WO2002068675A2

WO2002068675A2 - Product, method and system for identifying collagenase-cleaved type ii collagen

Info

Publication number: WO2002068675A2
Application number: PCT/US2002/005790
Authority: WO
Inventors: R. Clark Billinghurst; C. Wayne Mcilwraith
Original assignee: Colorado State University Research Foundation
Priority date: 2001-02-23
Filing date: 2002-02-25
Publication date: 2002-09-06
Also published as: AU2002254035A1; WO2002068675A3

Abstract

A polyclonal anti-neoepitope antibody is described that recognizes collagenase-cleaved, type II collagen fragments in a horse, such antibody able to detect increases in type II collagen cleavage in diseased equine articular cartilage. Preferably, the antibody comprises 234 CEQ antibody. Using such antibody, a method is provided for detecting cartilage degradation in the joints of horses.

Description

PRODUCT, METHOD AND SYSTEM FOR DDENTIFYTNG COLLAGENASE-CLEAVED TYPE II COLLAGEN

FIELD OF THE INVENTION The present invention is directed to a antibody and method for detecting cartilage degradation, particularly in the joints of horses and/or dogs, and in one embodiment comprises a polyclonal anti-neoepitope antibody that recognizes collagenase-cleaved type II collagen fragments in a horse and/or dog.

BACKGROUND AND SUMMARY OF THE INVENTION

The two main structural components of articular cartilage are collagen and the proteoglycan. aggregate called aggrecan. Fibrillar type LT collagen is the major collagen of articular cartilage (~90-95%) and provides its tensile strength. It has recently been shown that during cartilage degradation induced by the proinflammatory cytokine interleukin-1 (IL-1), as may occur in arthritis, there is an initial loss of proteoglycans that, upon reaching a critical level, is followed by the collagenase-induced cleavage and denaturation of type II collagen. Due, in part, to its critical role in maintaining the integrity of articular cartilage, as well as its very slow rate of turnover, damage to the type II collagen network is believed to initiate the irreversible stages of cartilage degradation. Indeed, increased denaturation and cleavage of type II collagen have been identified in human articular cartilage in osteoarthritis (OA). A method to detect and monitor this process would allow for intervention at an earlier point in time than is currently possible, before the irreversible changes of OA.

The cleavage of fibrillar collagens, such as types I, II and HI, can occur in the individual α chains either outside or within the triple helical region of these molecules. The type II collagen molecule is composed of a triple helix of three identical α chains. The intrahelical digestion occurs at a specific and well-identified site (Gly 5-Leu/Ile₇₇₆) within each α chain of the triple helical collagen molecule (Figure 1) through the action of collagenases belonging to a family of zinc-dependent endopeptidases called matrix metalloproteinases (MMPs). CoUagenase 1 (MMP-1), coUagenase 2 (MMP-8) and coUagenase 3 (MMP-13), along with MMP-14, a membrane-type MMP (MT1-MMP), are the only mammalian enzymes known to be capable of initiating this intrahelical cleavage at neutral pH. These collagenases are believed to define the rate-limiting step in normal tissue turnover and in degradative processes, such as in arthritis. Antibodies have been developed to detect the fragments created by the cleavage of types I and II collagens by these MMP collagenases in humans. The principle behind the technology involved in producing these antibodies is that, as a result of the action of the collagenases on the individual collagen α chains, three-quarter (TC^A) and one quarter (TC^B) length fragments are created with new carboxy- (C-) and amino- (N-) termini, respectively. Based on terminal sequencing results for these fragments, antibodies are then prepared to react to the newly created termini or neoepitopes and not to react to intact collagen α chains. A previously generated anti-neoepitope antibody, COL2- 3/4C_Short₅ recognized not only the cleaved ³Λ (TC^A) fragments of type II collagen of humans, horses and cows, but, as well, the % α chain firagments of type I collagen in humans. This was not unexpected as the immunizing peptide used to create the COL2- 3/4C_Short antibody utilized the amino acid sequence for cleaved type II collagen (Gly-Pro- Hyp-Gly-Pro-Gln-Gly, where Hyp represents hydroxyproline) that differed by only one amino acid (highlighted below) from the C-terminal α chain sequence of the % fragments of type I collagen (Gly-Thr-Hyp-Gly-Pro-Gln-Gly; Figure 1).

To obtain an antibody that only recognized collagenase-cleaved type II collagen, it was hypothesized that the immunizing peptide needed to be lengthened in an amino- (N-) terminal direction to include more of the amino acids that differ between the collagen types at the C-termini of their % fragments. It had previously been shown that only the addition of the first amino acid N-terminal to the COL2-3/4C_Sho_rt neoepitope, (available from Shriner's Hospital) which is glutamic acid (Glu) in human type II collagen, was required to obtain a type II specific anti-neoepitope antibody called COL2-3/4Cι_ong (available from Shriner's Hospital). In preliminary studies, we obtained poor cross- reactivity of this antibody for cleaved equine type II collagen and that may be due to the substitution of Glu for aspartic acid (Asp) in horses at position -8 numbering from the C- terminus of the coUagenase generated type II collagen % α chain fragment. This Glu- Asp substitution is also seen in the type II collagens of cows, mice and rats (Figure 1).

The purpose of this study was to develop an antibody, using this technology, to identify the neoepitope that is the C-terminus of the type II collagen TC^A fragment produced by coUagenase digestion in horses. The published amino acid sequence for equine type II procollagen was used to identify the collagenase-cleavage site. Unlike the anti-neoepitope antibodies previously produced to recognize cleaved human types I and π collagens, our hypothesis was that by generating this new antibody according to an extended amino acid sequence specific for the equine type LI collagen cleavage site, immunoreactivity only to collagenase-digested type π collagen would result. Using this antibody in an immunoassay, type LT collagen cleavage in equine articular cartilage in culture could be identified and quantitated, and the effect of an MMP inhibitor on this degradation could be monitored. Finally, through the use of this antibody in immunohistochemistry, increased cleavage of type LT collagen in articular cartilage of horses with OA could be identified and localized.

Objective — To develop and characterize an antibody that specifically recognizes collagenase-cleaved type LI collagen in equine articular cartilage. Sample Population - All samples for collagen isolation, cartilage explant cultures and immunohistochemistry were from horses euthanized for problems unrelated to the musculoskeletal system.

Procedure - A peptide was synthesized representing the C-terminus (neoepitope) of the equine type II collagen fragment created by digestion with mammalian collagenases. This peptide, conjugated to ovalbumin, was used to immunize rabbits for the production of an antibody that was characterized by Western blotting for its reactivity to native and coUagenase 3 (MMP-13)-cleaved equine types I, LI and HI collagens. The antibody was evaluated as a true anti-neoepitope antibody in an ELISA using peptides with an amino acid added to or removed from the C-terminus of the immunizing peptide. Articular cartilage from the metacarpophalangeal joints of horses was cultured with interleukin-1 alpha (LL-lα) +/- a synthetic MMP inhibitor, BAY 12-9566 (available from Bayer Corporation), and collagen degradation was assayed using the antibody. Cartilage from an osteoarthritic joint was compared to that from a non-arthritic equine joint for staining with the antibody. Results - An affinity purified antibody, called 234CEQ antibody, recognized the

% (TC^A) f agments of purified equine type II collagen created by coUagenase 3, but not similarly generated fragments of equine types I and HI collagens. This was a true anti- neoepitope antibody, as the removal or addition of one amino acid at the C-terminus of the immunizing peptide resulted in a significant loss of competition for binding in an inhibition ELISA. There was a significant release from articular cartilage explants of type II collagen fragments bearing this neoepitope in response to LL-1 and this could be prevented with an MMP inhibitor. Finally, articular cartilage from an osteoarthritic joint of a horse showed increased staining with the 234CEQ antibody compared to cartilage from a normal joint.

Conclusions - We have generated a polyclonal anti-neoepitope antibody that recognizes collagenase-cleaved, type II collagen fragments in the horse. This antibody can detect increases in type II collagen cleavage in diseased equine articular cartilage.

Clinical Relevance - The 234CEQ antibody has the potential of detecting cartilage degradation in the joints of horses for the early diagnosis of arthritis and to monitor response to treatment.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 ~ The location and comparison of collagenase-generated % fragment C- terminal neoepitope of equine type II collagen with that of other species. Shown in the diagram are the ³A (TC^A) and % (TC^B) fragments resulting from the cleavage of a triple helical collagen molecule by mammalian coUagenase. The amino acid sequences for the newly created C-termini of the αl chain % fragments for the horse, cow, human, mouse and rat are aligned below the figure. Included are the aligned sequences for the αl and α2 chain ³A fragments of human type I collagen. Residues similar to the equine fragments are indicated for each of the other species by a dash below each amino acid in the horse sequence. The ONA-peptide conjugate used to generate the 234CEQ antibody is shown at the bottom of the figure. (Artwork by MarkLepik, Shriners Hospital, Montreal, QC)

Figure 2 — Neoepitope specificity of the 234CEQ antibody. Using an inhibition ELISA, varying dilutions of the immunizing peptide (•) were added to a 1:1000 dilution of a F(ab')₂ preparation of the 234CEQ antibody in 96-well round bottom microtiter plates. After 1 h at 37°C, aliquots were transferred to the wells of Immulon-2 microtiter plates precoated with the immunizing peptide conjugated to KLH. The plates were processed as described in Materials and Methods to generate the standard inhibition curve shown. This was repeated for similar dilutions of peptides that had either one amino acid added to (O) or removed from (D) the end of the immunizing peptide corresponding to the cleaved C-terminus of the ³A (TC^A) fragment. The sequences for these peptides are shown in the boxed inset. The percentage inhibition of binding by each peptide was calculated relative to the mean absorbance from the four maximum binding wells that contained 234CEQ antibody and no peptide, thereby representing 0% inhibition (100% binding). Figure 3 — Time course of cleavage of triple helical equine type II collagen of the horse by recombinant human MMP- 13 (rHuMMP-13). Type II collagen, purified from macroscopically normal appearing equine articular cartilage, was digested in solution over 24 h with APMA-activated rHuMMP-13 at a final molar ratio of 1:5 (enzyme:collagen). Aliquots were removed at 0, 0.5, 1, 3, 5, 18, 21 and 24 h and the MMP- 13 was inactivated with 20 mM EDTA. Each sample was separated under reducing conditions by SDS-PAGE (10%) and either stained with Coomassie Blue (A) or electrophoretically transferred to nitrocellulose and incubated with a 1:1000 dilution of the 234CEQ F(ab')₂ preparation (B). Digestion times are indicated above the corresponding sample lanes. The right margin indicates the positions of uncleaved αl(II) chains and the single (αTC^A) form of the ³A fragments produced by MMP- 13 cleavage.

Figure 4 ~ Immunoreactivity of the 234CEQ antibody for native and MMP- 13 cleaved equine types I, LT and HI collagens. Purified triple helical equine collagens, types I and HI from skin and type H from articular cartilage, were incubated with (+) and without (-) rHuMMP-13 for 24 h. Shown are the Coomassie Blue stained SDS-PAGE (10%) gel (A) and the Western blot for the 234CEQ antibody (B). The lanes are the molecular weight (MW) standards (lane 1), intact types I (lane 2), H (lane 4) and HI (lane 6) collagens, and cleaved types I (lane 3), H (lane 5) and HI (lane 7) collagens. The right margin is as described in the legend of Figure 3, with the exception that the two αTC^A fragments of type I collagen are also indicated. The left margin indicates the molecular weights, in kD, of the standards in lane 1.

Figure 5 — Time course of LL-1 induced proteoglycan and collagen degradation in equine articular cartilage explants and the dose-dependent effect of a synthetic MMP inhibitor. Media were replaced every other day in the cartilage cultures and the conditioned media that were removed were assayed /for degradation products of proteoglycan and type H collagen generated through the co-incubation with human IL- lα for 20 days. Shown are the media levels of proteoglycan fragments (A), measured as μg of sulfated glycosaminoglycans (sGAG) and collagenase-generated fragments of type H collagen (B), measured as nmoles of the 234CEQ neoepitope, both molecules per mg dry weight (DW) of cartilage. The effects of InM (λ), lOOnM (v) and 10μM(σ) of the nonpeptidic MMP inhibitor BAY 12-9566 (BAY 12) on the IL-1 induced generation of both metabolites (O) are indicated on each graph. Note that type H collagen cleavage follows the early peak in proteoglycan release and that there is a dose-dependent effect of BAY 12-9566 on the degradation of both molecules.

Figure 6 - Cumulative effect of the novel MMP inhibitor BAY 12-9566 on coUagenase cleavage of type H collagen in equine articular cartilage explant cultures. The conditioned media that were removed every other day from the cartilage cultures were assayed for levels of type H collagen fragments bearing the collagenase-generated neoepitope, 234CEQ. Shown are the total amounts of 234CEQ, expressed as nmoles of peptide per mg DW of cartilage, for the explants cultured over 20 days with IL-1 alone and those including InM, lOOnM and lOμM of BAY 12-9566. Significantly lower total levels of neoepitope in cultures with inhibitor compared to those with IL-1 alone are indicated as * (PO.05) and ** (P<0.01), as assessed using ANONA with Dunnett's post- test.

Figure 7 — Correlation between the 234CEQ and the COL2-3/4C_short assays measuring collagenase-generated cleavage products of collagen. Each media sample collected over the 20 day incubation period was assayed in both an ELISA utilizing the newly generated 234CEQ antibody that recognizes collagenase-cleaved type H collagen fragments and in the previously described COL2-3/4C_Sh0rt ELISA which measures cleaved types I and H collagen fragments. The values for each sample were plotted as ordered pairs and the line of best fit was plotted using linear regression analysis. Shown in the inset is the highly significant Pearson correlation r value.

Figure 8 ~ Immunostained sections of articular cartilage removed from an osteoarthritic (A, B) and a normal (C, D) joint of two horses. Articular cartilage sections in A and C were incubated with a 1:250 dilution of the 234CEQ coUagenase cleavage site antibody as an IgG F(ab')₂ preparation and immunoreactivity was detected using DAB substrate and a horse radish peroxidase labeled polymer conjugated with anti-rabbit secondary antibody. Controls for specificity of staining were serial sections for each cartilage sample (B and D) that were treated as above except that the primary 234CEQ antibody was preincubated for 1 h at 37°C with a 500 μg/ml solution of immunizing peptide before being added to the tissue section. Arrows in A indicate chondrocyte clusters (clones) in the osteoarthritic cartilage with the surrounding increase in immunostaining for the 234CEQ neoepitope. The articular surface in all sections is in the upper right hand corner of each photomicrograph. Scale = 100 μm Figure 9 ~ The graph shows the results of assaying the levels of 234CEQ fragments in serum samples drawn biweekly in 2 dogs (Dogs 9 and 34) after anterior cruciate ligament transection.

Figure 10 ~ The table compares the necropsy findings at Day 126 for both these dogs. Note that Dog 9 had lower pathology scores than Dog 34 and this correlated to the lower 234CEQ serum levels at most time points, and especially at Day 126, for Dog 9 versus Dog. 34. These are the only 2 dogs from this study that have been assayed for serum 234CEQ levels.

Figure 11 — The specific aims of this project were to compare the concentrations of 234CEQ in synovial fluids and sera of treadmill exercised and hand walked control horses. The overall goal of this study was to assess its usefulness as a diagnostic aid. Serum concentrations of 234CEQ were measured in exercised and control horses during the 6 month study. There was a significant increase in 234CEQ in the serum of both groups of horses during the study. Concentrations were higher at day 194 than at any other time (*), and higher at day 164 than at days 0, 14, 44 and 74.

Figure 12 — Synovial fluid concentrations of 234CEQ in the metacarpophalangeal joints of exercised and control horses. ** indicates significant differences between groups at day 194.

Figures 13a and Figures 13b — Cross-reactivity of 234CEQ for collagenase- cleaved (MMP- 13) canine type H collagen. Shown are the SDS-PAGE gel (a) and Western blot (b) for the sequential digestion of both canine and equine type H collagens with MMP-13. Both collagens were digested for 8 hours by human MMP-13, with aliquots being removed at 1, 4 and 8 hours of digestion. The aliquots were run on a SDS- PAGE gel as labeled in (a) and transferred to a nitrocellulose membrane for incubation with the 234CEQ antibody. The Western blot in (b) shows the staining of the individual MMP-13 cleaved collagen ³A fragments (and multimers) for both the equine type H collagen for which the antibody was developed (positive control) and for the canine type H collagen fragments.

DETAILED DESCRIPTION OF THE INVENTION

Experimental Design

A peptide was synthesized that consisted of the eight amino acid sequence forming the C-terminus of the ³/₄ α chain fragment of equine type H collagen, resulting from primary digestion with mammalian collagenases. After conjugation to a carrier protein, ovalbumin (OVA), the peptide-ONA was used to immunize rabbits for the production of a polyclonal antibody recognizing this neoepitope. Sera from the rabbits were affinity purified, and the resulting polyclonal antibody, hereinafter called 234CEQ, was tested by immunoblotting for reactivity to purified solutions of native and MMP-13 cleaved types I, H and HI equine collagens. An inhibition ELISA was developed utilizing the 234CEQ antibody, and competing peptides with an amino acid added to or removed from the C-terminus of the immunizing peptide were used to characterize 234CEQ as a true anti-neoepitope antibody. This ELISA was also used to detect collagenase-cleaved type H collagen fragments released from equine articular cartilage explants in response to culturing with interleukin-1 alpha (LL-lα) and to monitor the effect of a synthetic MMP inhibitor on the IL-lα induced collagen degradation. Finally, articular cartilages removed from a normal and an osteoarthritic joint of two horses were compared in terms of the degree and localization of immunostaining with the 234CEQ antibody.

Synthesis of the peptide representing the coUagenase cleavage site neoepitope for equine type II collagen

The peptide Cys-Gly-Gly-Asp-Gly-Pro-Hyp-Gly-Pro-Gln-Gly (C-G-G-D-G-P- P_OH-G-P-Q-G), representing the amino acid sequence for the collagenase-created C- terminus of the equine type H collagen α chain, was synthesized at a 0.25-mmol scale, using standard Fmoc (9-fluoroenylmethoxycabonyl) chemistry on a solid-phase peptide synthesizer by Macromolecular Resources (Colorado State University, Fort Collins, CO). The cysteine at the Ν-terminus of the peptide was added to allow for conjugation with the carrier proteins ONA and keyhole limpet hemocyanin (KLH) using the bifunctional reagent N-hydroxy-succinimidyl bromoacetate, as previously described. Two glycines were added to act as a spacer between the Ν-terminal cysteine and the neoepitope sequence, thereby improving the immunogenicity of the peptide when conjugated to ONA and the sensitivity of the assays using the peptide conjugated to KLH.

Polyclonal antibody preparation

Two female New Zealand White rabbits weighing 2.5-3.0 kg were immunized intramuscularly ( ) with 0.5 mg of neoepitope-ONA conjugate in 0.25 ml PBS and emulsified with 0.25 ml of complete Freund's adjuvant (CFA). Similar quantities of peptide-ONA emulsified with incomplete Freund's adjuvant (IF A) were administered LM every 2 weeks. Ten days after the second booster, test bleeds were performed and antibody titers measured with an ELISA, as previously described. Another booster injection of peptide-ONA in IF A was required to obtain good titers for both rabbits and the rabbits were then exsanguinated by cardiac puncture to obtain 50-60 ml of serum. To acquire affinity purified 234CEQ IgG antibody, aliquots of serum were initially applied to columns containing Protein A Sepharose High Performance to isolate polyclonal IgG. A column was then prepared by coupling the immunizing peptide, without carrier protein, to an activated affinity support, according to the manufacturer's instructions. The purified IgG was applied to the column and the bound 234CEQ IgG was eluted with glycine. For immunohistochemistry, F(ab')₂ preparations were obtained by pepsin digestion of the IgG, as previously described. The Fc portions and undigested IgG were removed by Protein A affinity chromatography, as described above.

Inhibition ELISA for antibody characterization and quantitation of cleaved collagen

The immunizing peptide, conjugated to KLH, was diluted to 0.1 μg/ml in Tris- buffered saline (TBS), pH 7.2, and 50 μl was added to each well of flat-bottom, 96-well, high binding tissue culture plates. After 24 hours at 4°C, the plates were washed three times with TBS containing 0.1% vol/vol Tween 20 (TBST). Νoncoated binding sites were blocked with 1% wt/vol BSA in TBS (TBS-1% BSA) for 30 min at room temperature. The plates were washed three times with TBST and stored at 4°C until used in an ELISA.

For antibody specificity analyses, peptides were synthesized that had the in- sequence amino acid residue either added to (+1) or removed from (-1) the end of the immunizing peptide corresponding to the cleaved C-terminus of the α chain ³/₄ fragment. Each of these peptides were prepared as 1 mg/ml stock solutions and serially diluted in TBS-1% BSA to compare standard curves obtained for each with that of the intact immunizing peptide. To determine non-specific binding, 100 μl consisting of 50 μl TBS- 1% BSA and 50 μl TBS-1% BSA-Tween was added to four wells of 96-well round- bottom polypropylene microtiter plates. The polyclonal F(ab')2 antiserum was diluted 1:1000 in TBS-1% BSA-Tween and 50 μl was added to all remaining wells of these preincubation plates. To four of these wells was added 50 μl of TBS-1% BSA for determination of maximum binding in the absence of the inhibitory epitopes. To the rest of the wells containing antiserum was added 50 μl/well of each of the peptide dilutions, in duplicate, and the plates were covered.

Following incubation for 1 h at 37°C, 50 μl from each well was transferred to the equivalent wells of the plates precoated with immunizing peptide conjugated to KLH. These plates were incubated for 1 h at room temperature and then washed three times with TBST. A goat anti-rabbit IgG antibody conjugated to alkaline phosphatase was diluted 1:20,000 in TBS-1% BSA-Tween and 50 μl was added to each well. After a 1 h incubation at 37°C, the plates were washed three times each with TBST and distilled water. An ELISA amplification system was used, according to the manufacturer's instructions, and the color development was measured as absorbance at 490 nm. For each plate, the mean absorbance from the four nonspecific binding wells was subtracted from the absorbance values of all other wells. The percentage inhibition of binding by each peptide was calculated relative to the mean absorbance from the four maximum binding wells, which represented 0% inhibition. For the measurement of cleaved type H collagen levels in samples, this ELISA was repeated, as described above, with the following exceptions. The samples were added in duplicate and not diluted. The concentration of 234CEQ neoepitope in each sample was determined from the standard curve generated using dilutions of the immunizing peptide, without carrier protein. Results were expressed on a molar basis using the molecular weight for the 234CEQ peptide of 837.

Immunoblottingfor antibody specificity using cleaved types I, II and III collagens

Equine types I, H and HI collagens were isolated by pepsin digestion and differential salt precipitation. Normal articular cartilage was the source of type H collagen and fetal skin was used to obtain types I and HI collagens. Solutions of each were prepared by dissolving the lyophilized collagens in 0.5 M acetic acid and diluting to a final concentration of 2.5 mg/ml of digestion buffer consisting of 50 mM Tris, 10 mM CaCl₂, 0.5 M NaCl, 0.01% Brij 35 and 0.02% NaN₃, pH 7.6. Recombinant human proMMP-13 was activated by incubation with 2 mM (final concentration) aminophenyl mercuric acetate (APMA) in the same digestion buffer for 90 min at 37°C. The activated MMP-13 was added to the collagen solutions at a final molar ratio of 1:5 (MMP- 13 :collagen). Samples were incubated for 24 h at room temperature and the coUagenase was inactivated with 20 mM (final concentration) EDTA. For studies of time-dependent coUagenase cleavage of native type H collagen, aliquots were removed during the 24 h incubation at the times indicated in the figure legend (Fig. 3) and the MMP-13 in each aliquot was inactivated with EDTA.

The MMP-13 cleaved fragments of the purified collagens were separated using sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE), as previously described. The 10%, 1-mm thick, mini-Protean gels were stained with Coomassie Blue R-250. Electrophoretic transfer to nitrocellulose membranes and immunodetection of the TC^A collagen fragments were performed as previously described, with the following exceptions. The BSA-blocked membranes were incubated overnight with the 234CEQ antibody at 1:1000 dilution in PBS-3% BSA-Tween, and after three 10 min washes in PBS-1%BSA-Tween, alkaline phosphatase-conjugated goat anti-rabbit IgG F(ab')₂ fragment diluted 1:1000 with PBS-3% BSA-Tween was added for 1 h at room temperature.

Equine articular cartilage explant cultures

Full-depth articular cartilage slices were removed from the macroscopically normal appearing metacarpophalangeal joints of 6 horses within 3 h of death. The cartilages were pooled and placed for 1 h in serum-free Dulbecco's modified Eagle's medium (DMEM) containing 3.6 mg/ml HEPES, 3.7 mg/ml sodium bicarbonate, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2.5 μg/ml amphotericin B. After another hour in this basic medium (without amphotericin B) and with 10 X the antibiotic concentration added, the cartilage pieces were washed and 40-50 mg wet weight of cartilage was randomly transferred to each well of 48-well tissue culture plates. The explants were cultured in 1 ml/well of basic medium supplemented with 10 μg/ml BSA, 50 μg/ml L-ascorbic acid, and 5 μg/ml insulin, 5 μg/ml transferrin, and 5 ng/ml sodium selenite. After a four day preculture period in a humidified incubator at 37°C and 5% CO₂, the media were removed and triplicate samples were cultured in the same basic medium, with or without (control) the addition of 10 ng/ml recombinant human interleukin-1 alpha (IL-lα). As well, triplicate samples were cultured with or without IL- lα and 1, 100 or 10000 nM of BAY 12-9566, a nonpeptidic biphenyl MMP inhibitor, in 0.1% dimethyl sulfoxide (DMSO). Exposure of equine articular cartilage explants to this concentration of DMSO should not affect proteoglycan metabolism. The media, LL-lα and inhibitor were replaced every 2 d and the conditioned media were frozen at -70°C until assayed for proteoglycans, using the previously described dimethylmethylene blue (DMMB) spectrophotometric assay, and for cleaved collagen using both the previously described COL2-3/4C_Sh_0rt assay and the new 234CEQ ELISA, as described above. The cultures were terminated after 20 days.

Immunohistochemistry

Full depth sections of articular cartilage were removed from the proximal surface of the third carpal bone of an osteoarthritic and a normal intercarpal joint of two horses within 3 h of death. The determination of OA in the one horse was based on clinical signs of joint effusion and pain, gross examination of the joint tissues during autopsy, and eventually confirmed with the histological appearance of the middle carpal bone as described in the results. The tissue blocks were mounted in OCT embedding media, snap frozen in liquid nitrogen and 6-μm thick sections were cut at -18°C using a cryostat. Sections were picked up on glass microscope slides, fixed for 10 min in acetone, and rehydrated with 0.05M TBS. AU incubations were performed at room temperature, except where noted. After a 1 h incubation with 5% wt/vol normal fetal bovine serum in TBS, slides were rinsed with TBST, and 0.5% H O₂ in methanol was added for 10 min to block endogenous peroxidase activity. Before adding the primary antibody, sections were rinsed in TBST and a 0.25 U/ml solution of chondroitinase ABC was added for 90 min at 37°C. This was followed by a TBST rinse, the addition of 0.2M EDTA in 50 mM Tris, pH 7.6 for 1 h and a final TBST rinse. The 234CEQ IgG F(ab')₂ was added at a 1:250 dilution in antibody diluent for 30 min followed by 3 X 10 min washes with TBST. A detection system containing a horse radish peroxidase labeled polymer conjugated with anti-rabbit secondary antibody was added to the sections for 30 min and rinsed with 3 X 10 min with TBST. Diaminobenzidene (DAB) substrate was then added for 7.5 min, followed by 2 x 10 min rinses with dH O. The sections were then counterstained with hematoxylin and permanently mounted. Serial sections were stained with the COL2- 3/4C_Short primary antibody as positive controls, and with 234CEQ antibody preincubated with a 500 μg/ml solution of the immunizing peptide for 1 h at 37°C as specific negative controls. Statistical analysis

The data requiring statistical analysis was that relating to the experiments measuring neoepitope levels in the articular cartilage explant cultures. All statistical analyses were performed using a commercial software program. Deviation of the neoepitope levels from Gaussian distribution was determined using the Kolmogorov- Smirnov (KS) test (a normality test). One-way ANONA was used to compare the neoepitope levels in the IL-1 treated cartilages cultured with 0 nM, 1 nM, 100 nM and 10 μM concentrations of the MMP inhibitor, BAY 12-9566. Means for the groups with different concentrations of the inhibitor were compared to the group cultured with IL-1 alone using Dunnett's post test. A test for linear trend was performed to determine if there was a significant decrease in the means for neoepitope release from the groups of IL-1 stimulated articular cartilages with an increase in BAY 12-9566 concentration. Correlation between neoepitope levels measured with the 234CEQ and the COL2- 3/4C_short ELISAs were determined using Pearson correlation analysis. Linear regression was used to plot the best fit line for the correlation data. For all the tests performed, statistical significance was established at PO.05.

RESULTS

Characterization of the 234CEQ antibody as an anti-neoepitope antibody for collagenase-generated fragments of equine type II collagen

With just one amino acid added to or removed from the end of the immunizing peptide, there was almost complete loss of reactivity with the 234CEQ antibody, as shown by a significant loss of inhibition of these competing peptides in an inhibition ELISA (Fig. 2). The recognition by the 234CEQ antibody of increasing levels of ³/ fragments generated during a 24 h digestion of purified equine type H collagen by the coUagenase MMP-13 was demonstrated by the immunoblotting of aliquots removed at various time points throughout the reaction (Fig. 3). The antibody recognized only the monomeric ³A α chain fragments (αTC^A) of the cleaved equine type H collagen and did not react with the uncleaved αl(IT) chains. Moreover, the 234CEQ antibody did not recognize either intact or cleaved α chains of types I and DT collagens isolated from equine skin (Fig. 4). Quantitation of proteoglycans and of collagen fragments bearing the 234CEQ neoepitope released from degrading equine articular cartilage in culture

Proteoglycan degradation peaked at 4 days of culture and steadily declined thereafter, as measured by sulfated glycosaminoglycan (sGAG) levels in the media of IL- 1 stimulated equine articular cartilage explants (Figure 5 A). The release of cleaved type

H collagen markedly increased immediately after the maximum sGAG release, peaking at day 8 of culture and steadily declining thereafter (Figure 5B).

The MMP inhibitor, BAY 12-9566, caused a dose-dependent decrease in both proteoglycan degradation and type H collagen cleavage. This reached statistical significance for both the lOOnm (PO.05) and lOμM (PO.01) concentrations of BAY 12-

9566, during peak release of these two major cartilage matrix components (Figure 5) and for total media levels of the 234CEQ neoepitope (Figure 6).

To confirm that the 234CEQ antibody was in fact measuring cleaved type H collagen fragments bearing the coUagenase cleavage site neoepitope, all media samples at each time point were also assayed using the previously described COL2-3/4C_Short antibody. The levels of each epitope, expressed as nmoles of peptide per μg dry weight of cartilage, were then plotted as ordered pairs for each media sample and the Pearson correlation coefficient determined. As shown in Figure 7, there was a highly significant correlation in the levels of these 2 neoepitopes (r = 0.377, p<0.0001), further qualifying the 234CEQ antibody as a useful tool for specifically detecting collagenase-induced type H collagen degradation.

Immunodetection of collagenase-cleaved type II collagen in equine osteoarthritic articular cartilage The articular cartilage removed from the proximal surface of a third carpal bone that grossly showed signs of degenerative changes, microscopically revealed extensive chondrocyte cloning throughout the matrix (Figure 8A,B). The staining for collagenase- cleaved collagen using the 234CEQ antibody was most intense immediately surrounding the chondrocyte clones located in the middle zone, with a decrease in staining towards the superficial and deep zones (Figure 8A). The specificity of this immunoreactivity was confirmed in a serial section stained with the 234CEQ antibody preincubated with the immunizing neoepitope peptide (Figure 8B). In normal equine articular cartilage there was no chondrocyte cloning and only minimal immunostaining in the deep zone of the articular cartilage (Figure 8C), that could be removed by preincubation of antibody with immunizing peptide (Figure 8D). Staining with the COL2-3/4C_Sh0rt antibody (not shown), which reacts with both types I and H collagenase-cleaved collagen fragments in horses, was similar to that with the 234CEQ antibody, indicating that the majority of cleaved collagen in the OA cartilage was that of type H collagen.

We have developed and characterized a polyclonal antibody, 234CEQ, that recognizes degradative fragments of equine type H collagen created by cleavage with the mammalian collagenases MMP-1, MMP-8, MMP-13 and or MMP-14 (MT1-MMP). Another antibody (COL2-3/4C_Sh_ort) has been produced that recognizes similarly produced collagen fragments in human and bovine cartilages, and although it also detects collagenase-cleaved type H collagen in the horse, this antibody reacts with cleaved fragments of type I collagen in all these species. Since type I collagen is found in bone, tendon, ligament and skin, to obtain a cartilage-specific antibody for collagen breakdown, a type H collagen-specific anti-neoepitope antibody is required. It was first necessary to prove that the 234CEQ antibody was a true anti- neoepitope antibody. This was accomplished by demonstrating decreased immunoreactivity of the antibody for peptides with alterations at the C-terminus representing the collagenase-cleaved end of type H collagen α chain fragments, as has been previously shown for both aggrecan and collagen anti-neoepitope antibodies. There was almost complete lack of recognition by the 234CEQ antibody for peptides having one amino acid either added to or removed from the C-terminus, as evidenced by the failure of these peptides to inhibit binding of the antibody to the immunizing peptide in an inhibition ELISA (Fig. 2). It was also shown, with immunoblotting, that the 234CEQ antibody did not detect intact, uncleaved α chains of purified equine type H collagen (Fig. 3). Together, these results confirm that 234CEQ is a true anti-neoepitope antibody.

The specificity of the 234CEQ antibody for collagenase-cleaved type H collagen was demonstrated by Western blotting of MMP-13 cleaved fragments of types I, H and HI equine collagens. An antibody has previously been generated that is specific for collagenase-cleaved human type H collagen (COL2-3/4Cι_on ), but it has weak cross- reactivity with collagenase-cleaved type H collagen of the horse. By substituting (in the immunizing peptide) the aspartic acid (D) found in the horse, for the glutamic acid (E) found 8 amino acids from the C-terminus of the type H collagen ³A α chain fragment in humans (Fig. 1), we have generated an antibody specific for cleaved type H collagen of the horse. There was no immunostaining by this antibody of either intact or collagenase- cleaved types I and HI collagens (Fig. 4). As the aspartic acid substitution is also found at this position in the cow, mouse and rat type H collagens (Fig. 1), it is likely, although untested at this time, that this antibody will also react with collagenase-generated type H collagen fragments in these species. The 234CEQ antibody is an anti-neoepitope antibody that specifically recognizes the C-termini of α chain fragments of equine type H collagen created by coUagenase digestion.

It has previously been shown that cleaved collagen fragments released from IL-lα stimulated equine articular cartilage can be detected in vitro, using the COL2-3/4C_sh₀rt antibody that detects both collagenase-cleaved types I and H collagens. Using these same media samples, we have now shown that the 234CEQ antibody, in an inhibition ELISA, is able to detect cleaved collagen fragments. In fact, the levels of the COL2-3/4C_Short and 234CEQ are significantly correlated (P < 0.0001), with a Pearson r value that could be improved from 0.377 to 0.526 with the removal of two high outlier values (22.78 and 10.52 nmoles/μg DW) for the COL2-3/4C_short epitope (Fig. 7). Moreover, the absolute levels of the two neoepitopes, expressed as nmoles of peptide per dry weight of cartilage, are very similar, confirming that the major collagenase-generated collagen fragment released from LL-lα stimulated articular cartilage of the horse is a product of type H collagen. This is important given that IL-1 has been shown to inhibit type H and stimulate type I collagen synthesis in cultured human chondrocytes.

To evaluate the 234CEQ antibody, in an ELISA, as a tool for monitoring the effect of novel anti-arthritic agents, the selective MMP inhibitor, BAY 12-9566, was added at different concentrations to the IL-lα stimulated equine articular cartilage explants. BAY 12-9566 was shown to induce dose-dependent and statistically significant inhibition of both the collagen and proteoglycan degradations caused by IL-1 in these cartilages. Using similar in vitro systems in other species, researchers have for some time been monitoring the effects of MMP inhibitors on IL-1 -induced cartilage degradation, and specifically, type H collagen degradation. Now the 234CEQ antibody can be utilized in assessing the efficacy of novel therapeutics in ameliorating cartilage collagen damage in the horse and/or dog.

A previous immunohistochemical study looking at human osteoarthritic cartilage identified a progression in collagen degradation from the superficial to deep zones with increasing severity of cartilage damage. The cartilage removed from the third carpal bone of a horse with advanced osteoarthritis in our study suggested a similar trend with a lack of staining in the superficial layer and intense staining for collagenase-cleaved type H collagen within the middle zone of the matrix. The localization of this staining around chondrocyte clusters implies a high level of coUagenase production and/or coUagenase activation by these clones to account for the marked increases in collagen degradation in these periclonal regions. This was confirmed by the minimal staining for cleaved type H collagen in normal equine articular cartilage that lacks these chondrocyte aggregates. Moreover, the removal of the majority of immunostaining by preincubating the antibody with its immunizing peptide demonstrated the specificity of the 234CEQ antibody for collagenase-cleaved collagen in equine articular cartilage. Articular cartilages from joints with varying degrees of osteoarthritis and other arthritides are being evaluated with the 234CEQ antibody in a similar fashion to further characterize type H collagen degradation as it relates to aging and the stage and type of joint disease in horses and dogs.

In summary, this study has characterized an antibody and an assay utilizing this antibody that can be used to identify and monitor the amount of type H collagen degradation occurring in diseased joints of horses and dogs. We have shown the usefulness of this anti-neoepitope antibody in an in vitro screening system to evaluate novel therapeutics designed to inhibit matrix metalloproteinase activities and specifically those of the mammalian collagenases, MMP-1, MMP-8 and MMP-13. Moreover, our immunohistochemical work using the 234CEQ antibody has allowed us to evaluate and localize type H collagen breakdown at the tissue level, aiding in our understanding of the changes occurring within articular cartilage at the molecular level in equine joint disease and similarly in canine joint disease. We are currently using the 234CEQ antibody in an ELISA to determine the levels of type H collagen fragments bearing the cleavage site neoepitope in the sera, synovial fluids and urines of normal horses and those with various arthritides. Preliminary results from this work indicate that this antibody will be able to identify type H collagen degradation fragments in equine body fluids, thereby introducing a potentially useful biochemical marker for the early detection and for monitoring the progression and response to treatment of joint disease in the horse, as well as other animals, such as canines. While various embodiments of the present invention have been described in detail, it will be apparent that further modifications and adaptations of the invention will occur to those skilled in the art. It is to be expressly understood that such modifications and adaptations are within the spirit and scope of the present invention.

Claims

What is Claimed is:

1. A polyclonal anti-neoepitope antibody that recognizes collagenase- cleaved, type H collagen fragments in a horse or a dog, said antibody able to detect increases in type H collagen cleavage in diseased articular cartilage.

2. The antibody as set forth in Claim 1, comprising 234 CEQ antibody.

3. A method for detecting cartilage degradation in the joints of horses or dogs, comprising providing a polyclonal anti-neoepitope antibody that recognizes collagenase-cleaved, type H collagen fragments and identifying the presence of antibodies bound to articular cartilage wherein the presence of said antibody indicates the presence of cartilage degradation in a j oint.

4. The method as set forth in Claim 3, wherein said polyclonal antibody comprises 234 CEQ antibody.

5. A polyclonal antibody consisting essentially of a 234 CEQ antibody.

6. A polyclonal anti-neoepitope antibody that recognizes collagenase- cleaved, type H collagen fragments in an animal selected from the group consisting of a horse, a dog, a cow, a mouse and a rat, said antibody able to detect increases in type H collagen cleavage in diseased articular cartilage.