(343a) Modeling LPS-Induced TNF-? Production in Macrophages

Sepsis results in about 200,000 deaths per year in United States alone, primarily due to the excessive immune response to the infecting pathogen [1]. Immune cells such as macrophages detect the presence of pathogen-derived molecules through pattern-recognition receptors (PRR). Toll-like receptor 4 (TLR4) is one such PPR that binds bacterial lipopolysaccharide (LPS), which is the major component of the cell wall in gram-negative bacteria, and initiates multiple intracellular signaling pathways to remove the pathogen [2]. Among different signaling pathways initiated by LPS binding to TLR4, the pathway leading to activation of the transcription factor NFκB is critical because NFκB mediates the production of several pro-inflammatory cytokines to eliminate the pathogens or PPR. One such cytokine is TNFα that stimulates pro-inflammatory signaling and cytokine production in other cells and leads to sustained inflammation. Since chronic inflammation can itself be detrimental to the organism, it is important to elucidate the molecular mechanisms underlying TNFα production and its regulation in the macrophage response to LPS.

Previous studies have constructed mathematical models to predict LPS-induced NFκB signaling pathway to study dynamic interactions among different components in the LPS-mediated NFκB signaling pathway [3, 4]. In this work, we updated the deterministic model developed by Caldwell et al. [3] with the in vitro measurements so that the model could simulate the NFκB dynamics accurately under the wide range of LPS concentrations. Data on the levels of IκBα, which binds and sequesters NFκB activation, and TNFα production were obtained in RAW264.7 macrophages in response to different concentration of LPS using intracellular staining and flow cytometry. A part of this data set was used to estimate parameters in the model while the remaining was used for validation. Since the number of data points was far less than that of parameters, the parameter estimation problem is ill-posed; that is, we could not estimate all parameters in the model. In order to acquire a robust solution from an ill-conditioning problem, the parameter selection method was employed [5]. Specifically, the parameter selection method, which involves the sensitivity analysis and the Gram-Schmidt orthogonalization, chose and estimated only some of the model parameters. In this study, six parameters, which were closely related to the transcription of IκBα and TNFα mRNA, were selected for the estimation. After estimating the selected parameters, the resultant deterministic model was able to qualitatively and quantitatively predict the dynamics of TNF and IκBα in macrophages upon LPS stimulation. Furthermore, the updated deterministic model can simulate cell-to-cell variability by differences in parameter values. First of all, a distribution matching methodology [6] is employed to estimate distributions of key parameter values with flow cytometry data. Next, a set of values is picked from the distribution corresponding to different parameters and assigned to the model to simulate the response of different cells in the population. As a result, the heterogeneous responses within cell population can be modeled by the collections of distinct responses from individual cells. Consequently, the final model will be capable of simulating both single-cell and population-average response to LPS.

Reference

1. Deng, M.; Scott, M.J.; Loughran, P.; Gibson, G.; Sodhi, C.; Watkins, S.; Hackam, D.; Billiar, T. R. Lipopolysaccharide clearance, bacterial clearance, and systemic inflammatory responses are regulated by cell type-specific functions of TLR4 during sepsis. J. Immunol. 2013, 190 (10), 5152-5160.
2. Kawasaki, T.; Kawai, T. Toll-like receptor signaling pathway. Front. Immunol. 2014, 5, 461.
3. Caldwell, A.B.; Cheng, Z.; Vargas, J.D.; Birnbaum, H.A.; Hoffmann, A. Network dynamics determine the autocrine and paracrine signaling functions of TNF. Genes & Dev. 2014, 28, 2120-2133.
4. Kellogg, R.A.; Tian C.; Lipniacki, T.; Quake, S.R.; Tay, S. Digital signaling decouples activation probability and population heterogeneity. eLife. 2015, 4, e08931.
5. Kravaris, C.; Hahn, J.; Chu, Y. Advances and selected recent devlopments in state and parameter estimation. Comp. & Chem. Eng. 2013, 51,111-123.
6. Hasenauer, J.; Waldherr, S.; Doszczak, M.; Radde, N.; Scheurich, P.; Allgower, F. Identification of models of heterogeneous cell populations from population snapshot data. BMC Bioinformatics 2011, 12(1), 125.