SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Objective

Intervertebral disc degeneration is linked to loss of extracellular matrix (ECM), particularly the early loss of aggrecan. A group of metalloproteinases called aggrecanases are important mediators of aggrecan turnover. The present study was undertaken to investigate the expression of the recognized aggrecanases and their inhibitor, tissue inhibitor of metalloproteinases 3 (TIMP-3), in human intervertebral disc tissue.

Methods

Twenty-four nondegenerated and 30 degenerated disc samples were analyzed for absolute messenger RNA (mRNA) copy number of ADAMTS 1, 4, 5, 8, 9, and 15 and TIMP-3 by real-time reverse transcription–polymerase chain reaction. Thirty-six formalin-fixed embedded intervertebral disc samples of varying grades of degeneration were used for immunohistochemical analyses. In addition, samples from 8 subjects were analyzed for the presence of matrix metalloproteinase (MMP)– and aggrecanase-generated aggrecan products.

Results

Messenger RNA for all the aggrecanases other than ADAMTS-8 was identified in intervertebral disc tissue, as was mRNA for TIMP-3. Levels of mRNA expression of ADAMTS 1, 4, 5, and 15 were significantly increased in degenerated tissue compared with nondegenerated tissue. All these aggrecanases and TIMP-3 were also detected immunohistochemically in disc tissue, and numbers of nucleus pulposus cells staining positive for ADAMTS 4, 5, 9, and 15 were significantly increased in degenerated tissue compared with nondegenerated tissue. Aggrecan breakdown products generated by MMP and aggrecanase activities were also detected in intervertebral disc tissue.

Conclusion

The aggrecanases ADAMTS 1, 4, 5, 9, and 15 may contribute to the changes occurring in the ECM during intervertebral disc degeneration. Targeting these enzymes may be a possible future therapeutic strategy for the prevention of intervertebral disc degeneration and its associated morbidity.

Chronic low back pain affects >70% of people at some point in their lives (1), with ∼10% being chronically disabled. The causes of low back pain are multifactorial, although ∼40% of all cases involve degeneration of the intervertebral discs (2). During degeneration, the matrix of the intervertebral disc undergoes structural, mechanical, and molecular changes resulting in a loss of demarcation between the outer annulus fibrosus and the inner nucleus pulposus. Additionally, alterations in collagen type and a decrease in proteoglycan content result in loss of tissue integrity, decreased hydration, and inability to withstand load (3). Importantly, the loss of proteoglycan, predominantly aggrecan, is considered to be an early indicator of intervertebral disc degeneration (4). Aggrecan molecules possess long core proteins with many chondroitin sulfate (CS) and keratan sulfate (KS) glycosaminoglycan (GAG) side chains (3). These GAG side chains are polyanionic due to the high content of carboxyl and sulfate groups, and thus they attract and bind water molecules, hydrating the tissue. Degradation of the proteoglycans, especially aggrecan, in disc degeneration results in decreased hydration and therefore in a reduced ability to resist compressive load.

As with articular cartilage, the extracellular matrix (ECM) in normal discs undergoes a process of remodeling and relies upon a delicate balance between matrix synthesis and degradation. In intervertebral disc degeneration there is a net increase in matrix-degrading proteinase activity compared with proteinase inhibitors, which disrupts the normal balance and leads to breakdown of ECM (5). Although the matrix metalloproteinases (MMPs) are thought to play a role in this process (6), it has been suggested that aggrecanases may also be involved, particularly since they participate in the degradation of aggrecan in articular cartilage and osteoarthritis (OA) (7, 8).

Aggrecanases are proteinases that cleave a particular glutamyl bond in the interglobular domain (IGD) of aggrecan, thereby releasing the bulk of the aggrecan molecule from the tissue (9). The first aggrecanase to be identified was termed aggrecanase 1 by Tortorella et al in 1999 (10) and is now known as ADAMTS-4. Later that year, the same research team reported the identification of a second enzyme, aggrecanase 2 (11), now known to be ADAMTS-5. The ADAMTS belong to a branch of the M12B adamalysin subfamily of metalloendopeptidases (12, 13) (http://merops.sanger.ac.uk/), and there are 19 representatives in the human genome. They are related to the MMP (M10) family but have different ancillary domains and are differentially expressed and regulated in model culture systems (14, 15). The common components of all ADAMTS family members are a signal peptide domain, a prodomain, a metalloproteinase domain, a disintegrin domain, a thrombospondin type 1 motif, a spacer domain, and a second thrombospondin module of a variable number of repeats at the C-terminal region (16). A phylogenetic subgroup of the ADAMTS enzymes (ADAMTS 1, 4, 5, 8, 9, and 15) possess aggrecanolytic properties (i.e., they are capable of cleaving aggrecan at the unique “aggrecanase” cleavage site in the IGD) (10, 11, 17–19). These ADAMTS are inhibited by tissue inhibitor of metalloproteinases 3 (TIMP-3) (20), which, like the ADAMTS, is bound in the ECM via interactions with sulfated GAGs.

To date, few studies have investigated the expression or activity of the aggrecanolytic ADAMTS in intervertebral discs and the role they may play in matrix degradation. Roberts et al (6) reported that aggrecanase-generated fragments increased with disease, their levels correlating with a higher degenerative grade of intervertebral discs, and Sztrolovics et al suggested a link with age (21). Hatano et al (22) showed expression of ADAMTS-4 messenger RNA (mRNA) and protein in herniated discs, while Le Maitre et al (23) demonstrated that native disc cells express ADAMTS-4 which increases with intervertebral disc degeneration. Recently, Patel et al (24) reported the presence of typical aggrecanase-generated aggrecan fragments in intervertebral discs, with ADAMTS-4 (but not ADAMTS-5) levels correlating with degeneration, particularly the p68 form of the enzyme. The aim of the present study was to investigate the mRNA and protein expression of the aggrecanolytic ADAMTS (ADAMTS 1, 4, 5, 8, 9, and 15) and their endogenous inhibitor TIMP-3 in nondegenerated human intervertebral discs and to investigate whether gene and protein expression and aggrecanase activity were altered during intervertebral disc degeneration.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Tissue samples.

Informed consent from the patients or relatives was given to obtain human intervertebral disc tissue at either postmortem examination or surgery. Local research ethics committee approval was also obtained.

Nondegenerated intervertebral disc samples were obtained postmortem within 18 hours of patient death. Previous studies have shown that intervertebral disc cells remain viable for at least 48 hours (23). The tissue consisted of full-thickness wedges of intervertebral disc of 120° arc removed anteriorly.

Samples of degenerated intervertebral disc were obtained from patients who had been diagnosed by magnetic resonance imaging as having intervertebral disc degeneration and who had undergone either anterior resection for disc replacement or spinal fusion to relieve chronic low back pain. Patients with classic sciatica were excluded.

Paraffin-embedded disc tissue.

We used archived blocks of tissue, incorporating annulus fibrosus and nucleus pulposus in continuity, fixed in 10% neutral buffered formalin and processed into paraffin wax. Since some specimens had contained bone, all samples had been radiologically decalcified in EDTA. Sections were taken for hematoxylin and eosin staining, and the degree of morphologic degeneration was scored according to previously reported criteria (25). The grading system generates a score ranging from 0 to 12: grades of 0–3 represent a histologically normal (nondegenerated) disc, grades of 4–6 indicate mild degeneration, grades of 7–9 indicate moderate degeneration, and grades of 10–12 indicate severe degeneration.

RNA extraction and reverse transcription (RT).

Disc cells were isolated from tissue as previously reported (26). RNA was extracted from 12 annulus fibrosus and 12 nucleus pulposus nondegenerated samples (3 lumbar disc levels from 4 individuals ages 37–61 years; mean age 50 years) and from 16 annulus fibrosus (from 15 individuals) and 14 nucleus pulposus (from 13 individuals) degenerated samples (ages 29–61 years; mean age 42 years) using TRIzol (Invitrogen, San Diego, CA) according to the manufacturer's instructions. RNA was treated with DNase using the Turbo DNA-free kit (Ambion, Austin, TX), and 500 ng was reverse-transcribed using Superscript II (Invitrogen) in accordance with the manufacturer's instructions.

Quantitative real-time RT–polymerase chain reaction (PCR).

Real-time RT-PCR for 18S ribosomal RNA (rRNA), ADAMTS 1, 4, 5, 8, 9, and 15, and TIMP-3 was performed on 24 nondegenerated disc samples and 30 degenerated disc samples using the genomic standard curve method of analysis to allow quantification of copy number (26, 27). Data for each target gene were normalized to 18S rRNA and presented as the number of copies per 6.25 ng RNA (equivalent to the amount of complementary DNA [cDNA] used per PCR reaction).

Primers and probes were designed using the Primer Express program (Applied Biosystems, Warrington, UK) within a single exon to allow detection of target genes in genomic DNA and cDNA samples. Total gene specificity was confirmed by BLAST searches (GenBank database sequences). Primers and probes were purchased from Applied Biosystems (Table 1).

Table 1. Real-time reverse transcription–polymerase chain reaction probes and primers*
TargetForward primer, 5′–3′Probe, 5′–3′Reverse primer, 5′–3′
  • *

    A predesigned amplification reagent was used instead of primers and probes for 18S ribosomal RNA. TIMP-3 = tissue inhibitor of metalloproteinases 3.

ADAMTS-1GGACAGGTGCAAGCTCATCTGCAAGCCAAAGGCATTGGCTACTTCTTCGTCTACAACCTTGGGCTGCAAA
ADAMTS-4ACTGGTGGTGGCAGATGACAATGGCCGCATTCCACGGTGTCACTGTTAGCAGGTAGCGCTTT
ADAMTS-5GGACCTACCACGAAAGCAGATCCCCAGGACAGACCTACGATGCCACCGCCGGGACACACGGAGTAC
ADAMTS-8GAGGTGGAGACGGGAGAGCTTGGCTCTCCTCCTCGCTGTCCTCCTAGCGCCTTCTGCCTCCT
ADAMTS-9GCATTAACTCTGCCACTGACCCTTCGCCTCCTCCTCTTCCTCCTCTACCTATAGAAACTGCTGGCCGAAGG
ADAMTS-15ATGTGCTGGCACCCAAGGTCCTGACTCCACCTCCGTCTGTGTCCACAGCCAGCCTTGATGCACTT
TIMP-3GCAGATAGACTCAAGGTGTGTGAAACCACTGCATGTCCCAACCAGACTGTGTTCCCTCACTCTTACATGCAGACA

Statistical analysis was performed using the Mann-Whitney U test to compare expression of the different target ADAMTS and TIMP-3 between cells from nondegenerated and degenerated intervertebral discs and between the different areas of the intervertebral disc.

Immunohistochemistry.

Immunohistochemistry was performed to localize ADAMTS 1, 4, 5, 9, and 15 and TIMP-3 in 36 human intervertebral discs (from a total of 18 individuals; mean age 46 years). These included 9 samples of grades 0–3, 8 samples of grades 4–6, 11 samples of grades 7–9, and 8 samples of grades 10–12. The immunohistochemistry protocol followed was reported previously (23). Antigen retrieval was performed using 0.01% (weight/volume) chymotrypsin and 0.1% (w/v) CaCl2 in Tris buffered saline for 20 minutes at 37°C. Primary antibodies diluted in 1% (w/v) bovine serum albumin were as follows: goat polyclonal antibody against ADAMTS-1 (1:100 dilution) (catalog no. SC-31080; Santa Cruz Biotechnology, Santa Cruz, CA), goat polyclonal antibody against ADAMTS-4 (1:25 dilution) (catalog no. SC-16533; Santa Cruz Biotechnology), rabbit polyclonal antibody against ADAMTS-5 (1:100 dilution) (catalog no. Ab13976; Abcam, Cambridge, UK), goat polyclonal antibody against ADAMTS-9 (1:10 dilution) (catalog no. SC-21500; Santa Cruz Biotechnology), rabbit polyclonal anti–ADAMTS-15 (1:500 dilution) (catalog no. ab28516; Abcam), or mouse monoclonal antibody against TIMP-3 (1:100 dilution) (catalog no. MAB 973; R&D Systems, Minneapolis, MN). Goat, mouse, and rabbit IgG (Dako, Carpinteria, CA) were used at equal protein concentrations as negative controls.

For analysis, each disc was morphologically separated into the 3 areas (nucleus pulposus, inner annulus fibrosus, and outer annulus fibrosus), and the number of immunopositive cells was expressed as a percentage of the total cell population for that area. Data were plotted as the mean ± SEM to represent 95% confidence intervals.

The proportions of immunopositive cells in each grade grouping were compared for statistical significance using the Mann-Whitney U test within each area of the disc. Additionally, numbers of cells immunopositive for the target protein in the different areas of the disc (i.e., nucleus pulposus versus inner annulus fibrosus, nucleus pulposus versus outer annulus fibrosus, and inner annulus fibrosus versus outer annulus fibrosus) were compared using the Wilcoxon paired sample test.

Detection of aggrecanolytic activity.

Western blotting using cleavage site–specific antibodies (28) was used to detect aggrecan core protein fragments generated by aggrecanase or MMP activity in nondegenerated and degenerated human intervertebral discs. Clinical samples from 8 subjects (4 with nondegenerated intervertebral discs and 4 with degenerated intervertebral discs) were analyzed.

Tissue samples were weighed and protein extracted by incubation in 20 volumes of 4M guanidinium hydrochloride in 50 mM acetate buffer, pH 5.0, containing proteinase inhibitors (0.1MN-6-aminohexanoic acid, 20 mM benzamidine HCl, 10 mM EDTA, 5 mMN-ethylmaleimide, 0.5 mM phenylmethylsulfonyl fluoride) at 4°C for 48 hours. Insoluble residues were removed by centrifugation at 2,000g for 10 minutes at 4°C. Samples were dialyzed against 50 mM Tris HCl, 50 mM NaCl, pH 8.0, containing proteinase inhibitors. Protein concentrations were then determined with the Micro Bicinchoninic Acid Protein Assay Kit (Pierce, Rockford, IL) using manufacturer protocols. Deglycosylation of the aggrecan core proteins was achieved enzymatically as previously described (29). The samples were then concentrated by precipitation with 10% (w/v) trichloroacetic acid and subjected to 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes.

Membranes were blocked with 10% (w/v) nonfat dried milk powder in phosphate buffered saline (PBS) for 1 hour, and primary antibody was added in 3% (w/v) nonfat dried milk powder in PBS containing 0.1% Tween for 16 hours at 4°C. The primary antibody was either mouse monoclonal antibody BC-3 (1:100 dilution) (catalog no. ab3773; Abcam), which recognizes the new N-terminus produced by aggrecanase activity in the IGD, or mouse monoclonal antibody BC-14 (1:100 dilution) (catalog no. ab3776; Abcam), which recognizes the new N-terminus produced by MMP action on the IGD (30). Membranes were then washed 3 times in PBS–Tween, followed by incubation for 1 hour with goat anti-mouse IgG (1:2,500 dilution; Dako) and washing in PBS–Tween. Cleavage products were visualized with the ECL detection kit (Amersham Biosciences, Roosendaal, The Netherlands).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Gene expression of ADAMTS and TIMP-3 in human intervertebral discs.

ADAMTS-1, -4, -5, -9, and -15 mRNA and TIMP-3 mRNA were expressed in the annulus fibrosus and nucleus pulposus of both nondegenerated and degenerated human intervertebral discs. However, no ADAMTS-8 gene expression was seen in any intervertebral disc tissues investigated, although expression was seen in the positive control (standard genomic DNA; data not shown). ADAMTS-1, -4, -5, -9, and -15 gene expression was observed in a higher number of degenerated disc samples than nondegenerated disc samples (Figure 1). TIMP-3 gene expression was seen in 100% of the nondegenerated samples, but despite 100% of the degenerated annulus fibrosus samples expressing TIMP-3, only 86% of the degenerated nucleus pulposus samples exhibited expression. For the ADAMTS, there was no difference between the nucleus pulposus and the annulus fibrosus in the number of samples exhibiting expression.

thumbnail image

Figure 1. Percentage of nondegenerated and degenerated human intervertebral disc (IVD) samples expressing mRNA for the aggrecanolytic ADAMTS and tissue inhibitor of metalloproteinases 3 (TIMP-3).

Download figure to PowerPoint

When we compared the expression levels of the target genes, a statistically significant increase in the gene expression of ADAMTS 1, 4, 5, and 15 was seen in degenerated disc samples compared with nondegenerated disc samples (P < 0.05) (Figures 2A–C and E). Gene expression for ADAMTS-9 also increased in degenerated disc samples compared with nondegenerated disc samples, although this did not reach statistical significance (P > 0.05) (Figure 2D). The increase in absolute copy number in degenerated disc samples compared with nondegenerated disc samples was not only limited to the pooled data sets. When the nucleus pulposus and annulus fibrosus regions of the intervertebral disc were analyzed separately for ADAMTS-1, -4, -5, and -15 mRNA expression, increases in the degenerated nucleus pulposus compared with the nondegenerated nucleus pulposus and in the degenerated annulus fibrosus compared with the nondegenerated annulus fibrosus were also statistically significant (P < 0.05).

thumbnail image

Figure 2. Spread plots illustrating expression levels of the aggrecanolytic ADAMTS and tissue inhibitor of metalloproteinases 3 (TIMP-3) in human nondegenerated and degenerated intervertebral discs. Horizontal bars represent the median. The absolute copy number is the number of mRNA molecules per 6.25 ng RNA. ∗ = P < 0.05.

Download figure to PowerPoint

No corresponding increase was seen in TIMP-3 gene expression in degenerated intervertebral discs compared with nondegenerated samples (Figure 2F). The overall mRNA expression level for TIMP-3 was higher (∼100-fold) in both nondegenerated and degenerated disc samples than for the ADAMTS (Figure 2). No significant differences were seen between the levels of expression in the annulus fibrosus and nucleus pulposus for any of the target genes in nondegenerated or degenerated discs except for ADAMTS-4 expression in degeneration, where mRNA expression was higher in the degenerated annulus fibrosus than in the degenerated nucleus pulposus (P < 0.05).

Coexpression of ADAMTS in intervertebral discs.

Spearman's rank correlation analysis was performed to investigate whether any of the ADAMTS or TIMP-3 were coordinately expressed. All of the aggrecanolytic ADAMTS were coordinately expressed with one another (P < 0.005), but not with TIMP-3.

Immunohistochemical localization.

Immunoreactivity for ADAMTS 1, 4, 5, 9, and 15 and TIMP-3 was observed in both nondegenerated and degenerated human intervertebral discs. Cytoplasmic staining of native disc cells was observed for ADAMTS 1, 4, 5, 9, and 15 and TIMP-3 (Figure 3), with generally more intense staining for the ADAMTS, but not for TIMP-3, being seen in degenerated as compared with nondegenerated tissue. ECM staining was seen for ADAMTS 9 and 15 and for TIMP-3. In severely degenerated tissue, staining for ADAMTS-15 in the ECM was particularly strong (Figure 3E). For the ADAMTS in general, staining was more prominent in the nucleus pulposus and inner annulus fibrosus than in the outer annulus fibrosus. For ADAMTS-4, a significantly higher proportion of immunopositive cells was observed in the nucleus pulposus than in the inner annulus fibrosus (P < 0.05). However, no difference in the number of immunopositive cells was seen between the nucleus pulposus and inner annulus fibrosus for ADAMTS 1, 5, and 9 and TIMP-3. Cells in the outer annulus fibrosus exhibited expression of all targets; however, this was at significantly lower levels than in the nucleus pulposus and inner annulus fibrosus (P < 0.05). No staining was observed in the blood vessels for any of the target molecules except for ADAMTS-1. All IgG controls were negative.

thumbnail image

Figure 3. Representative immunohistochemical staining for ADAMTS-1 (A), ADAMTS-4 (B), ADAMTS-5 (C), ADAMTS-9 (D), ADAMTS-15 (E), and tissue inhibitor of metalloproteinases 3 (TIMP-3) (F) in normal and degenerated intervertebral disc samples. Bars = 280 μm. NP = nucleus pulposus.

Download figure to PowerPoint

Changes with histologic grade of degeneration.

In the nondegenerated samples (grades 0–3), low proportions of cells in the nucleus pulposus and inner annulus fibrosus were immunopositive for ADAMTS 4 and 5 (20% and 15%, respectively), while ∼30% of disc cells exhibited immunopositivity for ADAMTS 1 and 9 and TIMP-3. Increased numbers of immunopositive cells were seen in the degenerated discs, with ∼40% of cells staining for ADAMTS-4 and 50% of cells staining for ADAMTS-5 in the nucleus pulposus of severely degenerated discs. The number of immunopositive cells in the nucleus pulposus staining for ADAMTS 4 and 5 increased significantly with the severity of degeneration (P < 0.05) (Figures 4B and C), with staining for ADAMTS-5 in the inner annulus fibrosus also increasing with severity of degeneration. For ADAMTS-1 there was no difference in immunopositivity as degeneration increased (Figure 4A), and for ADAMTS-9 there was no change in immunopositivity until severe degeneration in the nucleus pulposus (grades 10–12), where immunopositivity increased from 30% to 50% (Figure 4D). ADAMTS-15 exhibited the greatest change in percentage of immunopositive cells. In the nucleus pulposus, for instance, <10% of cells stained in nondegenerated tissue, while this increased to ∼65% in severely degenerated tissue (Figure 4E).

thumbnail image

Figure 4. Histograms illustrating the percentage of cells in the nucleus pulposus, inner annulus fibrosus, and outer annulus fibrosus staining for the ADAMTS aggrecanases and tissue inhibitor of metalloproteinases 3 (TIMP-3), classified according to grade of degeneration (see Materials and Methods). Values are the mean ± SEM. ∗ = P < 0.05.

Download figure to PowerPoint

There was no significant change in the numbers of cells staining positive for TIMP-3 in the different grades of degeneration either in the nucleus pulposus or in the inner annulus fibrosus. However, TIMP-3 decreased significantly in the severely degenerated outer annulus fibrosus samples (P < 0.05) (Figure 4F), although overall immunopositivity was low in the outer annulus fibrosus for all target genes.

Aggrecanolytic activity in intervertebral disc tissue.

Three aggrecanase-generated fragments were detected using antibody BC-3. The most intensely stained was ∼30 kd and was present in 3 of the 4 degenerated nucleus pulposus samples analyzed, but in only 1 of the 4 nondegenerated nucleus pulposus samples (Figure 5A). The remaining 2 bands were much less strongly stained, at ∼60 kd and 70 kd, and were present in both nondegenerated and degenerated samples.

thumbnail image

Figure 5. Western blots illustrating in vivo degrading enzyme activity. A, Detection of aggrecanase-generated fragments using antibody BC-3. Western blotting compares aggrecanase activity in normal (lanes 1–4) and degenerated (lanes 5–8) human intervertebral disc samples. B, Detection of matrix metalloproteinase (MMP)–generated fragments using antibody BC-14. Western blotting compares MMP activity in normal (lanes 1–4) and degenerated (lanes 5–8) human intervertebral disc samples. Positions of the molecular weight markers are indicated at left. Arrows at right indicate the position of the aggrecanase- or MMP-generated aggrecan fragments visualized.

Download figure to PowerPoint

One main MMP-generated fragment of ∼30 kd was detected with antibody BC-14 (Figure 5B). This was present in the same 3 degenerated samples and in the nondegenerated sample that stained with antibody BC-3 (Figure 5). The relatively low molecular weight of the breakdown products detected by cleavage site–specific antibodies directed at the IGD of aggrecan demonstrates that the aggrecan core protein had already been cleaved C-terminally to the IGD. Cleavage at the CRFG656–ISAV site which is undertaken by the MMPs (31) would generate a fragment of ∼30 kd upon cleavage of the core protein at the aggrecanase site (E373–A) in the IGD.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

The ECMs of articular cartilage and the nucleus pulposus of the intervertebral disc share many similarities, being rich in type II collagen and proteoglycans, and these matrices both undergo a continuous process of homeostatic turnover driven by their resident populations of chondrocytic cells. Although nucleus pulposus cells and chondrocytes show a similar phenotype (25), these cells also differ in several aspects, including expression of certain key genes such as type IIA procollagen (32) and regulation of intracellular pH (33). One key feature that distinguishes the 2 tissues is the relative proportions of proteoglycan and collagen in their matrices, with nucleus pulposus tissue having a much higher ratio of proteoglycan:collagen (∼27:1) than that in cartilage (∼2:1) (34). During diseases such as OA and intervertebral disc degeneration, an imbalance occurs between anabolic matrix production and catabolic matrix degradation, resulting in an overall loss of both proteoglycans and collagens and eventual failure of the tissues. The role of MMPs in both homeostatic turnover and pathologic breakdown of articular cartilage and intervertebral disc ECMs has been studied for a number of years and is relatively well established. However, while the importance of members of the aggrecanase enzyme family in ECM breakdown is still under debate, the high proportion of proteoglycans in nucleus pulposus tissue suggests that aggrecanases may play a fundamental role in the initiation and progression of intervertebral disc degeneration.

A range of studies have demonstrated the importance of both ADAMTS-4 and ADAMTS-5 in the breakdown of articular cartilage in arthritis diseases (7, 35–37), but at present it is not known whether this is matched in degeneration of the intervertebral disc. ADAMTS-4 expression has been demonstrated in herniated intervertebral discs (38), and increased expression of the same enzyme was seen in disc degeneration (23). The same enzyme, particularly the processed p68 form, was found to be associated with degenerated intervertebral disc tissue and with increased levels of aggrecanase-generated aggrecan fragments (24). However, the present study is the first to investigate the gene and protein expression of all of the recognized aggrecanolytic ADAMTS (1, 4, 5, 8, 9, and 15) in a large group of nondegenerated and degenerated human intervertebral disc samples.

The results demonstrate for the first time that ADAMTS-1, -4, -5, -9, and -15 and TIMP-3 mRNA and protein are present in both nondegenerated and degenerated human intervertebral discs. The fact that expression is seen in nondegenerated discs could indicate a possible role for the ADAMTS enzymes in the normal turnover of aggrecan and other matrix molecules in the healthy disc matrix.

Interestingly, ADAMTS-8 was the only “aggrecanolytic” ADAMTS not expressed in either nondegenerated or degenerated discs despite exhibiting expression in both normal and OA articular cartilage (19). However, ADAMTS-8 is thought to be only weakly aggrecanolytic compared with other ADAMTS enzymes (37), and a study by Demircan and coworkers using a chondrosarcoma cell line and human chondrocytes failed to find expression of ADAMTS-8 and demonstrated that expression was not inducible by interleukin-1β (IL-1β) or tumor necrosis factor α (38). Since these 2 cytokines are thought to be key regulators of matrix enzyme–mediated catabolism in human intervertebral disc degeneration (39, 40), these findings may help explain the lack of ADAMTS-8 expression in either the nondegenerated or degenerated samples examined in the current study. The study by Demircan et al (38) also demonstrated that IL-1β was capable of increasing expression of ADAMTS 4, 5, and 9, and in our current study there was an increase in the number of cells expressing these genes and a significant increase in the mRNA and protein levels of ADAMTS 4, 5, and 9 in degenerated nucleus pulposus compared with nondegenerated samples. Since IL-1 is known to be increased in intervertebral disc degeneration (39), this may also explain the increased levels of these aggrecanases.

When we compared nucleus pulposus and annulus fibrosus regions, no differences were observed between the levels of mRNA expression, except for ADAMTS-4 expression in degeneration, where mRNA expression was higher in the annulus fibrosus than in the nucleus pulposus. Since the annulus fibrosus samples used for RNA analysis incorporated both inner annulus fibrosus and outer annulus fibrosus and since our data showed increased protein expression in the inner annulus fibrosus, this would suggest that the majority of aggrecanase mRNA expression detected in the annulus fibrosus may be derived from the inner annulus fibrosus. At the protein level, significantly lower expression of all targets was seen in the outer annulus fibrosus compared with the nucleus pulposus and inner annulus fibrosus regions. Thus, these data suggest that the ADAMTS play a more dominant role in the central regions of the disc, which consist predominantly of proteoglycans that are substrates for the aggrecanolytic ADAMTS.

The ECM of the annulus fibrosus region of the intervertebral disc is predominantly composed of type I collagen, and the amount of proteoglycan in the annulus fibrosus, particularly in the outer annulus fibrosus region, is minimal compared with that in the nucleus pulposus. Degeneration has been shown to start within the nucleus pulposus, and as the degradation becomes more severe, the inner annulus fibrosus and outer annulus fibrosus regions become affected. In the present study, increased mRNA and protein levels of ADAMTS expression were observed in the nucleus pulposus and inner annulus fibrosus as degeneration progressed, suggesting that the ADAMTS enzymes are likely to play a prominent role in the degradation process and loss of aggrecan.

In addition to increases in ADAMTS mRNA and protein expression, the present study has demonstrated, although in a small number of samples, that there is aggrecanase and MMP activity in vivo in the disc, and indeed these findings appear to suggest an up-regulation in degeneration. While the antibody used in the present study was directed toward a neoepitope in the IGD of aggrecan, this is not the only site susceptible to cleavage by the aggrecanases (41, 42). However, this site is arguably the most damaging to aggrecan function, since cleavage in this region removes the entire CS- and KS-rich regions, completely removing all GAG side chains, which would be detrimental to the hydration of the intervertebral disc. There is mounting evidence demonstrating that in degraded cartilage, aggrecan is primarily cleaved by aggrecanases and then later by MMPs (7, 9, 43, 44). This theory is supported by the report by Mercuri et al (45) that aggrecanase-generated G1 fragments are substrates for MMPs, but the MMP-generated 342FFGVG fragments are resistant to cleavage by aggrecanases. Further investigation would be necessary to establish whether MMPs or aggrecanases play the dominant role in intervertebral disc degeneration; however, the data presented here strongly suggest an important role for aggrecanases in the breakdown of aggrecan in the nucleus pulposus during intervertebral disc degeneration.

TIMP-3 mRNA expression was substantially higher in nondegenerated nucleus pulposus and annulus fibrosus samples than was mRNA expression of any of the ADAMTS enzymes studied, although levels of expression did not increase significantly with degeneration. While TIMP-3 is known to inhibit the ADAMTS, it also inhibits many other matrix proteinases such as the MMPs and ADAMs (46, 47) by binding noncovalently to the active sites of the target enzymes in a 1:1 stoichiometry (48). It can also be sequestered into the matrix via interaction with GAGs, and the high levels of TIMP-3 in healthy tissues may help to maintain the homeostatic balance of matrix turnover by inhibiting a wide range of catabolic enzymes. However, it is not known whether the static level of TIMP-3 in the degenerated state would be able to cope with the substantial increases in the ADAMTS, combined with the known increases of MMPs reported in intervertebral disc degeneration (23, 49). An increased number of degenerated samples (compared with nondegenerated samples) exhibiting expression of ADAMTS 4, 5, and 15 at the mRNA level, and increased mRNA copy numbers of all ADAMTS target genes in degeneration combined with no corresponding increase in TIMP-3 gene expression, could indicate a role for the ADAMTS aggrecanases in degeneration of intervertebral discs. This theory is supported by the fact that (although in a limited number of samples) we also demonstrated aggrecanase and MMP activity, suggesting that both these groups of enzymes are indeed active in the intervertebral disc, particularly in degenerated tissue.

While there has been much debate over the last decade about the relative importance of MMPs and aggrecanases in the degradation of aggrecan in cartilage, several studies have highlighted the importance of a number of aggrecanases, predominantly ADAMTS-4 and ADAMTS-5, in both the initiation and progression of OA (8, 9, 35, 36). Furthermore, inhibition of ADAMTS-4 and ADAMTS-5 in OA cartilage explants prevented the breakdown of aggrecan (8). While the results presented here do not give a definitive answer as to which enzyme or group of enzymes are the most important for intervertebral disc degeneration, they demonstrate that a wide range of aggrecanases (ADAMTS 1, 4, 5, and 15) show significant increases in degenerated tissue and therefore suggest a role for the ADAMTS enzymes in intervertebral disc tissue breakdown and degradation of aggrecan. Results of Western blotting using aggrecan neoepitopes also suggest that MMPs and aggrecanases may be working simultaneously to break down aggrecan, and the combined increase in their expression levels, without a coordinate increase in TIMP-3, may be sufficient to cause an imbalance in the normal homeostatic mechanism, leading to tissue breakdown. Further studies are therefore required to establish whether any one particular ADAMTS is vital for matrix degradation, or whether matrix degradation is a combined effect of multiple enzymes.

Evidence from imaging studies suggests that loss of aggrecan from the nucleus pulposus is an early and reversible process in degeneration which precedes the irreversible breakdown of the collagen network and eventual loss of disc height and function. Identification of a key enzyme, or enzymes, could lead to the development of a therapy aimed at preventing aggrecan loss in the early stages of degeneration, which would remove the need for surgical intervention.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Prof. Hoyland had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Buttle, Freemont, Hoyland.

Acquisition of data. Pockert, Lyon, Deakin.

Analysis and interpretation of data. Pockert, Richardson, Le Maitre, Lyon, Deakin, Buttle, Freemont, Hoyland.

Manuscript preparation. Pockert, Richardson, Buttle, Freemont, Hoyland.

Statistical analysis. Pockert, Le Maitre, Hoyland.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES