Dissertations / Theses on the topic 'Hemodynamic effects of insulin'

Hemodynamic effects of insulin resistance (IR) are thought to be largely dependent on its relationship with body mass index (BMI) and blood pressure (BP) levels. The first part of the present thesis was aimed at exploring whether IR is associated with hemodynamic indices of cardiovascular function in a large sample of non-diabetic individuals from the general population (n=731) and if so, to explore if such relationship is continuous across different categories of BMI (lean, overweight and obese), and BP (normal BP, high-normal BP and hypertension). IR was assessed with the homeostasis model assessment of IR (HOMA-IR). Based on a value of HOMA-IR of 2.09 (75th percentile of distribution curve), subjects were classified as insulin-sensitive (IS, HOMA<2.09) or insulin-resistant (IR, HOMA≥2.09). Synchronized beat-to-beat recordings of stroke volume (impedance cardiography) and R-R interval (ECG), along with repeated BP measurements were performed over 5 minutes. Stroke index (SI), cardiac index (CI), systemic vascular resistance index (SVRI), left cardiac work index (LCWI), pre-ejection period (PEP) and left ventricular ejection time (LVET) were computed and averaged. In analysis of co-variance allowing for confounders, IR subjects showed significantly higher BP levels and SVRI, and reduced R-R interval, SI, CI, LCWI, PEP and LVET. These differences remained significant when analyses were performed within each BMI and BP category. Overall, these results indicate that effects of IR on hemodynamic indices of cardiovascular function are continuous across different BMI and BP categories, reinforcing the importance of IR in the pathogenesis of cardiovascular alterations beyond its association with obesity and hypertension. The finding of a significant association between IR and hemodynamic alterations even in lean and normotensive subjects was the rationale to explore potential mechanisms for these alterations in this selected group of subjects. Specific objectives of this second part of the thesis were: 1) To explore the relationship between insulin resistance and systemic hemodynamics, cardiac baroreflex sensitivity and indices of autonomic CV modulation. 2) To explore the relationship of insulin resistance with 24h heart rate, average blood pressure levels and blood pressure variability over the 24h; and 3) To explore the relationship of insulin resistance with central blood pressure levels and with measures of large artery stiffness and wave reflections. The study population for these analyses was constituted by subjects who were below the 30th percentile of diastolic blood pressure (DBP) distribution curve (DBP ≤72 mmHg) and who had no elevation in systolic BP levels. In addition, subjects were excluded in case of diabetes mellitus (fasting blood glucose ≥126 mg/dL or use of medications for previously diagnosed type 2 diabetes) obesity (BMI≥30) or taking medications with effects on BP. A total of 90 subjects fulfilling inclusion criteria were considered for the present analysis and underwent further assessments. Insulin resistance was assessed with HOMA-index and subjects classified into IR tertiles, based on the distribution of HOMA-index values. 24h Ambulatory BP monitoring was performed. Mean SBP and DBP were averaged for the day, night and 24h, and the respective day-to-night dipping was calculated. BPV was assessed for SBP and DBP as 24h standard deviation (SD), weighted 24h SD (wSD), daytime and night-time SD. Recordings of pulse waveform were obtained by means of a previously validated oscillometric device for ambulatory BP monitoring with in-built transfer-function like method. Aortic pulse wave velocity (PWV, m/s) and other measures derived from pulse wave analysis such as augmentation index (AIx, %), central SBP (cSBP), central DBP (cDBP) and central pulse pressure (cPP) were computed. Peripheral SBP and DBP, and heart rate (HR) were recorded and pulse pressure (PP) calculated as the difference between SBP and DBP. Non-invasive assessment of beat-to-beat BP, R-R interval (ECG) and stroke volume (by means of impedance cardiography) were performed during 10 min in supine position and specific hemodynamic indices associated with their measurement were computed and averaged: RRI (msec), heart rate (HR, bpm), stroke volume index (SI, mL/beat/m2), cardiac index (CI, L/min/m2), SBP (mmHg) and DBP (mmHg), systemic vascular resistance index (SVRI, dyn/sec/cm-5/m2), left cardiac work index (LCWI, Kg/m/m2), pre-ejection period (PEP, msec), left ventricular ejection time (LVET, msec) and PEP/LVET ratio were calculated. Cardiac autonomic modulation was assessed by computer analysis of 10 min beat-to-beat BP and ECG recordings in resting supine position. Cardiac baroreflex sensitivity (BRS) was estimated by sequence method. Total variance, low-frequency (LF) and high-frequency (HF) spectral components of HR variability (HRV) were assessed by autoregressive analysis. LF/HF ratio was calculated. After multiple regression analysis, adjusting for common confounders such as age, sex, HR and BMI, increasing values of HOMA-IR were associated with reduced RRI, SI, CI, and with increased SVRI, SBP and DBP. IR was also associated with reduced BRS (up, down, and total slopes), decreased parasympathetic indices of autonomic CV modulation (SDRRI, HF-power, total power) and a predominance of sympathetic component of HRV (increased LF/HF ratio). Increasing values of HOMA-IR were also associated with increased HR and average SBP levels (during day, night and 24-h period), with augmented BP variability (Day SBP SD, and SBP wSD) and with a reduced dipping of HR. Finally, insulin resistance was shown to be associated with increasing values of aortic PWV, and with higher central and peripheral SBP and DBP levels. Overall, these results support significant associations between insulin resistance and changes in hemodynamic and autonomic indices of cardiovascular function, even after accounting for common confounders. These findings suggest that in normotensive healthy adults, increases in insulin resistance may promote alterations in autonomic cardiovascular modulation, in systemic hemodynamics and in arterial stiffness, all of which are known contributors to the pathogenesis of hypertension.

Type 2 diabetes mellitus is a chronic condition caused by a deficiency in the secretion of insulin from the islets of Langerhans and/or impaired insulin signalling, resulting in hyperglycaemia. The role of the endocannabinoid system is well-recognised in the CNS and immune system, but its role in glucose homeostasis is poorly understood. The aim of this study was to define the roles of cannabinoids in insulin secretion, to provide insights into their therapeutic potential (or limitation) in the treatment of type 2 diabetes. Isolated islets were used, from Wistar rats, in static incubation studies measuring changes in insulin secretion rates. The endocannabinoid anandamide (AEA) was found to inhibit insulin secretion in a glucose- and concentration-dependent manner, with an IC50 of 1.6μM (95% CI: 227nM to 4.0μM; n= 10). Upon further analysis of the concentration-response data islet sensitivity to AEA appeared to vary, with islets either appearing to be sensative (IC50 220nM; 95% CI: 21.9nM to 2.2μM; n= 5) or less sensative (IC50 12.3μM; 95% CI: 6.8μM to 19.4μM; n= 5) to AEA. Pre-incubation of islets with a fatty acid amide hydrolase inhibitor did not affect islet responsiveness to AEA. AEA-mediated inhibition of insulin secretion was not consistently affected by cannabinoid receptor 1 (CB1) or CB2 antagonism. Surprisingly, the CB1 receptor antagonist AM251 was found to inhibit insulin secretion in a glucose- and concentration-dependent (IC50 1.6μM; 95% CI: 507nM to 3.3μM; n= 6) manner. Results from this study suggest that differences in CB-receptor signalling pathways, rather than endocannabinoid metabolism, could be responsible for the variations in the potency of AEA between islet preparations. Characterisation of cannabinoid signalling in islets was hindered as the CB receptor antagonists used in this study also affected insulin secretion. This study highlights the dynamics of endocannabinoid signalling in islets, which may be linked to their physiological function.

The preadipocyte is crucial for healthy adipose tissue (AT) remodeling, and insulin resistance in these cells may contribute to AT dysfunction. Chronic exposure to insulin and high glucose induces insulin resistance in the 3T3-L1 mouse adipocyte cell line in vitro, however, whether this occurs in human preadipocytes is not known. To investigate this, human preadipocytes were isolated from subcutaneous AT obtained from 6 female patients undergoing elective surgery (Research Ethics Board-approved). Human preadipocytes were incubated in 5 mM glucose or 25 mM glucose in the presence or absence of 0.6 nM insulin for 48 hours, followed by acute 100 nM insulin stimulation. 25 mM glucose + 0.6 nM insulin inhibited insulin-stimulated tyrosine phosphorylation of IR-β (77%) and IRS-1 (81%) compared to NG (p<0.01), however, insulin-stimulated Ser473 Akt phosphorylation was not affected. 25 mM glucose and/or 0.6 nM insulin did not significantly change levels of pro-inflammatory adipokines. 25 mM glucose and/or 0.6 nM, prior to and/or during 14 days of adipogenic induction, did not affect levels of adipogenic markers or intracellular triglyceride accumulation.

Suspension of the body in the head-down posture (90° below the horizontal) for traction and added resistance to exercise has generated considerable interest in recent years. However, recent investigators of inversion have cautioned individuals not to participate in such activities until further research could be performed on the effects of exercise in the head-down position. The purpose of this investigation was to examine the acute metabolic and hemodynamic responses of men at rest and during exercise in the inverted posture (90° head-down tilt) versus the supine and standing postures. The parameters investigated were oxygen consumption(V̇O₂), heart rate(HR), systolic blood pressure (SBP), and diastolic blood pressure(DBP). Eleven male recreational athletes underwent 6 sessions of postural change. The baseline posture was sitting and the critical positions were supine, standing, and inverted. The subjects were asked to remain in each of these postures for three minutes. In the first 2 sessions, oxygen consumption(V̇O₂), was measured at rest and during 45° hip-f1exion respectively, The V̇O₂ in the inverted posture at rest was found to be 1.7% greater than the V̇O₂ in the standing and the supine postures. V̇O₂ in the inverted posture during exercise was 7% and 36.5% greater than in the supine and standing postures, respective1y. A statistical significance in HR at rest in the standing posture versus the inverted and supine postures was observed. During 45° hip-flexion activity, the HR in the standing posture was found to be significantly faster than the supine posture. The HR in the inverted posture was significantly faster than the supine posture as well. At rest, there was no significant increase in SBP as re1ated to posture. During exercise, the SBP was significant1y greater; at rest in each of the postures. Both postural and exercise factors significantly affected the DBP. The post-hoc analyses showed supine resting DBP was significantly lower than in the other two resting postures. During exercise, the standing BP was significantly greater than the supine and inverted DBP. These data demonstrate: A) V̇O₂ in the supine and inverted postures is significantly greater than in the standing postures. B) a statistically significant increase in HR occurs in the standing posture as compared to the supine and inverted postures; however, it does not appear to be clinically significant, C) with the arms maintained in the anatomical position for all postural changes, the SBP was not significantly affected by the change of posture, but was significantly increased with exercise. D) and DBP in the standing posture was statistically greater than in the other two postures and DBP in the inverted posture was significantly elevated above that found in the supine posture.
Master of Science

Chapter one provides an overview of the regulation of insulin secretion/action, details normal insulin action, and, describes the concept of insulin resistance and its assessment. It specifically focuses on obesity (as the most important contributor to insulin resistance) and reviews dietary effects on insulin resistance and plasma lipids. Chapter two describes methods which are common to chapters 3, 4 and 5: assessment of habitual diet; formulation/practical administration of intervention diets; blood sampling; and assessment of body composition, insulin sensitivity, and vascular compliance. Chapter three compares the effects of a high versus low sucrose diet (25 vs. 10 %, respectively, oftotal energy intake) in 13 healthy subjects, in a randomised crossover design with sequential 6 week dietary interventions separated by a 4 week washout. Diets were weight maintaining with identical macronutrient and fibre content. The results demonstrated that a high sucrose intake, as part of an isocaloric weight maintaining diet, had no detrimental effect oninsulin resistance. Chapter four investigates the effects of a low carbohydrate versus low fat weight reduction diet (0.5 kg/week) on insulin sensitivity and cardiovascular risk in overweight/obese subjects (n = 24) using a parallel group randomised controlled trial design. Following similar weight loss, both diets were equally effective in improving insulin sensitivity. The low fat diet also had beneficial effects on augmentation index (a measure of arterial stiffness), a finding which was not evident within the low carbohydrate group. Chapter five explores whether adipokines (leptin, adiponectin and retinol binding protein 4) potentially mediate the effects of weight loss (induced by hypocaloric dieting in study 2) on insulin sensitivity. Results demonstrated that diet induced changes in adipokines did not correlate with the change in insulin sensitivity. Chapter six provides a general discussion and concluding remarks for the main body of the thesis.

Diabetes mellitus is not only a very serious disease, causing discomfort and pain to millions across the world, but with the aging of the population and the prevalence of a sedentary lifestyle, it is assuming the proportion of a real epidemic, becoming a public health and social emergency. In answer to this call, research on diabetes has been intensely carried out in the past decades and the knowledge and understanding of its etiology have been significantly improved. However, investigation is still ongoing, many important questions are still unanswered, and the causes eliciting the pathogenesis and progression of the diseases are not yet fully understood. During all these years of research, sophisticated tools have been developed to study the glucose-insulin metabolic system in vivo, and cope with the inaccessibility to direct measurement of some of the key phenomena underlying the glycemic control. Such tools, as complex test protocols and model-based approaches used to interpret the experimental data, have proven powerful weapons of investigation, but, considering the epidemic proportion of the disease, there is great demand for approaches that are less invasive, less expensive and therefore more suitable for large clinical studies. Mathematical and statistical techniques that collectively go under the name of “population approaches” have already been developed and are largely employed for pharmacokinetics and pharmacodynamics studies in drug development. However, in spite of their interesting potential, they have not found significant application yet in the context of metabolism research. Thus, investigation is required to probe the feasibility and relevance of such approaches in the study of diabetes. The research presented here addresses these issues, and is aimed at applying these sophisticated techniques to the modeling of glucose metabolism, first assessing the applicability of these approaches and tailoring them to the problem under investigation, then employing them to the analysis of data from population studies. First, a simulated but physiologically plausible dataset is created based on previous real data and employed as a benchmark to assess the applicability of population approaches to the Intra Venous Glucose Tolerance Test (IVGTT) minimal model of glucose disappearance. Various population algorithms have been proposed in the literature, therefore a thorough comparison of the available methodologies is performed, and a sparse data situation is replicated to test the robustness of these methods in such cases. The results select the First-Order Conditional Estimation as method of choice and show its robustness to poor sampling. Then, a larger real dataset is employed and analyzed with the same techniques, this time assessing the quality of the results with a Monte Carlo sampling approach to profile the likelihood function. Then the population model is optimized, to provide a base model for the following covariate analysis. In fact, at the time of the experiments, demographic data about the subjects has been collected, and the purpose of the covariate analysis is to determine whether some of these variables are significantly correlated with the model parameters and can be successfully used to explain part of the differences among the subjects. After a first exploratory regression analysis, different models are tested, integrating the most significant covariates directly as predictors into the model. In agreement with previous findings in literature, basal insulinemia, age and visceral abdominal fat are shown to be good predictors of insulin sensitivity and their introduction in the model is able to account for about a third of the between-subject variability of the values of this parameter. The use of covariates enhances the explanatory power of the model and opens the way for devising new lighter experimental protocols. One of the main benefits of the population approaches, in fact, consists in their ability to borrow information across the population and use it to improve the individual parameter estimates. As a result, the experimental protocols can be less demanding, both in invasiveness and economic cost, allowing in this way a broader use in large clinical studies. The results presented here, in fact, show that population approaches are very robust and able to cope with sparse data situations. In addition, the use of covariates in the model enhances even further the power of such techniques and makes them very appealing approaches to the study of glucose-insulin metabolism. In addition, a population approach is proposed to solve the problem of the estimation of the Disposition Index (DI) of glucose tolerance in a population. Since both insulin sensitivity and beta-cell response must be taken into account to assess the actual efficiency of the glucose disposal system, the DI was proposed to condense the information conveyed by both these parameters in a single value. Traditionally, approaches based on a geometrical fit are used to determine the value of DI in a population of subjects characterized by the same degree of glucose tolerance. However, all these methods rely on the assumption that all the subjects in the population share exactly the same value of DI and are therefore not able to account for the population variability, which is inevitably inherent to biological data. In this work, a NonLinear Mixed-Effects Approach is proposed to analyze the distribution of the insulin sensitivity and beta-cell response indices across a population, and then obtaining the information on the DI from the population features thus estimates. Comparisons on simulated datasets between the newly proposed method and its competitors prove that a proper model of the variability structure is essential to avoid severe bias in the estimates.
Il diabete mellito è non solo una patologia molto seria, che causa disagi e sofferenze a milioni di persone nel mondo, ma, anche a causa dell’affermarsi di uno stile di vita sedentario e dell’invecchiamento della popolazione, negli ultimi decenni ha raggiunto proporzioni epidemiche, diventando una vera e propria emergenza sanitaria e sociale. Per fronteggiare questo problema, molte risorse sono state dedicate all’attività di ricerca scientifica, che ha permesso una più profonda conoscenza dell’eziologia del diabete. Tuttavia, il diabete è a tutt’oggi ancora inguaribile e molte questioni rimangono aperte, fra cui la completa comprensione dei fattori che causano e fanno progredire la malattia. Anni di ricerca hanno permesso di sviluppare molti sofisticati strumenti per studiare il sistema metabolico glucosio-insulina in vivo e poter così fronteggiare il problema dell’inaccessibilità diretta di alcuni dei fenomeni chiave che controllano la glicemia. Tali strumenti, fra cui protocolli di studio e approcci basati su modello usati per interpretare i dati sperimentali, si sono rivelati armi molto potenti nelle mani dei ricercatori, ma le proporzioni epidemiche della malattia e il parziale cambiamento delle strategie e obiettivi della ricerca hanno sollevato l’esigenza di poter disporre di metodologie meno invasive, più economiche, e quindi più adatte ad essere applicate ad estesi studi clinici. Alcuni strumenti matematici e statistici che sono collettivamente conosciuti con il nome di “approcci di popolazione” sono già stati sviluppati e vengono largamente impiegati in studi di farmacocinetica e farmacodinamica, per lo sviluppo di farmaci. Tali approcci si prefiggono come obiettivo primario di stimare la distribuzione dei parametri di un modello all’interno di una popolazione e pertanto si avvalgono, per la stima individuale, delle informazioni disponibili sull’intero gruppo di soggetti. Sono particolarmente adatti a situazioni in cui il campionamento intensivo in un singolo soggetto non è possibile, e quando l’interesse del ricercatore è focalizzato sulla variabilità inter-individuale. Tuttavia, nonostante le loro interessanti potenzialità, gli approcci di popolazione non sono ancora apprezzati all’interno dell’ambiente di ricerca sulle malattie metaboliche, e la loro applicazione in tali studi è stata molto limitata. Pertanto è necessaria dell’attività di ricerca per saggiare l’effettiva fattibilità e rilevanza dell’utilizzo di tali approcci nello studio del diabete. La ricerca qui presentata risponde a queste esigenze, proponendosi come obiettivo l’applicazione di queste sofisticate tecniche ai modelli di metabolismo del glucosio, prima testandone la fattibilità e adattandole al problema in esame, e poi impiegandole nell’analisi di dati raccolti in studi di popolazione. Poiché in letteratura sono stati proposti molti diversi algoritmi, come primo passo, un dataset simulato è stato utilizzato per effettuare un confronto delle metodologie quando applicate al modello minimo del glucosio per il Test di Tolleranza IntraVenosa al Glucosio (IVGTT). First-Order Conditional Estimation (FOCE) si è rivelato come l’algoritmo più soddisfacente, in quanto ha fornito i risultati più accurati e robusti in caso di scarsità o rumorosità dei campioni. Successivamente, per validare i risultati trovati su dati reali, l’analisi è stata ripetuta su un dataset più esteso, relativo a 204 soggetti sani testati con IVGTT. Per poter saggiare la bontà delle soluzioni fornite dai vari algoritmi, è stato impiegato un sistema di stima della likelihood function basato su campionamento Monte Carlo. Questa analisi, non solo ha permesso di confermare la scelta di FOCE come metodo preferenziale, ma si è anche rivelata come un potente strumento per valutare la precisione delle stime dei parametri di popolazione. Successivamente, è stato messo a punto e ottimizzato un modello di popolazione, conservando nella matrice di covarianza solo i termini di correlazione fra i parametri SI-P2 e SG-VOL. Questo modello è servito come base per la successiva integrazione di covariate nel modello. Al momento dell’esecuzione degli esperimenti, infatti, sono stati raccolti alcuni dati sui pazienti, fra cui altezza, peso, sesso, età, glicemia e insulinemia basali, informazioni sul grasso corporeo. È stata effettuata una analisi per determinare quali fra queste variabili potessero essere usate per spiegare parte della variabilità nei valori dei parametri del modello minimo fra i diversi soggetti. Il risultato è un modello che integra queste informazioni direttamente nelle sue equazioni, mentre i coefficienti di regressione per ognuno dei predittori diventano veri e propri parametri del modello e il loro valore viene ottimizzato insieme agli altri parametri di popolazione. L’analisi effettuata ha trovato come buoni predittori per SI e P2 l’ età, l’insulinemia basale e il grasso addominale, che in ambo i parametri riescono a spiegare una buona fetta della variabilità inter-individuale. Sia l’impiego di metodologie di popolazione, sia l’introduzione delle covariate nel modello, permettono di aumentarne il potere predittivo, e sono in grado di usare informazioni indipendenti dai soli dati sperimentali. Questo permette di mettere a punto dei protocolli di studio meno invasivi, meno costosi, e pertanto più adatti ad un impiego su larga scala: ulteriore ricerca potrebbe avere come obiettivo l’ottimizzazione di una sampling schedule ridotta, che si avvantaggi dell’utilizzo degli approcci di popolazione. Ad ogni modo, il dataset utilizzato in questa analisi comprende solo soggetti sani, ed è quindi caratterizzato da una quantità limitata di variabilità di popolazione. Pertanto, sarebbe necessario ripetere l’analisi su altri dataset, per poter confermare questi risultati, in particolare sulle covariate. Inoltre, in una sezione successiva, un metodo di popolazione è stato applicato anche ad un altro problema diverso, la stima del Disposition Index (DI) del glucosio. Questo è un indice calcolato combinando sensitività e responsività all’insulina, che serve per testare l’effettiva efficacia del sistema di controllo della glicemia. Ci sono due versioni proposte per la formula, una semplificata, che consiste semplicemente nel prodotto (da cui il nome di Legge Iperbolica), e una con un parametro aggiuntivo ad esponente della sensitività all’insulina. Per poter calcolare il DI medio in una popolazione, e per poter saggiare quale delle due formule sia effettivamente più adatta, in letteratura si trovano alcuni approcci basati su un fit geometrico. Tuttavia, alcune approssimazioni sono utilizzate per semplificare il fit, e sono molte le questioni metodologiche spesso sottovalutate. Pertanto viene presentato qui un nuovo metodo Total Least Squares (TLS) che affronta il problema senza l’impiego di approssimazioni. Grazie ad alcune simulazioni, si è effettuato un paragone fra i vari metodi disponibili, e il nuovo algoritmo è risultato migliore rispetto ai predecessori. Tuttavia, tutti gli algoritmi basati su fit si fondano sull’ipotesi che i soggetti appartenenti alla popolazione abbiano lo stesso valore di DI, e l’unica fonte di incertezza nei dati sia dovuta alla stima degli indici di secrezione e sensitività. Questa ipotesi sembra una forte semplificazione e, in effetti, l’analisi di un dataset reale sembra confermare la presenza di variabilità di popolazione nei valori del DI. Ulteriori simulazioni hanno confermato che tutti metodi basati su fit, TLS compreso, falliscono quando la variabilità di popolazione è presente. Pertanto, è stato ideato un altro metodo basato su approcci di popolazione e, in particolare, su NonLinear Mixed-Effects Models (NLMEM), che è in grado di separare la variabilità nei dati, poiché fondato su ipotesi meno restrittive. Tale algoritmo stima i parametri della distribuzione di probabilità congiunta degli indici di secrezione e sensitività, e poi estrae le informazioni sul DI dalla matrice di covarianza. NLMEM si è rivelato equivalente a TLS quando non c’è variabilità di popolazione, ma di gran lunga più affidabile quando le ipotesi per il fit geometrico non sono rispettate, pertanto si è deciso di utilizzarlo sul dataset reale per testare la validità della legge iperbolica. Anche se una validazione su altri dataset è auspicabile per validare i risultati qui presentati, il modello con il parametro aggiuntivo sembra spiegare i dati in maniera più soddisfacente, e il valore del parametro sembra dipendere dalla coppia di parametri usata per la definizione del DI, più che dalla popolazione in esame (anziani piuttosto che giovani). Inoltre, nello studio qui proposto, il punto di partenza sono stati i valori degli indici di secrezione già calcolati, insieme con la loro precisione, grazie ad un metodo tradizionale; un approccio ancora più potente consisterebbe nell’utilizzare un modello di popolazione per stimare contemporaneamente sia gli indici di secrezione che sensitività, sia i parametri della loro distribuzione di popolazione, da cui ricavare le informazioni sul DI. Riassumendo, in questo lavoro si sono messi in luce i vantaggi dell’applicazione di approcci di popolazione nello studio nel diabete. Le potenzialità sono molte, dal miglioramento delle stime dei parametri individuali grazie all’uso dei prior di popolazione o di covariate e la relativa possibilità di mettere a punto protocolli di studio più leggeri, fino all’analisi di situazioni in cui la struttura gerarchica della variabilità è un aspetto cruciale.

The omega-3 polyunsaturated fatty acids (n-3 PUFAs) are a unique class of fatty acids that cannot be manufactured by the body, and must be acquired via dietary sources. In the UK, as well as in other Western nations, these ‘essential’ fatty acids are consumed in quantities that fall below government guidelines. This thesis examined the relationship between n-3 PUFAs and cognitive function and mood in healthy children (8-10 years) and adults (18-35 years), with a view to evaluate their efficacy for cognitive and mood enhancement in these populations. A second aim was to evaluate the effects of n-3 PUFAs on cerebral hemodynamics, a novel line of enquiry. Chapters 2 and 4 describe novel intervention studies that assessed the effects of n- 3 PUFA supplements on cognitive function and mood in healthy children and adults, respectively. In Chapter 3, the relationship between peripheral PUFA concentrations, a correlate of dietary PUFA intake, and cognitive and function and mood was examined for the first time in healthy adults. Chapter 5 describes a pilot trial in which Near Infrared Spectroscopy (NIRS) imaging technique was applied to investigate the cerebral hemodynamic effects of n-3 PUFA supplements. The results of this study were explored in more detail in Chapter 6, with the additional inclusion of parallel cognitive measures. Most notably, the behavioural data from the intervention studies described herein do not support the use of n-3 PUFA supplements for cognitive and mood enhancement in healthy children and adults not consuming appreciable amounts of oily fish. However, the results do suggest that supplementation with dietary n-3 PUFAs has an impact on peripheral fatty acid status and cerebral hemodynamics in healthy adults. Taken together, these findings suggest that, in healthy, cognitively intact individuals, short-term use of n-3 PUFA supplements has a minimal effect on behaviour; the impact of long-term n-3 PUFA dietary intake or supplement use over the course of the entire lifespan on behaviour should be addressed further.

Primary cultures of skeletal muscle satellite cells were induced to proliferate by exposure to physiologic levels of somatomedins and pharmacologic levels of insulin. Dexamethasone inclusion in serum containing medium facilitated the ovine somatomedin (oSm) (P < 0.05), but that both were different than the proliferation induced by MSA/rIGF-II (P < 0.05). In the presence of insulin concentrations that promote maximum proliferation, addition of oSm did not produce an additive effect, whereas the addition of MSA/rIGF-II did produce a significant increase in satellite cell proliferation above that induced by insulin. A more, in depth, analysis of the interaction of MSA/rIGF-II with its satellite cell receptor under a variety of experimental conditions revealed that binding of ¹²⁵I-MSA/rIGF-II was inhibited by oSm and MSA/rIGF-II, but not by insulin. Migration, and localization of ¹²⁵I-MSA/rIGF-II-receptor complexes in 7% sodium dodecyl sulfate polyacrylamide gels suggest that these complexes are Type II IGF receptors. In addition, this receptor system of satellite cells was shown to be modulated by other hormones; notably, pre-exposure of cells with insulin increased ¹²⁵I-MSA/rIGF-II binding, while oSm, or MSA/rIGF-II preincubation decreased the binding of ¹²⁵I-MSA/rIGF-II. Therefore, the proliferative effects of MSA/rIGF-II appeared not as a consequence of MSA/rIGF-II induction of other receptor types such as the insulin, or Type I IGF receptor systems. Concommitant to the previous experimentation, oSm was further examined in an initial attempt to elucidate its biologic binding mechanism in myogenic satellite cells. Binding of ¹²⁵I-oSm was inhibited by MSA/rIGF-II, insulin and IGF-I; thus these data suggest that oSm may be the ovine analog to human IGF-I. In addition, pre-exposure of cells to MSA/rIGF-II and oSm down-regulated the ability of satellite cells to bind oSm, while only concentrations of insulin greater than 550 ng insulin had this ability. Collectively, these data support the hypothesis that somatomedins play an important role in the control of postnatal muscle growth by providing a link between these hormones and satellite cells, one of the significant target cells involved in the growth process.

The effect of varying concentrations of glucose or the gastrointestinal hormones, gastric inhibitory polypeptide (GIP) and somatostatin (SS-14), on the in vitro immunoreactive insulin (IRI) response to the parasympathetic neurotransmitter, acetylcholine (ACh) was investigated. The isolated, vascularly perfused rat pancreas was used in all experiments. Acetylcholine (1.0 µM) did not stimulate IRI secretion in the presence of 2.2 mM glucose. However, in the presence of 4.4, 6.6, or 8.9 mM glucose, ACh (1.0 µM) potently stimulated IRI secretion (approximately fourfold). At a higher glucose concentration (17.8 mM), the IRI response to ACh was reduced. GIP also potentiated the IRI response to 1.0 µM ACh. This potentiation was most marked in the presence of 1.0 nM GIP, whereas the effect of concomitant infusion of 0.2 nM GIP and 1.0 µM ACh was only slightly greater than additive. SS-14 potently inhibited ACh-stimulated IRI secretion. These results demonstrated the glucose dependency of cholinergically stimulated IRI secretion, and that physiological levels of glucose and GIP increased B-cell sensitivity to cholinergic stimulation. It was suggested that the parasympathetic stimulation of IRI secretion associated with food intake could be affected by postprandial increases in glucose, GIP, and SS-14. The idea that endogenously released somatostatin may have influenced glucose or GIP-stimulated IRI secretion was not supported by the present experiments, since neither glucose (8.9 mM) nor GIP (2.0 nM) were found to have a significant effect on the release of pancreatic somatostatin-like immunoreactivity (SLI). Both atropine (1.0 µM) and hexamethonlum (100 µM) inhibited the IRI response to ACh. This suggested that the parasympathetic stimulation of IRI secretion was mediated not only by muscarinic receptors on the B-cell, but also by nicotinic receptors on intrapancreatic ganglia. Neither atropine nor hexamethonlum had a significant effect on glucose- or GIP-stimulated IRI secretion, indicating that the IRI response to these stimuli was not mediated by cholinergic receptors. Both SS-14 and the synthetic somatostatin analog SMS 201-995 (SMS; Sandostatin®) inhibited IRI secretion stimulated by 8.9 mM glucose, 2.0 nM GIP, or 1;.0 µM ACh, but not 17.8 mM glucose. The most potent inhibition by both SS-14 and SMS was observed in the presence of the weakest IRI stimuli (8.9 mM glucose and 1.0 µM ACh). These results suggested that the inhibitory effects of somatostatin on the B-cell could be overcome by the presence of strong stimuli. In addition, the inhibitory effects of the native hormone and the analog were found to be approximately equipotent (weight basis), indicating that the increased potency of SMS previously observed in vivo was due to its longer half-life in plasma, and not due to a more potent direct effect on the B-cell.
Medicine, Faculty of
Cellular and Physiological Sciences, Department of
Graduate

Insulin causes uptake of glucose into cells and also causes increases in skeletal muscle blood flow by vasodilatation. In order for insulin to act at receptors on the skeletal muscle membrane, it must move through the capillary endothelium. In conditions associated with insulin resistance, insulin-induced glucose uptake and vasodilatation are often blunted and disordered transport across endothelium has been suggested. Firstly, studies were conducted using the hyperinsulinaemic euglycaemic (HE) clamp in rats to assess the relationship between insulin-induced muscle vasodilatation and glucose uptake into skeletal muscle. Insulin-induced vasodilatation was abolished by nitric oxide (NO) synthase inhibition in Wistar rats, implicating NO. Nutritional status and hence availability of glucose also affected vasodilatation in Wistar rats, leading to the proposal that vasodilatation was linked to glucose metabolism. Vasodilator response to insulin was significantly higher in lean versus obese Zucker rats. Nicotinic acid (NAc) did not decrease plasma FFA concentrations versus HE clamp, nevertheless, vasodilator response and glucose uptake were enhanced in lean and obese rats by some mechanism other than lowering FFA. Optimisation of the microdialysis technique allowed measurement, for the first time, of interstitial insulin concentrations in lean and obese Zucker rats and showed a significant difference under basal conditions.

Mestrado em Biologia Molecular e Celular
Este estudo avaliou se o tratamento de ratos diabéticos induzidos com estreptozotocina, com insulina ou factor de crescimento derivado da insulina (IGF1), em doses que não revertem a hiperglicemia, afectam os sinais comportamentais da neuropatia diabética e a ativação neuronal na medula espinhal. Foi também avaliada a participação de algumas das principais áreas do tronco cerebral (VLPAG), envolvidas na modulação descendente da dor. Uma semana após a indução da diabetes, foi iniciado o tratamento dos animais, 3 vezes por semana durante 3 semanas, com soro fisiológico, insulina (2 IU) ou IGF1 (2,5 mg / Kg). O tratamento com insulina ou IGF1 preveniu sinais comportamentais de neuropatia diabética, denominada alodínia mecânica. A avaliação comportamental dos animais através do teste de formol evidenciou que, quer a insulina quer o IGF1, previnem a elevada frequência de espasmos observados nos ratos diabéticos, para valores semelhantes aos dos controlos. Quanto à activação nociceptiva da expressão de c-fos no corno dorsal da medula espinal, esta foi inibida por ambos os tratamentos. A melhoria das acções comportamentais e da activação nociceptiva dos neurónios espinhais mediada pelo tratamento com IGF1 é susceptível de ser devida aos efeitos na modulação dolorosa descendente proveniente do tronco cerebral, mediada pela serotonina e noradrenalina. Ratos diabéticos apresentaram elevados números de neurónios imunorreactivos para TpH (marcador de neurónios serotoninérgicos) no RVM, ou para TH (marcador de neurónios noradrenérgicos) no núcleo celular noradrenérgico A5 (pontine). Estes números foram normalizados para níveis controlo, mas apenas quando tratados com IGF1, uma vez que a insulina não afecta estes parâmetros. Observou-se que os níveis de serotonina e noradrenalina na medula espinhal estavam aumentados, bem como os ratos tratados com insulina. Estes resultados evidenciaram que a insulina e o IGF1 possuem diferentes efeitos no sistema nervoso, sendo que os efeitos centrais devem-se essencialmente ao IGF1.
This study evaluates if the treatment of Streptozotocin-induced diabetic rats with insulin or insulin growth factor 1 (IGF1), in doses that do not reverse hyperglicemia, affect behavioural signs of diabetic neuropathy and neuronal activation at the spinal cord. The participation of main brainstem areas involved in descending modulation of pain was also evaluated (VLPAG). One week after diabetes induction, the animals were injected, 3 times per week, with saline, insulin (2 IU) or IGF1 (2.5 mg/Kg) during 3 weeks. Treatment with insulin or IGF1 prevented behavioural signs of diabetic neuropathy, namely mechanical allodynia. Behavioural evaluation of the animals by the formalin test showed that insulin and IGF1 strongly prevented the higher frequency of flinching behavior to values similar to controls. Nociceptive activation of c-fos expression induced by formalin at the spinal dorsal horn was inhibited by both treatments. The improvement of behavioural actions and nociceptive activation of spinal neurons mediated by IGF1 treatment are likely to be due to effects in descending pain modulation from the brainstem, mediated by serotonin and noradrenaline. Diabetic rats presented higher numbers of neurons immunoreactive for TpH (marker of serotoninergic neurons) at the rostroventromedial medulla or for TH (marker of noradrenergic neurons) at the pontine A5 noradrenergic cell group. These numbers were normalized to control levels only after IGF1 treatment, but not after insulin. The levels of serotonin and noradrenaline at the spinal cord were increased in non treateddiabetic rats and insulin treated-rats. These results show that insulin and IGF1 appear to have different effects on the nervous system, with more central effects being ascribed to IGF1.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior
The aim of this study was to evaluate the hemodynamic and metabolic responses of L-Glutamine in sedentary subjected to physical exertion. The sample consisted of 11 sedentary volunteers (05 men and 06 women between 30 - 45 years). We conducted a clinical study, crossover, randomized, double-blind. The protocol was divided into two stages. The protocols used participant (0.5 g / kg) of L-glutamine or calcium caseinate added to a milk drink Nescau Light (200ml) for seven consecutive days, with an interval of one week of completion of a protocol and early other. The volunteers were evaluated for cardiopulmonary and metabolic response during cardiopulmonary exercise testing on a treadmill. Were determined in each individual oxygen consumption (VO2 Max), respiratory quotient (RQ), the workload performed, heart rate (HR), systolic and diastolic blood pressure (SBP / DBP) in phases: pre - effort, the anaerobic threshold, maximal effort and during the recovery effort. Were also evaluated serum concentrations of glucose and lactate in four stages: fast (at least 8 hours), immediately before the effort at the end of exercise and 30 minutes after exercise. The prior offer of L-glutamine (0.5 g / kg in sedentary individuals, orally for seven days did not cause hemodynamic changes during physical exertion. On the other hand modified the metabolic response by reducing the lactacemia before the physical activity (3, 68 Â 0.8 versus 2.2 Â 0.9, p = 0.01) in the presence of maximal (12.4 Â 5.1 versus 10.3 Â 3.2, p = 0.04), and at the end of the exercise stress test (6.9 Â 3.6 versus 5.02 Â 1.77, p = 0.01). L-glutamine in nutraceutical doses is recommended for use to practitioners of physical activity and not hemodynamic features.
The aim of this study was to evaluate the hemodynamic and metabolic responses of L-Glutamine in sedentary subjected to physical exertion. The sample consisted of 11 sedentary volunteers (05 men and 06 women between 30 - 45 years). We conducted a clinical study, crossover, randomized, double-blind. The protocol was divided into two stages. The protocols used participant (0.5 g / kg) of L-glutamine or calcium caseinate added to a milk drink Nescau Light (200ml) for seven consecutive days, with an interval of one week of completion of a protocol and early other. The volunteers were evaluated for cardiopulmonary and metabolic response during cardiopulmonary exercise testing on a treadmill. Were determined in each individual oxygen consumption (VO2 Max), respiratory quotient (RQ), the workload performed, heart rate (HR), systolic and diastolic blood pressure (SBP / DBP) in phases: pre - effort, the anaerobic threshold, maximal effort and during the recovery effort. Were also evaluated serum concentrations of glucose and lactate in four stages: fast (at least 8 hours), immediately before the effort at the end of exercise and 30 minutes after exercise. The prior offer of L-glutamine (0.5 g / kg in sedentary individuals, orally for seven days did not cause hemodynamic changes during physical exertion. On the other hand modified the metabolic response by reducing the lactacemia before the physical activity (3, 68 Â 0.8 versus 2.2 Â 0.9, p = 0.01) in the presence of maximal (12.4 Â 5.1 versus 10.3 Â 3.2, p = 0.04), and at the end of the exercise stress test (6.9 Â 3.6 versus 5.02 Â 1.77, p = 0.01). L-glutamine in nutraceutical doses is recommended for use to practitioners of physical activity and not hemodynamic features.

The dairy cow experiences a period of immunosuppression around the time of calving that contributes to the increased incidence and severity of infectious diseases observed during this period. This reduction in immune capacity is due in part to the impairment of neutrophil function, a key component of the innate immune system. In fact, the success of the host defense mechanisms against infection depends on the ability of neutrophils to reach the site of the infection, recognize, engulf and ultimately destroy the pathogen using several mechanisms such as the generation and release of reactive oxygen species (ROS) and the recently described neutrophil extracellular traps or NETs. The alteration in some of these functions and the overall killing ability of neutrophils during the periparturient period has been widely described. However, the physiological mechanisms underlying the period of immunosuppression are not completely elucidated. Interestingly, the impairment of these immune defense mechanisms coincides with the profound metabolic changes associated with parturition and lactogenesis. Changes in several hormones and metabolites have been proposed to be the cause of the reduction in neutrophil function, but the effect of insulin on the functional capacity of these cells has not been investigated. Not only does the concentration of plasma insulin fall as parturition approaches, but also the animal experiences a period of impaired insulin action, termed insulin resistance, during this same time-frame. Therefore, we isolated circulating neutrophils from periparturient and midlactating cows and incubated them with insulin alone or in combination with the insulin-sensitizing agent 2,4- thiazolidinedione (TZD). Subsequently, we measured the total, extracellular, and intracellular generation of ROS, NETs release, phagocytic and killing ability. Insulin did not improve any of the parameters used to assess neutrophil function. In contrast, TZD had a potent inhibitory effect on the total ROS generation, despite an increase in extracellular superoxide anion production. Surprisingly, TZD did not alter the ability of neutrophils to phagocytose and/or kill Staphylococcus aureus during an in vitro coculture. Results suggest that TZD can reduce the oxidative stress that neutrophils experience during their respiratory burst and diminish the damage that ROS cause to the surrounding tissue without compromising the capacity of neutrophils to eliminate the invading pathogen.

The metabolic disorder diabetes; is a global epidemic affecting people in developed countries and increasingly in developing countries. In two decades time, 350 million people will be diabetic at the current rate of prevalence. In a preliminary study, insulin resistant rats were treated with Prunus Africana (plant A) for 28 days. Plasma samples obtained from P. africana treated rats had increased insulin levels compared to normal and untreated insulin resistant rats (Karachi, 2009). The treatment of insulin resistant rats with P. africana also showed increased glucose uptake in rat adipose tissue (Karachi, 2009), suggesting that P. africana had anti-diabetic properties. The aim of the study was to investigate the mechanism of the anti-diabetic properties of P africana extract. Increased insulin secretion was confirmed by the increased Cpeptide concentration in plasma samples of rats treated with P. africana. In order to explain the high insulin levels, several hypothesis’ were investigated: (1) P. africana may increase insulin secretion in β cells, hence the effect of P. africana on insulin secretion by INS-1 cells was investigated; (2) P. africana may increase insulin secretion by prolonging the half-life of glucagon like peptide-1 (GLP-1) by decreasing dipeptidyl peptidase IV (DPP IV) activity; the effect of P. africana on DPP IV activity was determined spectrophotometrically, (3) P. africana may increase the half-life of insulin in the plasma by decreasing the activity of insulin degrading enzyme (IDE); the effect of P. africana on IDE in rat muscle and spleen samples was investigated. To explain the increased glucose uptake in adipose tissue observed in the previous study two parameters were investigated: (1) increased GLUT4 expression in P. africana treated rats; the effect of P. africana treatment on the expression of glucose transporter 4 (GLUT4) was determined using real-time polymerase chain reaction (RT-PCR), (2) P. africana may increase glucose utilization; the effect of P. africana on glucose utilization was determined in 3T3-L1 cells. The plant extract did not significantly increase insulin secretion by INS-1 cells in the absence of glucose. P. africana decreased DPP IV activity in rat plasma when compared to the untreated insulin resistant rats and this could be a mechanism by which insulin secretion is increased during plant treatment. P. africana decreased IDE activity (however not significantly) when compared to the untreated insulin resistant The effects of a Kenyan antidiabetic plant on insulin homeostasis KY Suleiman VII rats. P. africana appeared to have no effect on GLUT4 expression. The plant appeared to increase glucose utilization in 3T3-L1 cells in the absence of insulin suggesting that P. africana may have insulin like activity. In summary, this study indicates that P. africana is indirectly involved in inhibiting DDPIV. This in turn can increase the half life of GLP-1, which in turn can enhance the secretion of insulin. P. africana increases glucose utilization although there was no evidence that the GLUT 4 transporter has a higher expression in the plant treated rats. Further studies should be conducted to investigate the expression of GLUT1 under the same conditons.

There are numerous parallels between aging and obesity, and insulin may play a crucial role in modulating both conditions. For instance, elevated insulin levels are closely associated with obesity, although the causal role of insulin hypersecretion in the development of obesity remains controversial. Interestingly, genetically reducing components of insulin/insulin-like growth factor (IGF)-1 signaling can increase lifespan in invertebrates and mammals. However, impaired insulin-stimulated glucose disposal is a form of decreased insulin signaling that is paradoxically a detrimental feature of mammalian aging, whereas long-living mammals often show enhanced responsiveness to insulin stimulation. Therefore, the role of insulin/IGF-1 signaling for mammalian longevity, and the relative functions of the insulin and IGF-1 ligands, are still unclear. The lifelong effects of moderately decreasing insulin production in mammals had not been directly tested. In this dissertation, the goal was to further delineate effects of lowering insulin levels on obesity and metabolic health across the lifespan of a mammalian model organism, culminating in an evaluation of longevity. We used a model in which the rodent-specific insulin gene was fully inactivated (Ins1-null), and compared mice with full or partial expression of the ancestral insulin gene (Ins2). Male and female Ins1-/-:Ins2+/- and Ins1-/-:Ins2+/+ littermates were fed a chow diet or high fat diet, and were evaluated across their lifetime to determine long-term effects of reducing insulin gene dosage on obesity, glucose homeostasis, and other physiological parameters. The studies herein show that murine insulin levels and metabolic homeostasis might be regulated in a sex-specific, environmentally-dependent manner, since inactivating one Ins2 allele unexpectedly did not cause a consistent reduction of circulating insulin in Ins1-null male mice, and we observed cross-cohort hyper-variability in circulating insulin of male mice. However, limiting insulin hypersecretion in young, growing female mice can confer long-term protection against obesity. Furthermore, we found that lowering circulating insulin has the potential to improve glucose homeostasis and insulin sensitivity in advanced age, as well as lead to lifespan extension in mammals. To our knowledge, these studies are the first to demonstrate that a targeted, moderate reduction of insulin may be sufficient to promote healthier aging and extend lifespan in mammals.
Medicine, Faculty of
Graduate

An increased concentration of insulin-like growth factor-1 (IGF-1) is recognized to be an independent risk factor for pre-menopausal breast cancer. Tamoxifen is thought to initially reduce IGF-1 concentrations and increase levels of the IGF binding proteins 1 and 3. Isoflavones are oestrogen-like plant compounds, which may, because of their structural similarities to tamoxifen, also alter IGF status. This thesis first compares IGF-1, IGF binding protein 1(BP-1) and IGF binding protein 3 (BP-3) levels in breast cancer patients (n=14) compared with control subjects (n=23) and then assesses the effect of tamoxifen on IGF status after 9, 18 and 27 months of treatment. Concentrations of IGF-1, BP-1 and BP-3 at baseline did not differ between cases and controls. However on tamoxifen treatment, BP-1 and BP-3 were both significantly increased after 18 and 27 months while the IGF-1/BP-3 ratio was significantly decreased after 9 and 18 months. A feasibility study was then conducted to compare the effects of acute (single 80 mg load) versus chronic (80 mg/day for 7 days) administration of isoflavone- containing soy and linseed cereal bars on IGF-1, IGFBP-1 and IGFBP-3 in healthy female volunteers (n=10). Assays were established for the analyses of serum IGF-1, BP1 and BP3 concentrations and GCMS analysis of isoflavones in urine. Concentrations of IGF-1 were unchanged following the single 80 mg load. However, IGF-1 and BP-3 concentrations were significantly elevated after a week of supplementation with the soy and linseed bar. A larger study was then carried out to assess the effect of one-month isoflavone supplementation (80 mg/d) in tablet form, on IGF status in healthy pre- (n=16) and post-menopausal (n=7) women. This was a randomized, placebo-controlled crossover study with a minimum two-month washout period. Hormonal, antioxidant and lipid altering effects of the supplement were also examined, as was background diet. For pre-menopausal subjects, while there was an effect of time on IGF-1 (p=0.005), BP-1 (p=0.004), and BP-3 (p<0.001) confirming that IGF profile is influenced by menstrual cycle, this did not differ between placebo and isoflavone supplement. When the change in IGF-1 over the whole supplementation period was compared between the supplement and placebo phases, there was a non-significant reduction in change in IGF-1 (p=0.06) on isoflavone supplement compared to placebo. However, this may have been due to the non-significantly higher baseline concentrations of IGF-1 during the supplement phase. In post-menopausal subjects, there was no effect of isoflavone supplementation in comparison with placebo on IGF-1, BP-1 or BP-3. Finally, experiments were done in vitro to study the effect of isoflavones on DNA synthesis and proliferation in ERa positive MCF-7 and ERa negative MDA-MB 231 human breast cancer cells. In MCF-7 cells, low concentrations of isoflavones (0.1uM-10uM) stimulated DNA synthesis while high concentrations of genistein and equol were able to inhibit DNA synthesis with IC50 values of 93uM and 55uM respectively compared with 1.17uM tamoxifen. In MDA-MB231 cells the Isoflavones did not stimulate DNA synthesis at any concentrations but significantly inhibited DNA synthesis with IC50 values of 114uM, 14.1 uM and 14.55uM for daidzein, genistein and equol respectively compared with 1.35uM tamoxifen. This work suggests that isoflavone supplementation may alter IGF profile in pre menopausal women, and could suggest a role for these dietary compounds in breast cancer prevention. This should be further investigated in long term intervention studies with isolated isoflavone supplements.

Insulin resistance, the primary metabolic abnormality underpinning type 2 diabetes mellitus and obesity, is an important risk factor for the development of cardiovascular disease. Endothelial dysfunction represents one of the earliest phases in the natural . history of atherosclerosis and is normally offset in health by various endogenous repair processes. Circulating endothelial progenitor cells (EPCs) participate in endothelial repair following arterial injury. Type 2 diabetes is associated with fewer circulating EPCs, EPC dysfunction and impaired endothelial-repair. I set out to determine whether insulin-resistance per se adversely affects EPC mediated endothelial-regeneration using mice hemizygous for knockout of the insulin receptor (IRKO) and wild-type (WT) littermate controls. The metabolic phenotype of IRKO mice was consistent with compensated insulin resistance. Flow cytometry demonstrated that IRKO mice had fewer circulating EPCs than WT mice. Culture of mononuclear-cells confirmed that IRKO mice had fewer EPCs in peripheral-blood, but not in bone-marrow or spleen, suggesting a mobilization defect. Defective VEGF-stimulated EPC mobilization was confirmed in IRKO mice, consistent with reduced eNOS expression in bone marrow and impaired vascular eNOS activity. Paracrine angiogenic activity of EPCs from IRKO mice was impaired compared to those from WT animals. Endothelial-regeneration of the femoral artery following denuding wire-injury was delayed in IRKO mice compared to WT. Transfusion of mononuclear-cells and c-kit bone-marrow cells from WT mice normalized the impaired endothelial-regeneration in IRKO mice. However, transfusion of c-kit+ cells from IRKO mice was less effective at improving endothelial-repair. My data suggest that insulin-resistance impairs EPC function and delays endothelial-regeneration following arterial injury. These findings support the hypothesis that insulin-resistance per se is sufficient to jeopardise endogenous vascular repair. Defective endothelial-repair may be normalised by transfusion of EPCs from insulin- sensitive animals but not from insulin-resistant animals. These data may have important implications for the development of therapeutic strategies for insulin- resistance associated cardiovascular disease.

In Drosophila, the insulin signaling pathway is at the interface between dietary conditions and control of growth and development, reproduction, stress responses and life span. Eight insulin like peptides (Dilp1-8), an insulin tyrosine kinase receptor (dInR) and its downstream components, as well as a relaxin-like receptor type (Lgr3) form the core of this signaling. Here we showed that the dInR mediates post-mitotic cell growth specifically in about 300 peptidergic neurons expressing the basic helix loop helix (bHLH) transcription factor Dimmed (Paper I).  Overexpression of dInR in Dimm positive neurons leads to increased size of cell body, Golgi apparatus and nucleus, whereas dInR knockdown causes an opposite effect. Manipulation of downstream components of insulin signaling induces similar changes in Dimm positive neurons. This mechanism is nutrient dependent. In Paper II, we further investigate the relation between Dimmed and dInR for regulation of cell growth. Coexpressing Dimm and dInR in a range of Dimm negative neurons results in increased cell size in both larval and adult stages. We provide further evidence that dInR regulates cell growth in a Dimm dependent manner and that DILP6 from glia cells is involved in this regulation. In addition, we find that Dimm alone is capable of triggering cell growth in certain neuron types at different developmental stages. Furthermore, ectopic Dimm alone can block apoptosis.  Dimm is a known master regulator of peptidergic cell fate. In paper III we find that ectopic expression of Dimm in Dimm negative motor neurons results in transformation the neurons towards a neuroendocrine phenotype. They acquire enlarged axon terminations and boutons, lose both pre- and postsynaptic markers, and display diminished levels of wingless and its receptor dFrizzled. Furthermore they show increased expression of several Dimm targets. Finally, combined ectopic Dimm and dInR expression gives rise to stronger phenotypes. In paper IV we studied another DILP possibly involved in growth regulation, the under-investigated DILP1. We generated Dilp1-Gal4 lines and anti DILP1 antibodies and found that DILP1 is transiently expressed in brain insulin producing cells (IPCs) from pupal stages to newly hatched adult flies. Diapausing virgin female flies display a high expression level of dilp1/DILP1 over at least 9 weeks of adult life. DILP1 expression is also correlated with the persistence of larval/pupal fat body and its expression is regulated by other DILPs and short neuropeptide F (sNPF). Flies mutant in dilp1 display increased food intake, but decreased stress resistance and life span. We found no obvious role of DILP1 in growth regulation.

At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Manuscript. Paper 3: Manuscript. Paper 4: Manuscript.

Thesis (Ph.D. in Epidemiology) -- University of Colorado at Denver and Health Sciences Center, 2006.
Typescript. Includes bibliographical references (leaves 168-182). Free to UCDHSC affiliates. Online version available via ProQuest Digital Dissertations;

Insulin resistance (IR) leads to the development of type 2 diabetes mellitus and predisposes to cardiovascular disease (CVD) through its link with endothelial dysfunction. Cardiovascular risk factors and iatrogenic damage lead to biochemical and structural injury to the endothelium. Endogenous repair mechanisms are in place to regenerate injured endothelium. Insulin resistance has recently been shown to impair endothelial repair. The endogenously produced circulating insulin-like growth factor binding protein-1 (IGFBP-1) is potentially protective in the vasculature by stimulating nitric oxide production and enhancing insulin signalling in the endothelium. Cross-sectional studies have shown an association between low IGFBP-1 levels and CVD. This raises the possibility of exploiting IGFBP-1 therapeutically to prevent CVD in patients with diabetes. This project investigated whether IGFBP-1 enhances vascular endothelial repair in insulin resistant mice in vivo and probed potential molecular mechanisms by examining the effects of IGFBP-1 on human endothelial cells (EC) and angiogenic progenitor cells (APCs) in vitro. Endothelial regeneration was enhanced following arterial endothelium-denuding injury in IRKO mice by over-expressing human IGFBP-1. This was not explained by altered abundance or function of APCs. Incubation with IGFBP-1 significantly enhanced the ability of human EC to adhere to and regenerate denuded human vein ex vivo. In EC, IR was mimicked by the pro-inflammatory cytokine tumour necrosis factor-alpha (TNF-α) which significantly inhibited EC migration and proliferation in vitro. Co-incubation with IGFBP-1 restored the migratory and proliferative capacity of EC. IGFBP-1 significantly increased FAK phosphorylation, induced rapid activation of RhoA, and increased expression of α5β1 and αVβ3 integrins in EC. These multifactorial effects of IGFBP-1 on EC responses and acceleration of endothelial regeneration in mice raise the possibility that manipulating IGFBP-1 could be a strategy to enhance endothelial repair in humans with IR.

The purpose of this study was to determine if α1- adrenergic receptor blockade alters the hemodynamic response to exercise in young (<25 yr) male adult borderline hypertensives differently than in young normotensives. Five hypertensive (HTN, MAP>105 mmHg) and 7 normotensive (NTN, MAP<95 mmHg) college-age males underwent two 30 min bouts of cycle ergometry exercise at 50% V02Pk in a warm (25°C, 50% rh) environment; one bout occurred followed α1-receptor blockade with prazosin (HTN-α, NTN-α) and the other following placebo administration (HTN-p, NTN-p). At rest, HTN-p exhibited an elevated cardiac output (Q, p=.024) and MAP (p=.007). Resting Q was similar for HTN-α and NTN-α. Resting heart rate (HR) was elevated more in HTN-α than NTNα (p=.013) and not different for placebo. Resting and exercise forearm blood flows were similar between groups and altered similarly with prazosin. Exercise resulted in greater (p=.035) Q for HTN vs NTN (HTN-u > NTN-α; HTN-p = NTN-p). HR was higher (p=.043) with prazosin for both groups. Regardless of drug treatment, MAP was stable for NTN while it declined after 10 min of exercise in HTN. Rectal temperatures rose above baseline after 10 min. since Q was similar between groups with placebo but not with α1- blockade, and FBF, MAP, and HR were similarly altered between drug trials, it was concluded that young male hypertensives have an elevated blood pressure due to an elevated Q. In this group, α1-blockade may reduce Q by reducing central venous return.
Ph. D.