1. Top of page
  2. Abstract
  6. Acknowledgements


To determine the importance of the enzymatic activity of ADAMTS-4 in normal growth and development and to evaluate the role of ADAMTS-4 in the progression of osteoarthritis (OA).


We generated catalytic domain–deleted ADAMTS-4–transgenic mice and performed extensive gross and histologic analyses of various organs. The mice were challenged by surgical induction of joint instability leading to OA, to determine the importance of the enzymatic activity of ADAMTS-4 in the progression of the disease. The response of wild-type (WT) and ADAMTS-4–knockout (ADAMTS-4–KO) articular cartilage to interleukin-1 and retinoic acid challenge in vitro was also evaluated.


ADAMTS-4–KO mice up to 1 year of age exhibited no gross or histologic abnormalities in 36 tissue sites examined. Despite evidence of ADAMTS-4 expression and activity in growth plates of WT mice, catalytic silencing of this proteinase caused no abnormalities in skeletal development, growth, or remodeling. There was no effect of ADAMTS-4 knockout on the progression or severity of OA 4 weeks or 8 weeks after surgical induction of joint instability. Enzymatic cleavage of aggrecan at the TEGE373–374ARGS site was clearly evident after exposure of articular cartilage from ADAMTS-4–KO mice to inflammatory cytokines.


Although expression of the ADAMTS-4 gene has been found in many tissues throughout the body, deletion of enzymatic activity did not appear to have any effect on normal growth and physiology. Our study provides evidence that ADAMTS-4 is the primary aggrecanase in murine growth plates; however, deletion of its enzymatic activity did not affect normal long bone remodeling. Our results also lead to the hypothesis that, in the mouse, ADAMTS-4 is not the primary enzyme responsible for aggrecan degradation at the TEGE373–374ARGS site. The elucidation of the relative importance of ADAMTS-4 in the pathologic process of human OA will require examination of human OA tissues and evidence of disease modification in patients following therapeutic intervention.

Articular cartilage is the avascular tissue that forms the articulating surface of all joints. The extracellular matrix (ECM) of this tissue imparts the biomechanical characteristics that are essential for the painless, friction-free gliding of bone against bone during articular movement. A necessary and plentiful component of cartilage ECM is aggrecan, a large proteoglycan consisting of a protein core backbone substituted with many highly sulfated glycosaminoglycans. Aggrecan provides cartilage with the ability to resist compressive forces. While many enzymes have been demonstrated to be capable of cleaving the protein backbone of aggrecan (1), analysis of the degradation products in naturally occurring osteoarthritis (OA) reveals that a major cleavage site is within the interglobular domain between amino acids 373 and 374 (2). ADAMTS-4 (aggrecanase 1) and, subsequently, ADAMTS-5 (aggrecanase 2) have been identified as the known enzymes that are most efficiently capable of cleaving aggrecan at that particular site (3, 4).

ADAMTS-4 is a member of the “disintegrin and metalloproteinase with thrombospondin-like repeat” family of proteins. Evidence for the involvement of ADAMTS-4 in joint disease comes from several reports of increased expression after stimulation of articular tissues with inflammatory cytokines (5–9), as well as in vitro findings indicating that ADAMTS-4 is one of the few enzymes that can efficiently cleave aggrecan at the site that is cleaved in naturally occurring disease (10, 11). The complete characterization of the function of ADAMTS-4 is still unclear, and there is evidence that it may be involved in pathologic or physiologic processes in nonarticular organs. Abbaszade et al (4) demonstrated expression of ADAMTS-4 in many normal human tissues, including heart, brain, lung, and skeletal muscle. Matthew and colleagues (12) showed that ADAMTS-4 is expressed in glioma cells and is capable of cleaving brevican, a brain-specific ECM protein. Sandy et al (13) reported that ADAMTS-4 could cleave versican, a widely distributed proteoglycan. Whether these expression patterns and activities implicate ADAMTS-4 in nonarticular cartilage processes remains to be determined.

The growth plate is another area that undergoes dramatic cartilage remodeling. During the physiologic process of endochondral ossification, cells within the cartilaginous growth plates proliferate and lay down significant ECM, the composition of which is similar to that of articular cartilage. This cartilaginous matrix is then enzymatically removed concurrent with subsequent vascular invasion and replacement with bone. Regulation of the ECM during growth plate remodeling has been linked to several enzymatic processes, primarily involving matrix metalloproteinases (MMPs), and animals with disruption of specific MMPs have exhibited growth plate abnormalities (14). The involvement of ADAMTS-4 in normal growth plate remodeling has not been fully investigated, but has been implicated due to identification of the aggrecanase-specific cleavage site both in growth plates and in secondary centers of ossification (15, 16).

ADAMTS-4 is, in summary, a zinc-dependent metalloproteinase that is expressed in many tissues throughout the body, but its function in these tissues is largely unknown. Aggrecanase activity in cytokine-stimulated articular cartilage and human OA articular cartilage has been identified, with ADAMTS-4 and/or ADAMTS-5 implicated as potential mediators of this important pathologic process. This same activity has been identified in normal growth plates, implying that this protein may exert physiologic as well as pathologic activity in skeletal tissues. Herein we describe the generation of a mouse line in which the genomic DNA coding the catalytic domain of ADAMTS-4 was deleted, creating a murine model with which to study the effect of lack of enzymatic activity of this protein throughout development and during growth, maturation, and aging. To further evaluate the contribution of this enzyme to the process of joint disease, joint instability was surgically induced in these mice, and the rate of progression and severity of OA was compared in wild-type (WT) versus genetically manipulated (knockout [KO]) mice. In addition, the degradative process in the cartilage of these KO mice was characterized by analysis of the aggrecan degradation products inducible in the animals.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Generation of ADAMTS-4–KO mice.

ADAMTS-4–KO mice used in these studies were generated from inbred 129SvEv-Brd mice carrying a cre/loxP-type conditional KO (CKO) allele of the ADAMTS-4 gene (Lexicon Genetics, The Woodlands, TX). The conditional allele contained loxP sites flanking exon 4 so that cre recombination resulted in deletion of exon 4, which encodes the majority of the enzyme active site (Figure 1A). Mice carrying the ADAMTS-4 CKO allele were created by homologous recombination in ES cells followed by blastocyst injection to generate chimeric mice. The ADAMTS-4–KO mouse line was produced by crossing CKO mice with protamine-cre (prot-cre)–transgenic mice (Lexicon Genetics), which resulted in cre-mediated deletion of exon 4 in the sperm of male offspring carrying both the mutant ADAMTS-4 and prot-cre alleles (Figure 1B).

thumbnail image

Figure 1. Targeted disruption of the murine ADAMTS-4 gene. A, Map of the amino acid sequence of ADAMTS-4, demonstrating the amino acids encoded by exon 4. The majority of the amino acids required for metalloproteinase activity (in shaded boxes) are within exon 4. The consensus sequence HEXXH is divided between exon 3 and exon 4 and therefore disrupted by deletion of exon 4 and insertion of a loxP site. Also encoded within exon 4 is the invariant methionine-containing “met-turn” (asterisk). B, Genomic map of the murine ADAMTS-4 gene and the predicted genomic map of the conditional knockout (CKO) and KO after cre-mediated deletion of exon 4. The locations of exons 1–5 are shown. C, Allele characterization of the ADAMTS-4 knockout, showing the map of the location of the primers and predicted basepairs of polymerase chain reaction (PCR) products generated by using primers 1, 2, and 3 to characterize the genomic DNA from wild-type (WT), CKO, and KO mice. D, PCR products generated using primers 1 and 2 depicted in C. The WT PCR product was 420 bp, the CKO product was 480 bp due to insertion of 1 loxP site, and no product from KO was expected, due to excision of the region surrounding exon 4. E, PCR products generated using primers 1 and 3 depicted in C. The KO allele was recognized as a 460-bp band. Lane M contains basepair markers.

Download figure to PowerPoint

Genotyping for ADAMTS-4 WT, CKO, and KO alleles.

Genotyping for all 3 alleles of the ADAMTS-4 gene was performed by polymerase chain reaction (PCR) using DNA template from proteinase K digests of tail biopsy specimens. The WT and conditional alleles were identified by PCR using a forward primer of the sequence 5′-GACAGGGTTTCTCTGTGTAGCG-3′ (primer 1) and a reverse primer of the sequence 5′-GCTGTCGATGATCAGATCTACC-3′ (primer 2) (Figures 1C and D). The KO allele was identified using the forward primer 5′-GACAGGGTTTCTCTGTGTAGCG-3′ (primer 1) and the reverse primer 5′-AAGGGATGCGCCACCATGCC-3′ (primer 3) (Figures 1C and E).

Preparation of RNA for reverse transcriptase–PCR (RT-PCR).

Total RNA was prepared from WT and KO mouse lung using the RNeasy kit according to the instructions of the manufacturer (Qiagen, Valencia, CA). The quality of the RNA samples was confirmed in an RNA 6000 Nano assay using a Bioanalyzer (Agilent Technologies, Palo Alto, CA). PCR primers (BioSource International, Camarillo, CA) were designed and synthesized as follows: primer 4 5′-CCTGACCACTTTGACACAGC-3′, primer 5 5′-CAAGGTGAGTGCTTCGTCTG-3′, primer 6 5′-GGCTCTTCAGGGTCCACAT-3′. RT-PCR was performed using a Qiagen OneStep RT-PCR kit with a 600-nM concentration of each primer and 200 ng of total RNA in a DNA Engine Dyad thermal cycler (MJ Research, Waltham, MA). The RT-PCR was run under the following conditions: 30 minutes at 50°C, 15 minutes at 95°C, and 35 cycles of 15 seconds at 95°C, 1 minute at 59°C, and 30 seconds at 72°C.

Histologic and pathologic assessment of young and aged mice.

Complete necropsies were performed on 9 male 14-week-old ADAMTS-4–KO mice 4 weeks postoperatively and on 7 male 18-week-old ADAMTS-4–KO mice 8 weeks postoperatively. Ten age-matched male WT 129-SvEvBrd mice were also examined at both time points, for comparison. In addition, 7 ADAMTS-4–heterozygous mice were examined at 18 weeks of age. Examinations at necropsy included macroscopic observations, determination of body and organ weights, and hematologic, serologic, and microscopic examination of brain, joints, vertebrae, spinal cord, spleen, kidneys, liver, heart, duodenum, pancreas, lung, thymus, testes, and eyes. Complete necropsies with microscopic examination of 36 tissues per mouse were subsequently performed on 6 male and 6 female ADAMTS-4–KO and 5 male and 5 female WT mice at 1 year of age. The tissues examined included brain (multiple cross-sections which included all quadrants), pancreas, salivary glands, submandibular lymph node, mesenteric lymph node, thymus, adrenal glands, pituitary gland, thyroid, trachea, heart, aorta, stomach, duodenum, jejunum, ileum, colon, cecum, liver, gallbladder, spleen, skeletal muscle, peripheral nerve, skin, mammary gland, lungs, kidneys, urinary bladder, ovaries, uterus, vagina, sternum, stifle joint, paw, thoracolumbar spine, and eyes. All tissues were fixed in 10% neutral buffered formalin, decalcified if mineralized, processed routinely, paraffin embedded, and prepared as routine hematoxylin and eosin–stained sections. All slides were evaluated by a board-certified veterinary pathologist (BS) for the presence or absence of pathologic lesions.

An additional 5 ADAMTS-4–KO and 5 WT mice (3 female and 2 male animals in each group) were killed at 1 year of age, and Faxitron images (Faxitron X-ray, Wheeling, IL) of disarticulated femurs were measured to determine femoral length.

In situ hybridization of murine articular and growth plate cartilage and embryos.

For analysis of ADAMTS-4 expression in joints, WT C57BL/6 mice were killed at 6 weeks of age and whole femoral tibial joints were dissected, fixed in 4% paraformaldehyde for 24 hours, and decalcified in 20% EDTA. The tissues were deparaffinized, rehydrated, and permeabilized. Digestion with proteinase K (5 μg/ml; Sigma, St. Louis, MO) was performed for 17 minutes at 37°C. For synthesis of in situ hybridization probes, complementary DNA (cDNA) templates of mouse ADAMTS-4 were obtained by PCR using an internally generated library. Sense and antisense probes were labeled with digoxigenin-UTP (Roche Molecular Biochemicals, Alameda, CA). Bound digoxigenin-labeled probe was detected using an alkaline phosphatase–tagged antidigoxigenin antibody (Roche Molecular Biochemicals). Slides were counterstained with hematoxylin.

Analysis of ADAMTS-4 gene expression was performed on murine embryos harvested daily from day 6 to day 17 postcoitus. Embryos were fixed in 4% paraformaldehyde–phosphate buffered saline (PBS) overnight at 4°C, washed in PBS and saline, dehydrated, embedded in paraffin, and sectioned longitudinally. For synthesis of in situ hybridization probes, 2 different areas of the murine ADAMTS-4 gene were PCR amplified using a murine cDNA library as a template and the following sets of primers: set A, forward 5′-GTGACTGCTCCAGGACTTGT-3′ and reverse 5′-TCACAGCCAGCATGGATACA-3′; set B, forward 5′-TGTATCCATGCTGGCTGTGA-3′ and reverse 5′-GCAGCCTCAGAGACACTGTC-3′. Embryo sections were deparaffinized, rehydrated, permeabilized, and hybridized with both 35S-labeled probes, and autoradiography in emulsion was carried out.

Surgical induction of OA.

All studies were performed with approval of the Wyeth Institutional Animal Care and Use Committee. Mice were anesthetized with 250 mg/kg intraperitoneal tribromoethanol (Sigma), and knees were prepared for aseptic surgery. A longitudinal incision medial to the patellar ligament was made, the joint capsule was opened, and the meniscotibial ligament, anchoring the medial meniscus to the tibial plateau, was identified. In a subset of animals, no further manipulation was performed, and this group was considered sham operated. In the experimental group, the medial meniscotibial ligament was transected, resulting in destabilization of the medial meniscus (DMM). In both sham and DMM animals, the joint capsule and subcutaneous layer were sutured closed separately and the skin was closed by application of Nexaband S/C tissue adhesive (Abbott, North Chicago, IL). Buprenorphine (Buprenex; Reckitt & Colman, Kingston-upon-Hull, UK) was administered pre- and postoperatively. Ten KO and 10 WT animals were killed by CO2 administration 4 weeks postoperatively, and 7 KO and 10 WT mice were killed 8 weeks postoperatively. In addition, 7 heterozygous mice were killed 8 weeks after DMM.

Assessment of the progression and severity of OA.

Intact knee joints were kept in 4% paraformaldehyde for 24 hours, then decalcified in EDTA/polyvinylpyrrolidone for 5 days. Joints were embedded in paraffin and 6-μm frontal sections obtained through the entire joint. Slides were stained with Safranin O–fast green and graded at 70-μm intervals through the joint, using a modification of a semiquantitative scoring system described by Chambers et al (17) in which 0 = normal cartilage; 0.5 = loss of Safranin O without structural changes; 1 = roughened articular surface and small fibrillations; 2 = fibrillation down to the layer immediately below the superficial layer and some loss of surface lamina; and 3 = mild (<20%), 5 = moderate (20–80%), and 6 = severe (>80%) loss of noncalcified cartilage. Scores of 4 (erosion to bone) were not a feature of this model. All quadrants of the joint (medial tibial plateau, medial femoral condyle, lateral tibial plateau, and lateral femoral condyle) were scored separately. A minimum of 12 levels were scored by 3 observers, under blinded conditions, for each knee joint. Scores were expressed as the maximum histologic score for each joint or the summed histologic score. The summed score represented the additive scores for each quadrant of the joint on each histologic section through the joint. This method of analysis enabled assessment of severity of lesions as well as reflecting the surface area of cartilage affected with OA lesions.

Preparation and culture of cartilage explants.

Femoral heads were harvested from 4-week-old WT and KO mice, and the cartilage was separated from the underlying subchondral bone. Cartilage samples were cultured as explants for 48 hours at 37°C in a humidified atmosphere of 5% CO2 and 95% air in Dulbecco's modified Eagle's medium (DMEM) containing 1% antibiotic/antimycotic solution (Sigma), 2 mM glutamine, 10 mM HEPES, 50 mg/ml ascorbate, and 10% fetal bovine serum. Explants were then washed 3 times and cultured for an additional 72 hours in serum-free DMEM plus 10 ng mouse interleukin-1α (IL-1α)/ml (Sigma) and 10−5M retinoic acid (Sigma). Conditioned medium was collected and cartilage harvested at the end of the culture period.

Quantitation of proteoglycan.

The proteoglycan content in the medium was measured as sulfated glycosaminoglycan by a colorimetric assay using dimethylmethylene blue and chondroitin sulfate C from shark cartilage (Sigma) as a standard, according to a previously reported procedure (18). Harvested cartilage samples were digested with proteinase K for 16 hours, centrifuged, and supernatant collected. Proteoglycan content in the digested cartilage was also measured, to determine the total proteoglycan content in the cartilage and enable calculation of the percent release of proteoglycan during the experimental protocol.

Western blot analysis.

Aggrecan fragments in conditioned medium were analyzed by Western blot analysis as previously described (19), using neoepitope monoclonal antibody AGG-C1 (0.04 μg/ml), which recognizes the aggrecanase-generated C-terminal interglobular neoepitope TEGE373.

Immunohistochemical identification of TEGE373 neoepitope in murine articular cartilage.

The polyclonal antibody anti-NITEGE373 was produced by immunizing naive rabbits with the synthetic peptide Ac-CNITEGE-COOH coupled to the carrier protein keyhole limpet hemocyanin (BioSource International). Serum that reacted with the synthetic, immunizing peptide by enzyme-linked immunosorbent assay was screened against intact aggrecan as well as ADAMTS-4–digested aggrecan by Western blotting. The antibody reacted with G1-TEGE373 generated by ADAMTS-4 digestion of bovine aggrecan but did not react with intact (undigested) aggrecan, validating it as a “neoepitope” antibody. Positive serum was purified by affinity chromatography using the immunizing peptide.

For immunostaining of TEGE373 neoepitope in murine growth plates and joints, stifle joints were harvested from 18-week-old WT or ADAMTS-4–KO mice. For immunostaining of TEGE373 neoepitope in murine articular cartilage, femoral heads from KO and WT animals were harvested after 3 days of tissue culture in the presence or absence of 10 ng IL-1α/ml and 10−5M retinoic acid. Tissues were frozen in OCT and 5-μ sections were cut. Endogenous peroxidase activity was blocked with hydrogen peroxidase (DakoCytomation, Carpinteria, CA), and the sections were deglycosylated with 0.1 units chondroitinase ABC/ml (Sigma), 0.1 unit keratanase I/ml (Seikagaku, Tokyo, Japan), and 0.1 unit keratanase II/ml (Seikagaku) for 1 hour at 37°C. Primary antibody or normal rabbit serum was added to sections for 12 hours, and secondary antibody (donkey anti-rabbit; Rockland, Gilbertsville, PA) was added for 30 minutes. Sections were incubated with ABC–peroxidase followed by diaminobenzidine substrate (Vector, Burlingame, CA). The sections were counterstained with hematoxylin. A polyclonal anti-TEGE antibody was used in these tissue immunostaining assays rather than the monoclonal AGG-C1 antibody, to eliminate background staining of endogenous mouse immunoglobulin by the secondary anti-mouse IgG antibody.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Generation of exon 4–deleted ADAMTS-4 mice.

Genotyping of KO, CKO, and WT mice was performed to ensure deletion of exon 4 (Figures 1C–E). The PCR products generated by primers 1 and 2 or primers 1 and 3 confirmed deletion of exon 4. Total RNA was isolated from lungs of WT and KO animals to characterize ADAMTS-4 messenger RNA (mRNA) in the WT mice and to determine whether or not it was being transcribed in the KO mice (Figure 2). RT-PCR using primer sets spanning exon 4 demonstrated amplified products in both the WT and the KO animals, with a reduction in size in the KO mice corresponding to the deleted exon 4, suggesting the expression of genetically modified ADAMTS-4 mRNA.

thumbnail image

Figure 2. Deletion of exon 4 in ADAMTS-4–KO mouse mRNA. A, Map of the site of primers used in reverse transcriptase–PCR (RT-PCR) reactions to identify mRNA in WT and KO animals. B, RT-PCR using primers 4 and 5, generating products of 548 bp in the WT animals and 377 bp in the KO animals (indicating deletion of exon 4). C, RT-PCR using primers 4 and 6, generating products of 290 bp in the WT animals and no PCR product in the KO (since primer 6 was designed to recognize a site within the deleted exon 4). Lane M contains basepair markers. See Figure 1 for other definitions.

Download figure to PowerPoint

Gross and histologic assessment of exon 4–deleted ADAMTS-4 mice.

Analysis of gross appearance, body weight, and serum chemistry and hematology revealed no significant differences between WT and ADAMTS-4–KO animals at either 14–18 weeks of age or 1 year of age. In addition, all tissues examined were histologically normal, and histologic analysis of a wide variety of tissues revealed no difference between the KO and WT animals.

Localization of ADAMTS-4 mRNA in normal murine joints and embryos.

In situ hybridization of whole joints from WT C57BL/6 mice, using an ADAMTS-4 cDNA probe, demonstrated the presence of ADAMTS-4 mRNA in the growth plates only (Figure 3A). No evidence of ADAMTS-4 mRNA was observed in the articular cartilage of normal animals (results not shown).

thumbnail image

Figure 3. Expression and activity of ADAMTS-4 mRNA in murine growth plates. A, In situ hybridization of normal murine proximal tibial growth plates with antisense GAPDH cDNA, sense ADAMTS-4 cDNA, and antisense ADAMTS-4 cDNA probes. B, Localization of the aggrecanase-generated TEGE373 neoepitope in growth plates of 14–18-week-old WT and ADAMTS-4–KO mice. See Figure 1 for definitions. (Original magnification × 20.)

Download figure to PowerPoint

In situ hybridization using cDNA probes to detect ADAMTS-4 mRNA expression in developing embryos from 6 days to 17 days postcoitus did not reveal any expression of ADAMTS-4 mRNA (results not shown). This lack of identification of ADAMTS-4 mRNA was consistent despite the concurrent use of 2 different radiolabeled cDNA probes to amplify any existing expression.

Identification of aggrecanase activity in growth plates and cartilage of WT and KO mice.

Immunohistochemical analysis of WT and KO mouse growth plates was performed using a polyclonal neoepitope antibody designed to bind the new C-terminus (TEGE373) in the interglobular domain of aggrecan following cleavage by ADAMTS-4 or other aggrecanases (Figure 3B). In WT mice, analysis of tissues following immunohistochemical staining demonstrated significant evidence of TEGE373, indicating aggrecanase activity in the growth plates. In extreme contrast, there was no staining by the TEGE373 neoepitope in KO mouse growth plates. Immunohistochemical analysis of WT and KO mouse joints 8 weeks post-DMM demonstrated spotty staining in areas where cartilage ECM remained. The frequency of TEGE373-positive areas was similar in OA cartilage from WT and KO mice, and the predominant sites of staining were intracellular and pericellular.

Examination of long bone length in WT and KO mice.

To determine whether lack of aggrecanase activity in the KO animals had any effect on growth plate development or remodeling, long bone length was compared in WT and KO animals by measuring the length of femur images obtained by Faxitron. This analysis revealed no difference in femoral length between WT and KO animals (data not shown). This finding, coupled with the normal histologic appearance of femoral, tibial, and sternal growth plates, confirmed the lack of physiologic consequences following deletion of ADAMTS-4 activity and apparent elimination of aggrecanase-mediated cleavage of the aggrecan interglobular domain.

Rate and severity of OA after surgical induction of joint instability.

Blinded histologic evaluation of the knee joints of untreated mice and mice that had undergone sham surgery revealed minimal OA at baseline in both the WT and the ADAMTS-4–KO animals. Four weeks after surgery to induce instability, the WT and KO mouse knees exhibited comparable, moderate progression of OA. Lesions were most apparent on the medial side of the joint on both the femur and the tibia. Eight weeks after surgery, all of the unstable knees exhibited moderate-to-severe OA. The scores were analyzed by obtaining the grade for the worst lesion in each animal and then calculating the average worst grade for each experimental group (maximum histologic score) (Figure 4A). Scores were also evaluated by adding the grades for each section through the whole joint for each animal and then calculating the average summed score for each experimental group (summed histologic score) (Figure 4B). With this method of analysis, scores reflected both the severity of the pathologic process and the surface area of joint involvement. Neither analysis revealed significant differences between experimental groups.

thumbnail image

Figure 4. Histologic scores of joints from WT mice (open bars), ADAMTS-4–KO mice (solid bars), and ADAMTS-4–heterozygous mice (shaded bars) 4 weeks and 8 weeks after surgical induction of joint instability. A, Scores expressed as the mean maximal score from each joint. B, Scores expressed as the mean of the sum of the scores from each histologic section through the joints. With both scoring methods, sham surgery did not induce changes different from those observed in nonmanipulated animals. Similarly, after surgical destabilization of the medial meniscus (DMM), there were no significant differences in the severity of osteoarthritis between WT, ADAMTS-4–KO, and ADAMTS-4–heterozygous mice at either time point. See Figure 1 for other definitions.

Download figure to PowerPoint

Characterization of aggrecan degradation in ADAMTS-4–KO mice.

Cartilage was harvested from femoral heads of untreated WT and ADAMTS-4–KO mice and cultured with or without the addition of IL-1 plus retinoic acid. Analysis of proteoglycan content of the media demonstrated comparable proteoglycan release by articular cartilage from WT and ADAMTS-4–KO mice (Figure 5A). Further analysis of this media using a monoclonal neoepitope antibody recognizing the new C-terminus of the aggrecanase-mediated cleavage site of aggrecan (TEGE373) demonstrated that the cartilage from the KO mice was responding to the inflammatory stimuli by increasing activity of an “aggrecanase” other than ADAMTS-4 (Figure 5B). To further confirm these findings, the cartilage in these cultures was evaluated by immunostaining with a polyclonal antibody developed against this same C-terminal neoepitope generated at the aggrecanase site. The results of these studies confirmed that there was a significant and equivalent increase in the TEGE373 neoepitope in articular cartilage from both the WT and the ADAMTS-4–KO mice (Figure 6).

thumbnail image

Figure 5. Proteoglycan release from articular cartilage of WT and ADAMTS-4–KO mice. A, Percent total proteoglycan release from cultured articular cartilage after 72 hours in the presence or absence of interleukin-1 (IL-1) and retinoic acid. The percent release was not significantly different between WT and ADAMTS-4–KO mouse cartilage. B, Analysis of TEGE373 neoepitope release by articular cartilage after 72 hours in culture in the presence or absence of IL-1 and retinoic acid. The presence of the neoepitope was increased by cytokine stimulation in both WT cartilage– and ADAMTS-4–KO cartilage–conditioned media. ADAMTS-4–generated G1-TEGE373 from purified bovine aggrecan (Agg) was run as a standard (asterisk). See Figure 1 for other definitions.

Download figure to PowerPoint

thumbnail image

Figure 6. Immunostaining of femoral head articular cartilage from WT (A) and ADAMTS-4–KO (B) mice after 72 hours in culture in the presence of interleukin-1 and retinoic acid. The antibody used recognized the TEGE373 neoepitope (original magnification × 40). See Figure 1 for definitions.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  6. Acknowledgements

There has been excitement following the reporting of a putative “aggrecanase” capable of cleaving articular cartilage aggrecan at the site within the core protein that is cleaved in human OA (3). Discovery of this enzyme, ADAMTS-4, was quickly followed by the identification of a second, closely related, protein also capable of cleaving aggrecan at the same site, ADAMTS-5 (4). The findings of studies undertaken to positively implicate these enzymes as present and culpable in the disease process have been more difficult to interpret, however. While levels of expression of ADAMTS-4 enzyme can be quite low in nonmanipulated, diseased tissue, several studies have shown ADAMTS-4 to be inducible by inflammatory stimuli. Yamanishi et al (9) found that synovial fibroblasts from humans with OA or rheumatoid arthritis constitutively expressed both ADAMTS-4 and ADAMTS-5, but only ADAMTS-4 expression could be increased by cytokine stimulation. Bau and colleagues (5) observed minimal levels of ADAMTS-4 mRNA in human OA chondrocytes, but demonstrated increased expression with cytokine stimulation. Malfait et al (20) presented evidence that both enzymes were expressed in human OA cartilage, but the levels of expression appeared quite low.

Questions regarding the true function of these enzymes have also been raised as a result of evidence of widespread expression patterns in human tissues. In particular, ADAMTS-4 was found in lung, skeletal muscle, heart, and brain (4). Matthew and colleagues explored the substrate specificity of the ADAMTS enzymes and found that ADAMTS-4 was capable of cleaving brevican, a brain-specific ECM protein (12), and Sandy et al (13) demonstrated that ADAMTS-4 was also capable of cleaving versican, a widely distributed ECM constituent.

We attempted to investigate the physiologic function of ADAMTS-4 by generating a mouse line in which the catalytic portion of the ADAMTS-4 gene was deleted. Although ADAMTS-4 has several domains with potential nonenzymatic functions, its primary role has thus far been hypothesized to be extracellular proteoglycan degradation. Mice in which ECM-degrading metalloproteinase genes have been deleted have been described by several authors (14, 21–28). Deletion of the MMP-2, MMP-7, MMP-11, and MMP-12 genes caused no abnormalities in unchallenged mice (21–24). It is important to note, however, that despite the lack of abnormalities in the MMP-2–KO mouse, obvious abnormalities have been found in humans with mutations in the MMP-2 gene, in several Saudi Arabian families with inherited osteolysis (29). Mice with deletion of the MMP-9 gene did exhibit some growth plate pathology (14), and both MMP-14–KO and ADAMTS-1–KO mice had significant abnormalities, implicating these metalloproteinases in several physiologic processes (25, 28).

In none of these cited studies was OA induced in the mice until a recent study by Clements et al (27), in which joint instability was created in MMP-3–KO mice as well as in mice with deletion of other genes implicated in the inflammatory cascade. The progression of OA was accelerated in the MMP-3–KO animals, and the authors speculated that perturbation of the degradative pathways caused disruption of regulatory mechanisms. It is notable that deletion of none of these metalloproteinases resulted in embryonic death.

The ADAMTS-4–KO mice in our study had no apparent developmental dysfunction, and their gross appearance at birth was normal. Our failure to detect any expression of ADAMTS-4 mRNA during murine development, despite analysis of embryos each day throughout development and the use of 2 radiolabeled probes applied concurrently, provides evidence that protein expression, if present, must be at a very low level during development. Extensive histologic evaluation of up to 36 tissues, including joints and the brain, harvested from 14–18-week-old animals and 1-year-old animals, revealed no indication of developmental or growth deficiencies. Apparently, ADAMTS-4 enzymatic activity was not necessary for normal development, growth, or function of any organ examined, although this does not take into consideration potential requirements for such activity in response to physiologic challenge. There are several examples of KO mice that have had normal gross appearance but have responded abnormally to challenge (22, 24, 30, 31).

ADAMTS-4–KO mice can still produce mRNA of the truncated form of the gene, and therefore may potentially secrete protein with no enzymatic function. We were unable to identify ADAMTS-4 protein in tissue extracts from KO or WT mice despite obvious evidence of activity. If the inactive protein were produced in these animals, findings in the KO mice would mimic the result of inhibition of ADAMTS-4 enzymatic activity by therapeutic intervention.

An attempt to identify the mRNA in normal, WT murine femorotibial joints by in situ hybridization techniques revealed that the ADAMTS-4 message was present in the proximal tibial growth plate, but there was a lack of hybridization in the articular cartilage. Aggrecanase activity was confirmed by appearance of the TEGE373 neoepitope in WT, but not KO, mouse growth plates. This implicates ADAMTS-4, and not ADAMTS-5 or any other aggrecanase, as the source of aggrecanase activity in these growth plates. It was therefore surprising that lack of this activity in the KO animals did not result in abnormal skeletal remodeling. At 1 year of age, WT and KO animals were the same size, and sensitive quantification of long bone length did not reveal any differences between WT and KO animals. One possible explanation is that ADAMTS-4–mediated aggrecan degradation during skeletal remodeling is only a minor fraction of the remodeling activity and can easily be compensated for by the activity of other proteinases such as MMPs.

Data generated by others has demonstrated that ADAMTS-4 mRNA and/or activity is inducible in articular tissues following inflammatory stimuli (5–9, 31). It is therefore possible that ADAMTS-4 activity is not required for normal ECM turnover, but is induced in inflammatory diseases such as arthritis. In order to create an inflammatory milieu and precipitate the OA process, we surgically induced joint instability in the mice. OA developed as expected, and the disease progressed over the course of 8 weeks. There was no difference in the rate or severity of disease progression in the ADAMTS-4–KO mice compared with WT mice.

Three hypotheses that could explain the apparent lack of importance of ADAMTS-4 in murine OA are 1) that aggrecanases are not active in the articular cartilage of the mouse, 2) that MMPs take over the degradation of aggrecan if aggrecanases are not active, and 3) that aggrecanases other than ADAMTS-4 are the primary mediators of aggrecan destruction in OA. To test these hypotheses, we removed articular cartilage from the femoral heads of KO and WT mice and cultured them in the presence of inflammatory cytokines. Destruction and release of proteoglycan into the media was equivalent with specimens from WT and KO mice. We then evaluated the site of cleavage of the aggrecan in the cartilage, using immunohistochemical techniques and Western blot analysis of released proteoglycan. Comparison of the TEGE373 neoepitope left within the cartilage matrix demonstrated that this aggrecanase-specific neoepitope was present in both KO and WT mice, with relatively equal staining intensity. We also identified this cleavage product released into the media.

The TEGE373 antibody was used for analysis based on evidence that the G1 fragment of aggrecan with the TEGE373 C-terminus is released from articular cartilage following IL-1–induced aggrecanase induction (32–35). It was apparent from the results that aggrecan was still being cleaved by an aggrecanase in the operated articular cartilage from ADAMTS-4–KO mice. Similar immunostaining of cartilage from the murine OA knees demonstrated TEGE373 neoepitope at disparate and spotty sites within the joint, depending on the stage of the OA process (results not shown). It is notable that the TEGE373 neoepitope appeared throughout the ECM in the cytokine-stimulated cartilage but was largely intracellular in the growth plates and OA tissue. The profound induction of aggrecanase activity by cytokines, and the choice of the optimal time for analysis, resulted in significant TEGE373 remaining within the ECM of the in vitro–cultured cartilage; the TEGE373 would subsequently disappear. The internalization of the G1 domain of aggrecan has been described (36, 37) and could explain the obvious intracellular staining during the more normal processes of growth plate remodeling and OA. It is likely that TEGE373 within the ECM had disappeared from these tissues by the time of analysis.

In summary, the results of the analysis of the ADAMTS-4–KO mice demonstrated that ADAMTS-4 enzymatic activity is not required for normal development, growth, or homeostasis. Because we do not know whether the nonenzymatic portions of this protein are translated in the KO mice, it is possible that the function of the ADAMTS-4 protein in nonarticular tissues may be through the thrombospondin-like or disintegrin domains. Although there was evidence of significant ADAMTS-4–mediated aggrecan degradation in growth plates, deletion of this gene did not affect skeletal growth or remodeling. When the joints of the ADAMTS-4–KO mice were subjected to mechanical instability to induce OA, there was no reduction in the rate or severity of aggrecan loss or disease progression. Challenge of cartilage from these mice with inflammatory cytokines also initiated aggrecan cleavage, apparently by a non–ADAMTS-4 “aggrecanase”; the site of cleavage of the aggrecan core protein was the site designated the “aggrecanase” cleavage site, TEGE373–374ARGS.

Based on the results of the current study, it is apparent that ADAMTS-4 is not the aggrecanase responsible for aggrecan degradation in murine osteoarthritis. Whether this observation extends to human OA is yet to be determined; it is possible that expression and activity of ADAMTS-4 in the mouse joint differs from that in other species. This is certainly the case with other matrix-degrading enzymes, most notably MMP-1 (38, 39). ADAMTS-5 has clearly been implicated as a potent aggrecanase (4). Others have also reported that ADAMTS-1 can cleave aggrecan at the TEGE373–374ARGS cleavage site, albeit with greatly reduced efficiency (40, 41). ADAMTS-9 has also been shown to be capable of inefficient cleavage of aggrecan at a putative “aggrecanase” site (42). Elucidation of which gene encodes the enzyme responsible for human degradative diseases of articular cartilage will undoubtedly require examination of human diseased tissue.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors would like to acknowledge the valued input and efforts of Macy Jin, Jennifer Tavares, Wen Kuang, and the Department of Bioresources staff at Wyeth Research, as well as the input of Dr. Chris Little.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Bigg HF, Rowan AD. MMPs in rheumatoid arthritis and osteoarthritis. Curr Opin Pharmacol 2001; 1: 31420.
  • 2
    Lohmander LS, Neame PJ, Sandy JD. The structure of aggrecan fragments in human synovial fluid: evidence that aggrecanase mediates cartilage degradation in inflammatory joint disease, joint injury, and osteoarthritis. Arthritis Rheum 1993; 36: 121422.
  • 3
    Tortorella MD, Burn TC, Pratta MA, Abbaszade I, Hollis JM, Liu R, et al. Purification and cloning of aggrecanase-1: a member of the ADAMTS family of proteins. Science 1999; 284: 16646.
  • 4
    Abbaszade I, Liu RQ, Yang F, Rosenfeld SA, Ross OH, Link JR, et al. Cloning and characterization of ADAMTS-11: an aggrecanase from the ADAMTS family. J Biol Chem 1999; 274: 2344350.
  • 5
    Bau B, Gebhard PM, Haag J, Knorr T, Bartnik E, Aigner T. Relative messenger RNA expression profiling of collagenases and aggrecanases in human articular chondrocytes in vivo and in vitro. Arthritis Rheum 2002; 46: 264857.
  • 6
    Curtis CL, Hughes CE, Flannery CR, Little CB, Harwood JL, Caterson B. n-3 fatty acids specifically modulate catabolic factors involved in articular cartilage degradation. J Biol Chem 2000; 275: 7214.
  • 7
    Tortorella MD, Malfait AM, Deccico C, Arner E. The role of ADAM-TS4 and ADAM-TS5 (aggrecanase-2) in a model of cartilage degradation. Osteoarthritis Cartilage 2001; 9: 53952.
  • 8
    Little CB, Hughes CE, Curtis CL, Jones SA, Caterson B, Flannery CR. Cyclosporin A inhibition of aggrecanase-mediated proteoglycan catabolism in articular cartilage. Arthritis Rheum 2002; 46: 1249.
  • 9
    Yamanishi Y, Boyle D, Clark M, Maki RA, Tortorella MD, Arner EC, et al. Expression and regulation of aggrecanase in arthritis: the role of TGF-β. J Immunol 2002; 168: 140512.
  • 10
    Tortorella M, Pratta M, Liu RQ, Austin J, Ross OH, Abbaszade I, et al. Sites of aggrecan cleavage by recombinant human aggrecanase-1 (ADAMTS-4). J Biol Chem 2000; 275: 1856673.
  • 11
    Tortorella M, Liu RQ, Burn T, Newton RC, Arner E. Characterization of human aggrecanase 2 (ADAM-TS5): substrate specificity studies and comparison with aggrecanase 1 (ADAM-TS4). Matrix Biol 2002; 21: 499511.
  • 12
    Matthew RT, Gary SC, Zerillo C, Pratta M, Solomon K, Arner EC, et al. Brain-enriched hyaluronan binding (BEHAB)/brevican cleavage in a glioma cell line is mediated by a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) family member. J Biol Chem 2000; 275: 22695703.
  • 13
    Sandy JD, Westling J, Kenagy RD, Iruela-Arispe ML, Verscharen C, Rodriguez-Mazaneque JC, et al. Versican V1 proteolysis in human aorta in vivo occurs at the glu441-ala442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem 2001; 276: 133728.
  • 14
    Vu T, Shipley JM, Berger JE, Helms JA, Hanahan D, Shapiro SD, et al. MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 1998; 93: 41122.
  • 15
    Plaas AH, Sandy JD. A cartilage explant system for studies on aggrecan structure, biosythesis and catabolism in discrete zones of the mammalian growth plate. Matrix 1993; 13: 13547.
  • 16
    Lee ER, Lamplugh L, Davoli MA, Beauchemin A, Chan K, Mort JS, et al. Enzymes active in the areas undergoing cartilage resorption during the development of the secondary ossification center in the tibiae of rats ages 0-21 days. I. Two groups of proteinases cleave the core protein of aggrecan. Dev Dyn 2001; 222: 5270.
  • 17
    Chambers MG, Cox L, Chong L, Suri N, Cover P, Bayliss MT, et al. Matrix metalloproteinases and aggrecanases cleave aggrecan in different zones of normal cartilage but colocalize in the development of osteoarthritic lesions in STR/ort mice. Arthritis Rheum 2001; 44: 145565.
  • 18
    Farndale RW, Buttle DJ, Barrett AJ. Improved quantitation and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue. Biochim Biophys Acta 1986; 883: 1737.
  • 19
    Collins-Racie LA, Flannery CR, Zeng W, Corcoran C, Annis-Freeman B, Agostino MJ, et al. ADAMTS8 exhibits aggrecanase activity and is expressed in human articular cartilage. Matrix Biol. In press.
  • 20
    Malfait AM, Liu RQ, Ijiri K, Komiya S, Tortorella MD. Inhibition of ADAM-TS4 and ADAM-TS5 prevents aggrecan degradation in osteoarthritic cartilage. J Biol Chem 2002; 277: 222018.
  • 21
    Itoh T, Ikeda T, Gomi H, Nakao S, Suzuki T, Itohara S. Unaltered secretion of β-amyloid precursor protein in gelatinase A (MMP-2) deficient mice. J Biol Chem 1997; 272: 2238992.
  • 22
    Wilson CL, Ouellette AJ, Satchell KP, Ayabe T, Lopez-Boado S, Stratman JL, et al. Regulation of intestinal α-defensin by the metalloproteinase matrilysin in innate host defense. Science 1999; 286: 1137.
  • 23
    Shipley JM, Wesselschmidt RL, Kobayashi DK, Ley TJ, Shapiro SD. Metalloelastase is required for macrophage-mediated proteolysis and matrix invasion in mice. Proc Natl Acad Sci U S A 1996; 93: 39426.
  • 24
    Masson R, Lefebvre O, Noel A, El Fahime M, Chenard MP, Wendling C, et al. In vivo evidence that the stromelysin-3 metalloproteinase contributes in a paracrine manner to epithelial cell malignancy. J Cell Biol 1998; 140: 153541.
  • 25
    Holmbeck K, Bianco P, Caterina J, Yamada S, Kromer M, Kuznetsov SA, et al. MT1-MMP-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen turnover. Cell 1999; 99: 8192.
  • 26
    Mudgett JS, Hutchinson NI, Chartrain NA, Forsyth AJ, McDonnell J, Singer II, et al. Susceptibility of stromelysin 1–deficient mice to collagen-induced arthritis and cartilage destruction. Arthritis Rheum 1998; 41: 11021.
  • 27
    Clements KM, Price JS, Chambers MG, Visco DM, Poole AR, Mason RM. Gene deletion of either interleukin-1β, interleukin-1β–converting enzyme, inducible nitric oxide synthase, or stromelysin 1 accelerates the development of knee osteoarthritis in mice after surgical transection of the medial collateral ligament and partial meniscectomy. Arthritis Rheum 2003; 48: 345263.
  • 28
    Shindo T, Kurihara H, Kuno K, Yokoyama H, Wada T, Durihara Y, et al. ADAMTS-1: a metalloproteinase-disintegrin essential for normal growth, fertility and organ morphology and function. J Clin Invest 2000; 105: 134552.
  • 29
    Martignette JA, Al Aqeel A, Al Sewairi W, Boumah CE, Kambouris M, Al Mayouf S, et al. Mutation of the matrix metalloproteinase 2 gene causes a multicentric osteolysis and arthritis syndrome. Nat Genet 2001; 28: 2615.
  • 30
    Hautamaki RD, Kaboyashi DK, Senior RM, Shapiro SD. Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice. Science 1997; 277: 20024.
  • 31
    Corry DB, Rishi K, Kanellis J, Kiss A, Song LZ, Xu J, et al. Decreased allergic lung inflammatory cell egression and increased susceptibility to asphyxiation in MMP2-deficiency. Nat Immunol 2002; 3: 34753.
  • 32
    Ratcliffe A, Tyler JA, Hardingham TE. Articular cartilage culture with interleukin-1: increased release of link protein, hyaluronic-binding region and other proteoglycan fragments. Biochem J 1986; 238: 57180.
  • 33
    Fosang AJ, Tyler JA, Hardingham TE. Effect of interleukin-1 and insulin like growth factor-1 on the release of proteoglycan components and hyaluronan from pig articular cartilage in explant culture. Matrix 1991; 11: 1724.
  • 34
    Sztrolovics R, Recklies AD, Roughley PJ, Mort JS. Hyaluronate degradation as an alternative mechanism for proteoglycan release from cartilage during interleukin-1β-stimulated catabolism. Biochem J 2002; 362: 4739.
  • 35
    Yasumoto T, Bird JL, Sugimoto K, Mason RM, Bayliss MT. The G1 domain of aggrecan released from porcine articular cartilage forms stable complexes with hyaluronan/link protein. Rheumatology (Oxford) 2003; 42: 33642.
  • 36
    Embry J, Knudson W. G1 domain of aggrecan cointernalizes with hyaluronan via a CD44-mediated mechanism in bovine articular chondrocytes. Arthritis Rheum 2003; 48: 343141.
  • 37
    Fosang AM, Last K, Stanton H, Weeks DB, Campbell IK, Hardingham TE, et al. Generation and novel distribution of matrix metalloproteinase-derived aggrecan fragments in porcine cartilage explants. J Biol Chem 2000; 275: 3302737.
  • 38
    Balbin M, Fueyo A, Knauper V, Pendas AM, Lopez JM, Jimenez MG, et al. Collagenase 2 (MMP-8) expression in murine tissue remodeling processes. J Biol Chem 1998; 273: 2395968.
  • 39
    Balbin M, Fueyo A, Knauper V, Lopez JM, Alvarez J, Sanchez LM. Identification and enzymatic characterization of two diverging murine counterparts of human interstitial collagenase (MMP-1) expressed at sites of embryo implantation. J Biol Chem 2001; 276: 1025362.
  • 40
    Kuno K, Okada Y, Kawashima H, Nakamura H, Miyasaka M, Ohno H, et al. ADAMTS-1 cleaves a cartilage proteoglycan, aggrecan. FEBS Lett 2000; 478: 2415.
  • 41
    Rodriquez-Manzaneque JC, Westling J, Thai SN, Luque A, Knauper V, Murphy G, et al. ADAMTS-1 cleaves aggrecan at multiple sites and is differentially inhibited by metalloproteinase inhibitors. Biochem Biophys Res Commun 2002; 293: 5018.
  • 42
    Somerville RP, Longpre JM, Jungers KA, Engle JM, Ross M, Evanki S, et al. Characterization of ADAMTS-9 and ADAMTS-20 as a distinct ADAMTS subfamily related to Caenorhabditis elegans GON-1. J Biol Chem 2003; 278: 950313.