MMP-2 mediates local degradation and remodeling of collagen by annulus fibrosus cells of the intervertebral disc

Abstract

Introduction: Degeneration of the intervertebral disc (IVD) is characterized by marked degradation and restructuring of the annulus fibrosus (AF). Although several matrix metalloproteinases (MMPs) have been found to be more prevalent in degenerate discs, their coordination and function within the context of the disease process are still not well understood. In this study, we sought to determine whether MMP-2 is associated with degenerative changes in the AF and to identify the manner by which AF cells use MMP-2.

Methods: Two established animal models of disc degeneration, static compression and transannular needle puncture of rodent caudal discs, were examined for MMP-2 immunopositivity. With lentiviral transduction of an shRNA expression cassette, we screened and identified an effective shRNA sequence for generating stable RNA interference to silence MMP-2 expression in primary rat AF cells. Gelatin films were used to compare gelatinase activity and spatial patterns of degradation between transduced cells, and both noninfected and nonsense shRNA controls. The functional significance of MMP-2 was determined by assessing the ability for cells to remodel collagen gels.

Results: Both static compression and 18-g annular puncture of rodent caudal discs stimulated an increase in MMP-2 activity with concurrent lamellar disorganization in the AF, whereas 22-g and 26-g needle injuries did not. To investigate the functional role of MMP-2, we established lentivirus-mediated RNAi to induce stable knockdown of transcript levels by as much as 88%, and protein levels by as much as 95% over a 10-day period. Culturing transduced cells on gelatin films confirmed that MMP-2 is the primary functional gelatinase in AF cells, and that MMP-2 is used locally in regions immediately around AF cells. In collagen gels, transduced cells demonstrated an inability to remodel collagen matrices.

Conclusions: Our study indicates that increases in MMP-2 observed in human degenerate discs are mirrored in experimentally induced degenerative changes in rodent animal models. AF cells appear to use MMP-2 in a very directed fashion for local matrix degradation and collagen remodeling. This suggests that MMP-2 may have a functionally significant role in the etiology of degenerative disc disease and could be a potential therapeutic target.

Figure 1

Schematic representation of collagen gel mechanical testing. Collagen gels were clamped on each end at the mesh, and creep testing was performed in tension by using hanging weights. An initial weight of 1.5 g was added, followed by increments of 0.5 g at 2-minute intervals until failure. The cross-sectional area, measured before loading, was used to calculate applied engineering stress, and analysis of deformations in individually captured frames was used to calculate the engineering strain. Mechanical viscoelastic behavior was quantified by fitting data to the standard linear viscoelastic solid-model equation for creep loading.

Figure 2

Static compression induces matrix metalloproteinase 2 (MMP-2) activation. Immunohistochemical staining using an antibody (CA719E3C; Labvision, Fremont, CA, USA) that recognizes only the latent MMP-2 zymogen in nonloaded (left) and loaded (right) IVDs. These data are unpublished, but presented in a poster [15]. The loss of positive staining from cells of loaded discs coincided with previously observed increases in activation of MMP-2, as determined with enzyme-linked immunosorbent assay (ELISA) [10].

Figure 3

Transannular puncture injury upregulates matrix metalloproteinase 2 (MMP-2). Immunohistochemical staining of MMP-2 in rat caudal discs subjected to transannular punctures by using hypodermic needles (18-g, 22-g, and 26-g) or subjected to sham surgery (no puncture). Although AF cells in all tissue sections exhibited some immunopositivity, more positive cells were found, and the qualitative intensity of staining was markedly higher in punctured discs, particularly for 18-g needles. We previously showed that 18-g hypodermic needle punctures are associated with a greater incidence of degenerative changes in the AF [16].

Figure 4

Stable silencing of matrix metalloproteinase 2 (MMP-2) by using shRNA-mediated RNAi. Silencing of MMP2 in primary rat caudal annulus fibrosus (AF) cells at 4 and 10 days after lentivirus infection, with and without treatment with blasticidin to select for pure populations of transduced cells. (a) Relative expression of treatment groups compared with noninfected controls, as quantified with real-time RT-PCR. At 4 and 10 days, MMP2 transcript levels were significantly decreased by 70% and 88%, respectively, in blasticidin-treated cultures (*P < 0.03). (b) Total protein levels for MMP-2 in cell lysates and (c) in conditioned media. At 10 days, total MMP-2 is 95% and 90% lower (*P < 0.05) in blasticidin-treated cell lysates and culture media, respectively, compared with that in noninfected controls (Cells).

Figure 5

shRNA construct against matrix metalloproteinase 2 (MMP-2) (shMMP2) inhibits localized gelatin degradation surrounding annulus fibrosus (AF) cells. Microscopy images of gelatin films seeded with (a) no cells, (b) noninfected AF cells, (c) cells transduced with shMMP2, and (d) cells transduced with shNon. Films were stained at 4 days in culture with Ponceau S and visualized immediately. Scale bars represent 200 μm. All images were taken at 100×. Films with no cells (a) stained evenly red, whereas films seeded with noninfected AF cells (b) showed focal degradation of gelatin in the areas immediately surrounding cells, which is shown by the lighter-stained regions. Similar results are seen with cells transduced with shNon (d). Cells transduced with shMMP2 (c) show little to no focal degradation of gelatin, implying that MMP-2 is directly involved in gelatin breakdown. (e) Quantification of gelatin degradation. Gelatin films seeded with noninfected control AF cells and shNon-transduced cells were similar in percentage degradation of gelatin, which corroborated visual results. Compared with the shMMP2-transduced cells, degradation was significantly higher in the noninfected AF control group (P = 0.001) and shNon-transduced group (0.002), as observed visually.

Figure 6

shRNA construct against matrix metalloproteinase 2 (MMP-2) (shMMP2) impairs annulus fibrosus (AF) cell-mediated remodeling of collagen gels. Histology images of collagen gels containing (a) no cells, (b) nontransduced AF cells, (c) a pure population of cells transduced with shMMP2, (d) a pure population of cells transduced with shNon, and (e) a mixed population of shMMP2-transduced and nontransduced cells. All gels were cultured for 7 days. Images for (a) through (d) were taken at 400×; scale bars represent 50 μm. Image for (e) was taken at 100×; scale bar represents 200 μm. Noninfected AF cells (b) and shNon-transduced cells (d) were shown to restructure collagen gels. These cells appeared elongated and anchored to the fibers. Cells transduced with shMMP2 (c) appeared similar to gels without cells (a), with little to no restructuring of the collagen network and rounded cells. In gels containing mixed populations of cells (e), collagen fibers were reorganized but rounded cells were clearly visible without reorganization of fibers in their immediate surroundings.

Figure 7

shRNA construct against matrix metalloproteinase 2 (MMP-2) (shMMP2) inhibits strengthening of collagen gels by annulus fibrosus (AF) cells. Parameters obtained from mechanical testing of collagen gels. (a) Rupture stress of collagen gels seeded with wild-type noninfected cells was significantly higher than that of acellular gels, gels populated with shNon-transduced cells, and gels populated with shMMP2-transduced cells (P < 0.05). (b) Short-term and (c) long-term moduli were lower in cell-seeded collagen gels but not significantly different from acellular gels. (d) Collagen gels seeded with cells possessed significantly lower viscous coefficients than did acellular gels (P < 0.05).