Workshop 7: Insulin Secretion, Insulin Action, and Type 2 Diabetes: Abstracts and Lecture Materials
Insulin secretion from pancreatic islets is pulsatile, with period of 4-7 min. This reflects the oscillatory electrical activity and calcium dynamics of the beta-cells. Two candidate mechanisms for this oscillatory behavior have been proposed and studied. In one mechanism the oscillation is due to purely electrical or electrical/calcium processes, such as the feedback of calcium onto calcium-activated potassium channels. In another mechanism the oscillations are due to metabolic oscillations, such as oscillations in glycolysis. We present a mathematical model in which both electrical and metabolic oscillations play roles. This Dual Oscillator Model can account for a wider range of data than can other models, although there are still significant challenges. We discuss the rationale for the model, some successes of the model, and some remaining challenges.
Signal transduction networks often exhibit combinatorial complexity: the number of protein complexes and modification states that potentially can be generated during the response to a signal is large, because signaling proteins contain multiple sites of modification and interact with multiple binding partners. The conventional approach of manually specifying each term of a mathematical model is impossible. To avoid this problem, modelers often make assumptions to limit the number of species, but these are usually poorly justified. As an alternative, biomolecular interactions can be represented by rules specifying activities, potential modifications and interactions of the domains of signaling molecules. Rules are evaluated automatically to generate the reaction network. This approach is implemented in BioNetGen software (http://bionetgen.lanl.gov). I will discuss this approach for modeling of early events of signaling mediated by Epidermal Growth Factor and Insulin Growth Factor receptors.
Insulin controls plasma glucose level by stimulating glucose utilization by the tissues and inhibiting glucose production by the liver. The efficiency of this control is measured by insulin sensitivity (SI), i.e. the ability of insulin to enhance glucose utilization and inhibit glucose production.
A widely used techniques usable to assess this useful index is the glucose clamp technique, which is quite nonphysiological and invasive since it employs infusion of glucose and insulin, but allows then to calculate SI with a simple formula. Another possibility is to employ the simpler intravenous glucose tolerance test (IVGTT), which consists in the injection of a glucose bolus in vein and in the measurements of glucose and insulin concentration. However, in order to assess insulin sensitivity from IVGTT data, one needs to employ the minimal model of glucose kinetics which puts in relation plasma glucose and insulin measurements. The minimal model, proposed in 1979, has become a classic and approximately 60 papers employing it are published each year. In the past recent years, a new model has been proposed to quantify insulin sensitivity in a normal life conditions, e.g. during a meal; it has been validated against model-independent techniques and the gold standard euglycemic-hyperinsulinemic clamp, and its validity was proved also in reduced protocol studies.
Moreover, in order to segregate the action of insulin on glucose production and utilization, both during i.v. and oral perturbations, one can employ isotopic glucose tracers and a model tracer kinetics. In fact, glucose tracers have the same properties of unlabeled glucose for what concerns the utilization by the tissues, but are not endogenously produced by the body; thus their concentration depends on glucose disposal only.
Intuition based on static protein interactions is limited when signaling networks contain multiple feedback and crosstalk loops, as in the Insulin-like growth factor 1 (IGF 1) receptor pathway. Mechanistic modeling enables for examination of the role and importance of dynamic protein interactions and provides a foundation for developing targeted therapeutics.
The IGF 1 receptor pathway is known to play an important role in breast cancer. Stimulation of the IGF-1 receptor results in activation of multiple pathways that generate survival and proliferation cues, such as Erk and Akt. Given the known role of IGF-1 in cancer, we have taken a quantitative approach to understanding the receptor signaling pathways at the molecular level.
An ordinary differential equation based model of the signaling events post IGF-1 stimulation was built using quantitative experimental data obtained from IGF-1 stimulated MCF-7 cells. To challenge the model we generated experimental and simulated dose-response and time-dependent behavior of p-Erk and p-Akt in the presence of inhibitors targeting different positions within the IGF pathways.
We show that understanding subtle differences in network dynamics is crucial for predicting the unintended or counter-productive effects of targeted inhibitors. In addition, target optimization can be performed in silico to examine these effects prior to inhibitor design.
We have generated transgenic mice in which pancreatic beta-cells are specifically labeled with green fluorescent protein (GFP) under the control of mouse insulin I promoter (MIP). We have also developed a method to visualize GFP-expressing beta-cells together with surrounding tissues in the intact pancreas using reflected laser light confocal imaging. Three-dimensional reconstruction of beta-cells in the pancreas of neonatal mice show that beta-cell proliferation and islet formation occur contiguously in contrast to the adult pancreas where islets are distributed throughout the exocrine pancreas. This imaging technique also enabled us to visualize pathological changes in intraportally transplanted MIP-GFP islets and surrounding hepatic tissue. We show that early graft loss often observed clinically may be due to an islet-induced embolism that caused ischemic necrosis in the host hepatocytes, hence blocked blood supply to the graft. New imaging techniques coupled with mouse models are providing a new view of pancreatic islet development and pathophysiology of beta-cells.
Calcium release from intracellular calcium stores is known to play important roles in many cell types. Human pancreatic beta-cells contain at least three classes of intracellular calcium stores based on their mobilization through three classes of calcium-release channels: IP3-receptors, ryanodine-receptors (RyRs) and NAADP-receptors. NAADP- and IP3-receptors mediate calcium release in response to insulin, whereas glucose and GLP-1 have been implicated in activating RyRs. Although intracellular calcium stores do not appear to play an essential role in glucose-stimulated insulin release in primary beta-cells, it is becoming clear that these stores play critical roles in beta-cell survival. In order for beta-cells to remain viable, calcium flux via RyR must be maintained at an intermediate level. Blocking basal RyR activity leads to a hypoglycemia-related form of programmed cell death. On the other hand, inhibiting RyR can also attenuate certain modes of apoptosis associated with hyperstimulation. Together with the findings of others, our results point to a critical role for intracellular calcium release channels in beta-cell survival pathways.
Regeneration of the insulin-secreting beta cells of the islets of langerhans is a major goal in diabetes research, but the cell division patterns within the beta cell lineage are not known. To address this fundamental question we developed a novel DNA analogue-based lineage tracing technique to detect sequential cell division in vivo. We show that the gastrointestinal mucosa has an asymmetric lineage program, with proliferative progenitors that repeatedly divide to give rise to post-mitotic terminally differentiated cells. In comparison, differentiated adult beta cells symmetrically divide to expand cellular mass. Remarkably, beta cells exhibit a symmetric cell division pattern even during acute beta cell regeneration. Indeed, we detect no contribution of proliferative progenitors to any adult beta cells, not even during acute beta cell regeneration. Thus, our approach allows unbiased single cell resolution of developmental niches, and indicates that beta cells exhibit equal proliferation potential.
Work done in collaboration with Monica Teta, Matthew M. Rankin, and Simon Y. Long.
Mathematical models have been proposed to study insulin secretion since the early times of beta-cell function investigation. Landmark models have been in particular those by Grodsky and coworkers and Cerasi and coworkers, in the seventies. These unrivaled models were developed to propose hypotheses on the mechanisms governing the insulin response to glucose stimuli; however, they have not been widely applied to the study of beta-cell function in physiological and pathological conditions.
Based on these historical models, we have developed a new model that inherits three essential characteristics: a dose-response relating glucose concentration to insulin secretion, a mechanism representing the anticipation of the insulin response following a rapid rise in glucose concentration and time-dependent potentiation of insulin secretion. This simplified model, in contrast to the more elaborate predecessors, allows assessment of beta-cell function parameters from oral glucose tests or meals; therefore, wide application to a variety of studies has been possible, thus providing new knowledge on beta-cell function.
In our studies, we have in particular found that: 1) the single most relevant characteristic of beta-cell function is the beta-cell glucose sensitivity, i.e., the slope of the beta-cell dose-response. This parameter is impaired in diabetes and prediabetes; in a population of subjects with a wide range of glucose tolerance, glucose sensitivity is the most significant predictor of the glucose levels during an oral glucose test. Furthermore, prospectively, impaired glucose sensitivity is a predictor of the development of diabetes. 2) Time-dependent potentiation, i.e., a relative enhancement of the relationship between glucose concentration and insulin secretion during the oral test, is a significant contributor of the insulin response. Potentiation is a determinant of the postprandial glucose levels and is progressively blunted in hyperglycemic states. 3) In normal subjects, adaptation of the beta-cell response to insulin resistance does not involve all beta-cell function parameters. While basal insulin secretion increases when insulin sensitivity decreases, beta-cell glucose sensitivity is not affected. Therefore, normal insulin resistant subjects have higher basal insulin secretion compared to insulin sensitive subjects, but glucose sensitivity is the same. Furthermore, in insulin resistant prediabetic subjects basal insulin secretion adapts to insulin resistance but glucose sensitivity is impaired. This observation suggests the existence of distinct beta-cell mechanisms, of which those responsible for the dynamic secretory responses are selectively impaired in prediabetes. 4) In hyperglycemic states, and in particular in diabetes, a known secretion defect is the almost complete lack of the first phase insulin response. Application of the model to diabetic subjects has revealed that above a certain glucose threshold first phase insulin response is abolished, while the beta-cell glucose sensitivity, though impaired, is nonzero and remains a determinant the glucose levels.
In conclusion, the use of a model applicable to simple clinical tests has provided specific answers to a variety of problems concerning beta-cell function. In addition, the results obtained have shed light on the beta-cell mechanisms that are impaired in type 2 diabetes.
Metabolic actions of insulin to promote glucose disposal are augmented by vascular actions of insulin in endothelium that stimulate production of the vasodilator nitric oxide (NO). NO-dependent increases in blood flow to skeletal muscle account for 25% to 40% of the increase in glucose uptake in response to insulin stimulation. Phosphatidylinositol 3-kinase-dependent insulin-signaling pathways in endothelium related to production of NO share striking similarities with metabolic pathways in skeletal muscle that promote glucose uptake. Other distinct non-metabolic branches of insulin-signaling pathways regulate secretion of the vasoconstrictor endothelin-1 in endothelium. These vascular actions of insulin contribute to coupling metabolic and hemodynamic homeostasis under healthy conditions. Metabolic insulin resistance characterized by pathway-specific impairment in phosphatidylinositol 3-kinase-dependent signaling, which in endothelium may cause imbalance between production of NO and secretion of endothelin-1. This may lead to endothelial dysfunction and decreased blood flow, which worsens insulin resistance. Mathematical modeling of signal transduction pathways mediating the metabolic and vascular actions of insulin offers a useful approach for gaining insight into the complexities of reciprocal relationship between endothelial dysfunction and insulin resistance. Work done in collaboration with Michael J. Quon and Arthur Sherman.
Plasma insulin oscillates with a period of ~5 min, which is believed to result from pulsatile insulin secretion with a similar period. Individual beta-cells possess the necessary machinery for pulsatile secretion, but these individual pulses need to be synchronized in order for an overall oscillatory pattern to appear on the pancreatic level. Within single islets gap junctions are likely to provide the electrical and metabolic coupling leading to intra-islet synchrony. The mechanism providing synchrony between the ~1 million islets in the pancreas is however far from understood. Nerve signal have been proposed, as well as entrainment by an oscillatory glucose stimulus. I will discuss the latter idea based on theoretical investigations on entrainment by external glucose as well as a simplified whole-body model consisting of a model 'liver' and a population of islets coupled through plasma glucose.
Insulin and plasma free fatty acid (FFA) levels were obtained during insulin-modified frequently sampled intravenous glucose tolerance tests (IM-FSIGT). An approach similar to the minimal model of glucose disposal was adopted to quantify the influence of insulin on FFA. Six mathematical models were tested to predict the fall and rise of FFA levels in response to insulin. In the model that best balanced model complexity and fit to data, insulin acted nonlinearly through a remote compartment to suppress lipolysis and FFA was cleared from the plasma linearly. The model had four parameters that could be estimated reliably. A measure of the sensitivity of FFA suppression to insulin that is analogous to glucose sensitivity to insulin S| was derived from the parameters of the model. We call this new measure the FFA-insulin sensitivity index.
Hypothalamic glucose sensing neurons are exquisitely sensitive to changes in extracellular glucose. Glucose-excited (GE) neurons increase, while glucose-inhibited (GI) neurons decrease, their action potential frequency as glucose levels increase. In addition, hormones involved in energy homeostasis (e.g., insulin, leptin) regulate the activity of glucose sensing neurons. Our recent data suggest that glucose and insulin sensitivity of glucose sensing neurons is interdependent. That is, the effects of insulin on GE and GI neurons are dependent on ambient glucose concentrations. Conversely, insulin decreases the glucose sensitivity of GE neurons. Thus, the overall hormonal and nutrient composition of the hypothalamic extracellular milieu modulates glucose sensing neurons ability to respond to signals of peripheral energy homeostasis. This talk will focus on the interactions between glucose and insulin on the regulation of glucose sensing neurons.
Insulin is released from pancreatic islet beta cells in two distinct phases in response to glucose stimulation. The first phase of insulin release can be elicited by elevation of intracellular calcium levels to trigger fusion of granules pre-docked at the plasma membrane. The second phase of insulin release requires the amplifying action of glucose, and is presumed to require mobilization of storage pool granules to the cell surface. Thousands of storage granules exist behind a filamentous actin (F-actin) barrier in the pancreatic beta cell and F-actin remodeling is known to mobilize granules to the t-SNARE (target membrane soluble N-ethylmaleimide-sensitive factor attachment protein receptor) sites at the cell surface, yet the mechanisms involved in remodeling and granule mobilization are largely unknown. Current studies suggest the glucose-specific activation of Cdc42, a small Rho family GTPase, is a key proximal event essential for second-phase insulin release via F-actin remodeling and granule targeting to SNARE fusion sites. The t-SNARE protein Syntaxin 4 has also recently emerged as an essential distal regulator of second-phase secretion. New insights into this novel signaling cascade in the beta cell will be discussed, as well as the physiological impact of altered Syntaxin 4 abundance upon whole-body glucose homeostasis, given the action of Syntaxin 4 in facilitating both insulin granule exocytosis in the pancreatic islet and GLUT4 vesicle translocation in skeletal muscle.
Assessment of insulin secretion in humans under normal life conditions, and its interplay with insulin action and hepatic extraction, is important to define the alterations in glucose metabolism associated with different physiopathological states. We review data indicating that an OGTT or meal test is able to accomplish this goal, when interpreted with the oral beta cell minimal model. We begin by presenting the well established minimal model of C-peptide secretion and kinetics during an intravenous glucose tolerance test (IVGTT), highlight assumptions underlying it and describe the quantitative assessment of beta cell function it provides. Then, we discuss how the oral minimal model was developed and how the oral assessment parallels that of IVGTT. Comparison with the results from an organ model based on measurements in femoral artery and hepatic vein provides evidence of the minimal model validity, and its ability to assess different aspects of the beta cell secretory cascade is discussed. By combining C-peptide minimal model with an insulin minimal models, hepatic extraction can be concurrently assessed. Since insulin sensitivity can be measured as well, the minimal model approach enables to quantitatively portray the complex relationship between beta cell function, hepatic insulin extraction and insulin action. Finally, a variety of studies aiming to quantify the effect of age, gender, progression of diabetes and pharmacological treatments on these processes are reviewed.
Poster Presentations
Goforth et al. (2002) have reported that a calcium-dependent potassium current (K_slow) found in mouse beta cells is transiently potentiated and then inhibited by the SERCA blocker, thapsigargin. In order to model this behavior they postulated that the K_slow channels are localized in a narrow subspace located between the ER and the plasma membrane. The authors used a three-compartment ODE model in that study; here we revisit their simulations in a spatially detailed model.
We rely on a calcium bidomain model to effectively homogenize network ER and cytosol; the "subspace" is formed as a narrow region between the junctional ER and the plasma membrane. Our preliminary simulations with the PDE model effectively confirm the results of the Goforth et al. model: K_slow activation is mediated by the subspace gradient that reflects Ca2+ efflux from the ER; as the ER empties under SERCA block this gradient decays resulting in the inhibition of K_slow.
To improve the understanding of the processes that govern non-esterified fatty acid (NEFA) metabolism in the postprandial state, we have developed a physiology-based mathematical model of plasma NEFA dynamics. Known physiological mechanisms are quantified and used to describe NEFA dynamic.
Pancreatic beta-cells are clustered in islets of Langerhans which are typically a few hundreds of micrometers in a variety of mammals. In this study, we propose a theoretical model for the growth of pancreatic islets and derive the islet size distribution, based on two recent observations: First, the neogenesis of new islets becomes negligible after some developmental stage. Second, islets grow via a random process, where any cell in an islet proliferates with the same rate regardless of the present size of the islet. Our model predicts either log-normal or Weibull distributions of the islet sizes, depending on whether cells in an islet proliferate coherently or independently. To confirm this, we also measure the islet size by selectively staining islets, which are exposed from exocrine tissues in mice after enzymatic treatment. Indeed revealed are skewed distributions with the peak size of about 100 cells, which fit well to the theoretically derived ones. Interestingly, most islets turned out to be bigger than the expected minimal size (about ten or so cells) necessary for stable synchronization of beta-cells through electrical gap-junction coupling. The collaborative behavior among cells is known to facilitate synchronized insulin secretion and tends to saturate beyond the critical (saturation) size of about 100 cells. We further probe how the islets change as normal mice grow from young (6 weeks) to adult (5 months) stages. It is found that islets may not grow too large to maintain appropriate ratios between cells of different types. Our results implicate that growing of mouse islets may be regulated by several physical constraints such as the minimal size required for stable cell-to-cell coupling and the upper limit to keep the ratios between cell types. Within the lower and upper limits the observed size distributions of islets can be faithfully regenerated by assuming random and uncoordinated proliferation of each beta-cell at appropriate rates. Work done in collaboration with M. Y. Choi and D.-S. Koh.
In the glucose-insulin regulatory system, ultradian insulin secretory oscillations are observed to have a period of 50-150 minutes. After pioneering work traced back to the 1960s, several mathematical models have been proposed during the last decade to model these ultradian oscillations as well as the metabolic system producing them. These currently existing models still lack some of the key physiological aspects of the glucose-insulin system. Applying the mass conservation law, we introduce two explicit time delays and propose a more robust alternative model for better understanding the glucose-insulin endocrine metabolic regulatory system and the ultradian insulin secretory oscillations for the cases of continuous enteral nutrition and constant glucose infusion. With explicit delays, the model is more realistic in physiology, more accurate in mathematics, and more robust in potential applications. We study this model analytically and perform carefully designed numerical simulations by allowing one or two parameters to vary. Our analytical and numerical results confirm most current existing physiological observations and reveal more insightful information. The following factors are critical for ensuring the sustained oscillatory regulation and insulin secretion: (1) the time lag for insulin secretion stimulated by glucose and the newly synthesized insulin becoming 'remote insulin'; (2) the delayed effect of hepatic glucose production; (3) moderate insulin clearance rate; and (4) non-overwhelming glucose infusion.
Pulsatility is a hallmark of insulin secretion, and the loss of this rhythmic function may be an early indicator of pancreatic islet dysfunction. A reduction in the amplitude and regularity of insulin pulses has been found in type 2 diabetic patients, as well as their close relatives who do not show other early symptoms of the disease. Isolated islets also produce rhythmic patterns in insulin secretion and other parameters, such as intracellular calcium ([Ca2+]i), but the relevance of these rhythms to islet health and function has not been established.
We investigated whether the capacity of islets to generate oscillations in vitro correlates with islet health and function. Specifically, we used fluorescence-based imaging techniques to compare control 'healthy' islets and islets treated with pro-inflammatory cytokines by measuring the following: [Ca2+]i oscillations, glucose-stimulated [Ca2+]i responses to glucose (3 to 28 mM), and islet health determined by propidium iodine (PI) to label dead or dying cells. Approximately 80% of control islets were oscillatory, whereas less than 30% of cytokine-treated islets displayed [Ca2+]i oscillations. Among control and cytokine treated groups non-oscillatory islets had attenuated [Ca2+]i increases in response to glucose stimulation compared to oscillatory islets, suggesting that reduced oscillatory capacity is associated with reduced glucose sensitivity. Non-oscillatory islets also showed greater PI staining than oscillatory islets, indicating higher rates of cell death.
Changes in oscillatory properties have also been investigated in mouse models of type 2 diabetes. We have preliminary data that islets from leptin-receptor-deficient db/db mice show a nearly complete loss of [Ca2+]i oscillations by 8 weeks of age, yet maintain some capacity for glucose stimulation. In high-fat-fed mice that overexpress 12-lipoxygenase (a protein that metabolizes fatty acids and is associated with inflammatory damage), islets show reduced oscillatory capacity despite normal glucose stimulation. These data suggest that islet health and function are linked with oscillatory capacity and that the loss of oscillatory capacity may be an early indicator of beta-cell stress or dysfunction. This link between oscillatory capacity and islet function may be useful for comparing the overall health of cultured islets in response to therapies and/or for assessing islets in transplantation studies.
Work done in collaboration with P. Jahanshahi, R. Wu, and J.D. Carter.
Insulin is released from pancreatic beta-cells by calcium dependent exocytosis of insulin containing granules. Only a small fraction of these granules are immediately releasable and are believed to correspond to first phase insulin secretion. This pool has been investigated using short depolarizing voltage-clamp pulses combined with measurements of calcium currents and of cell membrane capacitance. Often, the data points are fitted to descriptive exponentials. The present work models this type of experiments biophysically by including 'hot-spots' below the Ca2+ channels where the Ca2+ concentration can reach sufficiently high levels to trigger exocytosis. This description includes only 4 parameters to be estimated, i.e., only one more than the phenomelogical, exponential description. It thus provides a biophysically sound, but yet simple, model of fast exocytosis in beta-cells. Moreover, it allows the distinction between different control points of cAMP-enhanced Ca2+ sensitivity of the exocytotic machinery.
The standard glucose signal to insulin release pathway in beta cells is augmented by coincident glucagon like peptide-1 (GLP-1) or gastrointestinal peptide (GIP) release from intestinal cells. This coincident signal works through G-protein modulation of adenylyl cyclase (AC8) to increase production of cAMP and subsequent PKA activation. Spatial orientation of cAMP activators and degraders induce a dynamic spatial compartmentalization restricting PKA signaling. Extent of the PKA signal then depends on initiating stimulus.
We create 3-compartment and continuum diffusion reaction models to explore mechanisms of catalytic PKA (cPKA) distribution shown in recent experimental results by Tengholm et al., PNAS (2006). With pulsatile inhibition of cAMP degradation rapid submembrane activation of PKA is registered. However, only with sustained cAMP elevation does cPKA seem to penetrate the nucleus at elevated levels. We discuss qualitative agreement of the 3-compartment model with the experiments and issues with the continuum model.
Glucagon and insulin are secreted from islets of Langerhans to regulate blood glucose levels. Secretion of both peptides is altered in type II diabetes, although the mechanisms leading to these alterations are unknown. We have developed a two-color capillary electrophoresis immunoassay to measure insulin and glucagon simultaneously. Insulin is measured by detection of a "green" fluorophore and glucagon is detected with a "red" fluorophore. Spectral resolution of the electrophoretic peaks allows quantitation of both compounds even though the peaks are not spatially resolved. This method has been used to quantify insulin and glucagon content in islets of Langerhans. In the future, this two-color assay will be used in combination with a microfluidic device for rapid and simultaneous monitoring of insulin and glucagon secretion from islets. The use of this device will facilitate studies on the paracrine interactions that regulate secretion of these hormones and on the signal transduction processes that become altered in diabetes.
Work done in collaboration with Christelle Guillo.
There is strong circumstantial evidence that ER calcium in beta-cells rises and falls with cytosolic calcium because calcium activates SERCA pumps. However, there is also evidence that the ER fills to a fixed level independent of cytosolic calcium, leading to the proposal that the pump activity is controlled primarily by ATP. We explore these alternatives and show that a model in which SERCA is activated by both cytosolic calcium and ATP but also inhibited by ER calcium can can account for both sets of observations. They occur in quantitatively distinct parameter regimes but are not mutually exclusive. This has possible ramifications for ER stress, which is partly linked to SERCA activity.
Work done in collaboration with Brad Peercy (LBM).
Diabetes mellitus, with its diverse clinical manifestations, requires trial and error even by diabetic specialists to obtain optimal therapeutic strategies for individual patients. To overcome this difficulty and to provide useful treatment solutions for general physicians, we devised the Diabetes Management System, aims at offering optimized therapeutic strategies for each patient by quantification of clinical condition.
The distinguishing feature of this system is its arithmetic quantification of the clinical conditions of patients with diabetes mellitus and provision of therapeutic solutions in daily clinical practice. Using a top-down approach, a diabetes model incorporated in the system is constructed to represent ability to secrete insulin, peripheral insulin sensitivity, hepatic glucose metabolism, and other factors that determine the patient's blood glucose levels. Entry of time-series data on blood glucose and insulin obtained during the oral glucose tolerance test (OGTT) into the model permits simulation of the patient's condition on the computer and estimation of internal parameters that enable reproduction of such data. Subsequently, the parameters are converted to indices, such as ability to secrete insulin, peripheral insulin sensitivity, and hepatic glucose metabolism.
Here, outcomes of treatment with medication were metabolically evaluated by the use of the system to non-insulin-dependent diabetes mellitus (NIDDM). Two profiles were obtained from a set of OGTT data "before" and "after" medication to the same subjects. The changes in each estimated patient's condition and system index were correlated very well.
These results support the validity of simulation of clinical condition by the present system and suggest the possibility of its usefulness in providing reasonable guidance to general physicians in managing patients with diabetes mellitus in daily clinical practice.
Work done in collaboration with T. Saito, T. Takahata, Y. Kouchi, Y. Naito, H. Nakajima, and K. Asano.
Insulin signaling pathway is a well characterized system, wherein most of the interactions are well known. The negative feedback regulation of phosphatase PTP1B by Akt has been demonstrated to exist in the insulin signaling pathway. Analysis in other systems has elucidated that positive feedback loops or double negative feedback loops embedded with nonlinearity exhibit switch-like bistable responses. In the current work, we present a dynamic model of the insulin signaling pathway. The threshold concentration of insulin required for glucose transporter GLUT4 translocation was studied with variation in system parameters and component concentrations. The dose response curves of GLUT4 translocation at various concentration of insulin obtained by steady state analysis were quantified in-terms of half saturation constant. The dynamic analysis demonstrated that insulin-stimulated GLUT4 translocation can operate as a bistable switch, which ensures that GLUT4 settles between two discrete, but mutually exclusive stable steady states. The detailed model was further reduced by considering a representation of the steady state response. The threshold concentration of insulin required for GLUT4 translocation changes with variation in system parameters and component concentrations, thus providing insights into possible pathological conditions. The model demonstrated that the dynamic response was sensitive to variations in phosphotase PTP1B concentration. The dynamic model can be further connected to the metabolic state of glucose uptake. The threshold concentration of insulin required for GLUT4 translocation and the corresponding bistable response at different system parameters and component concentrations was compared with reported experimental observations on specific defects in regulation of the system.
