Sabine Stübler (1,2), Charlotte Kloft (3), Wilhelm Huisinga (2)
(1) PharMetrX Graduate Research Training Program: Pharmacometrics & Computational Disease Modelling, Freie Universität Berlin/Universität Potsdam, (2) Mathematical Modelling and Systems Biology, Institut für Mathematik, Universität Potsdam, (3) Institut für Pharmazie, Freie Universität Berlin
Objectives: Inflammatory bowel diseases (IBD), characterised by chronic inflammation of the gut tissue, are caused by autoimmunity of T cells against commensal bacteria. As the therapeutic outcome of the currently used treatment regimens (e.g. immunomodulatory small molecule drugs or monoclonal antibodies targeting TNF-a) differs highly between drugs and patients, a better understanding of the mucosal immune system in the context of chronic inflammation is highly desirable. The objective of this work was to mathematically describe the most important cellular processes of the intestinal immune system to provide a basis for the analysis of drug effects and inter-individual variability.
Methods: Through an extensive literature query we identified important processes of the (innate and adaptive) mucosal immune system on the cellular level and described them (i) using literature values as parameters, (ii) by fitting parameters to literature data from human, mouse or in vitro studies, or (iii) by assuming reasonable parameter ranges resulting in adequate model behaviour, where literature data were not available. These processes were then combined into an ODE model.
Results: The developed systems biology model included different subsets and activation states of dendritic cells, macrophages, neutrophils, T cells and bacteria, described as concentrations (cells/mL) in two tissues (lamina propria (LP) and mesenteric lymph node (MLN)): Dendritic cells enter the LP as quiescent cells and can be activated by pro-inflammatory cytokines and/or take up antigen and then migrate to a MLN. Macrophages enter the LP as pro-inflammatory cells and are subsequently deactivated [1], and can take up antigen. Neutrophils enter the LP as inflammatory cells. Their apoptosis triggers resolution of the infection by producing specialised pro-resolving mediators (SPM) and stimulating macrophages to produce SPM, which inhibit innate immune cell recruitment and stimulate further neutrophil apoptosis. Recruitment, activation and inactivation of dendritic cells, macrophages and neutrophils are dependent on cytokine concentrations, which were implemented as weighted sums of the cytokine-producing cells due to their short half-lives. Activation of naive and memory T cells (mainly in MLN) is based on the contact rate between T cells and antigen-presenting cells (mainly dendritic cells), limited by the available surface of both cell types and the contact duration. Subsequently there is proliferation and apoptosis [2,3] and differentiation (to T helper cell types 1, 2 and 17 and regulatory T cells) with specific functions via cytokine production. Regulatory T cells are also able to inhibit antigen-presenting cells via cell-cell contact. Epithelial barrier, mucus layer and bacteria in the lumen are described as change from baseline. Pathogenic bacteria, which compete with commensal bacteria for nutrients in the lumen, are able to cross the epithelial barrier, thereby destroying it and allowing commensal bacteria to follow. Once in the LP, bacteria are eliminated by phagocytic cells, limited by the maximal killing rate per phagocyte [4].
The model was able to reflect the main characteristics of the mucosal immune response to infection with an intracellular or extracellular bacterial pathogen or damage of the epithelial barrier: A sharp increase (peak at 16 h) of the neutrophil concentration was followed by an inflow of macrophages. Effector T cell concentrations increased with a delay (peak at day 8). Bacteria up to a threshold were eliminated, and subsequently all cell concentrations returned to baseline (resolution of the inflammation).
Conclusions: The proposed model is a systematic combination of available knowledge on the local gut immune effects related to IBD. It allows for the first time to include different known IBD predispositions and stimuli and thereby to generate a virtual population of IBD patients, defined by different parameter sets accounting for different predispositions leading to disease. After combining the systems biology model with drug-specific PK models, we aim to identify to what extent the inter-individual variability in treatment outcome is PK- and/or PD-related.
References:
[1] Bain et al., Mucosal immunology 6.3 (2013): 498.
[2] Hawkins et al., Proceedings of the National Academy of Sciences 104.12 (2007): 5032-5037.
[3] Arias et al., Journal of theoretical biology 349 (2014): 109-120.
[4] Li et al., Journal of Experimental Medicine 200.5 (2004): 613-622.
Reference: PAGE 28 (2019) Abstr 8861 [www.page-meeting.org/?abstract=8861]
Poster: Drug/Disease Modelling - Other Topics