close

Вход

Забыли?

вход по аккаунту

?

Use of an antibody against the matrix metalloproteinasegenerated aggrecan neoepitope fvdipen-cooh to assess the effects of stromelysin in a rabbit model of cartilage degradation.

код для вставкиСкачать
ARTHRITIS & RHEUMATISM
Vol. 38, No. 10, October 1995, pp 1400-1409
0 1995, American College of Rheumatology
1400
USE OF AN ANTIBODY AGAINST THE
MATRIX METALLOPROTEINASE-GENERATED
AGGRECAN NEOEPITOPE FVDIPEN-COOH TO
ASSESS THE EFFECTS OF STROMELYSIN IN
A RABBIT MODEL OF CARTILAGE DEGRADATION
ELLEN KAHN BAYNE, KAREN L. MAcNAUL, SUSAN A. DONATELLI, AMY CHRISTEN,
PATRICK R. GRIFFIN, LORI A. HOERRNER, JIMMY R. CALAYCAY, JULIA M. AYALA,
KEVIN CHAPMAN, WILLIAM HAGMANN, JEFFREY R. WEIDNER, JOSEPH McDONNELL,
VERNON L. MOORE, RICHARD A. MUMFORD, MICHAEL W. LARK, and NANCY I. HUTCHINSON
Objective. To define the stromelysin cleavage site
in the interglobular domain of rabbit aggrecan, and to
determine whether the stromelysin-generated neoepitope
can be used as a marker of matrix metalloproteinase
(MMP) activity in vivo.
Methods. The carboxy-terminus sequence of the
stromelysin-generated hyaluronic acid-binding region
(HABR) of rabbit aggrecan was determined by reverse
transcription-polymerase chain reaction complementary DNA cloning and DNA sequence analysis, followed
by purification and mass spectral protein sequence
analysis of the HABR fragment. Active stromelysin was
injected into the stifle joints of rabbits, and a stromelysingenerated aggrecan neoepitope was analyzed by Western
blotting and localized in situ by indirect immunofluorescence. Proteoglycan fragments in joint fluids were quantified by a dimethylmethylene blue dye-binding assay.
Results. Stromelysin cleavage of rabbit aggrecan
generated a 55-kd HABR fragment that terminated in
the sequence FMDIPEN. An anti-FVDIPEN antibody
recognized the FMDIPEN neoepitope in situ in cartilage
from stromelysin-injected joints. The appearance of the
FMDIPEN neoepitope corresponded to the release of
Ellen Kahn Bayne, PhD, Karen L. MacNaul, BS, Susan A.
Donatelli, MS, Amy Christen, BS, Patrick R. Griffin, PhD, Lori A.
Hoermer, MS, Jimmy R. Calaycay, MS, Julia M. Ayala, BS, Kevin
Chapman, PhD, William Hagmann, PhD, Jeffrey R. Weidner, PhD,
Joseph McDonnell, BS, Vernon L. Moore, PhD, Richard A. Mumford, BA, Michael W. Lark, PhD, Nancy I. Hutchinson, PhD:
Merck Research Laboratories, Rahway, New Jersey.
Address reprint requests to Ellen K. Bayne, PhD, Merck
Research Laboratories, Room 80N-A38, P. 0. Box 2000, Rahway,
NJ 07065.
Submitted for publication December 13, 1994; accepted in
revised form April 18, 1995.
cartilage proteoglycan fragments into the joint fluid,
and could be inhibited by pretreatment of the rabbits
with a synthetic stromelysin inhibitor.
Conclusion. These results indicate that the antiFVDIPEN antibody can be used to assess the role of
MMPs in cartilage degradation in vivo.
Stromelysin is a matrix metalloproteinase
(MMP) that has been implicated in the turnover of
aggrecan from articular cartilage. Elevated levels of
stromelysin have been observed in the joint fluids and
synovia of patients with osteoarthritis (OA) and rheumatoid arthritis (RA) (1-3). Animal models of cartilage
destruction are often used to study the development
and progression of these diseases. Recently, in 2 rabbit
models of cartilage destruction, elevated levels of
stromelysin expression were demonstrated in affected
joints. In the first model, surgical removal of the
medial meniscus (4) resulted in elevated stromelysin
messenger RNA (mRNA) levels in the articular cartilage. In the second model, intraarticular injection of
human interleukin-1 (IL-1) into the stifle joint (5,6)
resulted in elevated stromelysin mRNA and protein
levels in both the synovium and cartilage. In the latter
studies, increased levels of stromelysin in the joint
fluids of the IL-1-injected rabbits correlated strongly
with increased release of proteoglycan fragments into
the joint fluids (5,6). This correlation suggested a
causal relationship between the expression of stromelysin and the digestion of cartilage aggrecan.
To more directly assess the role(s) of MMPs,
including stromelysin, in rabbit models of cartilage
destruction, it would be useful to have in situ markers
DETECTION OF STROMELYSIN-CLEAVED RABBIT AGGRECAN
of aggrecan cleavage. In the following study, stromelysin was used as a representative MMP for aggrecan
cleavage. The site at which stromelysin cleaved the
interglobular domain of rabbit aggrecan in vitro was
identified. Using an anti-FVDIPEN antibody, which
recognized the newly generated epitope FMDIPENCOOH, we demonstrate that intraarticular injection of
active stromelysin into the stifle joint of rabbits generated this neoepitope in situ within cartilage. Using a
synthetic stromelysin inhibitor, we also demonstrate that
inhibition of the generation of the neoepitope FMDIPEN
in vivo corresponded to a decreased release of cartilage
proteoglycan fragments into the joint fluid. These
results suggest that immunodetection of the neoepitope
F(M/V)DIPEN may be a useful method of assessing the
role of MMPs in cartilage degradation in vivo.
MATERIALS AND METHODS
Cloning and sequence analysis of rabbit aggrecan
complementary DNA (cDNA) encoding the G1 domain. Total
RNA was isolated from rabbit xiphoid or articular cartilage
as described previously (6), and the cDNA encoding the
GI and interglobular domains of aggrecan was generated
by reverse transcription-polymerase chain reaction (RTPCR). First-strand cDNA synthesis was performed using
random oligonucleotide hexamers and a Perkin Elmer (Emeryville, CA) RNA PCR kit. Three overlapping cDNA
clones spanning 1.2 kilobases of the rabbit aggrecan mRNA
were generated by PCR. The PCR primers used were as
follows: primer 1, GGACACACGGCTCCACTTGATTCTTGG; primer 2, GGAGAAGGGAGTCCAACTCTTCAAGGTG; primer 3, CCTGCAGAACAGTGCCATCATTGCCAC; primer 4, GGCGTGCACGTACACGGTCCTGACAC;
primer 5, GGTCACCTGCACGACGAGGTCCTCAC; and
primer 6, TCATCGACCCCATGCACCCTGTGACC. The
primer sequences corresponded to highly conserved regions
between the human and rat aggrecan sequences (7,s).
A 20-ng aliquot of the first-strand cDNA was diluted
into a 25-pl PCR mixture with 17 pmoles of upstream and
downstream primers. The PCR program was 19 cycles of
97°C for 30 seconds, 52°C for 30 seconds, and 72°C for 90
seconds, followed by I cycle of 97°C for 30 seconds, 52°C for
30 seconds, and 72°C for 120 seconds. The PCR products
were cloned by blunt-end ligation into Hinc II-digested
pBlueScript plasmid vector (Stratagene, La Jolla, CA) and
transformation into Escherichia coli DH5a. White colonies
grown on MacConkey agar were picked and grown for the
isolation of plasmid DNA by the alkaline-lysis procedure.
Plasmids containing inserts of the predicted size were sequenced using the Stratagene (La Jolla, CA) Sequenase kit.
Sequences from at least 2 each of the PCR products,
obtained from at least 2 independent PCR reactions, were
used to generate the final cDNA sequence. The oligonucleotide primers used for sequencing included PCR primers
1-6, as well as oligonucleotides derived from internal rabbit
1401
aggrecan cDNA sequences. The rabbit aggrecan sequences
have been supplied to GenBank (accession number L38480).
Isolation of intact aggrecan from rabbit cartilage. All
procedures with rabbits were approved by the Institutional
Animal Care and Use Committee, Merck Research Laboratories. Rabbit aggrecan was isolated from stifle joint articular
cartilage obtained from 6-8-week-old rabbits (9). Briefly, the
cartilage was removed from both femoral condyles and tibia1
plateaus and cut into 24-mm pieces in phosphate buffered
saline (PBS) containing 10 mM EDTA, 0.1M 6-aminohexanoic acid, 50 mM benzamidine hydrochloride, 1 mM phenylmethylsulfonyl fluoride, 50 mM N-ethylmaleimide, and 1
p g h l of pepstatin (proteinase-inhibitor cocktail). An equal
volume of 8M guanidine hydrochloride (GuHC1) was added
to the sample to bring the final GuHCl concentration to 4M.
The sample was homogenized in the 4M GuHCl using a
polytron homogenizer. The cartilage was extracted for 48
hours at 4"C, and the extract was clarified by low-speed
centrifugation.
The supernatant containing the aggrecan was
brought to 50 pg/ml with human umbilical cord hyaluronan
and dialyzed for 24 hours at 4°C against 0.1M sodium
acetate, pH 6.0, containing the proteinase-inhibitor cocktail.
After dialysis, the sample was clarified by centrifugation,
and the supernatant was fractionated through an associative
CsCl density gradient (starting specific density 1.5 g d m l ) .
The bottom fourth of the gradient (Al) was harvested, and
solid GuHCl was added to the sample to bring it to a final
concentration of 4M GuHCI. The sample was then fractionated on a dissociative CsCl density gradient (starting specific
density 1.5 g d m l ) . The bottom fourth of this gradient
(AlDl) was harvested, and the concentration of aggrecan
was determined using the dimethylmethylene blue (DMMB)
dye-binding assay as described previously (10).
Hyaluronan (7.7% weighdweight) was added to the
sample, and the sample was dialyzed for 24 hours at 4°C
against 0.1M sodium acetate, pH 6.8, to generate aggrecanhyaluronan complexes. The integrity of these complexes
was evaluated by gel filtration chromatography on an associative Sepharose CL-2B column (1 x 120 cm). Almost all of
the DMMB-positive material eluted in the void volume of the
column, indicating that the aggrecan bound the hyaluronan.
Activation of recombinant prostromelysin and digestion of isolated rabbit aggrecan. Recombinant rabbit prostromelysin (Celltech Ltd., Slough, UK) or recombinant
human prostromelysin (Celltech) (2 pM) was activated with
trypsin (80 nM) for 30 minutes at 37°C. The sample was then
incubated with excess soybean trypsin inhibitor (SBT1)agarose for 15 minutes at room temperature. The SBTIagarose-bound trypsin was removed from the sample
by centrifugation. Under these conditions, the active stromelysin migrated as a single band on sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) with a molecular weight of -42 kd. Enzyme activity could be
eliminated by incubation with 10 mM 1,lo-phenanthroline.
For purification of stromelysin-generated rabbit aggrecan HABR, the isolated aggrecan-hyaluronan complexes
(3 pM) were digested with rabbit stromelysin (0.3 pM) for 6
hours at 37°C in 0.1M sodium acetate, 10 mM CaCI,, pH 6.8,
buffer. The digested aggrecan (100 mg) was concentrated by
lyophilization and resuspended in 5 ml of water for chroma-
BAYNE ET AL
1402
tography on an associative Sepharose CL-2B column (5 x
120 cm). For Western blot analysis, isolated rabbit aggrecanhyaluronan complexes were digested with either rabbit or
human stromelysin for the times indicated in the text.
Gel filtration chromatography. Associative Sepharose CL-2B columns ( I or 5 x 120 cm) were eluted at 4°C
with 0.1M sodium acetate, 0.1% sodium azide, pH 6.8.
Dissociative Sepharose CL-2B columns (1 x 120 crn) were
eluted with 4M GuHC1, 0.1M sodium acetate, and 0.1%
sodium azide, pH 6.8, at 4°C.
Generation of tryptic map and mass spectrometry of
tryptic peptides. Purified rabbit aggrecan HABR was reduced
and alkylated with dithiothreitol and 4-vinylpyridine. The
resulting material was desalted using a TSK G3000 PW
column. The desalted protein was digested with sequencing
grade
(Boehringer Mannheim' Indianapolis' IN) and
was analyzed by capillary liquid chromatography-electrospray
mass spectrometry (LC-ESI-MS) using a Finnigan (San
Jose'
TSQ 700
quadrupole
spectrometer' The amino acid sequence Of the
peptide was
confirmed using capillary liquid chromatography-electrospray
tandem mass spectrometry (LC-ESI-MSIMS) (11).
Western blot analysis of in vitro-digested rabbit aggrecan. Rabbit aggrecan that had been digested in vitro was
evaluated with or without chondroitinase ABCikeratinase I1
digestion. For treatment with the glycosidases, 100-pl samples in 0.1M sodium acetate buffer, pH 8.0, were brought to
10 mM EDTA and treated overnight at 37°C with 0.02 units
of protease-free chondroitinase ABC. Keratinase I1 (0.1
units) was then added to each sample and incubated at 37°C
for 2 hours.
Samples were electrophoresed through 4 2 0 % SDSPAGE gels under reducing conditions and transferred to
nitrocellulose. The nitrocellulose filters were incubated in
5% nonfat dry milk for 1 hour at room temperature, followed
by incubation for 1 hour at room temperature with either 5
pg/ml of biotinylated anti-FVDIPEN IgG (121, or the monoclonal antibody 12/21/1-C-6, which recognizes the G1 domain within the HABR (13). The filters were washed and
incubated with either a 1:1,000 dilution of alkaline
phosphatase-streptavidin followed by BCIPMBT, or a
1:1,000 dilution of a biotinylated goat anti-mouse IgG followed by BCIPNBT.
Intraarticular injection of stromelysin. The stifle
joints of New Zealand white rabbits (6-8 weeks old, weight
1.1 kg) were injected intraarticularly with 100 pg of recombinant human stromelysin in 0.5 ml of buffer or with buffer
alone (5,14). In some cases, the animals were predosed (30
mgikg, intravenously) with the synthetic stromelysin inhibitor N-[l(R)-carboxyethyl]-cu-(S)-(2-phenylethyl)glycine-(~)leucine, N-phenylamide (15) 15 minutes prior to receiving
the intraarticular injections. The inhibitor was dissolved and
injected in vehicle consisting of 10% DMSO, 10% Cremophor E L (BASF, Ludwigshafen, Germany), and 80% phosphate buffer (500 mM) at neutral pH. Pharmacokinetic
studies had previously demonstrated that inhibitor was detectable in both blood and synovial fluid by 15 minutes after
intravenous injection (unpublished observations). Animals
were killed 1 hour after intraacticular injection, and the
joints were Iavaged with 2 ml of PBS. Cartilage for subsequent analyses was dissected from both femoral condyles
and tibia1 plateaus.
-
PCR-GENERATED cDNA ENCODING G1 DOMAIN
OF RABBIT AGGRECAN
1.2 kb cDNA
AA 65
AA460
SLN AGG
AT G
5'
3'
G1
G2
PREDICTED AMINO ACID SEQUENCES OF G1 DOMAINS
OF HUMAN, RAT, MOUSE AND RABBIT AGGRECAN
HUMAN 6 2
TAPSTAPLaPR~KWSRVSKEKEVVLLYliTECrVRYNSaY1PSD~~L~"Q~
RAT 6 2
T A P S T I P L ? . P R ~ K W S R V S K E K E Y V L L V R T E G q Y R V N S i Y T L E I Q "
HOUSE 6 2
TI\PST*PLTPRTKWSRYSKEKEYVLLVITEGqVRyNgiyI~SDATLEIQ~
consensus
RABBIT
=
*
HUMAN 1 2 3 ~ R S N D S G V Y R C E V M H G ~ E O S E I T L E Y V V K G I V F H Y R A ~ S T R Y T L D F ~ ~ A Q R A C L Q N S A I I A
RAT 1 2 3
L R S N D S G T Y R C E V M H G ~ E D S = A T ~ ~ V i " ~ G ~ V ~ " ~ ~ ~ ~ ~ = ~ ~ ~ ~ ~ ~ ~ ~ ~ Q R A C ~
MOUSE 1 2 3 L R S N D S G l Y R C E V M H G l E D S E I \ T L E Y i V ~ G ~ " ~ " ~ ~ A ~ S T ~ ~ ~ L D F D R ~ Q R A C L Q N S A ~ I A
RABBIT
LRsllosGIYRCsVMXGleosElTLeYvVXGvVFHYRa~S~RYT~DFD~~QRACLQNSA~~A
Consensus l R S N O S G l Y R C E V H l l O i E D S E i i T L E V v V K G i V F H Y R A ~ S ~ R Y T L D F ~ ~ ~ Q R A C L Q N S A ~ I A
HUMAN 1 8 4 T P E Q L Q l ~ Y E D G F H Q C D A G W L I \ D Q T V R Y P I H T P R E G C Y G D ~ D ~ F P G V R T Y G ~ ~ D T N E T Y D V
TPEQLQAlYEOGFHQCOliGWLhOaTVRYPIHTPREGCYG~~~=FPGV~~~GIRDTNETYDV
RI\T 1 8 4
HOUSE 1 8 4 T P E Q L Q ~ I Y E D G F X Q C D L i G ~ ~ * D Q ~ Y R ~ ~ ~ ~ T ~ ~ ~ G C ~ ~ ~ ~ ~ ~ ~ ~ G " ~ = = G I R D ~
RABBIT
TPEQLQAIYEOGFBQCDI\GWLI\DQTVRYPTHTPREGCYG~~~~FPGV~~~GIRDTNET~DV
HUMhN 2 4 5
RAT 2 4 5
MOUSE 2 4 5
RABBIT
YCFAEEMEGEVFY~TSPEKFTFOERRNECRRLGARLA~~Gh"Y~AWQ~G"0MCSAGWLA~~
Consensus T P E Q L Q l l l Y E D G F H Q C D * G W = A D Q ~ " * ~ ~ I " = ~ ~ ~ G C ~ ~ ~ ~ ~ = F ~ G " ~ ~ ~ G I R D ~ ~ ~ T Y D V
YCFAEEMEGEYFYRTSPEKFTFOEhllNECRTVCIRLATTGQLYLAWQqGMDMCSAGWLADR
YCFAEEIEGEVFYI\TSPEX~~~QE~ANECRRLGARLATTGQLKLAWQ~GMDMCSAGWLADR
YCFAEEMEGEVFYATSPEKF~~Q=~A~ECRRLGAR~ATTG~YL~WQ~GMDMCSAGWLADR
C o n s e n s ~ sY C F A E E M E G E V F Y A T S P E K F T F O E R I n E C R r l G l i R L A T T G ~ l Y L A W Q ~ G M D M C S A G W L A D R
Figure 1. Alignment of predicted amino acid (AA) sequence of
rabbit aggrecan G1 and interglobular domains with those of rat,
human, bovine, and mouse aggrecan. Top, Map of rabbit aggrecan
complementary DNA (cDNA) sequences, from which the predicted
amino acid sequence was obtained. Stromelysin (SLN) and aggrecanase (AGG) cleavage sites of the core protein, based on sequence
analogy to human aggrecan, are shown. Bottom, Alignment with rat,
human, bovine, and mouse aggrecan protein sequences. The rabbit
sequences which differ from human aggrecan are underlined.
Western blot analysis of cartilage from injected rabbit
joints. Cartilage from stromelysin- or vehicle-injected contralateral control joints of the rabbits was extracted with 4M
GuHCl in proteinase-inhibitor cocktail as described above.
The extract was dialyzed against water and evaluated directly by Western blotting after electrophoresis through
4 2 0 % SDS-PAGE.
Immunofluorescence microscopy of FMDIPEN neoepitope in rabbit articular cartilage. Cartilage from stromelysin- or vehicle-injected rabbit joints was frozen in OCT
embedding compound (Tissue-Tek, Miles Diagnostics,
Elkhart, 'IN) cooled with liquid nitrogen. Six-micrometer
cryostat sections were cut through the full-thickness cartilage samples. Cryosections were then fixed with Nakane
solution (16) followed by Triton X-100to increase permeability. The FMDIPEN neoepitope was labeled with antiFVDIPEN IgG (12), and bound antibodies were detected
using fluorescein isothiocyanate (FITCkonjugated secondary antibody. For specificity controls, the primary antibody
(50 &ml) was incubated with either the YVDIPEN peptide (1
DETECTION OF STROMELYSIN-CLEAVED RABBIT AGGRECAN
1403
in a volume of -2.0 ml, after subtracting the mass of
proteoglycan lavaged from the contralateral control joint.
The mean level of proteoglycan in the contralateral control
joints injected with vehicle alone were 17 f. 2 pgljoint (mean
& SEM; n = 6).
A
RESULTS
The predicted amino acid sequence of the G1 and
interglobular domains of rabbit aggrecan. Three overlapping RT-PCR cDNA products spanning approximately 1.2 kb of the rabbit aggrecan cDNA were
cloned and sequenced. Figure 1 (top) depicts a map of
Y D A I C (pe) Y T G E D F M D I P E N
(M+2H)" = 1051.4
+2
1051.3
100
80
60
1
2
1
2
40
Figure 2. Purified rabbit aggrecan hyaluronic acid-binding region
(HABR) generated by digestion with rabbit stromelysin. Lane 1,
Ten micrograms of purified rabbit HABR; lane 2, molecular weight
markers. A, Coomassie-stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis. B, Western blot with the 12/21/1-C-6
monoclonal antibody.
ng/ml) or with VDIPENFFGVG (1 ng/ml), a peptide that spans
the A ~ n ~ ~ ' - P cleavage
h e ~ ~ ' site, and clarified before staining.
Peptides were synthesized on an Applied Biosystems
(Foster City, CA) model 430A peptide synthesizer, purified
by reverse-phase high-performance liquid chromatography
(HPLC) on Waters C18 Deltapak columns. All peptides were
>95% pure by reverse-phase HPLC, and the structure of
each was confirmed by ESI-MS. Preimmune rabbit IgG
served as an additional negative control. Preparations were
viewed on a Zeiss fluorescence microscope equipped for
epifluorescence, with a narrow-band selective filter combination for FITC. Micrographs were prepared using Tri-X
film developed at 20°C for 6 minutes with Acufine developer.
Detection of proteoglycan fragments in joint fluids.
This procedure has been described previously (5). Briefly,
aliquots of synovial fluid samples were incubated for 24
hours at 37°C with 20 unitdm1 of Streptomyces hyaluronidase (Seikagaku Kogyo, Rockville, MD) to eliminate interference by high concentrations of hyaluronan present in
synovial fluids. Total sulfated proteoglycans were measured
by the DMMB dye-binding method described by Farndale et
a1 (10) and Dey et a1 (17). Results are expressed as the mass
of proteoglycan lavaged from the stromelysin-injected joint
20
4M)
600
800
1000 1200 1400 1600 1800 2000
MI2
-
STROMELYSIN CLEAVAGE SITE IN AGGRECAN
Rat
Human
Bovine
Rabbit
Mouse
Y
Y
Y
Y
Y
T
T
T
T
T
G
G
G
G
G
E
E
E
E
E
D F
D[F
D F
D F
D F
V
V
V
E
V
D I P E N
D I P E NI
D I P E S
D I P E N
D I P E N
F
F
F
F
F
F
F
F
F
F
G V G G E E D
G V G G E E D
G V G G
G V G G E E D
G V G G E Q D
Figure 3. Sequence determination of the stromelysin cleavage site
in rabbit aggrecan; alignment with homologous regions in rat,
human, bovine, and mouse aggrecan. Top, Electrospray mass
spectrometry sequence determination of tryptic peptide containing
the carboxy-terminus of stromelysin-generated rabbit aggrecan hyaluronic acid-binding region. Bottom, Alignment of the predicted
stromelsyin cleavage sites in rat, human, bovine, rabbit, and mouse
aggrecan. The boxed sequence, FVDIPEN, in human aggrecan was
used to develop an antipeptide antibody against this stromelysingenerated neoepitope.
1404
BAYNE ET AL
Figure 4. Anti-FVDIPEN antibody recognizes stromelysin-generated rabbit aggrecan hyaluronic acid-binding
region digested either in vitro or in vivo. A, Western blot analysis of rabbit aggrecan either undigested (lane 1) or
cleaved with rabbit stromelysin in vitro for 0 (lane 2), 5 (lane 3), 15 (lane 4), 30 (lane 3,60 (lane 6), 120 (lane 7), 180
(lane 8), and 360 minutes (lane 9). B, Immunoblot of cartilage extracts from rabbits injected intraarticularly with
buffer in one joint (lane 1) and with active stromelysin in the contralateral joint (lane 2). C, Immunoblot of
stromelysin-digested aggrecan (240 minutes) stained with anti-FVDIPEN (lane I), anti-FVDIPEN preincubated with
VDIPENFFGVG-NH, (lane 2), or anti-FVDIPEN preincubated with YVDIPEN (lane 3).
the amino-terminus of rabbit aggrecan and indicates
the region encoded by these cDNA clones. The predicted amino acid sequence of rabbit aggrecan was
highly homologous to that of human aggrecan and
aligned with amino acids 65-460 of human aggrecan.
Only 40 amino acid differences were observed within
the aligned regions of the rabbit and human protein
sequences. One of these changes was a valine-tomethionine change within 6 residues of the stromelysin
cleavage site in human aggrecan. An alignment of the
predicted amino acid sequence of the G1 and interglobular domains of rabbit aggrecan with those of rat,
human, and mouse aggrecan is shown at the bottom of
Figure 1.
Isolation of the stromelysin-generated HABR.
When isolated rabbit aggrecan complexed to hyaluronan was digested with rabbit stromelysin, the majority
of the DMMB-positive material shifted to an included
position on an associative Sepharose CL-2B column,
indicating that the aggrecan had been cleaved. When
the fractions from this column were evaluated by
Western blotting using a monoclonal antibody (12/21/
1-C-6) that specifically recognizes the G1 domain of
aggrecan (13), a stromelysin-generated HABRhyaluronan complex was identified in the void volume
of the column. These complexes were lyophilized,
resuspended in 1 ml of 4M GuHCl, and chromatographed on a dissociative Sepharose CL-2B column.
The resultant 55-kd HABR fragment included in the
dissociative column was again identified by Western
blotting. The purified HABR was homogeneous, as
assessed by Coomassie blue staining and Western
blotting (Figure 2). The identity as aggrecan HABR
was defined by the intense reactivity of the fragment
with the anti-HABR antibody 12/21/1-C-6(Figure 2B).
Identification of the stromelysin cleavage site in
rabbit aggrecan by mass spectrometry. Purified
stromelysin-generated rabbit aggrecan HABR was digested with trypsin and analyzed by capillary LC-ESIMS. The carboxy-terminal tryptic peptide was identified as YDAIC(pe)YTGEDFMDIPEN [(M+2H)” =
1051.41 (Figure 3, top panel), and the amino acid
sequence of the peptide was confirmed using capillary
LC-ESI-MS/MS. This result demonstrates that active
stromelysin cleaved rabbit aggrecan between the
amino acids asparagine and phenylalanine within the
DETECTION OF STROMELYSIN-CLEAVED RABBIT AGGRECAN
1405
Figure 5. Immunostaining of stromelysin-generated neoepitope in rabbit cartilage. Indirect immunofluorescence staining was performed using
anti-FVDIPEN IgG that was preincubated with A, VDIPENFFGVG peptide or C, YVDIPEN peptide. B and D show phase-contrast
micrographs corresponding to each fluorescence micrograph.
sequence FMDIPENFFGVGG, analogous to the
stromelysin cleavage site of human aggrecan (18)
(Figure 3, bottom panel).
Detection of the FMDIPEN neoepitope by Western blotting using an anti-FVDIPEN antibody. As previously described (12), an antipeptide antiserum was
generated against the peptide FVDIPEN, which represents the carboxy-terminal sequence of stromelysingenerated human aggrecan HABR. To determine if the
anti-FVDIPEN antibody recognized rabbit aggrecan
that had been cleaved with stromelysin in vitro, isolated rabbit aggrecan-hyaluronan complexes were incubated with activated rabbit stromelysin for 0-6
hours and analyzed by Western blotting (Figure 4A).
After stromelysin digestion, a 55-kd rabbit HABR
fragment was recognized by the anti-FVDIPEN antiserum, and the neoepitope was generated in a timedependent manner. Immunodetection with the anti-
FVDIPEN antiserum could be blocked by preincubation
of the antiserum with the peptide YVDIPEN, but not
with a peptide spanning the stromelysin cleavage site,
VDIPENFFGVG (Figure 4C). Similar results were
obtained when rabbit aggrecan was digested with human stromelysin (data not shown).
To determine if the neoepitope FMDIPEN was
generated upon stromelysin digestion of intact cartilage, rabbits were injected intraarticularly with
stromelysin in one joint and with vehicle in the contralateral control joint. Western blot analysis detected
a similar 55-kd HABR fragment in extracts of cartilage
from the stromelysin-injected joints (Figure 4B). In
addition, there are also several smaller molecular
weight FMDIPEN-positive G1 fragments in the cartilage from the stromelysin injected joint. These represent less than 10% of the total FMDIPEN-positive
material extracted from the joint, and their presence
1406
most likely reflects secondary cleavage of the aggrecan
molecules that are N-terminal to the FMDIPEN sequence, by either stromelysin or other enzymes within
the joint. It thus appears that a similar HABS fragment
could be generated by stromelysin cleavage of either
isolated aggrecan or intact cartilage. The antiFVDIPEN antibody did not detect significant levels of
the FMDIPEN epitope in cartilage extracts from the
contralateral control joints.
Immunostaining of cartilage from stromelysininjected rabbits. To determine if stromelysin-generated
rabbit HABR could be detected withiq intact cartilage
in situ, immunohistochemical experiments were performed on cartilage samples from rabbits injected
intraarticularly with either stromelysin or vehicle.
When cartilage from stromelysin-injected joints was
stained with anti-FVDIPEN IgG, followed by an
FITC-conjugated second antibody, immunofluorescence was observed throughout the matrix near the
articular surface of the cartilage, and was particularly
concentrated in pericellular and territorial regions
(Figures 5A and 6A). The specificity of the immunostaining was confirmed by preincubating the antiFVDIPEN IgG with either YVDIPEN peptide or a
peptide spanning the stromelysin cleavage site,
VDIPENFFGVG. The immunostaining was unaffected by preincubation with the spanning peptide
(Figure 5A), but was completely abolished by pretreatment with YVDIPEN (Figure SC).
In contrast to the intense immunostaining observed in the cartilage from the stromelysin-injected
joints, no immunostaining was observed in the vehicleinjected contralateral control joints (Figure 6E) or in
cartilage samples incubated with preimmune IgG (data
not shown). Systemic administration of a synthetic
stromelysin inhibitor 15 minutes prior to stromelysin
injection (n = 3) blocked generation of the FMDIPEN
neoepitope in all but the most-superficial zone of the
cartilage (Figure 6C).
Joint fluid levels of proteoglycan fragments in
rabbits injected intraarticularly with active stromelysin.
Proteoglycan levels in the joint fluids of stromelysininjected rabbits were compared between rabbits that
had been pretreated with either the vehicle or with a
synthetic stromelysin inhibitor. One hour after
stromelysin injection, in rabbits that had not received
inhibitor, 122 ? 10 pg/joint (mean 2 SEM; n = 3) of
glycosaminoglycan was released into the joint fluids.
Pretreatment with the synthetic stromelysin inhibitor,
resulted in a 92% reduction (10 k 4 pg/joint, n = 3) in
the level of proteoglycan fragments released. This
BAYNE ET AL
result, in combination with the findings of the
FMDIPEN immunolocalization studies, indicates that
the appearance of the FMDIPEN neoepitope corresponded to the stromelysin-generated loss of cartilage
proteoglycans.
DISCUSSION
The purpose of these studies was to develop a
method to more directly assess the role(s) of MMPs,
including stromelysin, in rabbit models of cartilage
destruction. This was done by defining the stromelysin
cleavage site in the interglobular domain of rabbit
aggrecan and demonstrating the ability of an antiFVDIPEN antibody to recognize the stromelysingenerated rabbit HABR in cartilage in situ. The
stromelysin cleavage site in rabbit aggrecan was determined by mass spectral analysis of tryptic peptides
derived from purified, stromelysin-generated rabbit
aggrecan HABR. Since such an analysis requires the
identification of peptides of known molecular mass,
and the sequence of rabbit aggrecan was unknown, it
was first necessary to determine the predicted amino
acid sequence of the G1 and interglobular domains of
rabbit aggrecan. This was accomplished by cloning
and sequencing a 1.2-kb segment of the rabbit aggrecan cDNA. Only 1 amino acid difference was observed
between the rabbit aggrecan sequence and the human,
rat, and mouse sequences within 6 amino acids of the
stromelysin cleavage site. Of the 395 amino acids
aligned, 40 amino acids differed between the rabbit and
human aggrecan sequences, and there was an overall
identity of 91%.
The carboxy-terminal sequence of the purified
stromelysin-generated HABR fragment was determined to be FMDIPEN. This sequence is homologous
to the carboxy-terminal sequence generated by
stromelysin digestion of human aggrecan, FVDIPEN,
and was recognized by an anti-FVDIPEN antibody.
Western blot analyses using the anti-FVDIPEN antibody demonstrated that the FMDIPEN neoepitope
was generated by stromelysin digestion of either isolated rabbit aggrecan or intact cartilage. In each case,
the size of the predominant FMDIPEN-containing
fragment was consistent with an intact amino-terminal
G1 domain. Immunodetection of FMDIPEN within
the cartilage of rabbits injected intraarticularly with
stromelysin demonstrated that the anti-FVDIPEN antibody also recognized the native neoepitope in situ
within the cartilage matrix. The distribution of neoepitope within cartilage was consistent with the pres-
DETECTION OF STROMELYSIN-CLEAVED RABBIT AGGRECAN
1407
Figure 6. Reduction of stromelysin-generated neoepitope in the cartilage of rabbits treated systemically with synthetic stromelysin inhibitor.
Indirect immunofluorescence staining in A, a stromelysin-injected joint, C, a stromelysin-injected joint from a rabbit that had received
stromelysin inhibitor 15 minutes prior to intraarticular injection of stromelysin, and E, the contralateral joint of a stromelysin-injected rabbit.
B, D, and F show phase-contrast micrographs corresponding to each fluorescence micrograph. Bar = 30p.
BAYNE ET AL
ence of stromelysin in the joint fluid of these rabbits,
and suggested that the stromelysin digested its way
into the matrix, starting from the articular surface. The
high concentration of neoepitope in pericellular and
territorial matrix most likely reflects the high aggrecan
concentration in these regions (19). The specificity of
the immunostaining was demonstrated by experiments
in which the immunofluorescence was blocked by
preincubation of the anti-FVDIPEN IgG with the
peptide YVDIPEN, but was not blocked by the spanning peptide VDIPENFFGVG.
The appearance of the FMDIPEN neoepitope
in cartilage corresponded to the release of high levels
of proteoglycan fragments into the joint fluid, suggesting that the appearance of the neoepitope was a
marker of cartilage destruction. This correlation was
extended when pretreatment of the rabbits with a
synthetic stromelysin inhibitor dramatically decreased
both the appearance of the stromelysin-generated
neoepitope within cartilage and the release of proteoglycan fragments into the joint fluid. Although stromelysin was used as a representative MMP, other members of the MMP family, including gelatinase A,
gelatinase B, collagenase, and PUMP, have also been
reported to cleave aggrecan in vitro at the same site
(20-22). Since MMPs have been shown to cleave at
this site, and a series of non-MMP aggrecan-degrading
enzymes, including human leukocyte elastase, cathepsin B, cathepsin G, and cathepsin D (12,23-25), do not
cleave at this site, we believe that this epitope may be
a specific marker of aggrecan cleavage by MMPs.
The ability to monitor generation of specific
aggrecan cleavage products in the rabbit provides an
opportunity to examine cartilage degradation in an
animal system where one can readily obtain sufficient
amounts of cartilage and synovial fluid for biochemical
as well as immunohistochemical analyses. Fluids and
cartilage can be assayed for matrix breakdown products, inhibitor levels can be monitored, and joints can
easily be injected as desired. The size of the rabbit also
facilitates various surgical techniques, such as meniscectomy. Recently, however, we have shown that the
same immunohistochemical techniques used in the
rabbit system can also be applied to other models.
Thus, using the same antibody described here, we
have been able to show that the FVDIPEN neoepitope is
induced during inflammatory arthritis in the mouse
(26). Similarly, studies of human cartilage demonstrate
concentrations of FVDIPEN neoepitope at sites of
cartilage damage (27). We believe that analysis of the
generation of the neoepitope F(M/V)DIPEN in in vivo
models of cartilage destruction may prove useful in
assessing the role of MMPs in arthritis.
ACKNOWLEDGMENTS
We would like to thank Dr. Jack Schmidt for his
scientific discussions and Drs. Alice Marcy and Irwin Singer
for their critical evaluation of the manuscript.
REFERENCES
1. Walakovits LA, Moore VL, Bhardwaj N, Gallick GS, Lark
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
MW: Detection of stromelysin and collagenase in synovial fluid
from patients with rheumatoid arthritis and posttraumatic knee
injury. Arthritis Rheum 3 2 3 5 4 2 , 1992
Lohmander LS, Hoermer LA, Lark MW: Metalloproteinases,
tissue inhibitor, and proteoglycan fragments in knee synovial
fluid in human osteoarthritis. Arthritis Rheum 36: 181-189, 1993
Wolfe GC, MacNaul KL, Buechel FF, McDonnell J, Hoermer
LA, Lark MW, Moore VL, Hutchinson NI: Differential in vivo
expression of collagenase messenger RNA in synovium and
cartilage: quantitative comparison with stromelysin messenger
RNA levels in human rheumatoid arthritis and osteoarthritis
patients and in two animal models of acute inflammatory arthritis. Arthritis Rheum 36:154&1547, 1993
Mehraban F, Riera H, Kuo SY, Chang C, Moskowitz RW:
Cartilage metalloproteinase genes are upregulated in experimental osteoarthritis and in isolated osteoarthritic chondrocytes in
culture (abstract). Trans Orthop Res SOC 18:682, 1993
McDonnell J, Hoermer LA, Lark MW, Harper C, Dey T,
Lobner J, Eiermann G, Kazazis D, Singer 11, Moore VL:
Recombinant human interleukin-1 p-induced increase in levels
of proteoglycans, stromelysin, and leukocytes in rabbit synovial
fluid. Arthritis Rheum 35:799-805, 1992
Hutchinson NI, Lark MW, MacNaul KL, Harper C, Hoerrner
LA, McDonnell J, Donatelli S, Moore V, Bayne EK: In vivo
expression of stromelysin in synovium and cartilage of rabbits
injected intraarticularly with interleukin-lp. Arthritis Rheum
35:1227-1233, 1992
Doege KJ, Sasaki M, Kimura J, Yamada Y: Complete coding
sequence and deduced primary structure of the human cartilage
large aggregating proteoglycan, aggrecan: human-specific repeats, and additional alternatively spliced forms. J Biol Chem
266:894-902, 1991
Doege K, Sasaki M, Horigan E, Hassell JR, Yamada Y:
Complete primary structure of the rat cartilage proteoglycan
core protein deduced from cDNA clones. J Biol Chem 262:
17757-1 7767, 1987
Hascall VC, Kimura JH: Proteoglycans: isolation and characterization. Methods Enzymol 82:769-800, 1982
Farndale RW, Buttle DJ, Barrett AJ: Improved quantitation and
discrimination of sulphated glycosaminoglycans by use of dimethyl-methylene blue. Biochim Biophys Acta 883: 173-177,
1986
Griffin PR, Coffman JA, Hood LE, Yates JR 111: Structural
analysis of proteins by capillary HPLC electrospray tandem
mass spectrometry. Int J Mass Spectrom Ion Processes 11I :
131-149, 1991
Lark MW, Williams H, Hoermer LA, Weidner J, Ayala J,
Harper C, Christen A, Olszewski J, Konteatis Z, Mumford R:
Quantification of a matrix metalloproteinase-generated aggrecan
GI fragment using monospecific anti-peptide serum. Biochem J
307:245-252, 1995
Caterson B, Calabro T, Donohue PJ, Jahnke MR: Monoclonal
antibodies against cartilage proteoglycan and link protein. In,
DETECTION OF STROMELYSIN-CLEAVED RABBIT AGGRECAN
14.
15.
16.
17.
18.
19.
20.
21.
Articular Cartilage Biochemistry. Edited by K Kuettner, R
Schleyerbach, VC Hascall. New York, Raven Press, 1986
Moore VL, Olszewski JM, McDonnell J, Williams H, Mumford
RA: Elicitation of a stromelysin-specific neoepitope in the IGD
region of aggrecan by the intraarticular injection of human
stromelysin into rabbits (abstract). Trans Orthop Res SOC 19:
312, 1994
Chapman KT, Kopka IE, Durette PL, Esser CK, Lanza TJ,
Izquierdo-Martin M, Niedzwiecki L , Chang B, Harrison RK,
Kuo DW, Lin T-Y, Stein RL, Hagmann WK: Inhibition of
matrix metalloproteinases by N-carboxyalkyl peptides. J Med
Chem 36:42934301, 1993
McLean IW, Nakane PK: Periodate-lysine paraformaldehyde
fixative: a new fixative for immunoelectron microscopy. J
Histochem Cytochem 22:1077-1083, 1974
Dey P, Saphos CA, McDonnell J, Moore VL: Studies on the
quantification of proteoglycans by the dimethylmethylene blue
dye-binding method: specificity, quantification in synovial lavage fluid, and automation. Connect Tissue Res 28:317-324,
1992
Flannery CR, Lark MW, Sandy JD: Identification of a stromelysin cleavage site within the interglobular domain of human
aggrecan: evidence for proteolysis at this site in vivo in human
articular cartilage. J Biol Chem 267:1008-1014, 1992
Knudson CB: Hyaluronan receptor-directed assembly of chondrocyte pericellular matrix. J Cell Biol 120:825-834, 1993
Fosang AJ, Neame PJ, Hardingham TE, Murphy G, Hamilton
JA: Cleavage of cartilage proteoglycan between GI and G2
domains by stromelysins. J Biol Chem 266:15579-15582, 1991
Fosang AJ, Neame PJ, Hardingham TE, Murphy G, Hamilton
22.
23.
24.
25.
26.
27.
1409
JA: Stromelysins and PUMP cleave cartilage proteoglycan
between G1 and G2 domains (abstract). Trans Orthop Res SOC
17:83, 1993
Fosang AJ, Last K, Knauper V, Neame PJ, Murphy G, Hardingham TE, Tschesche H, Hamilton JA: Fibroblast and neutrophi1 collagenases cleave at two sites in the cartilage aggrecan
interglobular domain. Biochem J 259:273-276, 1993
Fosang AJ, Neame PJ, Last K, Hardingham TE, Murphy G,
Hamilton JA: The interglobular domain of cartilage aggrecan is
cleaved by PUMP, gelatinases, and cathepsin B. J Biol Chem
267:19470-19474, 1992
Mok MT, Ilic MZ, Handley CJ, Robinson HC: Cleavage of
proteoglycan aggregate by leucocyte elastase. Arch Biochem
Biophys 292M2-447, 1992
Flannery CR, Sandy JD: Aggrecan catabolism in cartilage:
studies on the nature of a novel proteinase (aggrecanase) which
cleaves the Glu 373-Ala 374 bond of the interglobular domain
(abstract). Trans Orthop Res SOC18:190, 1993
Singer 11, Kawka DW, Bayne EK, Donatelli SA, Weidner JR,
Williams HR, Ayala JM, Mumford RA, Lark MW, Glant TT,
Nabozny GH, David CS: VDIPEN, a metalloproteinasegenerated neoepitope, is induced and immunolocalized in articular cartilage during inflammatory arthritis. J Clin Invest 95:
2178-2186, 1995
Bayne EK, Donatelli SA, Singer 11, Weidner JR, Hutchinson
NI, Hoermer LA, Williams HR, Mumford RA, Lohmander LS,
Lark MW: Detection of a metalloproteinase-generated aggrecan
HABR fragment within human OA and RA cartilage (abstract).
Trans Orthop Res SOC19:308, 1994
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 124 Кб
Теги
stromelysin, matrix, degradation, mode, fvdipen, assess, metalloproteinasegenerated, aggrecan, neoepitope, antibody, effect, rabbits, cooh, cartilage, use
1/--страниц
Пожаловаться на содержимое документа