of insulinon erythrocyte sodium transport mechanims, its possible role in thepathogenesis of arterial hypertention. Clin Exp Hypertens 10:545-559.101The University Group Diabetes Program (1975) A study of the effects ofhypoglycemic agents on vascular complications in patients with adult-onsetdiabetes. Evaluation of phenformin therapy. Diabetes 24(Suppl):165-184.Tsutsu N, Nunoi K, Kodama T, Nomiyama R, Iwase M, Fujishiana M (1990) Lack ofassociation between blood pressure and insulin in patients with insulinomas. JHperten 8:479-482.Tucker GT, Casey C, Phillips PJ, Conor H, Ward JD and Woods HF (1981)Metformin kinetics in healthy subjects and in patients with diabetes mellitus.Br J Clin Pharmacol 12:235-246.Uusitupa MIT, Mustonen JN and Airaksinen KEJ (1990) Diabetic heart muscledisease. Annals of Med 22:377-386.Vadlamudli RVSV, Rodgers RC and McNeill.JH (1982) The effect of chronic alloxanand streptozotocin-induced diabetes on isolated rat heart performance. Can JPhysiol 60:902-911.Vered Z, Battler A, Segal P et al. (1984) Exercise-induced left ventricular dysfunctionin young men with asymptomatic diabetes mellitus (Diabetic cardiomyopathy).Am J Cardiol 54:633-637.Vierhapper H, Waldhausl W and Nowontny P (1983) Effects of insulin on the rise inblood pressure and plasma aldosterone after angiotensin II in normal man. ClinSc 64:383-386.102Wales JK (1980) Treatment of the obese diabetic patient. In Recent Advances inObesity Research. Bjorntorp P, Cairelli M, Howard AN, Eds. London,Libby, p184-189.Weidman P, Beretta-Piccoli C, Trost BN (1985) Pressure factors and responsivenessin hypertension accompanying diabetes mellitus. Hypertension 7(Suppl II):32-42.Weinberger MIT (1986) Antihypertensive therapy and lipds. Paradoxical influences oncardiovascular disease risks. Am J Med 80(Suppl 2A):64-70.Welborn TA, Breckenridge A, Rubinstein AH, Dolley CT and Fraser TR (1966)Serum insulin in essential hypertension and in peripheral vascular disease.Lancet 1:1336-1337.Wilcock C and Bailey CJ (1990) Sites of metformin-stimulated glucose metabolism.Biochem Pharmacol 39:1831-1834.Wollen N and Bailey CJ (1988) Inhibition of hepatic gluconeogenesis by metformin:synergism with insulin. Biochem Pharmacol 37:4353-4358.Wollen N and Bailey CJ (1988) Metformin potentiates the antihyperglycemic actionof insulin. Diabete & Metab 14:88-91.Wu MS, Johnston P, Sheu WHH, Hollenbeck CB, Jeng CY, Goldfine ID, Chen Y-DIand Reaven GM (1990) Effect of metformin on carbohydrate and lipoproteinmetabolism in NIDDM patients. Diabetes Care 13:1-8.103Young JB, Rosa RM and Landsberg L (1984) Dissociation of sympathetic nervoussystem and adrenal medullary responses. Am J Physiol 247:E35-E40.Young JB, Saville ME, Rothwell NJ, Stock MJ and Landsberg L (1982) Effect of dietand cold exposure on norepinephrine turnover in brown adipose tissue of therat. J Clin Invest 69:1061-1071.Young JB, Kaufman LN, Saville ME and Landsberg L (1985) Increased sympatheticnervous system activity in rats fed a low protein diet. Am J Physiol 248:R627-637.Young JB (1988) Effect of experimental hyperinsulinemia on sympathetic nervoussystem activity in the rat. Life Sci 43:193-200.Yuen V, Orvig C, Thompson KH and McNeill JH (1993) Improvement in cardiacdysfunction in STZ-induced diabetic rats following chronic oral administrationof bis(maltolato)oxovanadium (IV). Can J Physiol Pharmacol (in press).Zarich SW and Nesto RW (1989) Diabetic cardiomyopathy. Am H J118:1000-1012.Zavaroni I, Dall Aglio E, Bruschi F, Alpi C, Coscelli C and Butturini U (1984)Inhibition of carbohydrate-induced hypertriglyceridemia by metformin. HormMetab Res 16:85-87.104Zavaroni I, Dall'Aglio E, Bonora E, Alpi 0, Passeri M and Reaven GM (1987)Evidence that multiple risk factors for coronary artery disease exist in personswith abnormal glucose tolerance. Am J Med 83:609-612.Zavaroni I, Sander S, Scott S and Reaven GM (1980) Effects of fructose feeding oninsulin secretion and insulin action in the rat. Metabolism 29:970-973.Zemel MB, Reddy S and Sowers JR (1991) Insulin attenuation of vasoconstrictorresponses to phenylephrine in Zucker lean and obese rats. Am J Hypert 4:537-539.105

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086319/manifest

Comment

Related Items