Expression OF membrane type 1 matrix metalloproteinase in human articular cartilage.

ARTHRITIS & RHEUMATISM
Vol. 40, No. 4, April 1997, pp 704-709
0 1997, American College of Rheumatology
704
EXPRESSION OF MEMBRANE TYPE 1 MATRIX METALLOPROTEINASE
IN HUMAN ARTICULAR CARTILAGE
FRANK H. BUTTNER, SUSAN CHUBINSKAYA, DANIEL MARGERIE, KLAUS HUCH,
JOHANNES FLECHTENMACHER, ADA A. COLE, KLAUS E. K U E n N E R , and ECKART BARTNIK
Objective. To detect the message for membrane
type 1 (MT1) matrix metalloproteinase (MMP) in articular cartilage and chondrosarcoma cells, to study its
expression in osteoarthritis (OA), and to determine
whether i nt e r l eu k in -lp (IL-1p) influences its
expression.
Methods. Reverse transcription-polymerase
chain reaction methods were used to detect message.
Cloning and sequencing were applied to confirm the
sequence. Northern blotting was used to quantify the
message for MT1-MMP and to compare its expression
in OA cartilage with or without IL-1p treatment. In situ
hybridization was utilized to show MT1-MMP transcripts in cartilage and to study the influence of IL-1p.
Results. The results show that MT1-MMP messenger RNA (mRNA) is expressed in chondrosarcoma
cells and OA chondrocytes. Results of the in situ
hybridization confirmed the expression in OA cartilage
as well as in normal cartilage. The level of mRNA was
not modulated following IL-1p stimulation.
Conclusion. This study shows that MT1-MMP is
expressed by chondrocytes. The similarities in mRNA
levels in OA and normal chondrocytes suggest that
regulation of MT1-MMP mRNA may not be a primary
factor in OA.
Proteolysis of the extracellular matrix of articular
cartilage is a hallmark of inflammatory and degenerative
joint diseases like rheumatoid arthritis and osteoarthritis
Frank H. Buttner, Diplom Biologe, Daniel Margerie, Diplom
Biologe, Eckart Bartnik, PhD: Hoechst AG, Kalle-Albert, Wiesbaden,
Germany; Susan Chubinskaya, PhD, Ada A. Cole, PhD, Klaus E.
Kuettner, PhD: Rush Medical College at Rush-Presbyterian-St.
Luke’s Medical Center, Chicago, Illinois; Klaus Huch, MD, Johannes
Flechtenmacher, MD: University of Ulm, Ulm, Germany.
Address reprint requests to Eckart Bartnik, PhD, Hoechst
AG, Kalle-Albert, Biomedical Research, H 528, Rheingaustrasse 190,
65174 Wiesbaden, Germany.
Submitted for publication August 1, 1996; accepted in revised
form October 25, 1996.
(OA). The process of matrix degradation is thought to
be mediated through the activity of matrix metalloproteinases (MMPs) (1). The most recent additions to this
expanding family include the membrane-type MMPs,
the first of which was identified by Sat0 et a1 (2) and
designated MT1-MMP. This MMP is bound to cellular
membranes. Strongin et a1 ( 3 ) were the first to describe
one possible function of MT1-MMP. Once activated,
MT1-MMP acts as a receptor for tissue inhibitor of
metalloproteinases 2 (TIMP-2), and the resultant dimer
then binds latent gelatinase A (MMP-2), activating the
gelatinase. Cao et a1 (4) showed that the transmembrane
domain of MT1-MMP plays an essential role in the
activation of progelatinase A. That study reported that
truncated forms of MT1-MMP containing no transmembrane domain were unable to activate gelatinase A,
whereas mutants of MT1-MMP with a transmembrane
domain from an unrelated protein were able to activate
gelatinase A. More recent reports show that
transmembrane-deficient mutants can still process
progelatinase A; however, different methods were used
for the two studies. Pei and Weiss ( 5 ) demonstrated
intrinsic matrix-degrading activity of transmembranedeletion mutants of MT1-MMP and suggested that there
may be other functions of MT1-MMP than activation of
gelatinase A at the cellular membrane ( 5 ) .
The current study concentrates on the expression
of MT1-MMP in articular cartilage, since it has been
previously reported that gelatinase A is present in that
tissue. In fact, activated gelatinase A as well as activated
collagenase have been described in extracts of articular
cartilage from O A patients in which tissue destruction is
evident (6). The expression of an activator of gelatinase
A in tissue in which the enzyme has been shown to be
active prompted the examination of whether MT1-MMP
is also expressed in articular cartilage. The results of
studies using amplification by reverse transcriptionpolymerase chain reaction (RT-PCR), Northern blot
analyses, and in situ hybridization show that MT1-MMP
MT1-MMP IN HUMAN ARTICULAR CARTILAGE
is expressed in both normal a n d OA articular cartilage,
as well as in human chondrosarcoma cells. While levels
of expression in the chondrosarcoma cells were relatively high, levels of expression in the OA cartilage were
much lower. N o detectable differences were evident
between levels of expression in t h e normal or the OA
chondrocytes beyond those observed between individual
cartilage samples. In addition, there was no detectable
change in MT1-MMP expression following stimulation by
the catabolic cytokine, interleukin 1p (IL-lp), which is
known to up-regulate t h e expression of other MMPs (1).
These data suggest that MT1-MMP expression in
chondrocytes may be constitutive. A n up-regulation of
the gelatinase A activator messenger RNA (mRNA)
may not be necessary for increased activity of the
enzyme protein during the destructive proteolytic events
associated with OA. It is also possible that the primary
function of MT1-MMP in cartilage is involvement in
some process other than the activation of gelatinase A.
PATIENTS AND METHODS
Chondrosarcoma cell culture. Human chondrosarcoma cells (SM-0219/JJ-012; a gift from Dr. J. Block, Rush
Medical College, Chicago, IL) were grown as monolayer
cultures in Dulbecco’s modified Eagle’s medium (DMEM)
with 25 Fl,g/mlof ascorbic acid (Sigma, Munich, Germany), 0.1
ngiml of insulin (Hoechst, Marburg, Germany), 100 mM
hydrocortisone (Sigma), 50 wg/ml of gentamicin (Gibco,
Grand Island, NY), 2 mM glutamine (Gibco), and 10% fetal
bovine serum (FBS; Sigma) in a humidified atmosphere of 5%
CO,, at 37°C.
Acquisition of human articular cartilage. Human hyaline articular cartilage was obtained from OA patients undergoing total joint replacement. These cartilage samples were
used for the RT-PCR and Northern blot analyses. For comparison with normal cartilage, samples were obtained from
donors with no joint disease through the Regional Organ Bank
of Illinois (ROBI). These normal cartilage samples were used
with the in situ hybridizations. For stimulation with IL-lp,
cartilage samples used in the explant cultures were also
obtained from normal donors and OA patients.
Tissue culture. To determine whether MT1-MMP
expression could be up-regulated in the presence of IL-lp,
cartilage explants (3 X 3 mm’) both from normal donors and
OA patients were cultured with recombinant human IL-lP
(Genzyme, Boston, MA) at concentrations of 0.05-50 pg/ml.
The media consisted of DMEM supplemented with 10% FBS,
25 &ml of ascorbate, and 50 pl,g/ml of gentamicin. Control
explants were cultured in identical media, except that the
IL-1p was omitted. Each explant was cultured in 1 ml of
medium in 24-well plates in a humidified atmosphere of 5%
CO, at 37°C (7,8).
Cell culture of human chondrocytes. Osteoarthritic
chondrocytes were isolated from the remaining articular cartilage from the OA patients and cultured in alginate beads for
705
72 hours as described by Hauselmann et a1 (9). The alginate
beads containing the chondrocytes were maintained in Ham’s
F-12DMEM (50/50) medium (Gibco), supplemented with
10% FBS, 25 pg/ml of ascorbic acid, and 50 &ml gentamicin
in a humidified atmosphere of 5% CO, at 37°C. For stimulation with IL-lp, cells were subdivided into 2 populations for a
further 3-day cultivation in the presence or absence of 50 pg/ml
of IL-lP.
RNA isolation. Total RNA was prepared according to
a modification of the protocol of Chomczynski and Sacchi (lo),
from chondrosarcoma cells, cultured 0.4 chondrocytes, and
intact OA cartilage. The RNA concentration was determined
spectrophotometrically.
PCR, cloning, and sequencing. Two sets of primers
(each 21 basepairs in length) were employed: set A for
nucleotides 112-1860 of MT1-MMP and set B for nucleotides
964-671 (2). RT-PCR was performed according to standard
procedures using Tuq polymerase. For generating the fulllength message for MT1-MMP, the PCR conditions were
modified by using 10% DMSO and increasing the annealing
temperature. Fragments generated by PCR were first analyzed
by restriction endonuclease digests. A full-length PCR message
for MT1-MMP was cloned into the pCDM8 vector (Invitrogen,
San Diego, CA) and sequenced.
Northern blot analyses. Total RNA was isolated either
directly from the OA cartilage (5 different samples; age range
64-80 years) or from cultured OA chondrocytes. The RNA
was resolved on agarose/formaldehyde gels, transferred to
Nylon membranes, and ultraviolet light-crosslinked. A probe
was generated from human chondrocyte RNA using RT-PCR
with primer set B, and was radiolabeled for hybridization with
w3’P-dCTP using random nonamer primers (Amersham,
Braunschweig, Germany). Hybridized blots were washed under
conditions of high stringency. The same blots were rehybridized with either p-actin or glyceraldehyde-3-phosphatedehydrogenase (GAPDH) probes to confirm that equal amounts of
RNA were loaded per lane. Quantitation was performed using
a phosphor image analyzer.
In situ hybridization. Articular cartilage from 7 normal
donors (age range newborn to 69 years) and 3 OA patients
(ages 59,69, and 74) as well as cultured cartilage explants were
fixed in 4% paraformaldehyde and paraffin-embedded. In situ
hybridization was also performed on sections of chondrosarcoma tissue from which the human chondrosarcoma cells
(JJ-012) were derived (a gift from Dr. J. Block). The probe
used for in situ hybridization was the reverse primer of set B.
The probe was 3’4abeled with 5’-(~t-thioI-~~S)-dCTP
using
terminal deoxynucleotidyl transferase (New England Nuclear,
Boston, MA). The radiolabeled probe was hybridized to
cartilage sections as previously described by Chubinskaya et a1
(11). Competitive inhibition controls were performed by mixing the radiolabeled primers with the same unlabeled probe in
ratios of 1:1, 1:2, 1:4, and 1%.
RESULTS
PCR identification of MT1-MMP. Two primer
sets were designed for the amplification of m R N A by
PCR. Primer set A corresponds to codons in the amino
BUTTNER ET AL
706
M
1749 bp
-
707 bp
-
1
2
3
4
M
Figure 1. Reverse transcription-polymerase chain reaction (RTPCR) products of RNA samples extracted from human chondrosarcoma cells (SM0219) or from human osteoarthritic articular cartilage,
employing primer sets A and B (see Patients and Methods for details).
Lanes 1 and 2, PCR products from human chondrosarcoma cells;
Lanes 3 and 4, PCR products from human articular cartilage; lanes 1
and 3, full-length message for membrane type 1 matrix metalloproteinase (MT1-MMP; 1,749 bp); lanes 2 and 4, amplified PCR fragment
of MT1-MMP (707 bp); lane M, molecular size markers (Boehringer
Mannheim nos. I11 and VI).
and the carboxyl terminals of MTl-MMP (nucleotides
112-1860) whereas set B corresponds to nucleotides
964-1671 (hinge to hemopexin domain). With RNA
from the chondrosarcoma cells, the amplification by
RT-PCR using both primer sets A and B resulted in
PCR fragments with the expected sizes of 1,749 bp and
707 bp, respectively (Figure 1, lanes 1 and 2). Fragments
of the same size were identified following amplification
of RNA originating from 4 different patients and 1
normal donor. RNA was either extracted directly from
cartilage (Figure 1, lanes 3 and 4, showing amplification
products from RNA directly isolated from the cartilage
of 1 patient) or from OA chondrocytes cultured in
alginate beads (data not shown). Restriction enzyme
analysis confirmed that the fragments indeed corresponded to MT1-MMP complementary DNA fragments
(data not shown).
The full-length PCR fragment of 1,749 bp derived with the use of primer pair A was subcloned.
Subsequent sequencing revealed a match to the sequence reported by Sat0 et al (2) (with the corrections
provided by Okada et a1 [12]), except for 10 nucleotides.
These differences may reflect true differences in the
sequence of cartilage MT1-MMP rnRNA; however, they
may also have been introduced by Taq polymerase
mistakes generated under the altered PCR conditions
needed for the successful amplification of the full-length
MT1-MMP mRNA. These data show that MT1-MMP is
expressed by cbondrosarcoma cells and chondrocytes
from OA cartilage.
Northern blot analyses. As an additional control
and in order to quantify MT1-MMP transcripts, Northern blot analyses were performed using total RNA
isolated directly from the cartilage of an additional 5 OA
patients. The Northern blots were incubated with a
radiolabeled probe that was generated from the same
primer pair B used for the PCR (Figure 2). The probe
hybridized to a 4.5-kilobase message, the same size as
the mRNA reported by Sat0 et a1 (2) and Takino et a1
(13) for MT1-MMP. Levels of mRNA were quantitated
using a phosphor imager and expressed as a percentage
of the detected p-actin or GAPDH signal after reprobing with 0-actin-specific or GAPDH-specific probes.
The results showed variations between 47% and 110%
(Figure 2, lanes 3 and 5 ) among the 5 OA patients
evaluated.
There were also no substantial differences between the mRNA levels in the OA chondrocytes cultured with and without IL-1p (Figure 2, lanes 6 and 7).
The quantification showed only a 15% difference.
Expression of MT1-MMP in cartilage as detected
by in situ hybridization. The results from the in situ
hybridization confirmed the Northern blot data, in that
all OA cartilage tested contained chondrocytes with
detectable levels of MT1-MMP mRNA (Figure 3A).
There was, however, variation in the levels of expression,
from weak to strong, within the chondrocytes from
different OA donors. In addition, no up-regulation of
MT1-MMP message was evident compared with controls following stimulation with IL-1p over the concentration range tested.
Within the cartilage sections following in situ
hybridization, mRNA was detectable in all zones, from
superficial to deep. In the OA cartilage where the
superficial layer was disrupted or absent, mRNA levels
for MT1-MMP did not appear to be elevated in the
middle and deep layers. When the levels of expression in
OA chondrocytes (from 3 patients) were compared with
the levels in chondrocytes from 7 normal donors, no
substantial differences could be found. As with the
expression in OA chondrocytes, levels of expression
varied among the normal chondrocytes, and upregulation by IL-1p could not be detected.
MT1-MMP IN HUMAN ARTICULAR CARTILAGE
1
2
3
4
4.5 kb -
5
707
6
7
- 4.5 kb
p-actin
GAPDH
Figure 2. Northern blot analysis of membrane type 1 matrix metalloproteinase (MT1-MMP) messenger
RNA derived from 5 different osteoarthritis (OA) patients or from OA chondrocytes cultivated in alginate
beads with or without interleukin-lp (IL-lp). Each lane contains 15 pg of total RNA. Lanes 1-5, RNA
extracted directly from 5 different OA patients; lanes 6 and 7, RNA extracted from cultivated articular OA
chondrocytes, without (lane 6) and with (lane 7) 50 pg/ml of IL-16. Hybridization with a labeled 707-bp
probe specific for MT1-MMP (lanes 1-7) resulted in a detection of a 4.5-kb MT1-MMP transcript. p-actin
or GAPDH was used as a control for loading equal amounts of RNA.
For all tissues, the variation in expression did not
appear to be related to heterogeneity of chondrocyte
density. While some of the OA cartilage had reduced
A
numbers of chondrocytes compared with the normal
cartilage, the observed variation related to the number
of grains over the chondrocytes. Within any 1 field, not
R
Figure 3. In situ hybridization. Dark field photomicrographs of A, osteoarthritic cartilage or B, chondrosarcoma tissue, showing hybridization with
radiolabeled antisense primer of primer set B (original magnification X 112.5).
BUTTNER ET AL
708
all chondrocytes were positive: of the positive chondrocytes in the same field, there appeared to be the same
approximate number of grains. But the number of grains
over the positive cells varied among patients and donors.
In situ hybridization with the chondrosarcoma tissue
showed strong expression of mRNA for MT1-MMP
(Figure 3B).
DISCUSSION
The results of this study clearly show that MT1MMP is expressed in human articular cartilage. Message
was detected by RT-PCR and Northern blot analyses in
RNA extracted from chondrosarcoma cells and from
OA chondrocytes. The fragments generated by the RTPCR corresponded to the size expected for the amplified
products of MT1-MMP. The full-length message for
MTl-MMP was cloned and sequenced. The sequence
was identical to that reported by Sat0 and Okada and
colleagues (2,12). Using Northern blot analyses, an
mRNA of 4.5 kb was detected-the same size for the
MT1-MMP mRNA reported by Sat0 and Takino et a1
(2,13). Finally, the technique of in situ hybridization was
employed to confirm the results obtained from Northern
blot analyses showing that mRNA for MTl-MMP was
present in OA chondrocytes. This technique also provided information that mRNA levels comparable to
those in OA chondrocytes were detectable in chondrocytes from normal donors. In addition, the mRNA levels
in cartilage both from OA patients and from normal
donors appeared similar in superficial-, middle-, and
deep-layer chondrocytes. In the OA cartilage where the
superficial layer was disrupted or absent, mRNA levels
for MT1-MMP did not appear to be elevated in the
middle and deep layers; an elevation has been reported
for other MMPs (11). In the chondrosarcoma tissue,
mRNA expression was stronger than those found in the
cartilage.
Stimulation with the catabolic cytokine, IL-lp,
did not appear to modulate the expression levels of
mRNA for MT1-MMP over the concentration range
tested (0.05-50 pg/ml) for the explant cultures using the
technique of in situ hybridization or at 50 pg/ml for the
Northern blot analyses. These concentrations have previously been shown to be effective in up-regulating
mRNA for MMP-8 (neutrophil collagenase) (11,14).
There are several possible explanations for the lack of
modulation by IL-1P. First, the concentrations may not
have been appropriate for gene stimulation. Second, it is
not clear whether there is an IL-1P-responsive element
in the regulatory sequences of this gene. Third, the
influence of IL-lP may be at the protein level, control-
ling the stability of the mRNA. Finally, the IL-1P
influence may be on the translation efficiency for the
MT1-MMP mRNA. The actual pathway site at which
IL-1p exerts its control is a subject for future studies.
There appeared to be no difference between the
level expression of mRNA for MT1-MMP in cartilage
from normal donors or from patients with end-stage OA.
However, if MT1-MMP is constitutivelyproduced by the
chondrocytes, the regulation of degradative enzymes
may reside at a point other than the regulation of the
MT1-MMP message, for example, at the level of the
activation of the proenzyme. Studies involving the trimolecular complex of MT1-MMP, TIMP-2, and gelatinase A ( 3 ) may be necessary before a full understanding
of the regulation of the proteolytic activity of MT1MMP is possible. The finding that MT1-MMP transmembrane domain-deficient mutants degrade extracellular matrix proteins such as gelatin, fibronectin,
laminin, vitronectin, and dermatan sulfate proteoglycan
( 5 ) further stresses the potential importance of MT1MMP in cartilage-degrading processes common to degenerative joint diseases like osteoarthritis.
REFERENCES
1. Woessner J F Jr: Matrix metalloproteinase and their inhibitor in
connective tissue remodeling. FASEB J 5:2145-2154, 1991
2. Sat0 H, Takino T, Okada Y, Cao J, Shinagawa A, Yamamoto E,
Seiki M: A matrix metalloproteinase expressed on the surface of
invasive tumour cells. Nature 370:61-65, 1994
3. Strongin AY, Collier I, Bannikov G, Marmer BL, Grants GA,
Goldberg GI: Mechanism of cell surface activation of 72-kDa type
1%' collagenase. J Biol Chem 270:5331-5338, 1995
4. Cao J, Sato H, Takino T, Seiki M: The C-terminal region of
membrane type matrix metalloproteinase is a functional transmembrane domain required for pro-gelatinase A activation. J Biol
Chem 2705301-805, 1995
5. Pei D, Weiss SJ: Transmembrane-deletion mutants of the
membrane-type matrix metalloproteinase-1 process progelatinase
A and express intrinsic matrix-degrading activity. J Biol Chem
271:9135-9140, 1996
6. Yu LP Jr, Smith GN Jr, Brandt KD, Capello W: Type XI
collagen-degrading activity in human osteoarthritic cartilage. Arthritis Rheum 33:1626-1633, 1990
7. Hauselmann HJ, Opplinger L, Michel BA, Stefanovic-Racic M,
Evans CH: Nitric oxide and proteoglycan biosynthesis by human
articular chondrocytes in alginate culture. FEBS Lett 352:361-364,
1994
8. Campbell IK, Piccoli DS, Hamilton JA: Stimulation of human
chondrocyte prostaglandin E2 production by recombinant human
interleukin-1 and tumor necrosis factor. Biochem Biophys Acta
1051:310-318, 1990
9. Hauselmann HJ, Aydelotte MB, Schumacher BL, Kuettner KE,
Gitelis SH, Thonar EJ: Synthesis and turnover of proteoglycans by
human and bovine adult articular chondrocytes cultured in alginate beads. Matrix 12:116-129, 1992
10. Chomczynski P, Sacchi N: Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroformextraction. Anal
Biochem 162:156-159, 1987
MT1-MMP IN HUMAN ARTICULAR CARTILAGE
11. Chubinskaya S, Huch K, Mikecz K, Cs-Szabo G, Hasty KA,
Kuettner KE, Cole AA: Chondrocyte matrix metalloproteinase-8:
up-regulation of neutrophil collagenase by interleukin-l/3 in human cartilage from knee and ankle joints. Lab Invest 74:232-240,
1996
12. Okada A, Bellocq JP, Rouyer N, Chenard MP, Rio MC, Chambon
P, Basset P: Membrane-type matrix metalloproteinase (MTMMP) gene is expressed in stromal cells of human colon, breast,
and head and neck carcinomas. Proc Natl Acad Sci U S A
92~2130-2734,1995
13. Takino T, Sato H, Yamamoto E, Seiki M: Cloning of a human
gene potentially encoding a novel matrix metalloproteinase having
a C-terminal transmembrane domain. Gene 155:293-298, 1995
14. Cole AA, Chubinskaya S, Schumacher B, Huch K, Cs-Szabo G ,
Yao J, Mikecz K, Hasty KG Kuettner KE: Chondrocyte matrix
metalloproteinase-8. J Biol Chem 271:11023-11026, 1996
Clinical Images: Severe destructive P,-microglobulin arthropathy after 28 years of hemodialysis
The patient, a 58-year-old man with a 28-year history of hemodialysis, was admitted to our hospital because of recurrent
hemarthrosis of the right shoulder (A). The joint had had to be aspirated with increasing frequency over the preceding months to
relieve pain, despite joint immobilization and antiphlogistic therapy. Blood loss from a 150-250-m1 hemorrhagic effusion at every
arthrocentesis had to be replaced. Radiography and magnetic resonance imaging (B) revealed erosions, hypertrophic pannus-like
material in synovium, and massive joint effusion (H = humerus; arrow 1 = erosion; arrow 2 = hypertrophic pannus). Synovectomy
was performed. Histologic evaluation of the synovium with hematoxylin and eosin staining showed marked hypervascularization,
cellular infiltrates, and hemorrhagic and hyaline structures (C). Congo red staining was diagnostic of amyloidosis. The green
birefringence seen on Congo red staining with polarizing microscopy (D) documented &-microglobulin, which was confirmed by
immunohistologic staining (not shown). After synovectomy, the patient was discharged and has had no recurrence of hemarthrosis
to date. We are indebted to Dr. B. Kreklau and Prot R. Ramanzadeh for per$orming the surgery, Dr. N. Prosenc for immunohistologic
analyses, and Dr. 0. Liangos for help with the clinical records.
H. Appel, MD
J. Sieper, MD
A. Distler, MD
J. Braun, MD
Klinikum Benjamin Franklin
Free University
Berlin, Germany