OSTEOARTHRIT S and CARTILAGE

Transcription

1 Osteoarthritis and Cartilage (1997) 5, ~) 1997 Osteoarthritis Research Society /97/ $12.00/0 OSTEOARTHRIT S and CARTILAGE Aggrecanase and metalloproteinase-specific aggrecan neo-epitopes are induced in the articular cartilage of mice with collagen IIinduced arthritis BY IRWIN I. SINGER*, SOL SCOTT*, DOUGLAS W. KAWKA*, ELLEN K. BAYNE*, JEFFREY R. WEIDNER*, HOLLIS R. WILLIAMS*, RICHARD A. MUMFORD*, MICHAEL W. LARK*S, JOSEPH McDONNELL'~, AMY J. CHRISTENt, VERNON L. MOOREt, JOHN S. MUDGETT* AND DENISE M. VIscot *Departments of Inflammation Research and -/-Pharmacology, Merck Research Labs., Merck & Co., Inc., Rahway, New Jersey 07065, U.S.A. Summary Objective: To analyze the roles of two classes of proteinases, 'aggrecanase', and matrix metalloproteinases (MMPs), in chondrodestruction during murine collagen-induced arthritis (CIA). Methods: Generation of the 'aggrecanase' neo-epitope (NITEGE~7~), and the MMP neo-epitope (VDIPEN 34~) within aggrecan was studied by immunoperoxidase microscopy using specific anti-peptide antibodies in normal and stromelysin-1 (SLN-1) deficient knockout mice with CIA. Results: High levels of NITEGE 37~ and VDIPEN TM neo-epitopes were observed in foci within CIA paw articular cartilage exhibiting depletion of glycosaminoglycans, in advance of significant cartilage erosion. The highest concentrations of NITEGE 373 and VDIPEN TM labeling were observed and often co-distributed in the chondrocyte pericellular matrix, suggesting that stimulated chondrocytes can synthesize and/or activate both enzymes. Other regions of the cartilage frequently exhibited either NITEGE :~73 or VDIPEN "~41 labeling, but not both neo-epitopes simultaneously, suggesting that 'aggrecanase' and MMP cleavages of aggrecan may be generated independently. No detectable differences were observed in expression or distribution of either neo-epitope in SLN-1 knockout versus wild-type mice. In addition, in vitro digestion of joint sections with SLN-1 did not alter the expression of cartilage NITEGE 37~, while markedly increasing VDIPEN TM labeling. Peripheral nerves and brains of naive mice also exhibited intense anti-nitege 37~ labeling. Conclusions: These data indicate that NITEGE st:~ and VDIPEN TM aggrecan neo-epitopes are sensitive and specific markers of early joint pathology, and are consistent with the hypothesis that SLN-1 does not have 'aggrecanase' activity, and that 'aggrecanase' is distinct from the MMPs which cleave aggrecan at the MMP site, Key words: Arthritis, Aggrecan neo-epitopes, Brevican, Nerves. Introduction REGULATED aggrecan turnover is critically important in connective tissue development, as well as in maintenance of connective tissue homeostasis. In diseases such as rheumatoid arthritis (RA) and osteoarthritis (OA), there appears to be accelerated aggrecan catabolism which ultimately results in complete destruction of the articular cartilage. The enzymes respon- Received 24 December 1996; accepted 27 June Address correspondence to: Irwin I. Singer, Ph.D., Merck Research Laboratories, Merck & Co., Inc., PO Box 2000, Rahway, NJ 07065, U.S.A. SPresent address: Department of Cellular Biochemistry, SmithKline Beecham Pharmaceuticals, King of Prussia, PA 19406, U.S.A. sible for aggrecan turnover have not yet been definitively identified, but several members of the matrix metalloproteinase (MMP) family have been implicated in this process. MMPs have been localized in regions of cartilage degradation [1-3] and have been shown to be elevated in joint fluids [4, 5], serum [6, 7] and cartilage extracts from patients with OA [8]. Using a combination of N-terminal sequence analysis and immunodetection with highly specific anti-peptide antibodies, two major catabolic sites within the interglobular domain of aggrecan have been identified: Asn341-Phe 342, a MMP site [9-12], and Glu37LAla 374, an 'aggrecanase' site [12-16]. Several MMPs have been shown to cleave at the Asn341-Phe 342 site including interstitial collagenase (MMP-1), gelatinase A (MMP-2), stromelysin 407

2 408 Singer et al.: NITEGE and VDIPEN neo-epitopes in CIA (MMP-3), matrilysin (MMP-7), neutrophil collagenase (MMP-8), and gelatinase B (MMP-9) [9-11, 17, 18]. However, MMP-8 is the only enzyme shown to cleave at the Glu 37;~ Ala ~74 site [10, 18, 19]; it was recently detected in both normal and osteoarthritic chondrocytes [20-22[. Although it is currently unclear whether 'aggrecanase' is a member of the MMP family, 'aggrecanase'- mediated release of aggrecan from cartilage explants can be blocked with MMP inhibitors [23]. Both 'aggrecanase' and MMP-generated fragments are found in cartilage from patients undergoing joint replacement for OA or RA, as well as in cartilage from individuals and animals with no known history of overt joint disease [24 28]. Furthermore, 'aggrecanase'-generated fragments are the predominant aggrecan fragment in joint fluids of patients with OA, RA or traumatic joint injury [4, 14]. Together, these data indicate that both 'aggrecanase' and the classical MMPs are active within articular cartilage and play a role in aggrecan turnover during normal cartilage remodeling, as well as in the uncontrolled aggrecan degradation in OA and RA. Several animal models have been used to study inflammatory arthritis, including the murine type II collagen-induced arthritis model (CIA) [29, 30], the proteoglycan-induced mouse arthritis model [31], and the streptococcal cell wall arthritis model iri rats [32]. We have previously demonstrated that the MMP-generated aggrecan fragment terminating in VDIPEN TM is present in areas of articular cartilage degeneration in these models ([33], unpublished observations). The appearance of this neo-epitope is consistent with cleavage of aggrecan by MMPs between Ash TM and Phe 342 in these inflammatory arthritis models. Antibodies to the 'aggrecanase'-generated aggrecan fragment terminating in the neo-epitope NITEGE at~ have. also been developed. They require the C-terminal Glu ata for optimal recognition, and do not recognize the NITEGE ~73 sequence when it is part of the aggrecan core protein [12]. The objective of the investigation reported herein is to determine if the 'aggrecanase'-generated neo-epitope NITEGE 37:3 is induced in articular cartilage in CIA, and to compare its distribution with that of the MMP-generated VDIPEN 34~ neo-epitope. These studies focused on mouse CIA because of the many features of this model in common with human inflammatory arthritis, and because it has enabled us to evaluate the distribution of these neo-epitopes in animals where one of the major MMPs, stromelysind, has been knocked-out (SLN-1 ko mice) [34, 35]. Methods ANIMALS CIA was induced in 8-10-week-old B10.RIII mice by immunization with a single intradermal injection of highly purified porcine type II collagen in complete Freund's adjuvant, as previously described [30], followed by subcutaneous administration of 327 ng human ril-l~3 on days 13, 14, and 15. SLN-1 knockout (ko) and wild-type (wt) mice [34] were back-crossed twice to B10.RIII mice to generate N2 B10.RIII SLN-1 ko and N2 B10.RIII wt back-cross mice. Animals were genotyped by Southern analysis to determine which animals carried the targeted SLN-1 allele as described [35]. N2 B10.RIII SLN-1 ko mice and N2 B10.RIII wt mice were immunized identically, as described above. Non-immunized mice of each genotype served as naive controls. Animals were euthanized 21 days later, when many exhibited stage 2 clinical symptoms (swelling of >2 digits, and/or the carpus and tarsus). Paws were removed proximal to the tarsus or carpus, embedded in OCT medium (Miles, Naperville, IL, U.S.A.), frozen in liquid nitrogen, and stored at -70~ Animals were maintained in American Association of Laboratory Animal Care-accredited facilities, and the exper-" iments were approved by the Merck Research Laboratory Institutional Animal Care and Use Committee in accordance with NIH and USDA guidelines. IMMUNOPEROXIDASE MICROSCOPY Sagittal 5 ~m cryosections were cut with carbide knives on a model OTF Bright Cryotome (Hacker Instruments, Inc. Fairfield NJ, U.S.A.) equipped with a cryostat frozen sectioning aid (Instrumedics, Inc., Teaneck, NJ, U.S.A.). Frozen sections were mounted on glass slides, treated with protease-free chondroitinase ABC (0.02 U, 30 mm Na acetate, 0.1 M Tris-HC1, ph 8.0; Seigagaku America, Rockville, MD, U.S.A.) as previously described [33], and mildly fixed for 20 min with Nakane solution (periodate/lysine/paraformaldehyde) [36]. NITEGE 373 neo-epitopes were labeled with rabbit anti-nitege 3~3 IgG [12] that was purified on a YPLPRNITEGE-peptide affinity resin. VDIPEN TM neo-epitopes were also labeled on adjacent cryosections with rabbit anti- VDIPEN TM IgG affinity-purified as previously described [33]. Bound antibodies were detected on the frozen sections via immunoperoxidase microscopy using the ABC technique (Elite kit; Vector Labs., Inc., Burlingame, CA, U.S.A.). Peroxidase reaction product was developed with

3 Osteoarthritis and Cartilage Vol. 5 No a glucose oxidase/dab/nickel method [37] with 1% orange G used as a counterstain. To control for specificity, the primary anti-nitege :~7:~ IgG (2 ~tg/ml) was incubated with 50 pg/ml of either the neo-epitope peptide (H2N-YPLPRNITEGE-COOH) or a peptide spanning the 'aggrecanase' cleavage site (H2N-YPLPRNITEGEARGS-COOH) prior to immunoperoxidase labeling. Similar controls were also conducted for the VDIPEN TM antibody [33]. To investigate whether the anti-nitege IgG is capable of interacting with a possibly related epitope present at an 'aggrecanase-like' cleavage site in brevican [38-40], similar competition experiments were conducted with the following peptides: H2N-LKLPQEAVESE-COOH, and H2N- LKLPQEAVESESRGA-COOH. Peptides were synthesized on an Applied Biosystems 430A peptide synthesizer and purified by reversed-phase high performance liquid chromatography (HPLC) on a Waters C18 Deltapak column. All peptides were >95% pure by reversed-phase HPLC, and the structure of each was confirmed by electrospray ionization mass spectroscopy. Pre-immune rabbit IgG was used as an additional negative control. To determine the effect of SLN-1 on expression of the NITEGE 373 neo-epitope, unfixed cryosections were digested with activated recombinant human SLN-1 (rhsln-1) (100~tg~/ml in 150mm NaC1, 25mM Tris-HC1, t0 mm CaC12, 0.05% Brij-35, ph 7.4) for 30min at 37~ prior to staining with either anti-nitege 373 'IgG or anti-vdipen :34~ IgG. The cartilage glycosaminoglycan (GAG) content of adjacent sections was also determined by toluidine blue staining as previously described [41]. Photomicrographs were made at with a Leitz Vario Orthomat microscope. Results DETECTION OF ANTI-NITEGE ~v3 LABELING IN CONTROL TISSUE High concentrations of anti-nitege 37:~ labeling were localized in peripheral nerves [Fig. l(a),(c)], and at the insertion sites of tendons [Fig. l(b)] in either naive [Fig. l(a),(b)], or IL-l[3-treated [Fig, 1(c)] control B10.RIII mice. Pre-incubation of the anti-nitege 373 IgG (2 pg/ml) with 50 ~g/ml YPLPRNITEGE 37:~, a peptide corresponding to the C-terminus of the 'aggrecanase'-generated aggrecan G1 fragment completely abolished this staining [Fig. l(d)]. In comparison, the control peptide a 84 b d. e~, "~ - Fro. 1. The 'aggrecanase' cleavage site (NITEGE ~73) is present in control B10.RIII mice. (a) Anti-NITEGE 373 is localized in the endoneurium of peripheral nerves exhibiting the characteristic snake-fence distribution of nerve fibers (arrows) in the basal reticular dermis and among muscle fibers, and (b) in the proximal Achilles' tendon (arrowheads) of non-boosted naive control mice. (c) Binding of NITEGE :~7:~ IgG to peripheral nerves occurs in the presence of the spanning peptide YPLPRNITEGEARGS, but is inhibited by the C-terminal peptide YPLPRNITEGE (d) in adjacent cryosections of IL-l[~-boosted but unimmunized control mice. Bar for (a)-(d)= 100 ~m.

4 410 S i n g e r et al.: N I T E G E a n d V D I P E N n e o - e p i t o p e s in CIA % P FIG. 2. Small NITEGE:373-positive foci are localized in the peripheral AC of naive control B10.RIII mice. S e q u e n t i a l cryosections of a tarsus stained with hematoxylin and eosin (a), toluidine blue (b), anti-nitege 373 IgG (c), and anti-vdipen TM IgG (d). (a) Central (arrowhead) and peripheral (arrow) regions of the AC are intact (S = synovium). (b) Central AC zone (arrowhead) is rich in GAGs, while the peripheral region (arrow) is GAG-depleted. (c) Peripheral AC exhibits focal NITEGE.7:~ staining (arrow), coincident with the GAG-depleted region shown in (b), whereas the central region (arrowhead) is unlabeled. (d) No VDIPEN TM staining is detectable throughout the AC of an adjacent section. Bar for (a)-(d) = 100 ~LM. s p a n n i n g this c l e a v a g e site ( Y P L P R N I T E G E A R GS; 50 gg/ml) did n o t i n h i b i t l a b e l i n g [Fig. 1(c)]. A n t i - V D I P E N TM s t a i n i n g was c o n s i s t e n t l y a b s e n t from t h e s e tissue sites ( d a t a n o t shown). T h e s e r e s u l t s c o n f i r m t h a t o u r a n t i - N I T E G E 37~ IgG specifically r e c o g n i z e s the N I T E G E 373 s e q u e n c e w h e n it is p r e s e n t e d as a C - t e r m i n a l n e o - e p i t o p e, b u t n o t w h e n it is a n i n t e r n a l c o m p o n e n t of t h e p e p t i d e sequence. I n a d d i t i o n, foci of N I T E G E a73 n e o e p i t o p e l a b e l i n g w e r e o b s e r v e d at t h e a r t i c u l a r s u r f a c e s of some p a w j o i n t s in b o t h n a i v e a n d IL-l~3-boosted n o n - i m m u n i z e d c o n t r o l B10.RIII mice. T h e s e j o i n t s s h o w e d no h i s t o l o g i c a l signs of d a m a g e or i n f l a m m a t i o n [Fig. 2(a)], b u t slight G A G d e p l e t i o n of the p e r i p h e r a l a r t i c u l a r c a r t i l a g e (AC) was o c c a s i o n a l l y d e t e c t a b l e by t o l u i d i n e b l u e s t a i n i n g [Fig. 2(b)]. S m a l l foci of a n t i - N I T E G E a73 l a b e l i n g were l o c a l i z e d in t h e p e r i c e l l u l a r a n d i n t e r t e r r i t o r i a l A C m a t r i x [Fig. 2(c)], c o i n c i d e n t w i t h this G A G d e p l e t i o n [ c o r r e s p o n d i n g a r r o w s in Figs 2(b) a n d (c)]. S l i g h t s t a i n i n g w a s also o b s e r v e d s u r r o u n d i n g c h o n d r o c y t e s of t h e adjac e n t calcified c a r t i l a g e [Fig. 2(c)]. T h e r e w a s no o b s e r v a b l e a n t i - V D I P E N TM s t a i n i n g at t h e s e NITEGEaT<positive foci [Fig. 2(d)]. N e i t h e r neoe p i t o p e w a s d e t e c t e d in t h e s y n o v i u m (Fig. 2). N o d e t e c t a b l e d i f f e r e n c e s w e r e o b s e r v e d b e t w e e n the j o i n t s of IL-113-boosted a n d n o n - b o o s t e d n a i v e c o n t r o l B 1 0. R I I I a n i m a l s. S i m i l a r r e s u l t s w e r e also o b t a i n e d w i t h n o n - i m m u n i z e d N2 B10.RIII SLN-1 Table I Capacity of 'aggrecanase' and 'aggrecanase-like' peptides to inhibit anti-nitege immunolabeling of peripheral nerves and chondrocytes in the naive mouse hind paw* Peptides Peripheral nerves Peripheral chondrocytes LKLPQEAVESE LKLPQEAVESESRGA 0 0 YPLPRNITEGE YPLPRNITEGEARGS 0 0 *Hind paws of untreated 8-week-old BI0.RIlI control mice were cryosectioned mid-sagittally, and the N I T E G E 373 neo-epitope was labeled in the presence of the indicated peptides by immunoperoxidase microscopy as described in Methods. Key: = nearly complete inhibition of anti-nitege 373staining; = total inhibition of anti-nitege 3v~staining; 0 = no inhibition of anti-nitege 3v3staining.

5 Osteoarthritis and Cartilage Vol. 5 No ko and N2 B10.RIII wt control mice (data not shown). NITEGE 37:~ IgG LABELS AN 'AGGRECANASE-LIKE' NEO-EPITOPE IN BREVICAN The intense anti-nitege a73 IgG labeling of peripheral nerves suggested that this antibody might also recognize a related epitope in nerve tissue proteoglycans. Additional competition experiments were therefore conducted to determine if the NITEGE 37~ antibody interacts with the C-terminal ESE epitope present at an 'aggrecanase-like' cleavage site in brevican [38-40]. Peptides containing either the native ESE C- terminus or the spanning sequence of the brevican 'aggrecanase-like' site were synthesized. The results of these experiments are presented in Table I. The brevican 'aggrecanase-like' C- terminal cleavage peptide LKLPQEAVESE inhibited the anti-nitege a7:3 labeling of chondrocytes equally, as well as YPLPRNITEGE, the aggrecan blocking peptide. LKLPQEAVESE also inhibited the anti-nitege '~7:~ staining of peripheral nerves nearly as well as YPLPRNITEGE. Alternatively, the control brevican spanning peptide LKLPQEAVESESRGA did not inhibit labeling of either chondrocytes or nerves. There was also intense anti-nitege at:~ IgG staining in naive control mouse brain cryosections, where labeling was particularly concentrated in the corpus collosum, and in most other myelinated nerve tracts (data not shown). INDUCTION OF NITEGE at:~ AND VDIPEN NEO-EPITOPES IN CIA High concentrations of both the NITEGE ata and VDIPEN TM neo-epitopes were induced and localized in large foci within the AC of tarsal and metatarsal joints of B10.RIII mouse paws with stage 2 CIA (Fig. 3). The NITEGE at* neo-epitope was particularly concentrated in the pericellular matrix of AC chondrocytes [Figs 3(a),(e),(g)]. The VDIPEN TM neo-epitope was also localized in the pericellular matrix of AC chondrocytes where it often appeared to co-distribute with NITEGE ata [corresponding arrows in Figs 3(a) and (b); (e) and (f); (g) and (h)]. Both neo-epitopes were induced in AC loci exhibiting GAG depletion (indicated by reduced toluidine blue staining; data not shown), in the absence of erosion of the articular surface. In addition to this codistribution, the 'aggrecanase' and MMP neo-epitopes also exhibited complementary staining patterns. The MMP neo-epitope was heavily concentrated at the TM articular surface and in the interterritorial matrix of the AC [Figs 3(b),(f),(h)]. NITEGE at~ labeling was usually not detectable at the AC surface, but tended to be localized around chondrocytes in the deeper zones of the AC adjacent to the calcifled cartilage [Figs 3(a),(e),(g)]. Some deeper chondrocytes were associated with both NITEGE 373 and VDIPEN TM staining, while others were NITEGEaT:~-positive and VDIPENa4~-negative [Figs 3(e)-(h)]. EXPRESSION OF THE NITEGE 373 EPITOPE IS UNAFFECTED BY SLN ACTIVITY To investigate the possible effects of SLN-1 activity on induction of the NITEGE 373 epitope, cryosections of B10.RIII mouse paws with clinical stage 2 CIA were exhaustively digested with rhsln-1 before to labeling the anti-nitege 373 or VDIPEN TM antibodies. Pre-digestion with exogenous SLN-1 did not alter the distribution of endogenous NITEGE 373 staining, and did not decrease or increase the intensity of NITEGE 373 labeling [Figs 4(a) and (b)]. In contrast, preincubation of adjacent sections with SLN-1 produced a massive increase in VDIPEN TM labeling in the AC [Figs 4(c) and (d)]. These data suggest that exogenous SLN can digest aggrecan within the cartilage matrix, but the 'aggrecanase'- generated G1 fragment appears resistant to this degradation. APPEARANCE OF THE 'AGGRECANASE' AND MMP CLEAVAGE SITES IN SLN-1 KO MICE with CIA High local concentrations of both the NITEGE at~ and VDIPEN TM neo-epitopes were also codistributed in the pericellutar matrix of AC chondrocytes in the paws of N2 B10.RIII SLN-1 ko mice with stage 2 CIA (data not shown). In addition, these neo-epitopes often exhibited non-coincident staining patterns at several AC foci in these N2 B10.RIII SLN-1 ko mice (Fig. 5), and in control B10.RIII animals exhibiting stage 2 CIA (data not shown). While both neo-epitopes were present in some metatarsal joints, other similar joints stained for either one or the other neo-epitope [Figs 5(a) and (b), (d) and (e)], in joints where the AC exhibited GAG depletion [Fig. 5(c) and (f)]. The relative abundance of these coincident and non-coincident patterns of NITEGE av~ and VDIPEN TM neo-epitope distribution in N2 BI0.RIII SLN-I ko mice, N2 BI0.RIII wt mice, and BI0.RI]I mice with stage 2 CIA is shown in Table If. All N2 Bt0.RIII SLN-1 ko, B10.RIII, and N2 B10.RIII wt mice with CIA exhibited either one or the other

6 412 S i n g e r e t a l. : N I T E G E a n d V D I P E N n e o - e p i t o p e s in C I A FIG. 3. Both the ' a g g r e c a n a s e ' and M M P neo-epitopes may be co-distributed in the AC of B10.RIII mice with CIA. A d j a c e n t cryosections of tibio-tarsal joints of paws with stage 2 CIA stained for NITEGE ~73 [(a), (c), (e), (g)] and VDIPEN 341 [(b), (d), (f), (h)]. (a) and (b) Both neo-epitopes may be co-localized at loci in certain AC regions, while other AC zones exhibit complementary distributions of these cleavage sites. (a) Intense NITEGE 373 labeling is present in the pericellular matrix (arrowheads). Long arrows indicate chondrocytes t h a t are also positive for VDIPEN 341 (b). Some regions with intense NITEGE 3w staining exhibit reduced VDIPEN 341 labeling [crosses; see (b)]. (b) VDIPEN TM staining is localized at the AC surface (short arrows), and in chondrocyte p e r i c e l l u l a r and i n t e r t e r r i t o r i a l matrices (arrowheads). Some foci with intense VDIPEN ~41 labeling exhibit little NITEGE 373 stain [asterisks; see (a)]. (c) and (d) Foci exhibiting both NITEGE 373 (c) and VDIPEN 341 (d) labeling. (e) and (f) H i g h e r magnification of the lower portions of (c) and (d) show NITEG E 37~(e) localized in the deep zone of the AC (arrowheads), while V D I P E N TM (f) is c o n c e n t r a t e d in the superficial zone of the AC. Corresponding arrows depict chondrocytes positive for N I T E G E 3n and VDIPEN3% (g) and (h) E n l a r g e m e n t of the upper regions of (c) and (d). NITEGE 373 (g) and VDIPEN TM (h) are localized in the same pericellular and i n t e r t e r r i t o r i a l matrices (arrows), but NITEGE 37~ labeling appears more widespread. Bar for (a) and (b)-- 20 pm; b a r for (c) and (d) = 100 pm; b a r for (e)-(h) = 50 ~tm.

7 Osteoarthritis and Cartilage Vol. 5 No ~iill iiiiiiiii ~ ~= FIG. 4. SLN-1 digestion does not remove existing NITEGE 3n labeling, and does not induce additional NITEGE 37:~ neo-epitope. Adjacent cryosections of a B10.RIII mouse tarsometatarsal joint with stage 2 CIA were stained for NITEGE 37a without (a) or with (b) SLN-1 predigestion, or for VDIPEN TM without (c) or with (d) SLN-1 predigestion. Pretreatment of the cryosection with SLN-1 induces intense VDIPEN TM labeling throughout the AC (d), but has no detectable effect on NITEGE 373 staining (b). Bar for (a)-(d) = 100 gm. neo-epitope localized in the AC in some of their affected joints. In addition, at least 75% of the mice examined regardless of genotype showed a coincident distribution of the NITEGE 37~ and VDIPEN TM neo-epitopes in the AC pericellular matrix in other affected joints. Discussion We have detected both the NITEGE :~73('aggre- canase') and VDIPEN TM (MMP) neo-epitopes in the AC of inflamed paws of mice with CIA via immunoperoxidase microscopy. The highest concentrations of immuno-specific labeling for both neo-epitopes were localized in the pericellular matrix of AC chondrocytes in paws with a clinical score of 2 or higher. This labeling appeared in portions of the AC exhibiting early stages of disease (as evidenced by local depletion of GAGs), before significant erosion of the AC occurred. Staining for the NITEGE ;~73 ('aggrecanase') neoepitope of aggrecan was performed using affinitypurified antibody and shown to be specific since: (1) it was inhibited by a peptide with NITEGE 373 C-terminus, but not with a peptide spanning the 'aggrecanase' cleavage site; (2) it stained only a specific proteoglycan fragment of the appropriate molecular weight extracted from human OA AC [26]; and (3) it required the entire NITEGE 373 sequence, including the free COO- of the C- terminal Glu for optimal recognition in a radioimmunoassay [12]. The NITEGE 373 antibody was also found to interact with a closely related 'aggrecanase-like' C-terminal ESE cleavage site of brevican [40], but not with ESE within the control peptide spanning this cleavage site. The immunospecificity of the VDIPEN TM antibody has been previously reported [33]. This is the first time that the NITEGE ~73 C-terminus has been identified and immunolocalized in AC of arthritic animals. Previous descriptions of the 'aggrecanase'-generated neo-epitope :~74ARGSV were derived from N-terminal sequencing of aggrecan fragments in culture media from AC explants stimulated with either IL-I~ or retinoic acid [15, 16], and synovial fluids from patients with traumatic joint injury, OA, or RA [13, 14]. More recently, aggrecan fragments with the NITEGE 373 C-terminus were detected in normal, OA, and RA human AC [25, 27]. Together these data are consistent with the proposal that 'aggrecanase' is an endoproteinase and that AC aggrecan fragments are not the result of proteolysis of the entire hyaluronanbinding G1 domain. In addition, these results indicate that aggrecan N-termini presenting either the NITEGE 373 or VDIPEN TM neo-epitopes remain bound within the cartilage matrix following cleavage. The patterns of VDIPEN TM staining are consistent with MMP activity being secreted by both AC chondrocytes, synoviocytes, and inflammatory cells from the synovial cavity and pannus. It appears that chondrocytes are the source of 'aggrecanase' in AC based on its localization in the pericellular matrix deep within the AC, and

8 414 S i n g e r et al.: N I T E G E a n d V D I P E N n e o - e p i t o p e s i n C I A Fro. 5. N I T E G E :~3 a n d V D I P E N TM expression may occur i n d e p e n d e n t l y in the AC of N2 B10.RIII SLN-1 ko mice w i t h CIA. C o n s e c u t i v e sections of a t i b i o t a r s a l j o i n t (a), (b), (c) a n d a t a r s o m e t a t a r s a l j o i n t (d), (e), (f) from a n N2 B10.RIII SLN-1 ko mouse with stage 2 CIA, were s t a i n e d with a n t i - N I T E G E 37a IgG (a), (d), a n t i - V D I P E N TM IgG (b), (e), or t o l u i d i n e blue (c), (f). I n t e n s e N I T E G E :~73s t a i n i n g (a) with little V D I P E N TM i n d u c t i o n (b) occurs in the GAG-depleted (c) t a r s a l AC (arrowheads). A l t e r n a t e l y, m a x i m a l V D I P E N TM l a b e l i n g (e) m a y occur w i t h o u t detectable N I T E G E 37:~ expression (d) in GAG-depleted (f) t a r s o m e t a t a r s a l AC. Bar for (a)-(f)= 100 ~LM. Table II Distribution of N I T E G E 373 and V D I P E N TM neo-epitopes in mice with CIA* Genotype Animals sampled C o i n c i d e n t N I T E G E ata a n d V D I P E N TM l a b e l i n g N o n - c o i n c i d e n t N I T E G E 37:~ a n d V D I P E N TM l a b e l i n g B10.RIII N2 B10.RIII SLN-1 ko N2 B10.RIII wt *Hind paws of mice exhibiting clinical stage 2 CIA (18 21 days after immunization) were cryosectioned mid-sagittally, and the NITEGE 37~ and VDIPEN 34j neo-epitopes detected in the sectioned articular cartilage of various joints by immunoperoxidase microscopy.

9 Osteoarthritis and Cartilage Vol. 5 No the paucity of anti-nitege 373 staining in the more superficial zones of the AC. Although both neo-epitopes were often co-localized surrounding chondrocytes at some foci within the AC, NITEGE 373 and VDIPEN TM were also localized independently of each other in other regions of the AC. Likewise, NITEGE37:~-positive G1 aggrecan fragments were generated in retinoic acidstimulated rat chondrosarcoma cells, without evidence of cleavage at the VDIPEN TM site [12]. This differential expression of the NITEGE 373 and VDIPEN TM neo-epitopes suggests that under some conditions, aggrecan molecules that are cleaved by 'aggrecanase' are not cleaved by classical MMPs, and vice versa. These results also imply that 'aggrecanase' activity is independent of MMP activity, and further suggest that MMPs and 'aggrecanase' may be regulated differently. This is not only evident in inflamed arthritic tissues, but also in the extracellular matrix of control tissues. Anti-NITEGE :~73 (but not anti- VDIPEN TM) labeling was detected in the peripheral AC, tendons, peripheral nerves, and brain of control mice. Similar results have previously been reported in the bovine system, in which 'aggrecanase'-generated G1 fragments were found in extracts of fetal and calf menisci and ligament, but not within growth plate or articular cartilage [42]. It appears that anti-nitege 373 is an early and sensitive marker of aggrecan or brevican catabolism in normal connective tissue. Together, these data suggest that 'aggrecanase' may be involved in normal proteoglycan metabolism whereas MMPs, in combination with 'aggrecanase', may play a more dominant role in aggrecan catabolism in the AC of the inflamed joint. The localization of strikingly high concentratons of anti-nitege :~73 IgG labeling in control B10.RIII mouse peripheral nerves and brain is a surprising observation. The nature of this interaction was investigated by performing competition experiments with the C-terminal and spanning peptides of an 'aggrecanase-like' cleavage site observed in brevican, an abundant aggrecan-like chondroitin sulfate proteoglycan present in adult brain [38 40]. The C-terminal brevican neo-epitope peptide LKLPQEAVESE was found to effectively inhibit the anti- NITEGE 373 immunolabeling of both peripheral nerves and chondrocytes in cryosections of the naive mouse paw, while the peptide spanning this cleavage site did not inhibit labeling. These observations suggest that in addition to the aggrecan -EGE site, the anti-nitege 373 IgG recognizes a closely related -ESE 'aggrecanase- like' neo-epitope in brevican. The generation of this brevican cleavage site appears to have a physiological basis, and presumably is not due to proteolytic cleavage which may occur during biochemical isolation of brevican [39], because the neo-epitope is abundant in freshly frozen, well-preserved peripheral nerves and brain. Aggregating proteoglycans of the aggrecan/versican family are abundant in the extracellular matrix of brain and peripheral nerves [39, 43, 44], and are believed to be involved in regeneration of the peripheral nervous system [45, 46]. The localization of anti-nitege 373 IgG in peripheral nerves and brain suggests that an 'aggrecanaselike' enzyme may regulate the catabolism of brevican, and possibly aggrecan or other aggrecan-like molecules, in normal nerves and during neural tissue remodeling. Interestingly, when SLN-l-treated mouse paw sections were stained with the neo-epitope antibodies, there was no change in the intensity of staining for the 'aggrecanase'-generated NITEGE 37~ neo-epitope, while VDIPEN TM (MMPgenerated neo-epitope) staining clearly increased. These data suggest that 'aggrecanase'-generated G1 neo-epitopes remain bound within the AC matrix even after subsequent exposure to SLN-1. This is a surprising result since theoretically, SLN-1 treatment should have converted the G1-NITEGE 373 fragments into GI-VDIPEN 341 species with a consequent loss of NITEGE 37:~ staining. Since treatment with SLN-l-generated a clear increase in VDIPEN TM labeling within the AC, G1 cleavage must be occurring under the conditions of the experiment. Therefore, SLN-1 probably cleaved the 'aggrecanase'-generated GI- NITEGE 373 fragments, leaving the cleaved C- terminal NITEGE 373 fragments bound within the AC matrix, possibly by non-covalent associations with other matrix molecules through keratan sulfate side chains. The additional observation that treatment of the cartilage with exogenous SLN-1 did not increase the level of NITEGE 3~3 labeling, suggests strongly that SLN-1 does not cleave at the 'aggrecanase' site, as shown previously with purified aggrecan [26]. It has previously been shown that SLN-1 cleaves aggrecan between Asn34'-Phe ~2 in vitro, generating an aggrecan fragment terminating in VDIPEN TM [47]. In contrast, SLN-1 does not cleave aggrecan at the 'aggrecanase' site under similar conditions [26]. To determine if SLN-1 was the only enzyme cleaving aggrecan at the MMP site in vivo, expression of the VDIPEN 341 neo-epitope was examined by immunolocalization in SLN-1 ko mice with CIA [34, 35]. No detectable differences in AC

10 416 Singer et al.: NITEGE and VDIPEN neo-epitopes in CIA loss, proteoglycan staining, or induction of aggrecan cleavage at VDIPEN TM were observed among B10.R111, N2 B10.R111 wt, and N2 B10.R111 SLN-1 ko mice with CIA [34, 35]. The current study addresses the question of whether deletion of SLN-1 gene influences aggrecan cleavage at the 'aggrecanase' site during CIA. There were no observable differences in the staining patterns or expression levels of NITEGE 373 within the AC of the N2 B10.Rl11 SLN-1 ko mice compared with N2 B10.Rlll wt siblings and B10.RIII animals expressing SLN-1. These data, together with the observation that digestion of mouse paw cryosections with rhsln-1 does not induce additional NITEGE 373 labeling in AC, suggest that SLN-1 does not have 'aggrecanase' activity, and does not seem to influence the expression level of NITEGE 37s in CIA. Furthermore, it appears that additional MMPs (possibly in combination with SLN-1), cleave aggrecan at the MMP site in the murine CIA model [35]. It has previously been shown that a number of MMPs can cleave aggrecan at the MMP site [9-11, 17, 19, 26], and that some of these enzymes are upregulated in paws of N2 B10.R111 SLN-1 ko animals with CIA [34, 35]. Therefore, in a complicated inflammatory process such as that seen in CIA, it is probable that multiple MMPs in addition to 'aggrecanase' may be contributing to aggrecan cleavage. The induction of both 'aggrecanase'- and MMP-generated aggrecan neo-epitopes appears to be a sensitive and speci.fic indicator of early AC pathology. Taken together with recent data localizing NITEGE 37~ and VDIPEN 34I in damaged regions of human RA and OA AC [24-27], our results suggest that the NITEGE 373 and VDIPEN TM neo-epitopes are relevant markers of aggrecan catabolism in arthritic disease. The data presented here are consistent with the proposal that 'aggrecanase' is a different enzyme than the MMPs which cleave aggrecan at the MMP site, and that both activities are differentially regulated. Unfortunately, the data do not resolve the question of which enzyme cleaves the aggrecan molecule first during the chondrodestructive process [26]. Using these neo-epitope antibodies in well-defined timecourse experiments may help to answer this question. In any case it seems that both 'aggrecanase', as well as additional 'classical' MMPs, are involved in aggrecan cleavage during CIA. Acknowledgments We thank Drs G. H. Nabozny and C. S. David (Department of Immunology, Mayo Clinic, Rochester, MN, U.S.A.) for their help in establishing the CIA mouse model at Merck Research Laboratories, Dr John W. Woods (Department of Inflammation Research, Merck Research Laboratories) for interpretation of the brain cryosections, and Dr John A. Schmidt (Department of Inflammation Research, Merck Research Laboratories) for productive discussions. References 1. Hembry RM, Bagga MR, Reynolds JJ, Hamblen DL. Immunolocalisation studies on six matrix metalloproteinases and their inhibitors, TIMP-1 and TIMP-2, in synovia from patients with osteo- and rheumatoid arthritis. Ann Rheum Dis 1995;54: Okada Y, Gonoji Y, Nakanishi I, Nagase H, Hayakawa T. Immunohistochemical demonstration of collagenase and tissue inhibitor of metalloproteinases (TIMP) in synovial lining cells of rheumatoid synovium. Virchows Arch B Cell Pathol Incl Mol Pathol 1990;59: Okada Y, Shinmei M, Tanaka O, Naka K, Kimura A, Nakanishi I, et al. Localization of matrix metalloproteinase 3 (stromelysin) in osteoarthritic cartilage and synovium. Lab Invest 1992;66:68(L Lohmander LS, Hoerrner LA, Lark MW. Metalloproteinases, tissue inhibitor, and proteog]ycan fragments in knee synovial fluid in human osteoarthritis. Arthritis Rheum 1993;36: Walakovits LA, Moore VL, Bhardwaj N, Gallick GS, Lark MW. Detection of stromelysin and collagenase in synovial fluid from patients with rheumatoid arthritis and posttraumatic knee injury. Arthritis Rheum 1992;35: Yoshihara Y, Obata K, Fujimoto N, Yamashita K, Hayakawa T, Shimmei M. Increased levels of stromelysin-1 and tissue inhibitor of metalloproteinases-1 in sera from patients with rheumatoid arthritis. Arthritis Rheum 1995;38: Manicourt DH, Fujimoto N, Obata K, Thonar EJ. Levels of circulating collagenase, stromelysin-1, and tissue inhibitor of matrix metalloproteinases 1 in patients with rheumatoid arthritis. Relationship to serum levels of antigenic keratan sulfate and systemic parameters of inflammation. Arthritis Rheum 1995;38: Dean DD, MarteLPelletier J, Pelletier JP, Howell DS, Woessner JF. Evidence for metalloproteinase and metalloproteinase inhibitor imbalance in human osteoarthritic cartilage. J Clin Invest 1989;84: Fosang AJ, Neame PJ, Hardingham TE, Murphy G, Hamilton JA. Cleavage of cartilage proteoglycan between gl and g2 domains by stromelysins. J Bio] Chem 1991;266: Fosang AJ, Last K, Knauper V, Neame PJ, Murphy G, Hardingham TE, et al. Fibroblast and neu-, trophil collagenases cleave at two sites in the cartilage aggrecan interglobular domain. Biochem J 1993;295: Flannery CR, Lark MW, Sandy JD. Identification of a stromelysin cleavage site within the interglobular domain of human aggrecan: evidence for proteolysis at this site in-vivo in human articular cartilage. J Biol Chem 1992;267:

11 Osteoarthritis and Cartilage Vol. 5 No Lark MW, Gordy JT, Weidner JR, Ayala J, Kimura JH, Williams HR, et al. Cell-mediated catabolism of aggrecan. Evidence that cleavage at the 'aggrecanase' site (Glu373-Ala374) is a primary event in proteolysis of the interglobular domain. J Biol Chem 1995;270: 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: Sandy JD, Flannery CR, Neame PJ, Lohmander LS. The structure of aggrecan fragments in human synovial fluid. Evidence for the involvement in osteoarthritis of a novel proteinase which cleaves the Glu373-Ala 'm bond in the interglobular domain. J Clin Invest 1992;89: Sandy JD, Neame PJ, Boynton RE, Flannery CR. Catabolism of aggrecan in cartilage explants. Identification of a major cleavage site within the interglobular domain. J Biol Chem 1991;266: Loulakis P, Shrikhande A, Davis G, Maniglia CA. N-terminal sequence of proteoglycan fragments isolated from medium of interleukin-l-treated articular-cartilage cultures. Putative site(s) of enzymic cleavage. Biochem J 1992;284: Fosang AJ, Neame PJ, Last K, Hardingham TE, Murphy G, Hamilton JA. The interglobular domain of cartilage aggrecan is cleaved by PUMP, gelatinases, and cathepsin B. J Biol Chem 1992;267: Fosang AJ, Last K, Neame PJ, Murphy G, Knauper V, Tschesche H, et al. Neutrophil collagenase (MMP-8) cleaves at the aggrecanase site E373- A374 in the interglobular domain of cartilage aggrecan. Biochem J 1994;304: Fosang AJ, Last K, Neame PJ, Hughes CE, Caterson B, Hardingham TE, et al. Neutrophil collagenase cleaves the aggrecanase site in the interglobular region of aggrecan. Trans Orthopedic Res Soc 1994;19:48 (Abstr.). 20. Chubinskaya S, Huch K, Mikecz K, Cs-Szabo G, Hasty KA, Kuettner KE, et al. Chondrocyte matrix metalloproteinase-8: up-regulation of neutrophil collagenase by interleukin-1 beta in human cartilage from knee and ankle joints. Lab Invest 1996;74: Cole AA, Chubinskaya S, Schumacher B, Huch K, Cs-Szabo G, Yao J, et al. Chondrocyte matrix metalloproteinase-8-~-human articular chondrocytes express neutrophil collagenase. J Biol Chem 1996;271: Huch K, Chubinskaya S, Harris AI, Mikecz K, Kuettner KE, Cole AA. Human osteoarthritic chondrocytes express message for neutrophil collagenase and stromelysin. Trans Orthopedic Res Soc 1995;20:338 (Abst.). 23. Buttle DJ, Saklatvala J, Tamai M, Barrett AJ. Inhibition of interleukin l-stimulated cartilage proteoglycan degradation by a lipophilic inactivator of cysteine endopeptidases. Biochem J 1992;281: Bayne EK, Donatelli SA, Singer II, Weidner JR, Hutchinson NI, Hoerner LA, et al. Detection of a metalloproteinase-generated aggrecan HABR fragment within human OA and RA cartilage. Trans Orthopedic Res Soc 1994;19:308 (Abstr.). 25. Bayne EK, Donatelli SA, Sargeant J, Singer II, Lark MW, Hoerrner LA, et al. Aggrecan fragments terminating in the sequence NITEGE '~7:~ are present within human OA and RA cartilage. Trans Orthop Res Soc 1995;20:328 (Abstr.). 26. Lark MW, Bayne, E.K., and Lohmander, L.S. Aggrecan degradation in osteoarthritis and rheumatoid arthritis. Acta Orthopaedica Scan 1995;66: Lark MW, Bayne EK, Flanagan J, Harper CF, Hoerner LA, Hutchinson NI, et al. Aggrecan degradation in human cartilage. Evidence for both matrix metalloproteinase and aggrecanase activity in normal, osteoarthritic and rheumatoid joints. J Clin Invest 1997;100: Olszewski J, McDonnell J, Stevens K, Visco D, Moore V. A matrix metalloproteinase-generated aggrecan neoepitope as a marker of skeletal maturation and aging in cartilage. Arthritis Rheum 1996;39: Moder KG, Nabozny GH, Luthra HS, David CS. Immunogenetics of collagen induced arthritis in mice: a model for human polyarthritis. Reg Immunol 1992;4: Wooley PH. Collagen-induced arthritis in the mouse. Methods Enzymol 1988;162: Glant T. Induction of cartilage degradation in experimental arthritis produced by allogeneic and xenogeneic proteoglycan antigens. Connect Tissue Res 1982;9: Case JP, Sano H, Lafyatis R, Remmers EF, Kumkumian GK, Wilder RL. Transin/stromelysin expression in the synovium of rats with experimental erosive arthritis. In situ localization and kinetics of expression of the transformationassociated metalloproteinase in euthymic and athymic Lewis rats. J Clin Invest 1989;84: Singer II, Kawka DW, Bayne EK, Donatelli SA, Weidner JR, Williams HR, et al. VDIPEN, a metalloproteinase-generated neoepitope, is induced and immunolocalized in articular cartilage during inflammatory arthritis. J Clin Invest 1995;95: Mudgett JS, Chartrain NA, Christen A, McDonnell J, Shen CF, Kawka DW, et al. Collagen-indeced arthritis in the stromelysin-1 (MMP-3) knock-out mouse. Trans Orthop Res Soc 1995;20:148 (Abstr.). 35. Mudgett JS, Hutchinson NI, Chartrain NA, Christen A, McDonnell J, Shen CF, et al. Stromelysin-1 deficient mice are susceptible to collagen induced arthritis and cartilage destruction. Arthritis Rheum 1997;in press. 36. McLean IW, Nakane PK. Periodate-lysineparaformaldehyde fixative. A new fixation for immunoelectron microscopy. J Histochem Cytochem 1974;22: Shu S, Ju G, Fan L. The glucose oxidase-dab nickel method in peroxidase histochemistry of the nervous system. Neurosci Lett 1988;85: Seidenbecher CI, Richter K, Rauch U, Fassler R, Garner CC, Gundelfinger ED. Brevican, a chondroitin sulfate proteoglycan of rat brain, occurs as secreted and cell surface glycosylphosphatidylinositol-anchored isoforms. J Biol Chem 1995;270:

12 418 Singer et al.: NITEGE and VDIPEN neo-epitopes in CIA 39. Yamada H, Watanabe K, Shimonaka M, Yamaguchi Y. Molecular cloning of brevican, a novel brain proteoglycan of the aggrecan/versican family. J Biol Chem 1994;269: Yamada H, Watanabe K, Shimonaka M, Yamasaki M, Yamaguchi Y. cdna cloning and the identification of an aggrecanase-like cleavage site in rat brevican. Biochem Biophys Res Commun 1995;216: Getzy LL, Malemud CJ, Goldberg VM, Moskowitz RW. Factors influencing metachromatic staining in paraffin-embedded sections of rabbit and human articular cartilage: a comparison of the Safranin O and Toluidine blue O techniques. J Histotechnol 1982;5:111~ Koob TJ, Hernandez DJ, Gordy JT, Sandy JD. Aggrecan metabolism in bovine meniscus: Role of aggrecanase in normal development. Trans Orthop Res Soc 1995;20:3 (Abstr.). 43. Bignami A, Hosley M, Dahl D. Hyaluronic acid and hyaluronic acid-binding proteins in brain extracellular matrix. Anat Embryol 1993;188: Hanemann CO, Kuhn G, Lie A, Gillen C, Bosse F, Spreyer P, et al. Expression of decorin mrna in the nervous system of rat. J Histochem Cytochem 1993;41: Braunewell KH, Martini R, LeBaron R, Kresse H, Faissner A, Schmitz B, et al. Up-regulation of a chondroitin sulphate epitope during regeneration of mouse sciatic nerve: evidence that the immunoreactive molecules are related to the chondroitin sulphate proteoglycans decorin and versican. Eur J Neurosci 1995;7: Tona A, Perides G, Rahemtulla F, Dahl D. Extracellular matrix in regenerating rat sciatic nerve: a comparative study on the localization of laminin, hyaluronic acid, and chondroitin sulfate proteoglycans, including versican. J Histochem Cytochem 1993;41: Lark MW, Williams H, I=Ioernner LA, Weidner J. Ayala JM, Harper CF, et al. Quantification of a matrix metalloproteinase-generated aggrecan G[ fragment using monospecific anti-peptide serum. Biochem J 1995;307: