The adipose tissue contributes to energy homeostasis by storing excess energy as triglycerides when the energy status is high, and by releasing fatty acids when the energy status is low. In addition to the involvement in energy homeostasis, the adipose tissue also has a function of hormonal control exerted by the release of adipokines such as adiponectin. Dysregulations in the signaling pathways of these two functions are involved in the development of type 2 diabetes and related complications such as cardiovascular disease. These signaling pathways are too complex to be fully unraveled without a systematic framework, such as mathematical modeling. Previous modeling works have investigated the insulin signaling pathways leading to glucose uptake in primary human adipocytes in response to insulin stimulation. Furthermore, experimental works have investigated how adrenergic stimuli and varying concentrations of intracellular mediators triggers the release of adiponectin from 3T3‐L1 adipocytes, and how insulin and adrenergic stimuli influence lipolysis in primary human adipocytes. Additionally, large‐scale phosphoproteomic data for insulin signaling in 3T3‐L1 adipocytes have become available. However, these experimental data had not been systematically investigated using mathematical modeling. In this thesis, I have used mathematical modeling to study three aspects of the adipocyte: 1) adiponectin release, 2) lipolysis, and 3) intracellular crosstalk between the pathways of glucose uptake, lipolysis, and adiponectin release. Finally, I have developed a new method for automatic model expansion.

In Paper I, we used mathematical modeling to test a hypothesis of the mechanisms controlling adiponectin exocytosis in 3T3‐L1 cells. We found that the hypothesis had to be revised in order to be in agreement with the available experimental data. We used the revised model to quantify the balance between the exocytosis and the endocytosis, and to predict the amount of released adiponectin in response to additional experiments.

In Paper II, we extended the adiponectin exocytosis model from Paper I with mechanisms for how extracellular adrenergic stimulation trigger adiponectin exocytosis. We also used the model to quantify the effect of a decreased amount of β₃‐adrenergic receptors on the adrenergically stimulated adiponectin exocytosis.

In Paper III, we tested a hypothesis of the impact of, and crosstalk between, insulin and adrenergic stimulation on the lipolysis. We used the model to test three different actions by insulin on the lipolysis, and to predict fatty acid release in vivo in response to stimulations with epinephrine and insulin.

In Paper IV, we combined the models from Paper I‐III with a previously published model for glucose uptake. We then used the connected model as a core model to which additional signaling data could be added using a new method for automatic model expansion. This new method incorporates prior‐knowledge and large‐scale data to expand a core model with thousands of additional phosphosites into a comprehensive model of the adipocyte. The comprehensive expanded model can propagate the effect of type 2 diabetes from the core model to a substantial part of the phosphoproteome, and could thus facilitate the finding of new drug targets or treatment regimens for type 2 diabetic patients.

Adipocyte cellular signaling, normally and in type 2 diabetes, is far from fully studied. We have earlier developed detailed dynamic mathematical models for some well-studied, and partially overlapping, signaling pathways in adipocytes. Still, these models only cover a fraction of the total cellular response. For a broader coverage of the response, large-scale phosphoproteomic data is key. There exists such data for the insulin response of adipocytes, as well as prior knowledge on possible protein-protein interactions associated with a confidence level. However, methods to combine detailed dynamic models with large-scale data, using information about the confidence of included interactions, are lacking. In our new method, we first establish a core model by connecting our partially overlapping models of adipocyte cellular signaling with focus on: 1) lipolysis and fatty acid release, 2) glucose uptake, and 3) the release of adiponectin. We use the phosphoproteome data and prior knowledge to identify phosphosites adjacent to the core model, and then try to add the adjacent phosphosites to the model. The additions of the adjacent phosphosites is tested in a parallel, pairwise approach with low computation time. We then iteratively collect the accepted additions into a layer, and use the newly added layer to find new adjacent phosphosites. We find that the first 15 layers (60 added phosphosites) with the highest confidence can correctly predict independent inhibitor-data (70-90 % correct), and that this ability decrease when we add layers of decreasing confidence. In total, 60 layers (3926 phosphosites) can be added to the model and still keep predictive ability. Finally, we use the comprehensive adipocyte model to simulate systems-wide alterations in adipocytes in type 2 diabetes. This new method provide a tool to create large models that keeps track of varying confidence.Competing Interest StatementThe authors have declared no competing interest.

Lipolysis and the release of fatty acids to supply energy fuel to other organs, such as between meals, during exercise, and starvation, are fundamental functions of the adipose tissue. The intracellular lipolytic pathway in adipocytes is activated by adrenaline and noradrenaline, and inhibited by insulin. Circulating fatty acids are elevated in type 2 diabetic individuals. The mechanisms behind this elevation are not fully known, and to increase the knowledge a link between the systemic circulation and intracellular lipolysis is key. However, data on lipolysis and knowledge from in vitro systems have not been linked to corresponding in vivo data and knowledge in vivo. Here, we use mathematical modelling to provide such a link. We examine mechanisms of insulin action by combining in vivo and in vitro data into an integrated mathematical model that can explain all data. Furthermore, the model can describe independent data not used for training the model. We show the usefulness of the model by simulating new and more challenging experimental setups in silico, e.g. the extracellular concentration of fatty acids during an insulin clamp, and the difference in such simulations between individuals with and without type 2 diabetes. Our work provides a new platform for model-based analysis of adipose tissue lipolysis, under both non-diabetic and type 2 diabetic conditions.

The beta-adrenergic response is impaired in failing hearts. When studying beta-adrenergic function in vitro, the half-maximal effective concentration (EC50) is an important measure of ligand response. We previously measured the in vitro contraction force response of chicken heart tissue to increasing concentrations of adrenaline, and observed a decreasing response at high concentrations. The classical interpretation of such data is to assume a maximal response before the decrease, and to fit a sigmoid curve to the remaining data to determine EC50. Instead, we have applied a mathematical modeling approach to interpret the full dose–response curvein a new way. The developed model predicts a non-steady-state caused by a short resting time between increased concentrations of agonist, which affect the dose–response characterization. Therefore, an improved estimate of EC50 may be calculated using steady-state simulations of the model. The model-based estimation of EC50 is further refined using additional time resolved data to decrease the uncertainty of the prediction. The resulting model-based EC50 (180–525 nM) is higher than the classically interpreted EC50 (46–191 nM). Mathematical modeling thus makes it possible to reinterpret previously obtained datasets, and to make accurate estimates of EC50 even when steady-state measurements are not experimentally feasible.