INTRODUCTION

Intervertebral disc degeneration (IDD) is a major contributor to the onset of low back pain. Within the Nucleus Pulposus (NP) a principal sign of IDD is an elevated presence of pro-inflammatory cytokines. Especially Interleukin 1beta (IL-1B) and Tumor Necrosis Factor Alpha (TNF-a) contribute to a catabolic shift in the NP homeostasis¹. The latter is characterized by an upregulation of proteases, such as MMP3 and ADAMTS4 that degrade the NP extracellular matrix. Primary inhibitors of those proteases are the tissue inhibitors of matrix metalloproteinases (TIMP). The objective of our work is to anticipate the dynamics among proteases, TIMP and pro-inflammatory cytokines within the human NP.

METHODS

The mRNA expressions of TIMP subgroups 1, 2 and 3, of MMP3, ADAMTS4, IL-1B and TNF-a, and the corresponding protein levels were estimated to anticipate protein-protein inhibitions between proteases and TIMP. At the mRNA level, the parallel network methodology (PN_t-Methodology)² was used to estimate relative mRNA expressions of TIMP and proteases under user-defined levels of glucose, pH, magnitude (intradiscal pressure) and frequency. Results were then coupled to a directed network modeling approach that predicted corresponding protein levels. Five human moving habits were simulated under optimal nutritional conditions (5 mM Glucose; pH 7.1): walking, sitting, jogging, hiking with 20 kg extra weight and exposure to vibration (sitting in/on a motor vehicle). Cell activities under the presence and absence of proinflammatory cytokines were estimated (Fig.1). A 1:1 stoichiometry and a selective binding affinity for TIMP1,2 for MMP3 and TIMP3 for ADAMTS4 was considered (Fig.1).

RESULTS

All activities but exposure to vibration could be related to anabolic activities. Thereby, especially TIMP3 levels were elevated, whereas TIMP1 and TIMP2 expressions were predicted to be lower. In absence of pro-inflammatory cytokines, cell activity of TIMP is pronounced, leading to a strong downregulation of MMP3 and ADAMTS4 proteins in anabolic activities. Catabolism due to presence of pro-inflammatory cytokines was characterized by more proteases and stagnant TIMP mRNA expression, leading to an impaired inhibition of MMP3, but not of ADAMTS4, in all anabolic human habits (Fig 2).

DISCUSSION

The model predicts higher amounts of TIMP3 compared to TIMP1,2, which agrees with experimental measurements, where lower percentages of TIMP1,2 immunopositive cells compared to TIMP3 immunopositive cells were found⁵. Under the influence of pro-inflammatory cytokines, stagnant TIMP and rising amounts of proteases were predicted, which agrees with observations made at higher grades of IDD⁵. Interestingly TIMP3 was predicted to be strongly expressed under anabolic activities, despite the presence of pro-inflammatory cytokines. The proposed combined network modelling approach successfully provided insights into the inhibitory regulations of TIMP. The utility of the methodology is to harness the potential of predictive in silico modelling to apprehend the key biological interactions that may contribute to the development of IDD. Future work should include time as a variable to dynamically estimate cell activity during human habits, and the corresponding roles of TIMP and proteases considering spatiotemporal effects of the tissue turnover.

ACKNOWLEDGEMENTS: European Commission & Research Council (ERC-2021-CoG-O-Health-101044828)