Human β-defensin 3 mediates tissue remodeling processes in articular cartilage by increasing levels of metalloproteinases and reducing levels of their endogenous inhibitors

ARTHRITIS & RHEUMATISMVol. 52, No. 6, June 2005, pp 1736–1745DOI 10.1002/art.21090© 2005, American College of Rheumatology

Human �-Defensin 3 Mediates Tissue Remodeling Processes inArticular Cartilage by Increasing Levels of Metalloproteinases

and Reducing Levels of Their Endogenous Inhibitors

Deike Varoga,1 Thomas Pufe,2 Jurgen Harder,3 Jens-Michael Schroder,3 Rolf Mentlein,2

Ulf Meyer-Hoffert,3 Mary B. Goldring,4 Bernhard Tillmann,2 Joachim Hassenpflug,3

and Friedrich Paulsen5

Objective. Beta-defensins are broad-spectrum an-timicrobial peptides (APs) that are components of in-nate immunity. Recent investigations showed the induc-tion of �-defensins in synovial membranes ofosteoarthritic (OA) joints and suggested that they havefunctions other than the ability to kill microbes. As aresult of these findings, we undertook this study toinvestigate the production of human �-defensin 3(HBD-3) in OA cartilage and to determine its influenceon chondrocyte function.

Methods. Healthy and OA cartilage were assessedfor HBD-3 expression by reverse transcriptase–polymerase chain reaction (RT-PCR) and immunohis-tochemistry. HBD-3 expression in C28/I2 chondrocytesafter administration of tumor necrosis factor � (TNF�)and interleukin-1 (IL-1) was determined by real-timeRT-PCR and immunodot blot. Enzyme-linked immu-nosorbent assay experiments were used to study the

effects of HBD-3 in cultured articular chondrocytes andin healthy and OA cartilage discs. Immunohistochemi-cal analyses were performed to study the expression ofmouse �-defensins (MBDs) in OA cartilage of STR/Ortmice.

Results. HBD-3 was induced in OA cartilagewithout bacterial challenge. Cytokines involved in thepathogenesis of OA, namely, TNF� and IL-1, werestrong inducers of HBD-3 in cultured chondrocytes.Application of the recombinant HBD-3 protein to cul-tured chondrocytes and cartilage discs resulted in in-creased production of cartilage-degrading matrix met-alloproteinases and in down-regulation of theirendogenous regulators, tissue inhibitors of metallopro-teinases 1 and 2. Furthermore, STR/Ort mice, which aregenetically predisposed to develop OA-like lesions in theknee joint, demonstrated an increased expression ofMBDs 3 and 4 in cartilage compared with that inhealthy animals.

Conclusion. These findings widen our knowledgeof the functional spectrum of APs and demonstrate thatHBD-3 is a multifunctional AP with the ability to linkhost defense mechanisms and inflammation with tissue-remodeling processes in articular cartilage. Moreover,our data suggest that HBD-3 is an additional factor inthe pathogenesis of OA.

Antimicrobial peptides (APs) are effector mole-cules of the innate immune system. They act as antibi-otics by directly killing microorganisms and are primarilyexpressed in epithelial tissues, where they help to limitinfections in the first hours after microbial colonization.The human innate defense system includes several dif-ferent subfamilies of APs, such as defensins, RNase 7,cationic antimicrobial protein (CAP-37), and the cathe-

Drs. Varoga and Pufe’s work was supported in part by a grantfrom the Hensel-Stiftung of the University of Kiel. Drs. Harder andSchroder’s work was supported by the SFB 617. Dr. Goldring’s workwas supported by the NIH (grants R01-AR-45378 and R01-AG-22021).

1Deike Varoga, MD: University Hospital Schleswig-Holstein,Campus Kiel, and Christian Albrechts University of Kiel, Kiel, Ger-many; 2Thomas Pufe, PhD, Rolf Mentlein, PhD, Bernhard Tillmann,PhD: Christian Albrechts University of Kiel, Kiel, Germany; 3JurgenHarder, PhD, Jens-Michael Schroder, PhD, Ulf Meyer-Hoffert, MD,Joachim Hassenpflug, PhD: University Hospital Schleswig-Holstein,Campus Kiel, Kiel, Germany; 4Mary B. Goldring, PhD: Beth IsraelDeaconess Medical Center, New England Baptist Bone and JointInstitute, and Harvard Institutes of Medicine, Boston, Massachusetts;5Friedrich Paulsen, PhD: Christian Albrechts University of Kiel, Kiel,and University of Halle, Wittenberg, Germany.

Address correspondence and reprint requests to DeikeVaroga, MD, Department of Orthopaedic Surgery, University of Kiel,Michaelisstrasse 1, 24105 Kiel, Germany. E-mail: [email protected].

Submitted for publication August 25, 2004; accepted inrevised form March 2, 2005.

1736

licidin LL-37 (1). Expression of some APs is constitutiveand contributes to the noninflammatory antimicrobialbarrier of epithelial surfaces, whereas other APs areinduced after appropriate stimulation (2–5). Defensinsrepresent an important peptide family among APs.These small (3–5 kd), cysteine-rich, cationic peptides aredivided into �- and �-defensins based on the locationand the connectivity of 6 conserved cysteine residues.

Human �-defensins 2 and 3 (HBDs 2 and 3) havebeen isolated from human lesional psoriatic scales (3,4),and various studies demonstrated an up-regulation inresponse to stimulation by tumor necrosis factor �(TNF�) (3), interleukin-1 (IL-1) (6), or bacteria (7).Moreover, �-defensins serve as a link between innateand adaptive immune responses by acting as chemotacticfactors for immature dendritic cells and T cells (8).Recent data revealed the induction of HBD-3 in kera-tinocytes by insulin-like growth factor 1 (9). The up-regulation by growth factors suggests the existence offunctions in addition to antimicrobial functions andmight reflect a potential influence of HBD-3 in remod-eling processes after tissue degradation.

Recently, Paulsen et al (10,11) demonstrated that“inner surfaces” such as synovial membranes of articularjoints are also protected from microbial invasionthrough the endogenous production of APs. In the caseof osteoarthritis (OA), the expression pattern of APs inthe synovial membranes changes. HBD-3 and LL-37messenger RNA (mRNA), which are not expressed inhealthy synovial membranes, are up-regulated withoutbacterial challenge in OA. OA in general is character-ized by a breakdown of extracellular matrix (ECM) ofarticular cartilage in the affected joints. The pathogen-esis involves multiple etiologies including mechanical,genetic, and biochemical factors.

Various in vitro and in vivo studies indicate thatIL-1 and TNF� are involved in the initiation and pro-gression of articular cartilage destruction (12,13). Matrixmetalloproteinases (MMPs) are thought to play a cen-tral role in cartilage degradation (14). The collagenases(MMPs 1, 8, and 13) are distinguished from other MMPsby their unique ability to cleave type II collagen, themajor component of articular cartilage ECM (15).MMPs may be induced in synoviocytes or chondrocytesafter induction of gene expression by numerous cyto-kines such as IL-1�, TNF� (16,17), or vascular endothe-lial growth factor (18). MMP activity is controlled in partby tissue inhibitors of metalloproteinases (TIMPs),which form inhibitory complexes in a 1:1 stoichiometry(19). The imbalance between proteinases and inhibitorsultimately leads to an altered net proteolysis of cartilage

components. Once damaged, articular cartilage has apoor intrinsic repair capacity.

STR/Ort mice, which are genetically predisposedto develop OA-like lesions in the knee before the age of6 months (20,21), have been used to study the pathogen-esis of OA in vivo. Comparable with the situation inhumans, the development of OA in mice correlates withthe up-regulation of proinflammatory cytokines such asTNF� and IL-1 (22). In order to investigate the role of�-defensins in vivo, genes for homologous peptides inlaboratory animal models should be identified. To date,more than 10 different mouse �-defensins (MBDs) havebeen found (23). Similar to the human homologs, theyare involved in innate, antibacterial defense mecha-nisms, although only MBD-2 (24) and MBD-3 (25,26)have been verified to be inducible. Controversies existregarding the murine homolog of HBD-3, since thesimilarity of the amino acid sequence to that of thehuman counterpart is low (23).

The induction of HBD-3 in synovial membranesof OA joints suggests that it has functions other than theability to kill microbes. This encouraged us to investigateits production in OA cartilage and to determine itsinfluence on chondrocyte function.

MATERIALS AND METHODS

Tissues. OA cartilage (n � 10 samples) was obtained,with approval of the Institutional Review Board, from patients(ages 38–78 years) who underwent knee joint replacement atthe Department of Orthopaedic Surgery, University of Kiel.Tissue samples were graded according to the Mankin scale(27), and only samples with moderate-to-severe OA (Mankinscore 6–14) were included. Healthy cartilage (n � 5 samples)was obtained from patients (ages 21–52 years) who underwentresection arthroplasty because of an extraarticular tumor.

Cell culture. The human chondrocyte cell line C28/I2was used to examine the regulation of �-defensin expression invitro. These cells, immortalized with SV40 large T antigen,express proteins such as aggrecan, type II collagen, and othermarkers that are typical of the differentiated phenotype (28)and have been used to study the regulation of gene expressionand signaling in response to cytokines and other factors(29,30). For experiments, 1 � 106 cells were seeded in 25-cm2

flasks and cultivated in Dulbecco’s modified Eagle’s medium(DMEM) with 10% fetal calf serum (FCS). When they reached80% confluency, the medium was changed to serum-freeDMEM containing 0.05% bovine serum albumin. IL-1� (10ng/ml; Tebu, Offenbach, Germany) or TNF� (10 ng/ml; Tebu)was added 3 hours later, and the incubation was continued for6 hours.

In addition, OA and healthy cartilage discs (3-mmdiameter � 1-mm thickness) were cultured under standardconditions as described by Kurz et al (31). After 24 hours of

HBD-3 IN OSTEOARTHRITIC CARTILAGE 1737

incubation with different amounts of HBD-3 peptide, condi-tioned medium was withdrawn and aliquots were assayed usingcommercial enzyme-linked immunosorbent assay (ELISA)kits.

Analysis of HBD-3 mRNA in human cartilage byreverse transcriptase–polymerase chain reaction (RT-PCR).Frozen tissue samples (20 mg) of healthy and OA cartilagewere crushed in an achate mortar under liquid nitrogen. TheRNA from the cultured chondrocytes and cartilage was ex-tracted and prepared for PCR as recently described (32).

For the PCR, 4 �l of complementary DNA (cDNA)was incubated with 30.5 �l of water, 4 �l of 25 mM MgCl2, 1 �lof dNTP, 5 �l of 10� PCR buffer, and 0.5 �l (2.5 units) ofPlatinum Taq DNA polymerase (Gibco, Karlsruhe, Germany)and an intron-spanning primer pair for HBD-3 (forward 5�-AGCCTAGCAGCTATGAGGATC-3�, reverse 5�-CTTCGGC-AGCATTTTGCGCCA-3�), which yielded a 206-bp amplifiedproduct at an annealing temperature of 60°C. A GAPDH-specificintron-spanning primer pair (forward 5�-CCAGCCGAGCCA-CATCGCTC-3�, reverse 5�-ATGAGCCCCAGCCTTCTCCAT-3�), which yielded a 360-bp amplified product, served as theinternal control. Thirty-five cycles were performed with eachprimer pair. All primers were synthesized by MWG Biotech(Ebersberg, Germany). The positive control cDNA includedsamples from human tonsils. For the negative control reaction,the cDNA was replaced with water.

Analysis of HBD-3 mRNA levels in cultured chondro-cytes by real-time RT-PCR. For real-time RT-PCR, RNA wasisolated from the cultured human chondrocytes with theRNeasy-Total RNA Kit (Qiagen, Hilden, Germany) accordingto the manufacturer’s instructions. Real-time RT-PCR wascarried out using a one-step system (QuantiTect SYBR GreenRT-PCR; Qiagen). For this purpose, 100 ng of total RNA wasadded. Real-time RT-PCR was used to monitor gene expres-sion using an iCycler (Bio-Rad, Munich, Germany) accordingto the standard procedure.

PCR was performed using Hot Star Taq DNA poly-merase, which is activated by an initial heating step whileOmniscript reverse transcriptase is deactivated. The tempera-ture profile included an initial denaturation at 95°C for 15minutes, followed by 37 cycles of denaturation at 95°C for 15seconds, annealing at 60°C for 30 seconds, elongation at 72°C(the elongation time depended on the size of the fragment; thenumber of basepairs divided by 25 yielded the time in seconds),and fluorescence monitoring at 72°C. Bio-Rad iCycler DataAnalysis software was used for PCR data analysis. The speci-ficity of the amplification reaction was determined by perform-ing a melting curve analysis. Relative quantification was per-formed by normalizing the signals of the different genesagainst those of �-actin (forward primer 5�-CTCCTT-AATGTCACGCAGGATTTC-3�, reverse primer 5�-GTGG-GGCGCCCCAGGCACCA-3�). The data were assessed from3 independent experiments performed in triplicate.

Immunohistochemistry. After fixation of the humancartilage in 4% paraformaldehyde, the tissue was embedded inparaffin, sectioned, and dewaxed. Endogenous peroxidases intissue sections were blocked with 3% H2O2, and tissue sectionswere subsequently incubated with normal serum (1:5 in Trisbuffered saline) from the species in which the primary anti-body was raised. Immunohistochemical staining was per-formed on 6-�m paraffin sections, using polyclonal primary

antibody against HBD-3 (diluted 1:50; Santa Cruz Biotechnol-ogy, Santa Cruz, CA). Prior to the incubation, the slides werepretreated either by microwave heating or by trypsinization.This was followed by incubation with the biotinylated second-ary antibody and by staining with a technique using peroxidase-labeled streptavidin–biotin (Dako, Glostrup, Denmark). Aftercounterstaining with hemalum, the sections were finallymounted with Aquatex (Boehringer, Mannheim, Germany).Negative control studies were carried out by absorption of theprimary antibody by recombinant protein (1:500 dilution).

HBD-3 immunodot blot. An immunodot blot assay wasperformed for HBD-3 protein detection. Standard curves weregenerated with a recombinant HBD-3 peptide (4). The peptidewas diluted serially from 100 ng/ml to 10 �g/ml. After 24-hourstimulation of 1 � 106 C28/I2 chondrocytes with IL-1 (10ng/ml) or TNF� (10 ng/ml), conditioned medium was with-drawn, and 3 �l of each sample was spotted in triplicate onnitrocellulose membrane. The membrane was blocked with10% nonfat dry milk and Tris buffered saline–0.1% Tween 20(TBST) at room temperature for 1 hour, then incubated at 4°Covernight with anti–HBD-3 antibody (diluted 1:50). The mem-branes were then incubated at room temperature for 1 hourwith rabbit anti-goat antibody (diluted 1:100,000 in blockingsolution). After washing in TBST, the membranes were devel-oped and visualized using the ECL chemiluminescence system(Amersham Biosciences, Freiburg, Germany) followed by ap-position with autoradiographic X-Omat AR film (EastmanKodak, Rochester, NY). For quantification of HBD-3, thesignal intensities of aliquots were compared directly with thoseof simultaneously prepared standards consisting of knownamounts of HBD-3 peptide, using PC BAS (LAS-1000; Fuji-film Medical Systems, Stamford, CT).

ELISAs. For ELISAs, aliquots of supernatant of carti-lage discs and cultured chondrocytes were analyzed. C28/I2cells (1 � 106) were seeded into fresh dishes and cultivated for72 hours in DMEM/Ham’s F-12 medium (1/1 [volume/volume]) plus 10% FCS. The cultures were then changed toserum-free medium, and after 3 hours of equilibration, thecells were treated for 24 hours with recombinant HBD-3 (4) at0.5, 1, 5, and 10 �g/ml. For ELISAs, conditioned medium waswithdrawn and aliquots were assayed using the followingcommercial kits from Amersham Biosciences: RPN 2610 (foractivated MMP-1), RPN 2613 (for MMP-3), RPN 2614 (forMMP-9), RPN 2621 (for MMP-13), RPN 2611 (for TIMP-1),and RPN 2618 (for TIMP-2). Signals were detected by chemi-luminescence reaction (ECL-Plus; Amersham Biosciences).The data were assessed from 3 independent experimentsperformed in triplicate. For ELISA experiments, 3 wells wereanalyzed for each sample.

Animals. STR/Ort mice are genetically predisposed todevelop an OA-like lesion of the medial tibial cartilage. Morethan 85% of male STR/Ort mice show signs of degenerativejoint disease in the tibia by age 6 months (20,21). After themice (ages 22–45 weeks) were killed, their knee joints wereremoved and demineralized by a procedure explained in detailin an earlier report (33). To study the in vivo expression ofMBDs in OA cartilage, we immunostained 8-�m serial histo-logic sections of joints with different stages of OA (grades0–IV) with anti–MBD-3 and anti–MBD-4 antibodies (diluted1:100; Santa Cruz Biotechnology). The OA-like lesions wereclassified in accordance with a recent report (34). Sections of

1738 VAROGA ET AL

mouse skin were used as positive controls. Age- and sex-matched BALB/c mice, which do not develop OA, were usedas control animals. The animal study was approved by theInstitutional Review Board.

Statistical analysis. Data are expressed as the mean �SD of tested samples. Statistical significance was evaluated byt-test.

RESULTS

Induction of HBD-3 in OA cartilage. To investi-gate the expression pattern of HBD-3 in human OAcartilage, tissue samples were analyzed by RT-PCR andimmunohistochemistry. Examination of OA cartilageshowed an induction of HBD-3 mRNA and protein byRT-PCR (lanes 1 and 2 in Figure 1A) and immunohis-tochemistry (Figure 1C), respectively, in the majority (8of 10) of the tissue samples. HBD-3 was absent in

healthy articular cartilage by RT-PCR (lanes 3 and 4 inFigure 1A) and immunohistochemistry (Figure 1B).Immunohistochemical labeling was found in the cyto-plasm of clustered chondrocytes and in the pericellularmatrix of all layers of articular cartilage.

Induction of HBD-3 expression in cultured artic-ular chondrocytes by IL-1 and TNF�. We next investi-gated whether IL-1 and TNF�, cytokines involved in thepathogenesis of OA, influence HBD-3 gene expressionin the C28/I2 chondrocyte culture model. Real-timeRT-PCR revealed induction of HBD-3 mRNA afterstimulation with 10 ng/ml of IL-1 or TNF� at physiolog-ically relevant concentrations. Six hours after exposureto TNF�, basal HBD-3 mRNA levels increased morethan 4-fold compared with the unstimulated samples,and addition of IL-1 nearly doubled the amount of

Figure 1. Reverse transcriptase–polymerase chain reaction (RT-PCR) analysis and immunohis-tochemistry showing human �-defensin 3 (HBD-3) in osteoarthritic (OA) cartilage. A, HBD-3mRNA was detected by RT-PCR in 2 different samples of OA cartilage (lanes 1 and 2). In contrast,healthy articular cartilage did not express HBD-3 transcripts (lanes 3 and 4). GAPDH mRNA wasassayed as the internal control for equal amounts of cDNA. B, By immunohistochemistry, there wasno staining for HBD-3 protein in samples of healthy articular cartilage. C, Expression of HBD-3was confirmed in OA cartilage. Immunoreactivity (arrows) was found in the cytoplasm andpericellular matrix of clustered chondrocytes in all layers of articular cartilage. The polyclonalantibody against HBD-3 was diluted 1:50. Negative control studies were carried out by absorptionof the primary antibody by recombinant protein (1:500 dilution). Bars � 10 �m. Color figure canbe viewed in the online issue, which is available at http://www.arthritisrheum.org.

HBD-3 IN OSTEOARTHRITIC CARTILAGE 1739

HBD-3 transcripts (Figure 2A). Immunodot blot assay-ing was done to estimate the quantities of HBD-3protein in chondrocytes after challenge with IL-1 andTNF�. Stimulation of cultured chondrocytes for 24hours revealed an increase in HBD-3 production (to 3.5�g/ml) upon treatment with TNF�. IL-1 increased theamount of chondrocyte-derived HBD-3 protein to3 �g/ml (Figure 2B).

HBD-3 stimulates the expression of MMPs incultured chondrocytes and cartilage discs and reducesthe production of their endogenous regulators, TIMPs 1and 2. To understand why HBD-3 is up-regulated in OAcartilage without bacterial challenge, we investigated theinfluence of recombinant human HBD-3 on metallopro-teinases that are involved in remodeling the cartilageECM, as well as on the major endogenous regulators oftheir activity, TIMPs 1 and 2. MMP-1, MMP-3, MMP-9,MMP-13, TIMP-1, and TIMP-2 protein could all bedetected in C28/I2 chondrocyte supernatants by specific

ELISAs. Exposure to different amounts (0.5, 1, 5, and 10�g/ml) of HBD-3 protein resulted in a clear up-regulation of MMPs 1 and 13. Compared with basalexpression levels, treatment with HBD-3 for 24 hours

Figure 2. Induction of human �-defensin 3 (HBD-3) in C28/I2 chon-drocytes by proinflammatory cytokines. A, Interleukin-1 (IL-1; 10ng/ml) increased HBD-3 transcription levels nearly 2-fold comparedwith those in untreated samples after 6 hours of stimulation. Tumornecrosis factor � (TNF�; 10 ng/ml) increased HBD-3 mRNA levelsmore than 4-fold. B, For immunodot blot, 1 � 106 cells were incubatedwith TNF� or IL-1 for 24 hours. Immunodot blot revealed up-regulation of HBD-3 protein to 3,500 ng/ml supernatant in response tostimulation. Values are the mean and SD. � � P � 0.01 versus controls.

Figure 3. Human �-defensin 3 (HBD-3) stimulates production ofmatrix metalloproteinases (MMPs) and reduces levels of their endog-enous inhibitors, tissue inhibitors of metalloproteinases 1 and 2(TIMPs 1 and 2), in cultured chondrocytes. C28/I2 cells were exposedto recombinant HBD-3 (at 0.5, 1, 5, and 10 �g/ml) for 24 hours. A,Clear up-regulation of MMP-1, up to 8-fold compared with untreatedcontrol (c), was revealed by specific enzyme-linked immunosorbentassay (ELISA). MMP-3 levels were much lower under standardconditions (2.1 ng/ml), but increased to a maximum level of 2.9 ng/mlupon addition of HBD-3 protein. B, Exposure to 1 �g/ml HBD-3resulted in a 30% down-regulation of the gelatinase MMP-9, whereashigher concentrations had no additional effect. Expression of thecollagenase MMP-13 increased nearly 3-fold after 24 hours of incuba-tion with HBD-3. C, To examine the effects of HBD-3 on theendogenous regulators of MMP activity, TIMPs 1 and 2, C28/I2chondrocytes were treated with HBD-3 for 24 hours. ELISA revealeddown-regulation of both proteins, with maximum down-regulation at 5�g/ml HBD-3, which reduced the level of TIMP-2 to one-sixth of thatat baseline. Values are the mean and SD. � � P � 0.01 versus controls.

1740 VAROGA ET AL

resulted in up to an 8-fold increase of MMP-1 (Figure3A) and in a nearly 3-fold induction of MMP-13 (Figure3B). The stromelysin MMP-3 was expressed at muchlower levels (2.1 ng/ml). After addition of 5 �g HBD-3into the culture medium, specific ELISA revealed amaximum peak of 2.9 ng/ml (Figure 3A). In contrast,MMP-9 levels in the culture medium were unchangedafter exposure to 5 and 10 �g/ml HBD-3, but a 30%down-regulation was observed after treatment with 1�g/ml HBD-3 (Figure 3B). Specific ELISAs for TIMPs 1and 2 revealed down-regulation following treatmentwith HBD-3 (0.5, 1, and 5 �g/ml) (Figure 3C). In

general, TIMP-2 was suppressed to a greater degreethan was TIMP-1.

To confirm the results obtained in the C28/I2monolayer culture using a cartilage organ culture modelwith an ECM structure, discs of healthy and OA carti-lage were incubated with 0.5, 1, and 5 �g/ml HBD-3protein for 24 hours. Similar to the results in the cellculture, MMPs 1 and 13 were strongly induced in healthyarticular cartilage discs with a normal matrix composi-tion (Figures 4A and D). In contrast, incubated OAcartilage discs were much more resistant to HBD-3protein. Compared with a 300% increase in healthy

Figure 4. HBD-3 induces expression of MMPs in cartilage explants and reduces production of their regulators, TIMPs 1 and 2. To evaluate theeffects of recombinant HBD-3 on cartilage explants with an extracellular matrix structure, different amounts of this protein (0.5, 1, and 5 �g/ml) wereincubated for 24 hours. Shown are expression of MMP-1 (A), MMP-3 (B), MMP-9 (C), MMP-13 (D), TIMP-1 (E), and TIMP-2 (F). Similar to theresults in chondrocyte monolayer culture, HBD-3 stimulated the expression of MMPs 1 and 13 while reducing that of TIMPs 1 and 2. Interestingly, healthycartilage explants showed higher levels of MMP-1 and MMP-13 expression than did osteoarthritic (OA) explants following treatment with HBD-3. Levelsof MMPs 3 and 9 were nearly unchanged following treatment with HBD-3. Down-regulation of TIMPs following treatment with HBD-3 was much morepronounced in OA explants. Values are the mean and SD. � � P � 0.01 versus controls. See Figure 3 for other definitions.

HBD-3 IN OSTEOARTHRITIC CARTILAGE 1741

tissue samples, levels of MMP-1 increased only 50%after 24 hours of stimulation with different amounts ofHBD-3 in OA tissue samples (Figure 4A). MMP-13was not inducible at all in the OA cartilage discs (Figure

4D). Further, levels of MMP-3 (Figure 4B) and MMP-9(Figure 4C) were nearly unchanged in the collectedsupernatants of HBD-3–incubated cartilage discs.

Moreover, ELISAs revealed a decrease in levels

Figure 5. Induction of mouse �-defensins (MBDs) 3 and 4 in osteoarthritic (OA) cartilage of STR/Ort mice. To investigate the in vivo expressionpattern of MBDs in cartilage, knee joints of BALB/c and STR/Ort mice were examined by immunohistochemistry. In contrast to BALB/c mice (A),STR/Ort mice with low-grade OA (grades I and II) demonstrated induction of MBD-3 (arrows in B) and MBD-4 (arrows in C) in articular cartilage.Surprisingly, mice with late-stage OA (grade IV) revealed no MBD expression in articular cartilage (D). Negative control studies were carried outby absorption of the primary antibody by recombinant protein (1:500 dilution). fc � femoral cartilage; m � meniscus; tc � tibial cartilage; sb �subchondral bone. Bars � 10 �m.

1742 VAROGA ET AL

of TIMP-1 (Figure 4E) and TIMP-2 (Figure 4F) follow-ing treatment with HBD-3 in the cartilage culture.Similar to the results in the monolayer culture, TIMP-2was suppressed more strongly than TIMP-1. The ob-served effects of HBD-3 were much more intense in OAcartilage explants, suggesting an influence of the ECMcomposition on the induction process. Taken together,these results suggest that HBD-3 causes a disruption of thebalance between destructive enzymes and their inhibitors.

Increased expression of MBDs 3 and 4 in carti-lage of STR/Ort mice. To investigate the in vivo expres-sion pattern of MBDs in cartilage with OA-like pathol-ogy, the knee joints of STR/Ort mice (ages 22–45 weeks)were examined by immunohistochemistry. In contrast tothe findings in age- and sex-matched BALB/c animals(Figure 5A), immunohistochemical staining of kneejoints of STR/Ort mice revealed increased expression ofMBD-3 (Figure 5B) and MBD-4 (Figure 5C) in cartilagewith low-grade OA (grades I and II). Immunoreactivitywas found in the cytoplasm and pericellular matrix ofchondrocytes in all layers of articular cartilage. More-over, cross-sections of articular cartilage from mice withlate-stage OA (grade IV) demonstrated no immunore-activity for MBDs 3 and 4 (Figure 5D).

DISCUSSION

OA is characterized by an imbalance betweenbiosynthesis and degradation of matrix components inarticular cartilage, ultimately leading to progressive de-struction of the tissue. In the present report, we demon-strate the expression of HBD-3 in mesenchymal OAcartilage. Similar to the findings of Paulsen and cowork-ers (11) in synovial membranes, our results show thatHBD-3 is present in OA cartilage in the absence ofbacterial challenge. These results suggest another func-tion of these antibacterial molecules, and they encour-aged us to examine the role of HBD-3 in OA.

First, we tested the capacity of the proinflamma-tory cytokines IL-1 and TNF�, which have roles in theinitiation and progression of OA (35), to induce HBD-3in cultured chondrocytes. After exposure to these stim-ulators, real-time RT-PCR and immunodot blot re-vealed a clear induction of HBD-3 mRNA and proteinin cultured chondrocytes. Our results showing up-regulation of HBD-3 expression after stimulation withproinflammatory cytokines are consistent with findingson defensin expression and regulation in major epitheliasuch as skin (3,36), respiratory tract (37,38), urogenitaltract (39), and gut (40); however, all of those investiga-tions focused on the antibacterial properties of de-

fensins. Recent data revealed more than just the cata-bolic functions of TNF� and IL-1. Aside from theirdestructive abilities, they are able to stimulate growthfactors such as bone morphogenetic protein 2 (BMP-2),osteogenic protein 1 (BMP-7), and transforming growthfactor �, which results in an increased synthesis ofaggrecan and type II collagen in articular cartilage(41,42). Regardless of their ultimately deleterious ef-fects on articular cartilage, TNF� and IL-1 could initiatethe repair response displayed by injured cartilage inearly stages of OA through their ability to enhanceanabolic pathways in chondrocytes.

To elucidate the possible involvement of HBD-3in cartilage destruction or repair, we assessed the effectsof recombinant proteins in vitro. We first analyzed theinfluence of HBD-3 on the major ECM-degrading met-alloproteinases and their endogenous inhibitors(TIMPs). ECM of articular cartilage consists mainly oftype II collagen, aggrecan and other proteoglycans,minor collagens (types V, VI, IX, X, and XI), and othernoncollagenous matrix proteins. MMP-1 and MMP-13(collagenases 1 and 3, respectively) are able to cleave thetriple-helical domain of collagens, including types II andX collagen (15), and they therefore play a decisive rolein cartilage degradation. MMP-3 (stromelysin) cleavesthe telopeptide regions or noncollagen domains of typesIX and XI collagen. As a consequence, incubation ofcartilage explants with MMP-3 results in the breakdownof the collagen network, a very early feature of OA (43).MMPs are known to be induced by IL-1 and TNF� inchondrocytes (16) and inactivated by their negativeregulators, the TIMPs (15). Up-regulation of the metal-loproteinases after HBD-3 stimulation of chondrocytessuggests that �-defensins may be involved in catabolicpathways in articular cartilage monolayer culture.

To evaluate the effects of cationic HBD-3 incartilage discs, where chondrocytes were surrounded bynegatively charged proteoglycans, organ culture experi-ments were performed. These experiments demon-strated that the difference in the charges of the ECMand the effector protein HBD-3 does not neutralize theinfluence of HBD-3 on ECM-degrading pathways inarticular cartilage. Interestingly, the observed effects ofHBD-3 were much more pronounced in the presence ofa normal matrix composition. It is reasonable to proposethat the grade of cartilage degradation might influencethe susceptibility of articular cartilage to HBD-3. More-over, TIMPs, the regulators of MMP activity, wereclearly down-regulated even in the presence of a normalECM structure. Taken together, our results indicate thatHBD-3 contributes to ECM-degrading processes in ar-

HBD-3 IN OSTEOARTHRITIC CARTILAGE 1743

ticular cartilage by activating MMPs 1 and 13 andsimultaneously reducing the expression of TIMPs 1 and2. Recent studies revealed the involvement of the chemo-kine receptor CCR6 in the immunomodulating processesof HBD-3 (44), but the putative receptor on chondrocytesremains to be determined in future investigations.

The in vivo induction of MBDs in the cartilage ofSTR/Ort mice supports the present in vitro results.Comparable with the situation in humans, the develop-ment of OA-like pathology is connected with the up-regulation of proinflammatory cytokines such as TNF�and IL-1 (45). Recent investigations revealed the induc-tion of MMPs and TIMP-2 in OA cartilage of STR/Ortmice (46). These observations demonstrate the similar-ities with the induction of OA in humans and confirmthe usefulness of this model for studying OA in vivo. Toour knowledge, there are no in vivo data concerning theinduction of �-defensins in sealed-off compartmentswithout contamination due to microbial colonization.The available reports are of investigations that havefocused on the expression of APs in epithelial surfacesdue to bacterial invasion (25) or in models of woundingwith an increased risk of infection (47,48). It is interest-ing to speculate why �-defensins are up-regulated in OAcartilage without microbial threat. Supported by the invitro results, we hypothesize that �-defensins augmentcatabolic pathways in articular cartilage, ultimately lead-ing to a timely breakdown of the ECM.

Until now, the expression of �-defensins wasconnected solely with antibacterial tasks involving directkilling of microorganisms and chemotactic activity. Thisreport shows a novel function of HBD-3. Exposure ofarticular chondrocytes and cartilage to HBD-3 results inincreased synthesis of ECM-degrading metalloprotein-ases and reductions in TIMPs 1 and 2. Taken together,our findings demonstrate that HBD-3 is a multifunc-tional AP with the ability to link host defense mecha-nisms and inflammation with tissue remodeling pro-cesses in articular cartilage. The role of HBD-3 in OArequires further elucidation.

ACKNOWLEDGMENTS

We thank F. Lichte, M. Lorenzen, I. Kronenbitter, R.Worm, and S. Seiter for their excellent technical assistance,and B. Muller-Hilke (Berlin, Germany) for providing theSTR/Ort mice.

REFERENCES

1. Zasloff M. Antimicrobial peptides of multicellular organisms.Nature 2002;415:389–95.

2. Bensch KW, Raida M, Magert HJ, Schulz-Knappe P, ForssmannWG. HBD-1: a novel �-defensin from human plasma. FEBS Lett1995;368:331–5.

3. Harder J, Bartels J, Christophers E, Schroder JM. A peptideantibiotic from human skin [letter]. Nature 1997;387:861.

4. Harder J, Bartels J, Christophers E, Schroder JM. Isolation andcharacterization of human �-defensin-3, a novel human induciblepeptide antibiotic. J Biol Chem 2001;276:5707–13.

5. Garcia JR, Krause A, Schulz S, Rodriguez-Jimenez V, Kluver E,Adermann K, et al. Human �-defensin 4: a novel inducible peptidewith a specific salt-sensitive spectrum of antimicrobial activity.FASEB J 2001;15:1819–21.

6. Singh PK, Jia HP, Wiles K, Hesselberth J, Liu L, Conway BA, etal. Production of �-defensins by human airway epithelia. Proc NatlAcad Sci U S A 1998;95:14961–6.

7. Harder J, Meyer-Hoffert U, Teran LM, Schwichtenberg L, BartelsJ, Maune S, et al. Mucoid pseudomonas aeruginosa, TNF�, andIL-1�, but not IL-6, induce human �-defensin-2 in respiratoryepithelia. Am J Respir Cell Mol Biol 2000;22:714–21.

8. Yang D, Chertov O, Bykovskaia SN, Chen Q, Buffo MJ, Shogan J,et al. �-defensins: linking innate and adaptive immunity throughdendritic and T cell CCR6. Science 1999;286:525–8.

9. Sorensen OE, Cowland JB, Theilgaard-Monch K, Liu L, Ganz T,Borregaard N. Wound healing and expression of antimicrobialpeptides/polypeptides in human keratinocytes, a consequence ofcommon growth factors. J Immunol 2003;170:5583–9.

10. Paulsen F, Pufe T, Petersen W, Tillmann B. Expression of naturalpeptide antibiotics in human articular cartilage and synovialmembrane. Clin Diagn Lab Immunol 2001;8:1021–3.

11. Paulsen F, Pufe T, Conradi L, Varoga D, Tsokos M, Papendieck J,et al. Antimicrobial peptides are expressed and produced inhealthy and inflamed human synovial membranes. J Pathol 2002;198:369–77.

12. Goldring MB. The role of the chondrocyte in osteoarthritis[review]. Arthritis Rheum 2000;43:1916–26.

13. Pelletier JP, Martel-Pelletier J, Abramson SB. Osteoarthritis, aninflammatory disease: potential implication for the selection ofnew therapeutic targets [review]. Arthritis Rheum 2001;44:1237–47.

14. Martel-Pelletier J, Pelletier JP. Neutral proteases in human osteo-arthritic synovium: quantification and characterization. J Rheuma-tol 1987;14:38–40.

15. Mitchell PJ, Magna HA, Reeves LM, Lopresti-Morrow LL, Yo-cum SA, Rosner PJ, et al. Cloning, expression, and type IIcollagenolytic activity of matrix metalloproteinase-13 from humanosteoarthritic cartilage. J Clin Invest 1996;97:761–8.

16. Vincenti MP, Brinckerhoff CE. Transcriptional regulation ofcollagenase (MMP-1, MMP-13) genes in arthritis: integration ofcomplex signaling pathways for the recruitment of gene-specifictranscription factors. Arthritis Res 2002;4:157–64.

17. Goldring MB. Osteoarthritis and cartilage: the role of cytokines.Curr Rheumatol Rep 2000;2:459–65.

18. Pufe T, Harde V, Petersen W, Goldring MB, Tillmann B, MentleinR. VEGF induces matrix metalloproteinase expression in chon-drocytes. J Pathol 2004;202:367–74.

19. Baker AH, Edwards DR, Murphy G. Metalloproteinase inhibitors:biological actions and therapeutic opportunities. J Cell Sci 2002;115:3719–27.

20. Walton M. Studies of degenerative joint disease in the mouse kneejoint; scanning electron microscopy. J Pathol 1977;123:211–7.

21. Mason RM, Chambers MG, Flannelly J, Gaffen JD, Dudhia J,Bayliss JT. The STR/ort mouse and its use as a model ofosteoarthritis. Osteoarthritis Cartilage 2001;9:85–91.

22. Chambers MG, Bayliss MT, Mason RM. Chondrocyte cytokineand growth factor expression in murine osteoarthritis. Osteoarthri-tis Cartilage 1997;5:301–8.

23. Yamaguchi Y, Nagase T, Makita R, Fukuhara S, Tomita T,

1744 VAROGA ET AL

Tominaga T, et al. Identification of multiple novel epididymis-specific �-defensin isoforms in humans and mice. J Immunol2002;169:2516–23.

24. Morrison GM, Davidson DJ, Dorin JR. A novel mouse � defensin,Defb2, which is upregulated in the airways by lipopolysaccharide.FEBS Lett 1999;442:112–6.

25. Bals R, Wang X, Meegalla RL, Wattler S, Weiner DJ, Nehls MC,et al. Mouse �-defensin 3 is an inducible antimicrobial peptideexpressed in the epithelia of multiple organs. Infect Immun1999;67:3542–7.

26. Burd RS, Furrer JL, Sullivan J, Smith AL. Murine �-defensin-3 isan inducible peptide with limited tissue expression and broad-spectrum antimicrobial activity. Shock 2002;18:461–4.

27. Mankin HJ, Dorfmaan H, Lipiello L, Zarins A. Biochemical andmetabolic abnormalities in articular cartilage from osteoarthritichips: correlation of morphology with biochemical and metabolicdata. J Bone Joint Surg Am 1971;53:523–37.

28. Goldring MB, Birkhead JR, Suen LF, Yamin R, Mizuno S,Glowacki J, et al. Interleukin-1�-modulated gene expression inimmortalized human chondrocytes. J Clin Invest 1994;94:2307–16.

29. Tan L, Peng H, Osaki M, Choy BK, Auron PE, Sandell LJ, et al.Egr-1 mediates transcriptional repression of COL2A1 promoteractivity by interleukin-1�. J Biol Chem 2003;278:17688–700.

30. Grall F, Gu X, Tan L, Cho JY, Inan MS, Pettit AR, et al.Responses to the proinflammatory cytokines interleukin-1 andtumor necrosis factor � in cells derived from rheumatoid synoviumand other joint tissues involve nuclear factor �B–mediated induc-tion of the Ets transcription factor ESE-1. Arthritis Rheum2003;48:1249–60.

31. Kurz B, Jin M, Patwari P, Chen DM, Lark MW, GrodzinskyAJ. Biosynthetic response and mechanical properties of articularcartilage after injurious compression. J Orthop Res 2001;19:1140–6.

32. Varoga D, Pufe T, Harder J, Meyer-Hoffert U, Mentlein R,Schroder JM, et al. Production of endogenous antibiotics inarticular cartilage. Arthritis Rheum 2004;50:3526–34.

33. Deng GM, Nilsson IM, Verdrengh M, Tarkowski A. Intraarticu-larly localized bacterial DNA containing CpG motifs inducesarthritis. Nat Med 1999;5:702–5.

34. Kurz B, Jost B, Schunke M. Dietary vitamins and seleniumdiminish the development of mechanically induced osteoarthritisand increase the expression of antioxidative enzymes in the kneejoint of STR/1N mice. Osteoarthritis Cartilage 2002;10:119–26.

35. Goldring MB. Anticytokine therapy for osteoarthritis. ExpertOpin Biol Ther 2001;1:817–29.

36. Liu L, Roberts AA, Ganz T. By IL-1 signaling, monocyte-derivedcells dramatically enhance the epidermal antimicrobial response tolipopolysaccharide. J Immunol 2003;170:575–80.

37. Bals R, Wang X, Wu Z, Freeman T, Bafna V, Zasloff M, et al.Human �-defensin 2 is a salt-sensitive peptide antibiotic expressedin human lung. J Clin Invest 1998;102:874–80.

38. Becker MN, Diamond G, Verghese MW, Randell SH. CD-14-dependent lipopolysaccharide-induced �-defensin-2 expression inhuman tracheobronchial epithelium. J Biol Chem 2000;275:29731–6.

39. Valore EV, Park CH, Quale AJ, Wiles KR, McCray PB, Ganz T.Human �-defensin-1: an antimicrobial peptide of urogenital tis-sue. J Clin Invest 1998;101:1633–42.

40. O’Neil DA, Porter EM, Elewaut D, Anderson GM, Eckmann L,Ganz T, et al. Expression and regulation of the human �-defensinshBD-1 and hBD-2 in intestinal epithelium. J Immunol 1999;163:6718–24.

41. Merrihew C, Soeder S, Rueger DC, Kuettner KE, ChubinskayaSE. Modulation of endogenous osteogenic protein-1 (OP-1) byinterleukin-1 in adult human articular cartilage. J Bone Joint SurgAm 2003;3 Suppl:67–74.

42. Andriamanalijaona R, Felisaz N, Kim SJ, King-Jones K, LehmannM, Pujol JP, et al. Mediation of interleukin-1�–induced transform-ing growth factor �1 expression by activator protein 4 transcriptionfactor in primary cultures of bovine articular chondrocytes: possi-ble cooperation with activator protein 1. Arthritis Rheum 2003;48:1569–81.

43. Thibault M, Poole AR, Buschmann MD. Cyclic compression ofcartilage/bone explants in vitro leads to physical weakening, me-chanical breakdown of collagen and release of matrix fragments.J Orthop Res 2002;20:1265–73.

44. Wu Z, Hoover DM, Yang D, Boulege C, Santamaria F, Oppen-heim J, et al. Engineering disulfide bridges to dissect antimicrobialand chemotactic activities of human �-defensin 3. Proc Natl AcadSci U S A 2003;100:8880–5.

45. Chambers MG, Cox L, Chong L, Suri N, Cover P, Bayliss MT, etal. Matrix metalloproteinases and aggrecanases cleave aggrecan indifferent zones of normal cartilage but colocalize in the develop-ment of osteoarthritic lesions in STR/ort mice. Arthritis Rheum2001;44:1455–65.

46. Flannelly J, Chambers MG, Dudhia J, Hembry JM, Murphy G,Mason RM, et al. Metalloproteinase and tissue inhibitor ofmetalloproteinase expression in the murine STR/ort model ofosteoarthritis. Osteoarthritis Cartilage 2002;10:722–33.

47. Dorschner RA, Pestonjamasp VK, Tamakuwala S, Ohtake T,Rudisill J, Nizet V, et al. Cutaneous injury induces the release ofcathelicidin anti-microbial peptides active against group A strep-tococcus. J Invest Dermatol 2001;117:91–6.

48. Koczulla R, von Degenfeld G, Kupatt C, Krotz F, Zahler S, GloeT, et al. An angiogenic role for the human peptide antibioticLL-37/hCAP-18. J Clin Invest 2003;11:1665–72.

HBD-3 IN OSTEOARTHRITIC CARTILAGE 1745