A computer simulation approach to assessing therapeutic intervention points for the prevention of cytokine-induced cartilage breakdown

Abstract

Objective: To use a novel computational approach to examine the molecular pathways involved in cartilage breakdown and to use computer simulation to test possible interventions for reducing collagen release.

Methods: We constructed a computational model of the relevant molecular pathways using the Systems Biology Markup Language, a computer-readable format of a biochemical network. The model was constructed using our experimental data showing that interleukin-1 (IL-1) and oncostatin M (OSM) act synergistically to up-regulate collagenase protein levels and activity and initiate cartilage collagen breakdown. Simulations were performed using the COPASI software package.

Results: The model predicted that simulated inhibition of JNK or p38 MAPK, and overexpression of tissue inhibitor of metalloproteinases 3 (TIMP-3) led to a reduction in collagen release. Overexpression of TIMP-1 was much less effective than that of TIMP-3 and led to a delay, rather than a reduction, in collagen release. Simulated interventions of receptor antagonists and inhibition of JAK-1, the first kinase in the OSM pathway, were ineffective. So, importantly, the model predicts that it is more effective to intervene at targets that are downstream, such as the JNK pathway, rather than those that are close to the cytokine signal. In vitro experiments confirmed the effectiveness of JNK inhibition.

Conclusion: Our study shows the value of computer modeling as a tool for examining possible interventions by which to reduce cartilage collagen breakdown. The model predicts that interventions that either prevent transcription or inhibit the activity of collagenases are promising strategies and should be investigated further in an experimental setting.

Figure 1

Simulation results showing the effect of interleukin-1 (IL-1) and/or oncostatin M (OSM) on the expression of matrix metalloproteinase 1 (MMP-1), MMP-13, and tissue inhibitor of metalloproteinases 1 (TIMP-1), using a simulated time period of 48 hours. Curves show the level of MMP-1 mRNA, MMP-13 mRNA, TIMP-1 mRNA, and collagen fragments. A, Effect of IL-1 alone. B, Effect of OSM alone. C, Effect of IL-1 plus OSM. Note that in B, MMP-1, MMP-13, and collagen levels are all zero; thus, the individual lines are not visible.

Figure 2

Simulation results showing the effect of an MMP-activating protease on the activation of MMPs and on collagen release, using a simulated time period of 14 days. Curves show the levels of active MMP-1 protein, active MMP-13 protein (scaled by a factor of 10), and collagen fragments. In each simulation, both IL-1 and OSM were added. A, Effect without MMP activator. B, Effect with addition of MMP activator. See Figure 1 for definitions.

Figure 3

Simulation results for interventions with interleukin-1 receptor (IL-1R) or oncostatin M receptor (OSMR) antagonists, using a simulated time period of 14 days. Simulated conditions consisted of IL-1 plus OSM plus matrix metalloproteinase activator. Curves show the percentage of aggrecan and collagen degraded. A, Effect of IL-1R antagonist. B, Effect of OSMR antagonist. Arrows in A and B show the direction of increase in the ratio of receptor antagonist to receptor (1, 10, 100, 1,000).

Figure 4

Simulation results showing the effect of inhibiting JAK-1, JNK, or p38 activity, using a simulated time period of 14 days. Simulated conditions consisted of IL-1 plus OSM plus MMP activator. Curves show the percentage of aggrecan and collagen degraded. A, Effect of JAK-1 inhibition. B, Effect of p38 inhibition. Arrows show the direction of increase (0–100%, in steps of 10; k_phoscFos = 5 × 10⁻⁷, 4.5 × 10⁻⁷, 4 × 10⁻⁷, …, 0) molecules⁻¹ seconds⁻¹). C, Effect of JNK inhibition. Arrows show the direction of increase (0–100%, in steps of 10; k_phoscJun = 1 × 10⁻⁴, 9 × 10⁻⁵, 8 × 10⁻⁵, …, 0 molecules⁻¹ seconds⁻¹). D, Effect of JNK inhibition on IL-1 plus OSM–induced cartilage breakdown. Bovine nasal cartilage discs were cultured in the presence of IL-1 (0.5 ng/ml) plus OSM (10 ng/ml) in the presence or absence of the JNK inhibitor SP600125 (SP; 30 μM) or DMSO control. Medium was removed on day 7, and fresh reagents added. On day 14, the medium was removed, and the remaining cartilage was papain digested. The hydroxyproline assay was used to measure the release of collagen into the medium on day 7 and day 14. Values are the mean ± SEM of data accumulated from a minimum of 2 different experiments of a total of 4 experiments conducted. ∗∗ = P < 0.01 for cytokine treatment versus cytokine plus inhibitor treatment, by t-test. See Figure 1 for definitions.

Figure 5

Simplified network diagram showing the involvement of JNK and p38 in the system. Interleukin-1 (IL-1) activates tumor necrosis factor receptor–associated factor 6 (TRAF6), which phosphorylates both p38 and JNK. JNK phosphorylates c-Jun, and p38 phosphorylates c-Fos, which has been up-regulated via the oncostatin M (OSM)/JAK-1/STAT-3 signaling pathway. Phosphorylated c-Fos binds to phosphorylated c-Jun to form the activator protein 1 (AP-1) complex. See Supplementary Figures 1 and 2 for diagrams showing all of the reactions (available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.38297/abstract).

Figure 6

Simulation results showing model predictions for the overexpression of TIMP-1 or TIMP-3 protein, using a simulated time period of 14 days. Simulated conditions consisted of IL-1 plus OSM plus MMP activator. Curves show the percentage of aggrecan and collagen degraded. A, Effect of TIMP-1 overexpression. B, Effect of TIMP-3 overexpression. Arrows in A and B show the direction of increase (2 × 10², 2 × 10³, 2 × 10⁴, 2 × 10⁵ molecules). See Figure 1 for definitions.