ARTHRITIS & RHEUMATISM

Vol. 58, No. 8, August 2008, pp 24202431

DOI 10.1002/art.23654
2008, American College of Rheumatology

Characterization of Metalloprotease Cleavage Products of

Human Articular Cartilage
Eugene Y. Zhen,1 Isabelle J. Brittain,1 Dennis A. Laska,1 Peter G. Mitchell,1 Eren U. Sumer,2
Morten A. Karsdal,2 and Kevin L. Duffin1
identification of novel potential substrates and cleavage
sites for individual enzymes under more physiologically
relevant conditions. Characterization of these cartilage
matrix peptides may help in the development of pharmacodynamic biomarkers of cartilage degradation, and
also may contribute to an understanding of the bioactive
peptides important in chondrocyte signaling.

Objective. To identify, characterize, and compare

proteolysis peptide products generated by metalloprotease digests of human articular cartilage.
Methods. Human articular cartilage was digested
by the addition of exogenous metalloproteases, including matrix metalloproteinases 2, 3, 8, 9, 12, and 13 and
aggrecanases ADAMTS-4 and ADAMTS-5. Proteolyzed
peptide products were identified by proteomics methods
using mass spectrometry.
Results. Complete sequences of the peptides proteolyzed from human articular cartilage, including Nand C-termini and hydroxylated posttranslational modifications, were determined. A wide variety of peptides,
originating from types I, II, and III collagen, biglycan,
prolargin, fibromodulin, fibronectin, decorin, cartilage
oligomeric matrix protein, cartilage intermediate-layer
protein, megakaryocyte-stimulating factor, mimecan,
aggrecan, and lumican, was analyzed following metalloprotease digestion. Release of peptides varied as a
function of time, enzyme specificity, and abundance.
Specific type II collagen peptide biomarkers, including
those containing the three-quarterlength fragment
cleavage site and those containing the domains for
helical peptide of type II collagen and C-telopeptide of
type II collagen, were observed after release by selected
proteases.
Conclusion. The use of intact cartilage instead of
purified protein substrates in the assay allowed for the

Osteoarthritis (OA) and rheumatoid arthritis

(RA) are typified by biologic changes in the articular
cartilage that lead to cartilage degradation and joint
space narrowing (1,2). Articular cartilage is composed of
2 primary matrix proteins, type II collagen and aggrecan,
as well as a number of other matrix proteins that provide
structural functions such as rigidity, flexibility, and compression damping. In addition, the cartilage matrix influences chondrocyte function through integrin receptor
signaling, initiated via ligation of either intact matrix
molecules or peptide fragments resulting from proteolytic activity (3).
In arthritis, an imbalance of chondrocyte matrix
synthesis and degradation leads to a net loss of matrix,
and eventually to joint impairment. In OA, the excessive
degradation is primarily due to the action of proteases
released by chondrocytes, synovial cells, and macrophages (4). Numerous serine proteases and metalloproteases are overexpressed by these cells during the disease process, but specific metalloproteases, including
the collagenase matrix metalloproteinase 13 (MMP-13)
and the aggrecanase ADAMTS-5, are thought to be
especially important for initiating and promoting cartilage matrix degradation in OA (5,6). MMP-13 cleaves
type II collagen at a three-quarterlength fragment
cleavage site in the triple-helical domain, which allows
helix unwinding and makes the protein more susceptible
to additional proteolysis (7,8).
Other metalloproteases are implicated in proteolyzing additional cleavage sites of type II collagen, which

1
Eugene Y. Zhen, PhD, Isabelle J. Brittain, MS, Dennis A.
Laska, BS, Peter G. Mitchell, PhD, Kevin L. Duffin, PhD: Eli Lilly and
Company, Greenfield, Indiana; 2Eren U. Sumer, MSc, Morten A.
Karsdal, MSc, PhD: Nordic Biosciences, Herlev, Denmark.
Dr. Zhen, Ms Brittain, and Mr. Laska own stock or stock
options in Eli Lilly. Drs. Mitchell, Karsdal, and Duffin own stock or
stock options in Eli Lilly and Pfizer.
Address correspondence and reprint requests to Kevin L.
Duffin, PhD, Eli Lilly & Company, 2001 West Main Street, Greenfield,
IN 46140. E-mail: Duffin_K@Lilly.com.
Submitted for publication December 10, 2007; accepted in
revised form April 25, 2008.

2420

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

thereby results in release of fragments with known

epitopes in the domains of helical peptide of type II
collagen (HELIX-II) and C-telopeptide of type II collagen (CTX-II) (9,10). ADAMTS-5 cleaves aggrecan at
5 known sites, which results in release of proteoglycan,
thus reducing the ability of cartilage to withstand compressive forces and allowing the access of additional
proteases to otherwise-inaccessible matrix proteins
(5,11).
The cell signaling that leads to cartilage degradation is mediated by a variety of pathways. Fibronectin,
an extracellular matrix protein, can undergo proteolysis,
yielding fragments that bind to integrin receptors and
initiate signaling pathways that regulate cell growth,
migration, and survival. Injection of fibronectin fragments directly into the synovial cavity leads to proteoglycan depletion and the induction of matrix-degrading
enzymes (12,13). This activity has been attributed to the
activation of the focal adhesionassociated tyrosine kinase pathway (14,15). Chondrocytes themselves express
interleukin-1, which acts in a catabolic role by inducing
cartilage damage through up-regulation of metalloproteases, reduction of proteoglycan synthesis, and stimulation of nitric oxide production (6,16).
Several peptides derived from cartilage have
been proposed as biomarkers of OA or as pharmacodynamic markers of disease-modifying OA drug
(DMOAD) activity. The HELIX-II754764 biomarker
was measured in serum using a polyclonal antibody
based enzyme-linked immunosorbent assay (ELISA),
and its serum levels correlated with increased incidence
of arthritis (10,17). The CTX-II11611166 biomarker, a
cross-linked peptide found near the C-terminus of type
II collagen, was measured in urine using a monoclonal
antibodyformatted ELISA (CartiLaps; Nordic Bioscience Diagnostics, Herlev, Denmark), and its levels
were found to be substantially decreased following metalloprotease inhibition (18,19). A three-quarterlength
peptide fragment of type II collagen was identified in the
urine of patients with OA, and was found to have
increased abundance (20,21). This predominant threequarterlength urine peptide of type II collagen was
later described as a 45-mer peptide, (TIINE)862906, the
levels of which were decreased after metalloprotease
inhibition (22).
Other cartilage fragments, including those originating from cartilage oligomeric matrix protein (COMP)
(23), YKL-40 (24), and fibromodulin (25), have also
been described as potential biomarkers of OA. Cleavage
products of aggrecan that originate from the activities of
ADAMTS-4 or ADAMTS-5 have been measured in

2421

synovial fluid, and these could also serve as cartilage

degradation biomarkers (26). A better understanding of
both the specific biochemical fragments that are being
measured and the pathways of production of these
fragments may increase the utility of the assays used for
detection and/or could help identify additional pharmacodynamic or disease markers.
Disease progression in arthritis is generally evaluated by radiographic imaging to determine joint space
narrowing related to cartilage thinning (27). Magnetic
resonance imaging is an advancing technology that offers the promise of a more sensitive and specific instrument for measuring disease progression over shorter
periods of time (28,29). The progression of cartilage
damage in arthritis is slow, and DMOAD trials may
require several years of study to demonstrate statistically
significant protection, via imaging end points, in patients
with OA. Biochemical biomarkers that can help select
patient populations, demonstrate mechanism proof-ofconcept, and establish appropriate dose levels will be
critical for decision-making in early clinical trials of
novel DMOADs.
In the present study, we measured the effects of
metalloproteases on human articular cartilage. Supernatants from digestions with these enzymes were analyzed
via proteomics methods to determine specific sites of
cleavage of cartilage proteins by different metalloproteases, for better understanding of their selectivities and
to measure the release of peptides that may serve as
useful biomarkers of cartilage degradation. The metalloproteases utilized in this study have been reported to
be overexpressed and active in arthritic cartilage
(11,18,30). Therefore, determination of their substrates
and products may lead to a better understanding of the
potential role of metalloproteases in arthritis, and also
may help in the identification of new biomarkers of
cartilage degradation.
MATERIALS AND METHODS
Materials. Human knee articular cartilage was obtained from the Clarion Tissue Bank (Indianapolis, IN)
through a protocol approved by the Indiana University Institutional Review Board. One cartilage sample originated from
a 64-year-old woman, and the other originated from a 35-yearold woman. Cartilage from the 64-year-old woman showed
pathologic signs of arthritis, whereas that from the 35-year-old
woman appeared normal. The tibial plateau region of the
knees of both subjects was excised without regard to pathologic
features, and samples were flash frozen in liquid nitrogen,
shipped on dry ice, and stored at 80C until dissection.
Human MMP proteases were obtained as full-length
proenzymes from EMD Biosciences, Inc. (San Diego, CA).

2422

APMA, obtained from Calbiochem (San Diego, CA), was used

for activation of the MMP proenzymes. Aggrecanases, including ADAMTS-4 and ADAMTS-5, were expressed and purified
in our laboratory. Both enzymes contained the catalytic domain, but were not full-length. The ADAMTS-4 sequence
contained amino acids 1579 (after cleavage of the prodomain,
the active construct is amino acids 213579, containing a
reprolysin domain, a cysteine-rich domain, and a thrombospondin type 1 repeat). The ADAMTS-5 sequence contained amino acids 1554 (after cleavage of the prodomain, the
active construct is amino acids 262554, containing a reprolysin
domain, ADAM spacer 1 domain, and thrombospondin type 1
repeats). In addition, full-length ADAMTS-4 was obtained
(Chemicon, Temecula, CA) and used as a comparator in this
study. Chemicals for digestion buffers, including 1M Tris
buffer (pH 7.4), NaCl, CaCl2, and zinc acetate, were obtained
from Sigma (St. Louis, MO).
Cartilage digestion. The frozen tibial plateau tissue
was thawed, and cartilage was prepared by dissecting 10
sections of each tissue sample, each weighing 7 mg. Cartilage
was shaved off the bone and cut into small sections with a
scalpel. Each section of cartilage was immediately placed in a
0.5-ml Eppendorf tube, followed by rinsing in 250 l of
digestion buffer for 2 hours with gentle agitation to remove
blood and other contaminants. Rinse buffer was decanted
prior to the digestion reactions.
MMP proenzymes required activation using 1-hour
incubations with 10 mM APMA at 37C. MMPs 2, 3, 8, 9, 12,
and 13 were included in the study. The digestion buffer for
MMP proteolysis contained 100 mM Tris (pH 8.0), 100 mM
NaCl, 10 mM CaCl2, and 2 mM zinc acetate. The digestion
buffer for aggrecanase proteolysis contained 50 mM Tris (pH
7.5), 100 mM NaCl, and 10 mM CaCl2. All buffer solutions
were prepared with Milli-Q purified water (Millipore, Billerica, MA).
An aliquot of 250 l of digestion buffer and 1 g of
enzyme was added to a section of articular cartilage from each
patient. A negative control containing digestion buffer without
the addition of exogenous protease was also included in each
study. Digestions were carried out at 37C with gentle agitation. Supernatants were removed and frozen in individual
tubes at 80C on days 1, 3, 7, 10, 14, 17, and 21. At the first
6 time points, each enzyme (1 g) in fresh digestion buffer
(250 l) was added to the cartilage samples after removing the
supernatants, and incubations were continued at 37C with
gentle agitation. Only the samples from days 1, 7, 14, and 21
were subjected to further analysis by high-performance liquid
chromatography (HPLC)tandem mass spectrometry (MS/
MS) analysis.
Proteomics analysis. A triple quadrupole/linear ion
trap mass spectrometer with a Surveyor HPLC system (ThermoFinnigan, Waltham, MA) was used for peptide identification. Peptides were separated on a ZORBAX 300SB-C18, 1
50mm column (Agilent Technologies, Santa Clara, CA) at a
flow rate of 50 l/minute. Solvent A contained 0.1% formic
acid in water and solvent B contained 0.1% formic acid in 50%
acetonitrile. The total column effluent was connected to the
electrospray interface of the mass spectrometer. The source
was operated in positive ion mode with 4.8 kV of source
potential, a sheath gas flow of 20 arbitrary units, and a capillary

ZHEN ET AL

temperature of 225C. The source lenses were set by maximizing the ion current for the 2 charge state of angiotensin.
Data were collected in the triple play mode, using the
following parameters: a centroid parent scan of 1 microscan
with a maximum injection time of 50 msec, a profile zoom scan
of 3 microscans with a maximum injection time of 500 msec,
and a centroid MS/MS scan of 2 microscans with a maximum
injection time of 2,000 msec. Dynamic exclusion settings were
set to a repeat count of 1, an exclusion list duration of 2
minutes, and rejection widths of 0.75 mass/charge (m/z) and
2.0 m/z. Collisional activation was carried out at a relative
collision energy of 35 eV and a quadrupole window of 3 amu
for the parent ion.
Peptide identification. MS/MS spectra were searched
against protein databases using both Sequest and X! Tandem
database search algorithms, as described by Higgs et al (31).
Only peptides with an X correlation (Xcorr) score of 2.5 and
with molecular weights from 300 amu to 3,500 amu were
selected for identification and quantification. Approximately
9% of the peptides that underwent MS/MS characterization in
this study were positively identified with an Xcorr score of
2.5.

RESULTS
Identification of peptides formed by metalloprotease cleavage of human articular cartilage. Human
articular cartilage was digested by a variety of metalloproteases, including gelatinases (MMPs 2 and 9), collagenases (MMPs 8 and 13), a stromelysin (MMP-3), a
macrophage metalloelastase (MMP-12), and 2 aggrecanases (ADAMTS-4 and ADAMTS-5). The newly released peptides were identified using HPLCMS/MS
methods. The cartilage samples used for digestion were
derived from 2 separate subjects; 1 sample exhibited
pathologic features consistent with OA, and the other
did not exhibit histologic signs of arthritis. The cartilage
samples were frozen immediately after collection, and
were kept frozen at 80C until used in laboratory
analyses.
Small tissue sections were excised from the tibial
plateau region, regardless of pathologic status. The
identities, numbers, and abundances of each peptide
were assessed in the samples from each subject, and
these values were found to be largely consistent across
samples. The numbers and abundances of the peptides
that were generated and positively identified in all of the
MMP digests were significantly increased over the levels
of endogenous peptide in control articular cartilage that
was incubated under identical conditions but without
added proteases.
Table 1 lists the 16 most prominent proteins
identified in articular cartilage in this study, based on
compiled proteolyzed peptide identities generated in all

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

Table 1. Total number of unique peptides identified from each

protein, compiled from all of the metalloprotease digests of articular
cartilage from 2 different subjects*
Protein

Cartilage
sample 1

Cartilage
sample 2

Type II collagen
Biglycan
Prolargin
Type III collagen
Fibromodulin
Clusterin
Type I collagen
Fibronectin
Decorin
COMP
CILP
CILP-2
Megakaryocyte-stimulating factor
Mimecan
Aggrecan
Lumican

110
105
117
46
30
51
19
58
21
18
23
10
6
18
8
5

118
140
133
77
11
53
12
29
14
28
28
12
0
57
0
9

* Values are the total number of unique peptides identified in each

protein (in all matrix metalloproteinase and ADAMTS digests of
articular cartilage from a 64-year-old woman [cartilage sample 1] and
a 35-year-old woman [cartilage sample 2]). COMP cartilage oligomeric matrix protein; CILP cartilage intermediate-layer protein.

of the digests. The number of unique peptides identified

from, and listed for, each protein was based on the
summed total number released by the different metalloproteases at 4 different digest time points over a
21-day period. The most abundant released peptides
originated from collagen proteins, including types I, II,
and III collagen, and from proteoglycans, including

2423

biglycan, prolargin, fibromodulin, decorin, mimecan,

and aggrecan. Additional abundant peptides originated
from megakaryocyte-stimulating factor, fibronectin, cartilage intermediate-layer protein, and COMP. Clusterin
was also one of the prominent proteins identified from
the protease digests and is known to be expressed in
both healthy and OA cartilage (32).
The metalloproteases degraded articular cartilage with varying preferences for specific extracellular
matrix proteins. Specific MMPs, including MMPs 2, 9,
and 13, generated the most numerous and most abundant peptides from proteolysis of various collagens. In
contrast, most of the proteases cleaved proteoglycans. In
fact, MMP-12 generated significantly more peptides
from the proteoglycans than from the collagens (Table
2). Moreover, MMP-12 appeared to be a preferred
protease for cleavage of COMP, fibronectin, decorin,
and fibromodulin.
Type II collagen and biglycan yielded the highest
number of peptides from human articular cartilage
under the conditions used in this study, suggesting that
these 2 proteins are abundant and susceptible to metalloprotease cleavage. Type II collagen was found to be a
substrate for all of the MMP enzymes tested in these
experiments. The numbers of unique type II collagen
peptides generated by each metalloprotease are listed in
Table 2.
The gelatinase MMP-2 was most active in proteolyzing type II collagen; however, all of the MMPs
were capable of generating many different peptide cleav-

Table 2. Number of unique peptides identified using specific matrix metalloproteinase (MMP) and ADAMTS proteinase digests of different
cartilage protein substrates*
Protein

MMP-2

MMP-3

MMP-8

MMP-9

MMP-12

MMP-13

ADAMTS-4

ADAMTS-5

Aggrecan
Biglycan
CILP
CILP-2
COMP
Clusterin
COL1A1
COL1A2
COL2
COL3
Decorin
Fibronectin
Fibromodulin
Lumican
Mimecan
Prolargin
Proteoglycan 4
Total

32
2
2

3
13
4
75
19
4

3
25
1
187

1
20

1
13
1
1
17
5
2
3
9

29

102

23

2
25
8
1
1
5

1
10

80

34

1
11
6
32
35

3
1

14

137

3
30
8
5
11
15
1

18
9
8
14
18
3

26
4
173

12
1

1
1

34
12
1
3
2

10
1
78

2
18
4

5
17

7
1
5
39
15
2
11
50
6
182

2
1
8
4
2
9
1

4
1
2
1
4

41

* Values are the number of unique peptides in each metalloprotease digest. Digestion buffer without the addition of exogenous protease was used
as a negative control. See Table 1 for other definitions.

2424

age products from this substrate. ADAMTS-4 was most

active in proteolyzing fibronectin and prolargin, but
neither of the aggrecanases, ADAMTS-4 or ADAMTS5, was very active in proteolyzing collagen.
All identified peptides originating from type II
collagen were released from the triple-helix region, and
the identified peptides accounted for more than 80% of
the sequence in that region. Many of the identified type
II collagen peptides had a similar sequence, but differed
in length by a few amino acids as a result of cleavage at
multiple nearby sites.
Biglycan peptides were identified in high numbers in the cartilage digests (Table 1), with the identified
peptides covering 60% of the protein sequence. Biglycan is a small proteoglycan that binds to collagen fibrils
that are located at the surface of cartilage and in
pericellular regions (33). In OA cartilage, a higher
concentration of biglycan was reported in the deeper
layers of the tissue (34).
Aggrecan, a very abundant extracellular matrix
component of articular cartilage, yielded only 8 peptide
fragments in this study, and these peptides were generally small, with molecular weights of 2 kd. None of
these 8 peptides resulted from known ADAMTS cleavage sites. Aggrecan is heavily glycosylated, which could
cause interference with peptide detection by MS/MS and
also may prevent the proteolysis of aggrecan to the
extent that it would affect the generation of peptides in
a molecular weight range (0.55 kd) that would be
easily detectable by MS/MS. In addition, glycosylated
peptides would not be identified with the protein identification software used in this study. Peptides resulting
from proteolysis of COMP were generated from the
C-terminal region of the protein.
Although many of the same peptides were generated in the cartilage digests by different MMPs,
unique peptides were also released by specific metalloproteases. For example, a type II collagen peptide with
a sequence of 886EKGEPGDDGPSGAEGPPGPQG906
was formed in high abundance by MMP-2, but not by
any other MMP used in this study (Figure 1A). Another
peptide generated from biglycan, with a sequence of
220
LTGIPKDLPET230, was only produced by digestion
with MMP-12 (Figure 1B). A type II collagen peptide
with a sequence of 834GRVGPPGSNGNPGPPGPPGP853 was released by ADAMTS-5 only. A prolargin
peptide, 250DSNKIETIPNGY261, was uniquely generated only by the 2 aggrecanases, ADAMTS-4 and
ADAMTS-5, and not by any of the other metalloproteases used in this study. These and other peptides may

ZHEN ET AL

Figure 1. Comparison of the relative abundances of the type II

collagen peptide EKGEPGDDGPSGAEGPPGPQG (A) and the biglycan peptide LTGIPKDLPET (B), generated by metalloprotease
digestion of articular cartilage on days 1, 7, 14, and 21. Supernatants
were analyzed by high-performance liquid chromatographytandem
mass spectrometry to generate the identities and relative responses of
individual peptides. Matrix metalloproteinases (M) 2 and 12 were
especially active in generation of the 2 peptides at all time points.
Nec negative control containing digestion buffer without the
addition of exogenous protease.

serve as useful pharmacodynamic biomarkers of specific

metalloprotease activity in articular cartilage.
Selected release of known biomarker epitopes as
peptides from metalloprotease digests. Previous reports
described a 45-mer peptide that resulted from cleavage
by collagenases at the three-quarterlength fragment
site of type II collagen, which was detected in human
and animal urine. This peptide was suggested to be a
biomarker of cartilage degradation in humans (22). The
results from the present study indicated that this peptide
was robustly formed from digestion of human articular
cartilage by the collagenases MMP-8 and MMP-13 at all
4 time points, and was also robustly generated by the
gelatinase MMP-2 across the entire 21-day digestion
period (Figure 2).
Significant levels of the 45-mer peptide were also
generated by MMP-3 and MMP-12 in the day 1 digests,
but not at later time points, even though the cartilage
discs remained largely intact. Similarly, digestion with
MMP-9 yielded decreased levels of this peptide over the
21-day digestion timeframe. Digestion with ADAMTS-4
and that with ADAMTS-5 did not yield the 45-mer
peptide at abundances significantly higher than the low
basal levels observed in the negative control samples.
Another type II collagen fragment, CTX-II,
which originates from the C-terminal telopeptide region
and contains 1161EKGPDP1166 as the recognition domain, has been described as a biomarker of cartilage
degradation. Monoclonal antibodies to this fragment
have aided in the development of a specific ELISA
(CartiLaps) for its detection in biologic samples (35). In
the present study, a peptide that contained the CTX-II

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

Figure 2. Comparison of the relative abundances of the type II

collagen 45-mer peptide TIINE, released by metalloprotease digestion
of articular cartilage over a 21-day time period. Supernatants were
collected on days 1 (1D), 7, 14, and 21, and then fresh buffer and
proteases were added for digestion during the next time period.
Supernatants were analyzed by high-performance liquid
chromatographytandem mass spectrometry to generate the identities
and relative responses of individual peptides. Matrix metalloproteinases (MMPs) 2, 8, and 13 were especially active in generating the type
II collagen 45-mer peptide at all time points. The aggrecanases
ADAMTS-4 (AD-4) and ADAMTS-5 did not generate this peptide in
abundances higher than control levels. Nec negative control containing digestion buffer without the addition of exogenous protease.

sequence was only observed by HPLCMS/MS as part of

a larger peptide, 1154FAGLGPREKGPDPLQY1169, in
MMP-2 digests of OA cartilage.
Because the peptide sequence . . .EKGPDP1166,
which is thought to be the substrate for the CartiLaps
assay, was either too polar or too low in molecular
weight to be identified by HPLCMS/MS in this study,
digest samples were analyzed by the CartiLaps ELISA.
Assay results on proteolytic digests of cartilage tissue
from 2 separate subjects showed that all MMP proteases
used in this study released the CTX-II epitope (Figure
3). MMP-2 was especially active in its generation. Neither of the aggrecanases, ADAMTS-4 or ADAMTS-5,
generated the epitope in detectable quantities (36,37).
Peptide fragments containing the HELIX-II domain epitope of type II collagen, 754ERGETGPP(OH)GTS764, have been described as potential biomarkers
of cartilage degradation. The HELIX-II epitope is located in the mid-section of the triple-helical region of
type II collagen, and was observed to be released at
increased levels in synovial fluid from patients with OA
and patients with early RA (10). In the present study,
digestion by MMPs 2, 8, 9, and 13 released a peptide
containing this epitope from articular cartilage.
Degraded proteins other than type II collagen
have also been advanced as biomarkers of arthritis.

2425

COMP, a member of the thrombospondin family of

proteins, has been shown to be increased in the serum of
patients with OA and those with RA (38,39). In this
study, MMP-12 was particularly active with respect to
generating COMP-derived peptide fragments. In comparison, other MMPs were less active at degrading
COMP.
Peptide fragments of biglycan, another extracellular matrix protein, have recently been cited as potential targets for cleavage by aggrecanases in OA (40).
Although biglycan is structurally unrelated to aggrecan,
it was shown to act as a substrate for ADAMTS-4 and
ADAMTS-5. In the present study, many biglycan peptides were generated by the aggrecanases and were also
generated by MMPs 2, 9, 12, and 13.
Kinetics of peptide formation by metalloprotease
cleavage of human articular cartilage. Many identical
(or similar in sequence) peptides, particularly from type
II collagen, were generated in cartilage digests by different metalloproteases; the 20 most abundant peptides
released by each of the MMPs and aggrecanases are
listed in Table 3. These peptides were produced at
multiple time points of cartilage digestion; however, the
time point of highest concentration was found to vary
depending on the particular enzyme. Many of the abun-

Figure 3. Results of the CartiLaps assay for C-telopeptide of type II

collagen (CTX-II) in proteolyzed human cartilage. Cartilage sections
were incubated with exogenous metalloproteases for 21 days and
analyzed for CTX-II by enzyme-linked immunosorbent assay. Fresh
protease was added to the incubated samples on days 4, 7, 10, 14, and
17. Supernatants were frozen and shipped on dry ice for analysis. The
data indicate that the matrix metalloproteinases (MMPs) were active
in generating the CTX-II epitope, but the aggrecanases ADAMTS-4
(AMTS-4) and ADAMTS-5 were not active. NEC negative control
containing digestion buffer without the addition of exogenous protease.

2426

ZHEN ET AL

Table 3. The 20 most abundant peptide fragments released by each metalloprotease and identified by high-performance liquid chromatography
tandem mass spectrometry in articular cartilage digests*
Peptide rank by abundance
Protein
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
Biglycan
CILP-1
CILP-1
CILP-1
CILP-1
CILP-1
CILP-1
CILP-2
CILP-2
CILP-2
CILP-2
Clusterin
Clusterin
Clusterin
Clusterin
Clusterin
COMP
Decorin
Decorin
Fibromodulin
Fibromodulin
Fibromodulin
Fibromodulin
Fibromodulin
Fibronectin
Fibronectin
Fibronectin
Fibronectin
Fibronectin
Mimecan
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin
Prolargin

Sequence
A.KLTGIPKDLPETLNELH.L
E.LHLDNNKLARVPSG.L
E.NSGFEPGAFDGLKLN.Y
E.NSGFEPGAFDGLKLNY.L
E.NSGFEPGAFDGLKLNYLRISEAK.L
G.LKSVPKEISPDTTLLDLQNNDISE.L
K.LTGIPKDLPET.L
K.LTGIPKDLPETLNELH.L
K.LTGIPKDLPETLNELHLDHNKIQAIE.L
K.RAYYNGISLFNNPVP.Y
K.SVPKEISPDTTLLDLQNNDISE.L
L.KSVPKEISPDTTLLDLQNNDISE.L
Y.LRISEAKLTGIPKDLPET.L
E.VVGEDPMAELEIPSRS.F
V.VGEDPMAELEIPSRS.F
W.SLNPDTGLWEEEGDFKFE.N
W.SLNPDTGLWEEEGDFKFEN.Q
W.SLNPDTGLWEEEGDFKFENQ.R
W.SLNPDTGLWEEEGDFKFENQRRN.K
A.TLGGEELEPAPSLPRPLPA.T
L.GGEELEPAPSLPRPLPA.T
L.GGEELEPAPSLPRPLPATV.G
T.ATLGGEELEPAPSLPRPLPA.T
D.IHFHSPAFQHPPTE.F
E.TVAEKALQE.Y
G.DQTVSDNELQEMSNQGSKYVN.K
R.RPHFFFPKSRIV.R
T.LIEKTNEERKTLLSN.L
D.FRAFQTVVLDPEGDAQIDPNWV.V
K.VPKDLPPDTTL.L
M.IVIELGTNPLK.S
E.HNNVYTVPDSY.F
E.LHLDHNQISRVPNN.A
N.LYLQGNRINEFS.I
Q.LQKIPPVNT.N
Q.LQKIPPVNTNLENLY.L
E.VFITETPSQPNSHP.I
E.VFITETPSQPNSHPIO.W
F.VTHPGYDTGNGIQLPGTSGQQP.S
H.LEANPDTGVLTVSWERSTTPDITGYR.I
P.SSSGPVEVFITETPSQPNSHPIQ.W
Y.LDHNALESVPLNLPE.S
A.FHDFSSDLENVPHLRY.L
D.LQHNRLSDGVFKPDT.F
D.LQHNRLSDGVFKPDTFHGLKN.L
D.SNKIETIPNGY.F
E.KNQLEEVPSALPRNLEQL.R
F.HDFSSDLENVPHLR.Y
H.DFSSDLENVPH.L
H.DFSSDLENVPHLRYL.R
I.DQRVLEKLPGLV.F
K.NQLEEVPSALPRNLEQL.R
L.DGNYLKPPIPLDLM.M
L.DLQHNRLSDGVFKPDT.F
L.DSNKIETIPNGY.F
L.SHNRISSVPAINNRLEH.L
L.YMEKNQLEEVPSALPRN.L
L.YMEKNQLEEVPSALPRNLEQ.L
M.EKNQLEEVPSALPRNLEQL.R
N.HISRIPPGVFSKLEN.L
N.LEQLRLSQN.H

M2

M3

M8

M9

M12

M13

AD4

AD5

20
14
19
10
5
10

19
2

17
20
13
5
12
12
3
5
9
10
18
19
7
6
13
14
9

7
8
16
18
11
3
2
5
4

15

14
18
20
13

18
8

14
9

15
4
17
10

13
9
15
15
7
12
5
8
8
7
2
19
14
6
1
17
14
19
17
16
11

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

2427

Table 3. (Contd)
Peptide rank by abundance
Protein
Prolargin
Prolargin
Prolargin
Tenascin
Tenascin
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL2
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3
COL3

Sequence
R.VLEKLPGLVFLYME.K
S.FPNLAFIRL.N
Y.MEKNQLEEVPSALPRNLEQL.R
C.SVDLESASGEKDLAPPSEPSESFQE.H
S.ASGEKDLAPPSEPSESFQE.H
P.GARGLTGRPGDAGPQGKVGPSGAPGEDGRPGPPGPQG.A
A.GRVGPPGSNGNPGPPGPPGP.S
A.GRVGPPGSNGNPGPPGPPGPS.G
D.AGAPGPQGPSGAPGPQGPTG.V
G.ARGDSGPPGRAGEPGLQGPAGPPGEKGEPGDDGPSG.A
G.ARGDSGPPGRAGEPGLOGPAGPPGEKGEPGDDGPSGAEGPPGPQG.L
G.ARGNDGOPGPAGPPGPYGPAGGPGFPGAPGAKGEAGPTG.A
G.DQGASGPAGPSGPRGPPGPVGPSG.K
G.ERGAPGNRGFPGQDGLAGPKGAPGERGPSG.L
G.FPGPRGPPGPQGATGPLGPK.G
G.IVGLPGQRGERGFPGLPGPSGEPGK.Q
G.KVGPSGAPGEDGRPGPPGPQG.A
G.LPGKDGETGAAGPPGPAGPAG.E
G.LPGTPGTDGPKGASGPAGPPGAQGPPG.L
G.LQGLPGPPGPSGDQGASGPAGPSGPRGPPGPVGPSG.K
G.LQGPRGLPGTPGTDGPKGASGPAGPPGAQGPPG.L
G.LRGLPGKDGETGAAGPPGPAGPAG.E
G.LRGLPGKDGETGAAGPPGPAGPAGERGEQGAPGPSG.F
G.LTGPAGEPGREGSPGADGPPGRDGAAG.V
G.LTGRPGDAGPQGKVGPSGAPGEDGRPGPPGPQG.A
G.NRGFPGQDGLAGPKGAPGERGPSG.L
G.RAGEPGLQGPAGPPG.E
G.RAGEPGLQGPAGPPGEKGEPGDDGPSGAEGPPGPQG.L
G.RSGETGPAGPPGNPGPPGPPGPPGPGIDMSA.F
G.SAGAPGIAGAPGFPGPRGPPGPQGATGPLGPK.G
G.VKGDRGETGAVGAPGAPGPPGSPGP.A
G.VKGDRGETGAVGAPGAPGPPGSPGPAGPTG.K
K.QGAPGASGDRGPPGPVGPPG.L
V.MQGPMGPMGPRGPPGPAGAPGPQG.F
G.ATGFPGAAGRVGPPG.S
A.IGSPGPAGPRGPVGPSGPPG.K
G.AIGSPGPAGPRGPVGPSGPPG.K
G.DKGEPGGPGADGVPGKDGPRGPTGPIGPPGPAG.Q
G.IAGITGARGLAGPPGMPGPRGSPGPQG.V
G.KNGETGPQGPPGPTGPGGDKGDTGPPGPQG.L
G.LRGGAGPPGPEGGKGAAGPPGPPG.A
G.QQGAIGSPGPAGPRGPVGPSGPPG.K
G.VAGPPGGSGPAGPPGPQG.V
G.YQGPPGEPGQAGPSGPPGPPG.A
K.GDPGPPGIPGRNGDPGIPGQPG.S
Q.GHAGAQGPPGPPGIN.G
A.IGSPGPAGPRGPVGPSGPPG.K
G.AIGSPGPAGPRGPVGPSGPPG.K
G.DKGEPGGPGADGVPGKDGPRGPTGPIGPPGPAG.Q
G.IAGITGARGLAGPPGMPGPRGSPGPQG.V
G.KNGETGPQGPPGPTGPGGDKGDTGPPGPQG.L
G.LRGGAGPPGPEGGKGAAGPPGPPG.A
G.QQGAIGSPGPAGPRGPVGPSGPPG.K
G.VAGPPGGSGPAGPPGPQG.V
G.YQGPPGEPGQAGPSGPPGPPG.A
K.GDPGPPGIPGRNGDPGIPGQPG.S
Q.GHAGAQGPPGPPGIN.G

M2 M3 M8 M9 M12 M13 AD4 AD5

16
10
11
11
1
15
12
15
19
10
8
9
5
11

1
13

17

17
1

6
6
20
10

20

13

19

9
10
11
16

4
6

20

8
9

3
6
4
2
12
18

14
2

19

5
9
8

1
2
6
15
18
20
17

4
2
6

16
16

18
4

4
12
7
3
16

18
7
11
13

15
12
3
4
16

13
1

17

7
3
11
14
12
2
20
16
13
3
4
16

12
17

7
3
11
14
12
2
20

* The peptides were ranked 120 according to their relative abundance in each digest. M matrix metalloproteinase; AD ADAMTS (see Table
1 for other definitions).
This peptide sequence is the same in type I collagen and type II collagen.

2428

dant peptides from type II collagen identified in MMP-2

digests were found to be most abundant at the longest
(21-day) time point (results not shown). In comparison,
similar peptides from digests with MMPs 3, 9, and 12
were most abundant in the first day of the digest. In
digestions with MMPs 8 and 13, the type II collagen
peptides were most abundant in the 7-day or 14-day
digests.
These findings suggest that MMPs 3, 9, and 12
may be acting on type II collagen molecules that are
already partially degraded, or are denatured in some
way. Once this source of substrate is depleted, then their
ability to degrade native helical collagen would be
limited. In contrast, the collagenases (MMPs 8 and 13)
will efficiently cleave native type II collagen in the
helical domain. The resulting denatured type II collagen
can then undergo further proteolysis by these proteases
into smaller peptides.
The timeframe of peptide generation was also
dependent on the cartilage protein substrate. For example, in MMP-3 digests, peptides from type II collagen
and biglycan dominated the list of most abundant peptides. The type II collagen peptides were most abundant
in the day 1 digests, whereas the biglycan peptides were
most abundant in the day 7 digests. In contrast, with
MMP-12 digestion, all peptides from type II collagen
and proteoglycans were most abundant in the day 1
digests.
Prolargin and biglycan cleavage products were
the most abundant peptides observed from ADAMTS-4
digestion of articular cartilage, and were observed in
highest abundance in the day 14 digests. One possible
contributor to some of these observations is the potential for the exogenously added enzyme to activate endogenous proenzymes that are already present in the
matrix. In vitro it has been demonstrated that several
MMPs are capable of activating other MMPs, leading to
cascades of enzyme activation. For example, MMP-3 is
an efficient activator of MMP-1 and MMP-13 (4143).
When MMP-3 is added to cartilage, it is possible that an
endogenous procollagenase is being activated and that
this leads to the initial high release of type II collagen
fragments under these conditions.
Posttranslational modifications of extracellular
matrix proteins present in human articular cartilage.
Two predominant posttranslational modifications of
type II collagen were observed in this study. Almost all
peptides originating from this protein contained hydroxylated proline and/or lysine residues. In fact,
searches of protein databases returned no collagen
peptide matches, except when hydroxyproline and hy-

ZHEN ET AL

droxylysine were included in the search criteria. No

other prominent posttranslational modifications of extracellular proteins were observed with the methods
used in this study.
DISCUSSION
Because arthritic degradation of articular cartilage is characterized by secreted metalloprotease proteolysis of extracellular matrix protein components of the
cartilage, we initiated a study to characterize the proteolyzed peptide products of cartilage after the addition of
exogenous metalloproteases that are overexpressed during disease progression. The proteomics methods used
in this study allowed identification of many of the
abundant peptide products generated from these digestions. Previous cartilage degradation studies have measured peptide products from specific matrix proteins
(44,45) or have used global proteomics approaches with
trypsin digestion (46) to measure cartilage proteins
present in intact cartilage. This study used global proteomics methods to identify peptide products released
by specific metalloproteases from all proteins present in
human articular cartilage.
A metalloprotease concentration of 4 g/ml was
used for all digestions in this study. Endogenous levels of
metalloproteases measured in synovial fluid vary widely
across patients, with levels lower than the ng/ml range up
to hundreds of g/ml, depending on the metalloprotease
(4749). MMP-3 levels were found at the g/ml range in
synovial fluid of OA patients, whereas MMP-13 levels
were found to be much lower or undetectable. We chose
to add 4 g/ml protease to cartilage digests to ensure
adequate peptide generation in the timeframe of the
study and to distinguish the peptides from those generated by any residual endogenous proteases remaining in
the cartilage.
The profile of peptides produced from the 2
available cartilage samples was generally consistent.
Some qualitative differences (i.e., identified peptides)
were noted between the 2 cartilage sources; however,
these were not evaluated further because of the limited
number of cartilage samples that were available for
analysis. Cartilage samples were only frozen once, at the
time of tissue collection, and thawed once, at the time of
tissue dissection, before their use in digestion experiments. MMP-2 was the most active protease used in the
study, and the aggrecanases were the least active in
generating peptides from cartilage. The MMPs were
especially active in cleaving type II collagen, the most
abundant protein component of articular cartilage. The

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

aggrecanases showed preference for cleaving

proteoglycan-containing proteins. However, all of the
proteases cleaved numerous different protein components of cartilage.
Four time points of digestion were selected for
peptide identification and quantification. In general, the
same peptides were produced at all 4 time points.
However, in some cases, peptides that were produced
abundantly at early digest time points were not generated or were generated less abundantly at later time
points. For example, the type II collagen 45-mer peptide
formed by cleavage at the three-quarterlength site was
abundantly generated at all 4 time points by MMPs 2, 8,
and 13, but was decreased in abundance at later time
points of digestion with MMPs 3, 9, and 12 (Figure 2).
Early digest time points may include peptides released
from partially cleaved proteins present in the original
cartilage samples or peptides released due to the activation of endogenous proMMPs. Unexpectedly, it is generally believed that MMP-13 and other collagenases are
responsible for cleavage at the three-quarterlength
fragment site of type II collagen; the gelatinase MMP-2
generated the 45-mer peptide in high abundances at all
4 time points (Figure 2). There has been a previous
report of MMP-2 acting as a true collagenase (50), and
our results are consistent with this prior observation.
A major proportion of the type II collagen peptides released from articular cartilage by the metalloproteases originated from the triple-helical region of the
molecule. Furthermore, 3 previously identified type II
collagen peptide domains described as potential biomarkers of OA were generated from the digestions: the
45-mer peptide released by MMPs 2, 8, and 13, as
described above, the previously reported HELIX-II sequence within the triple-helical domain, which was specifically generated by cleavage with MMPs 2, 8, 9, and
13, and the CTX-II telopeptide, located near the
C-terminus of type II collagen, which was abundantly
formed by cleavage with MMP-2 and, to a lesser extent,
by all of the MMPs used in this study.
Hydroxylation of proline and lysine residues is a
common feature of collagen and provides sites for
protein crosslinking. All collagen peptides identified in
this study were posttranslationally modified in this manner. However, crosslinked collagen was not identified
using the methods in this study. N- and O-glycosylation
is another common posttranslational modification of
cartilage matrix proteins, neither of which was detected
with the MS/MS methods used in this study.
Our study has identified many of the most abundant metalloprotease cleavage products of human artic-

2429

ular cartilage. The list of peptides generated by these

digestions provides insights into the broad number of
substrates susceptible to specific metalloprotease proteolysis in diseased cartilage. Many of these peptides are
potential biomarkers of arthritis or of metalloprotease
activity in articular cartilage. Further evaluation of many
of these peptides as biomarkers might be achieved by
quantitation in synovial fluid, blood, or urine.
ACKNOWLEDGMENTS
The authors would like to thank George Sandusky and
Leah Helvering for their help in procuring and dissecting the
articular cartilage samples used in this study, and Rick Higgs
for helping with data analysis.
AUTHOR CONTRIBUTIONS
Dr. Duffin had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy of the
data analysis.
Study design. Laska, Duffin.
Acquisition of data. Zhen, Brittain, Laska, Sumer, Karsdal.
Analysis and interpretation of data. Zhen, Brittain, Laska, Mitchell,
Sumer, Karsdal, Duffin.
Manuscript preparation. Zhen, Brittain, Laska, Mitchell, Sumer,
Karsdal, Duffin.
Statistical analysis. Zhen.

REFERENCES
1. Dieppe PA, Lohmander LS. Pathogenesis and management of
pain in osteoarthritis. Lancet 2005;365:96573.
2. Poole AR, Alini M, Hollander AH. Cellular biology of cartilage
degradation. In: Henderson B, Edwards JC, Pettipher ER, editors.
Mechanisms and models in rheumatoid arthritis. London: Academic Press; 1995. p. 163204.
3. Kempson GE, Muir H, Pollard C, Tuke M. The tensile properties
of the cartilage of human femoral condyles related to the content
of collagen and glycosaminoglycans. Biochim Biophys Acta 1973;
297:45672.
4. Malemud CJ, Islam N, Haqqi TM. Pathophysiological mechanisms
in osteoarthritis lead to novel therapeutic strategies. Cells Tissues
Organs 2003;174:3448.
5. Nagase H, Kashiwagi M. Aggrecanases and cartilage matrix degradation. Arthritis Res Ther 2003;5:94103.
6. Reboul P, Pelletier JP, Tardif G, Cloutier JM, Martel-Pelletier J.
The new collagenase, collagenase-3, is expressed and synthesized
by human chondrocytes but not by synoviocytes: a role in osteoarthritis. J Clin Invest 1996;97:20119.
7. Mitchell PG, Magna HA, Reeves LM, Lopresti-Morrow LL,
Yocum SA, Rosner PJ, et al. Cloning, expression, and type II
collagenolytic activity of matrix metalloproteinase-13 from human
osteoarthritic cartilage. J Clin Invest 1996;97:7618.
8. Billinghurst RC, Dahlberg L, Ionescu M, Reiner A, Bourne R,
Rorabeck C, et al. Enhanced cleavage of type II collagen by
collagenases in osteoarthritic articular cartilage. J Clin Invest
1997;99:153445.
9. Birmingham JD, Vilim V, Kraus VB. Collagen biomarkers for
arthritis applications. Biomarker Insights 2006;2:6176.
10. Charni N, Juillet F, Garnero P. Urinary type II collagen helical
peptide (HELIX-II) as a new biochemical marker of cartilage

2430

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.
21.

22.

23.

24.
25.

26.

27.

degradation in patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum 2005;52:108190.
Struglics A, Larsson S, Pratta MA, Kumar S, Lark MW, Lohmander LS. Human osteoarthritis synovial fluid and joint cartilage
contain both aggrecanase- and matrix metalloproteinase-generated aggrecan fragments. Osteoarthritis Cartilage 2006;14:10113.
Homandberg GA, Meyers R, Xie DL. Fibronectin fragments cause
chondrolysis of bovine articular cartilage slices in culture. J Biol
Chem 1992;267:3597604.
Xie DL, Hui F, Meyers R, Homandberg GA. Cartilage chondrolysis by fibronectin fragments is associated with release of several
proteinases: stromelysin plays a major role in chondrolysis. Arch
Biochem Biophys 1994;311:20512.
Gemba T, Valbracht J, Alsalameh S, Lotz M. Focal adhesion
kinase and mitogen-activated protein kinases are involved in
chondrocyte activation by the 29-kDa amino-terminal fibronectin
fragment. J Biol Chem 2002;277:90711.
Midwood KS, Mao Y, Hsia HC, Valenick LV, Schwarzbauer JE.
Modulation of cell-fibronectin matrix interactions during tissue
repair. J Investig Dermatol Symp Proc 2006;11:738.
Taskiran D, Stefanovic-Racic M, Georgescu H, Evans C. Nitric
oxide mediates suppression of cartilage proteoglycan synthesis by
interleukin-1. Biochem Biophys Res Commun 1994;200:1428.
Kong SY, Stabler TV, Criscione LG, Elliott AL, Jordan JM, Kraus
VB. Diurnal variation of serum and urine biomarkers in patients
with radiographic knee osteoarthritis. Arthritis Rheum 2006;54:
2496504.
Manicourt DH, Bevilacqua M, Righini V, Famaey JP, Devogelaer
JP. Comparative effect of nimesulide and ibuprofen on the urinary
levels of collagen type II C-telopeptide degradation products and
on the serum levels of hyaluronan and matrix metalloproteinases-3
and -13 in patients with flare-up of osteoarthritis. Drugs R D
2005;6:26171.
De Ceuninck F, Sabatini M, Renoux V, de Nanteuil G, Pastoureau
P. Urinary collagen type II C-telopeptide fragments are sensitive
markers of matrix metalloproteinase-dependent cartilage degradation in rat adjuvant-induced arthritis. J Rheumatol 2003;30:
15614.
Downs JT, Lane CL, Nestor NB, McLellan TJ, Kelly MA, Karam
GA, et al. Analysis of collagenase-cleavage of type II collagen
using a neoepitope ELISA. J Immunol Methods 2001;247:2534.
Woodworth TG, Otterness IG, Johnson K, Pickering E, Saltarelli
MJ. Urinary type II collagen neoepitope (uTIINE) in osteoarthritis (OA) patients is associated with disease severity [abstract].
Arthritis Rheum 1999;42 Suppl 9:S258.
Welsch DJ, Duffin KL, Nemirovskiy OV, Dufield DR, Sunyer T,
Howard CP, inventors; Pharmacia Corporation, assignee. Peptide
biomarker and method of identification. US patent 20030049715.
2003 Mar 13.
Di Cesare PE, Carlson CS, Stolerman ES, Hauser N, Tulli H,
Paulsson M. Increased degradation and altered tissue distribution
of cartilage oligomeric matrix protein in human rheumatoid and
osteoarthritic cartilage. J Orthop Res 1996;14:94655.
Register TC, Carlson CS, Adams MR. Serum YKL-40 is associated with osteoarthritis and atherosclerosis in nonhuman primates.
Clin Chem 2001;47:215961.
Heathfield TF, Onnerfjord P, Dahlberg L, Heinegard D. Cleavage
of fibromodulin in cartilage explants involves removal of the
N-terminal tyrosine sulfate-rich region by proteolysis at a site that
is sensitive to matrix metalloproteinase-13. J Biol Chem 2004;279:
628695.
Chockalingam PS, Zeng W, Morris EA, Flannery CR. Release of
hyaluronan and hyaladherins (aggrecan G1 domain and link
proteins) from articular cartilage exposed to ADAMTS-4 (aggrecanase 1) or ADAMTS-5 (aggrecanase 2). Arthritis Rheum 2004;
50:283948.
Bohndorf K, Schalm J. Diagnostic radiography in rheumatoid

ZHEN ET AL

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.
40.
41.
42.

43.
44.
45.
46.

arthritis: benefits and limitations. Baillieres Clin Rheumatol 1996;

10:399407.
Conaghan PG, Felson D, Gold G, Lohmander S, Totterman S,
Altman R. MRI and non-cartilaginous structures in knee osteoarthritis. Osteoarthritis Cartilage 2006;14 Suppl A:A8794.
Hoving JL, Buchbinder R, Hall S, Lawler G, Coombs P, McNealy
S, et al. A comparison of magnetic resonance imaging, sonography,
and radiography of the hand in patients with early rheumatoid
arthritis. J Rheumatol 2004;31:66375.
Tchetverikov I, Lohmander LS, Verzijl N, Huizinga TW, TeKoppele JM, Hanemaaijer R, et al. MMP protein and activity levels in
synovial fluid from patients with joint injury, inflammatory arthritis, and osteoarthritis. Ann Rheum Dis 2005;64:6948.
Higgs RE, Knierman MD, Gelfanova V, Butler JP, Hale JE.
Comprehensive label-free method for the relative quantification of
proteins from biological samples. J Proteome Res 2005;4:144250.
Connor JR, Kumar S, Sathe G, Mooney J, OBrien SP, Mui P, et
al. Clusterin expression in adult human normal and osteoarthritic
articular cartilage. Osteoarthritis Cartilage 2001;9:72737.
Wiberg C, Klatt AR, Wagener R, Paulsson M, Bateman JF,
Heinegard D, et al. Complexes of matrilin-1 and biglycan or
decorin connect collagen VI microfibrils to both collagen II and
aggrecan. J Biol Chem 2003;278:37698704.
Poole AR, Rosenberg LC, Reiner A, Ionescu M, Bogoch E,
Roughley PJ. Contents and distributions of the proteoglycans
decorin and biglycan in normal and osteoarthritic human articular
cartilage. J Orthop Res 1996;14:6819.
Christgau S, Garnero P, Fledelius C, Moniz C, Ensig M, Gineyts E,
et al. Collagen type II C-telopeptide fragments as an index of
cartilage degradation. Bone 2001;29:20915.
Oestergaard S, Chouinard L, Doyle N, Karsdal MA, Smith SY,
Qvist P, et al. The utility of measuring C-terminal telopeptides of
collagen type II (CTX-II) in serum and synovial fluid samples for
estimation of articular cartilage status in experimental models of
destructive joint diseases. Osteoarthritis Cartilage 2006;14:6709.
Sondergaard BC, Henriksen K, Wulf H, Oestergaard S, Schurigt
U, Brauer R, et al. Relative contribution of matrix metalloprotease
and cysteine protease activities to cytokine-stimulated articular
cartilage degradation. Osteoarthritis Cartilage 2006;14:73848.
Clark AG, Jordan JM, Vilim V, Renner JB, Dragomir AD, Luta
G, et al. Serum cartilage oligomeric matrix protein reflects osteoarthritis presence and severity: the Johnston County Osteoarthritis
Project. Arthritis Rheum 1999;42:235664.
Wislowska M, Jablonska B. Serum cartilage oligomeric matrix
protein (COMP) in rheumatoid arthritis and knee osteoarthritis.
Clin Rheumatol 2005;24:27884.
Melching LI, Fisher WD, Lee ER, Mort JS, Roughley PJ. The
cleavage of biglycan by aggrecanases. Osteoarthritis Cartilage
2006;14:114754.
Murphy G, Cockett MI, Stephens PE, Smith BJ, Docherty AJ.
Stromelysin is an activator of procollagenase: a study with natural
and recombinant enzymes. Biochem J 1987;248:2658.
Suzuki K, Enghild JJ, Morodomi T, Salvesen G, Nagase H.
Mechanisms of activation of tissue procollagenase by matrix
metalloproteinase 3 (stromelysin). Biochemistry 1990;29:
1026170.
Knauper V, Lopez-Otin C, Smith B, Knight G, Murphy G.
Biochemical characterization of human collagenase-3. J Biol
Chem 1996;271:154450.
Hughes C, Murphy G, Hardingham TE. Metalloproteinase digestion of cartilage proteoglycan: pattern of cleavage by stromelysin
and susceptibility to collagenase. Biochem J 1991;279(Pt 3):7339.
Smith GN Jr. The role of collagenolytic matrix metalloproteinases
in the loss of articular cartilage in osteoarthritis. Front Biosci
2006;11:308195.
Vincourt JB, Lionneton F, Kratassiouk G, Guillemin F, Netter P,
Mainard D, et al. Establishment of a reliable method for direct

METALLOPROTEASE CLEAVAGE PRODUCTS OF HUMAN ARTICULAR CARTILAGE

proteome characterization of human articular cartilage. Mol Cell

Proteomics 2006;5:198495.
47. Roy-Beaudry M, Martel-Pelletier J, Pelletier JP, MBarek KN,
Christgau S, Shipkolye F, et al. Endothelin 1 promotes osteoarthritic
cartilage degradation via matrix metalloprotease 1 and matrix metalloprotease 13 induction. Arthritis Rheum 2003;48:285564.
48. Tchetverikov I, Ronday HK, Van El B, Kiers GH, Verzijl N,
TeKoppele JM, et al. MMP profile in paired serum and synovial
fluid samples of patients with rheumatoid arthritis. Ann Rheum
Dis 2004;63:8813.

2431

49. Ishiguro N, Ito T, Ito H, Iwata H, Jugessur H, Ionescu M, et al.

Relationship of matrix metalloproteinases and their inhibitors to
cartilage proteoglycan and collagen turnover: analyses of synovial
fluid from patients with osteoarthritis. Arthritis Rheum 1999;42:
12936.
50. Aimes RT, Quigley JP. Matrix metalloproteinase-2 is an interstitial collagenase: inhibitor-free enzyme catalyzes the cleavage of
collagen fibrils and soluble native type I collagen generating the
specific 3/4- and 1/4-length fragments. J Biol Chem 1995;270:
58726.

DOI 10.1002/art.23727

Clinical Images: Synovial chondromatosis of the temporomandibular joint

The patient, a 49-year-old woman, had experienced right-sided earache and right temporomandibular joint (TMJ) pain for 2 years,
with occasional popping of the right TMJ upon mouth opening. Magnetic resonance imaging (A) revealed numerous low-signal
foci within the TMJ (arrows) (sagittal T2-weighted image [T temporal eminence; M mandibular condyle]), consistent with
synovial chondromatosis. The patient underwent arthrotomy and surgical excision of the intraarticular bodies (B), and the diagnosis
was confirmed. Synovial chondromatosis is an uncommon, benign, monarticular disorder of subsynovial cartilage neoplasia, which
results in the formation of hyaline cartilage nodules in a joint, tendon sheath, or bursa, and which often has a pathognomonic
radiologic appearance (Murphey MD, Vidal JA, Fanburg-Smith JC, Gajewski DA. From the archives of the AFIP: imaging of
synovial chondromatosis with radiologic-pathologic correlation. Radiographics 2007;27:146588).
Emily N. Vinson, MD
Thomas A. McGraw, DMD
Duke University Medical Center
Durham, NC