The Involvement of Protease Nexin-1 (PN1) in the Pathogenesis of Intervertebral Disc (IVD) Degeneration

Protease nexin-1 (PN-1) is a serine protease inhibitor belonging to the serpin superfamily. This study was undertaken to investigate the regulatory role of PN-1 in the pathogenesis of intervertebral disk (IVD) degeneration. Expression of PN-1 was detected in human IVD tissue of varying grades. Expression of both PN-1 mRNA and protein was significantly decreased in degenerated IVD, and the expression levels of PN-1 were correlated with the grade of disc degeneration. Moreover, a decrease in PN-1 expression in primary NP cells was confirmed. On induction by IL-1β, the expression of PN-1 in NP cells was decreased at day 7, 14, and 21, as shown by western blot analysis and immunofluorescence staining. PN-1 administration decreased IL-1β-induced MMPs and ADAMTS production and the loss of Agg and Col II in NP cell cultures through the ERK1/2/NF-kB signaling pathway. The changes in PN-1 expression are involved in the pathogenesis of IVD degeneration. Our findings indicate that PN-1 administration could antagonize IL-1β-induced MMPs and ADAMTS, potentially preventing degeneration of IVD tissue. This study also revealed new insights into the regulation of PN-1 expression via the ERK1/2/NF-kB signaling pathway and the role of PN-1 in the pathogenesis of IVD degeneration.

serpin domain, which is necessary for its inhibitory activity 21,22 . PN-1 can inactivate several proteases, including plasmin, plasminogen, and urokinase, preventing cartilage degradation [23][24][25] . As is known, the plasmin/plasminogen enzymatic cascades play an important role in cartilage catabolism, which is mediated by activated matrix metalloproteinases (MMP). IVD tissues share pathophysiological characteristics with osteoarthritis (OA) 26 . Various types of proteases are directly involved in ECM degradation; however, MMPs are considered the major enzymes 27 . Given the role of serine proteases in OA pathology, the endogenous serine protease inhibitor PN-1 could share a similar role in ECM degeneration involving the activated plasminogen/plasmin and MMP systems in IVD tissue.
Here, we hypothesize that the expression of PN-1 decreases during IDD, which is related to altered disc-cell function and subsequent characteristic features of IDD. Thus, this study aimed to investigate the expression of PN-1 during IDD, and to determine the induction of its regulation by pro-inflammatory cytokine TNF-α and IL-1β. Then, we examined the effects of PN-1 on the expression and activity of MMPs and ADAMTS in NP cells. Finally, the associated signaling pathway was investigated, focusing on the activation of ERK1/2/NF-κB.

Materials and Methods
This study complies with accepted ethical standards for human and animal research. The research has been approved by the Ethics Committee of Tongji medical college, and written informed consent was obtained from all participants.
Isolation and Culture of NP Cells. Normal human NP cells were isolated from the discs in younger scoliosis patients undergoing deformity correction surgery (Union Hospital, Tongji Medical College). Briefly, NP tissues were minced into small fragments and enzymatically digested in 0.2% type II collagenase and 0.25% trypsin for 3 hours. After being filtered and washed in PBS, the cells were seeded and cultured in growth medium (DMEM/F-12 supplied with 20% fetal bovine serum, 50 U/mL penicillin, and 50 μg/mL streptomycin) in a 5% CO 2 incubator. The cells were passaged two to three times for use in the following experiments.

Stimulation of NP cells.
NPs were seeded and cultured in 12-well dishes and allowed to reach 100% confluence. After being serum-starved for 2 hours, the cells were incubated in growth medium alone or in stimulatory medium containing IL-1β (10 ng/mL), TNF-α (50 ng/mL), or TGF-β1 (1 ng/mL) for 7, 14, and 21 days, and the medium was changed every 3 days. In control cultures, the medium was replaced at the indicated times. Each treatment was performed in three different wells. In another experiment, to test the signaling pathways involved in NP cells, PN-1 and/or IL-1β at different concentrations were used to treat cells for 24 h, and inhibitors of NF-kB ( Caffeic Acid Phenethyl Ester) and ERK (GDC-0994) were used.
ELISA. PN-1 is a secreted protein that acts in the ECM, and therefore, the protein analysis was performed on cell-conditioned media. The level of PN-1 was detected by ELISA from the stored supernatant of each group at different time points after induction, according to the manufacture's procedures (USCN, SED381H).
Immunohistochemistry. Immunohistochemistry (IHC) for PN-1 expression was performed on degenerated IVD tissue. To detect protein expression, antigen retrieval was performed on the IVD cryosections by incubation in 0.8% hyaluronidase at 37 °C for 60 min. The sections were washed gently with PBS for 5 min, and then blocked in 0.5% goat serum for 40 min at room temperature. Subsequently, the samples were incubated with polyclonal anti-PN-1 (1:100), or control rabbit IgG (2 μg/mL) at 4 °C overnight, followed by washing with PBS. Then secondary antibody conjugated to horseradish peroxidase (diluted 1:3000) was applied for 20~30 min at room temperature. After washing, the sections were incubated with diaminobenzidine (DAB; Solarbio, DA1010) reaction solution until color was detected, followed by counterstaining with hematoxylin, and images were captured using a microscope (Olympus).
Each sample was analyzed in triplicate for both, target and control genes.
Western Blot Analysis. The NP cells from each culture dish were collected and lysed in RIPA buffer. The lysates were centrifuged, and the protein concentrations were determined by the BCA protein assay. In each group, equivalent amounts of protein (50 μg) were used for electrophoresis. After electrophoresis, the proteins were electrotransferred to 0.45 μm-pore-diameter polyvinylidene difluoride (PVDF) membranes (Invitrogen). After being blocked in Tris-buffered saline containing Tween-20 with 5% milk powder, the membranes were immunolabeled with specific antibodies for PN-1 (1:1500), Gelatin zymography. Gelatin zymography is a sensitive and quantifiable method for analyzing proteolytic activity of enzymes, including MMP-2 (gelatinase A) and MMP-9 (gelatinase B). In brief, to detect the gelatinolytic activity of MMPs, cell extracts from different groups were incubated at 37 °C for 20 min in SDS sample buffer, and then electrophoresed on 8% polyacrylamide gels at 4 °C. Pre-stained SDS-PAGE markers and MMP-9 and MMP-2 standards were used to estimate the molecular weights of the bands. After electrophoresis, the gels were eluted twice with 2.5% Triton X-100, 50 mmol/L Tris-HCl, and 5 mmol/L CaCl2, pH 7.6 for 40 min. The gels were rinsed with wash buffer without Triton X-100. Subsequently, the gels were incubated with 50 mmol/L Tris-HCl, 5 mmol/L CaCl2, and 0.02% Brij-35, pH 7.6 at 37 °C for 42 h, which allows substrate degradation. The gels were fixed in 30% methanol and 10% acetic acid for 30 min, and stained with 0.5% Coomassie Brillant Blue for 3 h. Proteolytic bands were visualized using destaining solution A, B, and C (30, 20, and 10% methanol, and 10, 10, and 5% acetic acid, respectively). Finally, the gels were scanned, and MMP-9 and MMP-2 proteolytic activities were semiquantified based on computer-assisted image analysis. Results were expressed as the relative percentage of gelatinolytic activity (density of the active band in each group).
Statistical evaluation. All statistical analyses and plots were performed using the Prism GraphPad 5.0 software. Values are presented as mean ± SEM. All measurements were performed in triplicate. Changes in gene expression between the various treatment groups were evaluated by one-way ANOVA or Student's t-test. A p-value < 0.05 was considered statistically significant.

PN-1 gene expression in human NP tissue. PN-1 mRNA expression levels in human IVD tissue with
differing grades of disc degeneration were normalized to β-actin and presented as 2 −ΔΔCt . As seen in Fig. 1A, there was significant expression of PN-1 in non-degenerated IVD tissue (Grade 2), but expression was markedly weakened in degenerated IVD tissue. FN, a protein that induces intervertebral disc degeneration, was increased in the IVD samples with increased degrees of disc degeneration (Fig. 1B). Western blot analysis was performed to detect the protein expression of PN-1, FN, and FN-fs in IVD with varying degrees of disc degeneration (Fig. 1C).
As the photomicrographs show, the expression of PN-1 was similar to that seen by PCR (Fig. 1D), indicating that gene expression of PN-1 decreased during the process of disc degeneration. FN-fs play important roles in the development of human disc degeneration. In the study, FN and FN-f were analyzed by western blot, which showed that the highest amount of FN-fs emerged in the moderately degenerative discs (grades III) and retained a high level during IDD (Fig. 1E).
Histological examination for PN-1 expression in human NP tissue. The degree of degeneration of the IVD tissue was confirmed by Alcian blue staining. IHC was used to confirm the expression of PN-1 in human IVD tissue. As shown in Fig. 2, the extent of PN-1 expression by NP cells was clearly decreased in the degenerated IVD sample. Thus, the expression of PN-1 is inversely correlated with the grade of disc degeneration, in that a higher degenerative degree correlated to a lower expression level of PN-1 (Fig. 2).

PN-1 Expression in NP cells in response to IL-1β stimulation.
To determine the expression of PN-1 mRNA at the cellular level under degeneration, NP cells were cultured with IL-1β, and mRNA was detected using RT-PCR at different time points (10, 30, and 60 min, and 6, 12, and 24 h). As Fig. 3 shows, administration of IL-1β resulted in a significant decrease in PN-1 mRNA expression at 24 h (Fig. 3A,B). Cell lysates from different time points were collected and analyzed by western blot, which also confirmed a significant decrease of PN-1 expression at 24 h (Fig. 3C,D). The analysis of PN-1 expression levels on cell-conditioned media was performed by an ELISA, which indicated that PN-1 concentration increased significantly in NP cells in a time-dependent manner after IL-1β induction (Fig. 3E). Therefore, IL-1β induction could decrease the gene expression of PN-1 in NP cells, and the decrease in cytosolic PN-1 protein expression is due to an increase in protein secretion in the culture media.

Pro-inflammatory cytokines IL-1β and TNF-α decrease PN-1 expression in NP cells.
To detect the expression of PN-1 regulated by IL-1β and TNF-α, human NP cells were cultured with IL-1β (10 ng/mL), TNF-α (50 ng/mL), IL-1β+TGF-β1 (10 ng/mL), and TNF-α+TGF-β1, and detected by western blotting and immunofluorescent staining. The protein expression levels in different groups were examined at different time points (Fig. 4A). Western blot analysis showed a significant decrease in the PN-1 protein level in NP cells after administration of IL-1β and TNF-α on days 7 (Fig. 4B) and 14 (Fig. 4C). There was still a trend of decreased PN-1 expression at day 21 (Fig. 1D), but it was not as evident as on days 7 and 14. The results also indicated that TGF-β1

PN-1 inhibits IL-1β-induced MMP production in NP cells.
To investigate the effects of PN-1 on the protein expression of matrix degrading enzymes in IVD, NP cells were cultured in the presence of IL-1β with or without PN-1, and proteins were detected by western blotting. Changes in MMP-3, ADAMTS-4, MMP-13, MMP-9, ADAMTS-5, Agg, and COL2 levels were observed (Fig. 5A). Specifically, IL-1β treatment resulted in a significant up-regulation of MMP protein expression after 24 h stimulation in human NP cells. Interestingly, recombinant PN-1 was able to reverse the production of MMPs induced by IL-1β in human NP cells. Compared to IL-1β stimulation alone, PN-1 administration could reverse the expression of MMP-3 (Fig. 5B), ADAMTS-4 (Fig. 5C), MMP-13 (Fig. 5D), MMP-9 (Fig. 5E), and ADAMTS-5 (Fig. 5F) in a dose-dependent manner. The optimal concentration of PN-1 was 100 ng/mL. We detected the expressions of Agg and COL2. The results showed that IL-1β treatment down-regulated expressions of Agg and COL2, while PN-1 administration could attenuate the effects, reaching the highest levels at the optimal dose of 100 ng/mL.  (83 kDa) was reduced 0.36 and 0.29 fold after PN-1 administration, compared to after administration of IL-1β alone, at 30 and 60 min (Fig. 6B). Similarly, MMP-2 (62 kDa) was found to be reduced 0.54 and 0.22 fold (Fig. 6C). These results indicated that IL-1β causes a significant up-regulation of MMP-9/2 expression and gelatinolytic activity, while exogeneous PN-1 could invert the effects in human NP cells.

PN-1 inhibits IL-1β-induced MMP production through ERK1/2/NF-κB activation in NP cells.
To investigate the effects of PN-1 on the signaling pathway of IL-1β-induced IVD degeneration, we treated human NP cells with IL-1β, with or without specific inhibitors of NF-kB (Caffeic Acid Phenethyl Ester) and ERK (GDC-0994). Levels of phosphorylated and unphosphorylated P38, ERK1/2, and P65 proteins were determined by western blotting (Fig. 7). The levels of unphosphorylated P38 were similar in both groups at different time points. IL-1β stimulation induced increased levels of phosphorylated p38, which was significantly inhibited in NP cells treated by recombinant PN-1 addition (Fig. 7B). ERK1 and ERK2 contribute to cytokine dependent induction of intervertebral disc degeneration, and hence, activate the inflammatory-related signaling molecule NF-κB. To obtain further insights into the underlying mechanism, NF-κB signaling was also investigated. We observed that IL-1β induced increased levels of phosphorylated ERK1/2, which was significantly blocked by the addition of recombinant PN-1 (Fig. 7C,D). Hence, we sought to analyze whether PN-1 was able to modulate the phosphorylated levels of downstream effector P65, besides NF-κB. As shown in Fig. 7E, and as expected, IL-1β was able to increase the phosphorylated levels of P65, and the addition of recombinant PN-1 reduced the phosphorylated levels of P65. Finally, an ERK inhibitor was used, which showed that the phosphorylated levels of ERK1/2 were partially blocked by recombinant PN-1 in NP cells (Fig. 7G,H); an NF-kB inhibitor was also used, and as expected, the results showed that levels of P-P65 were significantly blocked by PN-1 addition. Therefore, the ERK1/2/NF-κB signaling pathways were involved in the regulation of PN-1 in the pathogenesis of IVD degeneration.

Discussion
Our findings from clinical samples provide evidence for our hypothesis that PN-1 is expressed in human IVD tissue and is highly regulated during the process of IDD in vivo and in vitro. These results also indicate that PN-1 could inhibit the activation of MMPs and ADAMTS through the ERK1/2/NF-κB signaling pathway. In all, these findings provide insights into the molecular mechanisms of IDD. Pro-inflammatory cytokines such as IL-1β and TNF-α, influence the degeneration of intervertebral discs, through protease pathways. PN-1, a 45-to 50-kDa glycoprotein, is encoded by the SERPINE2 gene on human chromosome 2q99-q35 29 . PN-1 is a member of the serine protease inhibitor (serpin) family, involved in tissue remodeling, cellular invasion, matrix degradation, and tumor growth 23 . PN-1 is barely detectable in plasma, but is found in many organs, such as the brain, heart, kidneys, lungs, spleen, muscle, and cartilage 30,31 , and is produced by most cell types, such as glial cells 32 , smooth muscle cells (SMCs) 33 , endothelial cells 34 , and fibroblasts 35 . PN-1 primarily localizes at the cell surface by binding to glycosaminoglycans (GAGs) 36 , which trap and potentiate PN-1 activity within the pericellular space, so that proteolysis does not lead to widespread protein destruction 21,37,38 . Therefore, the activity of proteases on cells is regulated by PN-1 bound to cell-surface GAGs. Furthermore, PN-1 plays an important role in the process of proteolysis in an irreversible and highly complex manner through the inhibition of proteases 39 .  Osteoarthritis (OA), a joint disease caused by cartilage loss, is strongly associated with a net loss of aggrecan and collagen breakdown caused by an imbalance in ECM homeostasis 40 . The function of articular cartilage depends on the molecular composition of the ECM, which mainly consists of collagen II/IX/XI fibrils and proteoglycan-glycosaminoglycan networks of aggrecan and hyaluronan. ECM components are essential in regulating chondrocyte metabolism and function, and play a crucial role in the ability of the tissue to withstand compressive forces and respond to mechanical loading 40 . In the cartilage of OA patients, there is a significant increase in the level of serine protease Htra1 (high temperature factor A-1), which mediates proteolysis of aggrecan and contributes to OA pathology 41,42 . Initially, PN-1 was identified as a serine protease inhibitor in cartilage anlagen during embryogenesis, and interacted with both ECM and intracellular cartilage components 30,43,44 , which is consistent with its role in the binding and inactivation of extracellular proteases followed by internalization 45 . The activity of PN-1 is regulated by several matrix components, such as FN 46 , collagen II, and collagen I 47 . Therefore, serine proteases play an important role in cartilage pathology, and endogenous serine protease inhibitors such as PN-1 could potentially be promising new therapeutic targets in OA 43,48,49 .
IVD tissues share pathophysiological characteristics with OA 26 ; however, little is known about the involvement of serine proteases in the pathogenesis of disc degeneration. Type I and II collagen fibrils make up almost 20% and 70% of the collagen network in the dry weight of the nucleus pulposus and annulus fibrosus, respectively 50 . Previous studies have confirmed that IDD is associated with fibrosis of the NP, in which the collagen meshwork is gradually replaced by fibrils of abnormal size and rigidity 6 , resulting in altered gene expression for matrix molecules, degradation enzymes, and catabolic cytokines 18,51 . In addition, alterations in PG lead to a decrease in hydration, loss of structural integrity, and an inability to withstand loads 52,53 . The molecular mechanisms of collagen fibrilogenesis are complicated, but highly related to mechanical stress, collagen genetic types, and aging 54 . Data are available for structural proteins of IVD, such as aggrecan 55 and collagen 56 ; however, little is known about the longevity of serine proteases from this tissue. The imbalance between serine proteases and serpins may have an effect on the degeneration of IVD; thus, expression of PN-1 is important to gain further insight into the mechanisms of IDD. The present study was designed to investigate the changes of PN-1 expression in degenerated IVD of different grades using RT-PCR and IHC. Investigations in human IVD tissue have demonstrated a decreased expression of PN-1 following IDD. Moreover, we found that PN-1 mRNA and protein levels were coordinately regulated during the progression of IDD in a stage-specific manner. Inflammatory cytokines, such as IL-1β and TNF-α, have been known to induce degeneration of IVD. Based on in vitro experiments, changes in MMP protein and mRNA expression levels have been detected by western blotting and RT-PCR, and activity of MMPs have been examined by gelatin zymography. IL-1β, TNF-α, and other pro-inflammatory cytokines have also been shown to induce MMP production in NP cells. As our results indicated, when stimulated with IL-1β, the secretion of MMPs and ADAMTS by human NP cells increased, whereas PN-1 administration could antagonize these proteases and potentially preserve IVD tissue from degeneration.
Fibronectin (FN) and its fragments (FN-fs) have been found to accumulate during disc degeneration and acceleration of IVD degeneration in rabbits 57 . Within human degenerated-disc tissue samples, a marked increase in FN and FN-fs has been observed; thus, they are considered an integral part of the underlying pathology of IDD 58-60 . PN-1 activity is at least partly regulated by the extracellular matrix component fibronectin 46,61 , and fibronectin and its fragments have been observed to be potent inducers of MMP expression during IVD degeneration 62 . In addition, serine proteases have been confirmed to regulate MMP expression on the generated fibronectin fragments 59 . Therefore, an endogenous serine protease inhibitor, such as PN-1, has potentially roles in modulating MMP expression in IVD cells. Nuclear factor kappa B (NF-κB) is a family of transcription factors, which becomes activated in response to inflammation, damage, and stress 63,64 . MMPs have been identified as NF-κB target genes in IVD cells responsible for ECM degradation, including MMP-1, MMP-2, MMP-3, MMP-9, and MMP-13 15,59,65 . NF-κB play a , and ADAMTS-5 (F) protein levels in NP cells treated with IL-1β, IL-1β + PN-1, IL-1β + ERK inhibitor, and IL-1β + NF-kB inhibitor. Densitometric analysis shows the promotive effect on Aggrecan (G) and COL2 (H) protein levels in NP cells treated with IL-1β, IL-1β + PN-1, IL-1β + ERK inhibitor, and IL-1β + NF-kB inhibitor. Data are representative of three independent experiments, and p-values are shown: *p < 0.05; **p < 0.01 compared to IL-1β stimulation alone.
Scientific RepoRts | 6:30563 | DOI: 10.1038/srep30563 central role in MMP and ADAMTS expression in NP cells stimulated by IL-1β 66 . Activation of the NF-κB signaling pathway results in increased matrix-degrading enzyme activity in the NP, which is an important catabolic pathway involved in IDD 67 . Therefore, NF-κ B pathway has been fully investigated in human NP cells stimulated with PN-1.
Disc degeneration is characterized by the loss of Agg and COL2 68 . Agg is a type of proteoglycan responsible for the normal structure of discs, while COL2 is an important component of the ECM, helping the IVD bear pressure. In this study, IVD degeneration induced by IL-1β lead to loss of Agg and COL2 in NP cells, which was associated with upregulation of MMPs and ADAMTS. PN-1 intervention reversed the downregulation of Agg and COL2 and the upregulation of MMPs, implying that PN-1 could partially antagonize NF-κB pathway-specific transcription factors and contribute to maintaining matrix integrity of the IVD. Moreover, we found that PN-1 could inhibit IL-1β-induced activation of ERK1/2 in NP cells. Therefore, PN-1 intervention can potentially reverse the expression of MMPs and ADAMTS through the ERK1/2/NF-κB signaling pathway in IDD. Agg and COL2 are necessary for the rebuilding of IVD, which also can be upregulated through the pathway by PN-1.
In conclusion, under inflammatory conditions, activated ERK1/2/NF-κB contributes to the pathogenesis of IDD, manifested by the upregulation of MMPs and ADAMTS and degraded disc matrix macromolecules Agg and collagen II. PN-1 is involved in pathogenesis, at least in part by suppressing IL-1β-induced activation of ERK1/2/ NF-κB and its downstream targets. Thus, decreased expression of PN-1 may serve as an important endogenous mechanism to increase the deleterious effects of excessive serine protease activity within the IVD.