Answers are grounded in this paper and cite their sources.
Numerous human diseases originate from an aberrant cellular metabolism. Persistent alterations in the concentrations of key metabolites like ATP, fatty acids or reactive oxygen species exceeding the normal physiological range may result in cellular dysfunction, activation of a damage response and ultimately cell death. Disease-causing deviations from normal metabolism may arise from different factors: Genetic alterations affecting the expression of metabolic enzymes, action of drugs or toxins diminishing the activity of metabolic enzymes or inadequate supply of substrates, in particular oxygen.1 The impact of genetic variations is impressively demonstrated by inborn monogenetic metabolic diseases like phenylketonuria, where the impaired activity of a single enzyme (phenylalanine hydroxylase) may derange the whole cellular metabolic network of brain cells in a way that intellectual disability, seizures, behavioural problems and mental disorders may occur without an early-onset strict phenylalanine-depleted diet. Insufficient supply of cellular nutrients such as glucose, fatty acids, amino acids and oxygen because of reduced tissue perfusion (ischaemia) prevents the adequate generation of energy-rich nucleotides such as ATP or GTP. This generates a latent cellular state of restricted energy production which may suddenly and unexpectedly progress to organ failure if substrate supply is further decreased (eg, because of a temporary heart insufficiency) or an extreme organ performance is demanded (eg, because of high muscle workload, intake of new drugs, long-term starvation or an infectious disease).2, 3 It is necessary to decipher the molecular mechanisms underlying disease-related metabolic changes at the cellular level to understand how malfunctions of tissues, organs and organisms develop. This will also help to develop efficient medical strategies to prevent or delay the disease manifestation and to improve recovery of the affected organ. Assessing to what extent the cellular metabolism of a specific cell of a patient already deviates from the normal is still a great challenge. Typically, the metabolic state of an organ is inferred from arterio-venous differences of oxygen and metabolites such as glucose, lactate, amino acids and fatty acids. Such measurements only assess the net outcome of the organ's metabolism without being able to quantify the contribution of various cellular pathways and the metabolic state of different cell populations within the organ. Currently available non-invasive techniques like Positron-Emission Tomography with radioactive tracers like fludeoxyglucose (FDG) or magnetic resonance spectroscopy can only capture a limited number of relevant metabolites and provide only static metabolic snapshots. Checking the metabolic reserve capacity by exposing the patient to a metabolic challenge is excluded for ethical reasons. Experiments with isolated cells in a test tube can hardly reproduce the same metabolic response as elicited in the physiological environment of the living organism. To overcome these hurdles, physiology-based mathematical models can be used to simulate the cellular response to metabolic perturbations.4 The architecture of the metabolic network governed by the model and the regulatory properties of the included enzymes and membrane transporters can be established and validated based on experiments with isolated tissues, cells and proteins. Once parametrized, the model can be used to predict dynamic metabolic changes in vivo for a variety of physiological conditions such as hypoxia, ischaemia, malnutrition, intake of drugs or presence of genetic enzyme variants. Recent work on modelling of liver metabolism may serve as a paradigm for such a model-based approach. One important goal of mathematical models in liver research is to gain a better understanding of the metabolic changes underlying non-alcoholic liver diseases (NAFLD), which are growing in epidemic proportions worldwide and now already start in the very young.5 Abnormal but clinically inconspicuous accumulation of triglycerides in hepatocytes (hepatic steatosis)6 is regarded as the initial step of NAFLD. A "fatty liver" increases the risk for developing an inflammation of the liver, the so- called non-alcoholic steatohepatitis, which may further progress to cirrhosis and eventually hepatocellular carcinoma (HCC).7, 8 Notably, disease progression occurs often silently and without clear clinical symptoms.9 Metabolic models of the liver range from genome-wide network models10 to small-scale metabolic models of single pathways.11, 12 Starting point of any metabolic model is the knowledge of the network architecture given in terms of the so-called stoichiometric matrix assigning to each biochemical reaction the metabolites that are consumed and produced. Knowledge of network stoichiometry is already sufficient to highlight possible routes that a given metabolite can take in a series of subsequent biochemical reactions. This allows to assess whether the network reconstruction is in principle capable of performing the many physiological functions reported for the liver such as conversion of ammonia to urea, synthesis and degradation of triglycerides or de novo synthesis of glucose from lactate and amino acids. However, for assessing the metabolic capacity of the liver in a given physiological or disease state it is indispensable to know the rates of the biochemical reactions and concentration of metabolites. The only reliable way to gather this information is kinetic modelling. In a kinetic model, the time-dependent changes of metabolite concentrations are the solutions of an ordinary differential equation system comprising the reaction rates for each biochemical reaction (if spatial concentration changes are not considered). The real challenge in kinetic modelling consists in the establishment of realistic enzymatic rate laws describing the relationship between reaction rates and metabolite concentrations. Thanks to hundreds of enzymatic studies on liver enzymes performed by experimental biochemists since the middle of the last century, physiologically reliable kinetic models were first established for small metabolic subsystems11-13 and recently even for the complete central metabolism of liver cells.14 The latter model allows simulating diurnal variations of the metabolic state of the liver at various perturbations caused by nutritional challenges (alcohol), drugs (valproate) and inherited enzyme disorders (galactosaemia). Regarding the development of a fatty liver, model simulations have provided evidence that long-term changes in the expression of lipogenic enzymes are required for the development of an alcoholic fatty liver. Using proteomics data to scale maximal enzyme activities, the model was also used to highlight differences in the metabolic functions of normal hepatocytes and malignant liver cells (adenoma and HCC). In order to include blood supply and the interstitial transport of nutrients as metabolic determinants, a multi-scale model of the so-called sinusoidal tissue unit was proposed, comprising a single liver capillary (sinusoid) which is surrounded by the space of Disse and a monolayer of hepatocytes.15 This tissue model takes into account the heterogeneous enzymatic endowment of hepatocytes along the porto-central axis of the sinusoid (metabolic zonation) and reveals intra-organic metabolic cycles, where glucose produced by hepatocytes of the portal region is partially consumed by hepatocytes of the central region.15 To study the dynamic adaptation of the protein abundance of liver enzymes to pathological conditions, the model in Ref.16 extended the multi-scale tissue model by treating the protein abundances as dynamic variables whose synthesis and degradation rates depend on the metabolic state. Based on this model it could be demonstrated that the control of enzyme abundances by nutrients, oxygen and hormones necessarily leads to the zonation of hepatic carbohydrate metabolism. The various kinetic modelling approaches to the study of liver metabolism are mutatis mutandis applicable to other organs such as the heart or the kidney. Acute kidney injury (AKI) is a serious clinical condition associated with systemic complications such as metabolic acidosis, high potassium levels, uraemia and changes in body fluid balance.17 Notably, the recovery rates from AKI have not improved significantly over the last 60 years,18 indicating a small increase in knowledge about the underlying pathogenetic processes. This is remarkable, given the numerous experimental studies dealing with the prevention and treatment of AKI (eg,4, 19-27). Investigation of kidney functionality focuses on the organ architecture and the molecular basis for permeabilities and electrophysiological properties of epithelial cells, essential for vasodilatation, electrolyte and fluid movement (eg,28-32). Hence, mathematical kidney models traditionally focus, in contrast to those of the liver, on the adequate description of the structural and spatial organization of the organ. Sophisticated structural models of the kidney have been developed to describe in a spatially resolved manner the arrangement of vessels, pressure gradients, osmolarities and the hormonal regulation of salt and water excretion.28 Multi-compartment models and models based on partial differential equations are available to compute the distribution and excretion of substances such as FDG,33 drugs,34, 35 glucose and lactate36 and to describe the regulation of renal salt and water excretion, glomerular blood flow, filtration rates and blood flow.37 While these models allow a detailed description of intra renal blood flow and water movement, an adequate mechanistic description of the cellular metabolism in various populations of renal tubular epithelial cells is lacking so far. Similar to the liver, renal epithelial enzyme patterns vary considerably, including reciprocal capacities of gluconeogenesis and glycolysis in tubule cells resident in the cortex and the medulla. This may fuel cortical cells with lactate formed by medullary cells as glycolytic product.38 Owing to the high similarities shared between the central metabolism of hepatocytes and renal tubule cells, re-parameterization of complex kinetic liver models14 with kidney-specific parameters, as has been done in Ref.39, represents an efficient way to generate metabolic cell models that can be integrated into existing tissue-scale models of the kidney. The functionality of liver and kidney depend on the plasma composition as well as the protein endowment. These are subjected to considerable diurnal variation and constitute a network, where individual habit, gene expression, hormone release and metabolite levels are interdependent.40-44 The ability of kinetic models to take into account time dependent input in plasma nutrient and hormone composition as well as protein abundances, enables the time-dependent computational assessment of organ functions even in such highly dynamic situations, something that cannot be achieved with other modelling methods like statistical modelling or flux balance models.45 None.
No takes yet. Share an insight, caveat, or question.
Berndt et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: