Computer model simulates liver cell membrane that secretes bile

The formation of membrane domains for bile secretion depends on the lipid composition of the outer leaflet of the apical membrane. To understand the process of the formation of bile-salt insoluble and bile-salt soluble lipid domains group of Prof. Holzhütter from Computational Systems Biochemistry Group in Charité (Berlin) simulated the canalicular membrane of a liver cell using a computer model of the movement of the lipids in the membrane.

  • The lipids of the canalicular membrane are 38 % cholesterol (blue), 47 % phosphatidylcholine (green) and 15 % sphingomyelin (red) and can be either in a bile-salt soluble (dark) or insoluble (light) state.
  • The simulations show how the lipids arrange on the membrane forming maze-like bile-salt soluble and insoluble domains growing over time. The bile-salt soluble domain shows a lipid composition similar to that of the bile fluid. With the computer model, it is possible to investigate the influence of changes in the membrane composition on the process of lipid secretion into the bile.
  • Author of text and image: Johannes Eckstein

  • Comprehensibility: 
    Average: 5 (5 votes)
    (1=bad 5=excellent)
    Reconstruction of the human hepatocyte is a basis for metabolic models

    The Holzhütter group from the Virtual Liver Network examined the biochemical literature to compile HepatoNet1 - a genome-scale metabolic network of the human hepatocyte. All biochemical reactions are recorded therein, enzyme catalyzed reactions and a few spontaneous reactions.

  • The hepatocyte is connected to the organism by two transport systems: the blood vessels (red in the sketch) and the bile ducts (olive in the image). HepatoNet1 distinguishes these transport processes. Included is also the information which genes encode the enzymes and transporters.
  • Hepatocytes are sub-structured in smaller units - the compartments. Among them are mitochondria (violet in the image). HepatoNet1 distinguishes 6 of them. Each reaction is attributed to a compartment. The transport processes between them are recorded in HepatoNet1. Many of them also need specific proteins, the encoding genes are also recorded.
  • In total, HepatoNet1 comprises 777 metabolites and 2539 reactions, divided in 1073 biochemical conversions and 1466 transport reactions. HepatoNet1 is finally a mathematical model that can predict, which biochemical processes are needed to fulfil a certain function and which amount is necessary. More specifically, it is a stoichiometric model. The computational method is called flux-balance analysis.
  • Author: Andreas Hoppe

    Publications expand more information

    Links expand more information
    Average: 3.8 (6 votes)
    (1=bad 5=excellent)
    Computational model describes self-organisation of polarized plasma membrane of hepatocyte

    Scientists from the Virtual Liver Network found that both types of syntaxins (syntaxin 3 and syntaxin 4) are transported via small vesicles to the plasma membrane of the hepatocyte. Upon arrival to the membrane, the vesicles fuse with the membrane.

    Vesicles which carry mostly syntaxin 3 fuse predominantly with those membrane sections that are already equipped with many syntaxin 3 and only few syntaxin 4.

  • Syntaxin 4-enriched vesicles behave in the opposite way. Using this knowledge and experimental data, a scientist Dr. Bernd Binder from the group of Prof. Holzhütter (Computational Systems Biochemistry Group in Charite, Berlin) generated a computational model. This model describes the self-organisation of the vesicle transport system together with segregation of syntaxin sections in the plasma membrane. This process is shown on the animated image.
  • The image depicts two phenomena: (1) Gradually larger organelles are formed within the hepatocyte by numerous fusion events among small vesicles. (2) The plasma membrane is equipped step by step with both types of syntaxin molecules. They are already segregated by the delivery of syntaxin 3 proteins to locations with high syntaxin 3 content and the delivery of syntaxin 4 to regions hosting many syntaxins of type 4.
  • Author of text and image: Bernd Binder

  • Comprehensibility: 
    Average: 4.5 (4 votes)
    (1=bad 5=excellent)
    Bridging experiments and modelling: The SABIO-RK reaction kinetics database
    SABIO-RK is a database providing information about biochemical reaction kinetics. Biochemical reactions are usually catalyzed by enzymes by lowering their activation energies. Enzymes are biological molecules (mainly proteins) that interact selectively with substrates and help them to react in a specific way leading to the products in a given time.

  • Without enzymes, cells would not be able to run all the metabolic processes from degradation of nutrients to synthesis of new cellular components quickly enough to sustain life.

  • Reaction kinetics is the study of the speed with which a biochemical reaction occurs and the factors that affect this speed. The speed of a reaction is the rate at which the concentrations of reactants and products change. There is a lot of data originating from lab experiments about the reaction rates of all kinds of biochemical reactions catalyzed by special enzymes.
  • In SABIO-RK such kinetic data of all sorts of organisms is collected from literature and lab experiments and offered in a structured format. The database contains information about the biochemical reaction, i.e. the participants (substrates, products, modifiers), enzyme details and the environmental conditions. The data is curated, standardized, annotated and linked to other databases. SABIO-RK is freely accessible via a web interface ( and search results can be exported in different formats. The aim of SABIO-RK is to support modelers and wet-lab scientists in their daily work. One example where SABIO-RK data was used within the Virtual Liver Network, is the creation of simulatable computer models that describe the glucose metabolism in the liver.
  • Authors: Renate Kania, Maja Rey, Ulrike Wittig, Martin Golebiewski

  • Image: Elina Wetsch

  • SABIO-RK Logo
    Average: 4 (1 vote)
    (1=bad 5=excellent)
    The Boolean logic behind cellular signal transduction

    AND, OR, NOT, those are the three basic functions of Boolean logic. With those functions entities that can just have two states, 1 (active) and 0 (inactive), are combined. The result of an AND function is 1, if each involved entity is 1.

    For the OR function to be 1 it suffices that only one entity is active. In contrast to that, the NOT function negates the value of an entity. Its result is 1, if the entity is 0, and vice versa.

  • Boolean logic is the basis for Boolean models in systems biology. In those models entities, here called species, are connected via Boolean functions. Those species often represent proteins and sometimes genes. Boolean models are used for explaining processes, in which the activities of the involved species show an ON/OFF pattern. This often comprises a signal cascade starting from outside the cell and going to the cell nucleus causing a changed activity of a gene there.
  • Logical models cannot explain the mechanisms behind this signal transduction. However, they can describe the order, in which the species are activated or inhibited. Apart from that the Boolean logic behind this activation mechanism is contained in Boolean models. A species can, e.g., be activated, if two other species are both active, which is represented with an AND function in the model. If in contrast to that a species is activated, if at least one of two other species is active, an OR function is used in the model.
  • A Boolean model is usually constructed by gathering information from many scientific publications. In order to show that the model can explain a certain cellular process, the model is simulated and the simulation results are compared to experimental data afterwards. Simulation of a Boolean model starts by fixing the states of some species. Based on this, the states of all other species are calculated until they do not change anymore. Then a so-called logical steady state is reached. The fixing of species states often corresponds to a specific treatment in an experimental data set. Such a treatment comprises the targeted stimulation and inhibition of some proteins. The states of the stimulated proteins are then set to 1, whereas the states of the inhibited proteins are set to 0.
  • In order to compare the Boolean model to the continuous experimental data, these data have to be transformed into 0 and 1 in advance. Then for each treatment the corresponding logical steady state of the model is calculated. After that, the determined states of the model species can be compared to the transformed experimental data.
  • Author: Roland Keller

  • Image: Stephanie Hoffmann

  • Comprehensibility: 
    Average: 5 (1 vote)
    (1=bad 5=excellent)