MT1-MMP M NEUTROPHILS: POTENTIAL MECHANISM FOR COLLAGENASE ACTNATION
Jim Yuan Lai
A thesis submined in conformity with the requirements for the degree of Master of Science (Periodontology)
Graduate Department of Dentistry University of Toronto
O Copyright by Jim Yuan Lai (2000)
National Library 191 of Canada Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bibliographie Services services bibliographiques
395 Wellington Street 395, rue Wellington Oîtawa ON K i A O N 4 Ottawa ON KI A ON4 Canada Canada
The author has granted a non- exclusive licence dowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of this thesis in rnicroforrn, paper or eleclronic formats.
The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be printed or otherwise reproduced without the author's permission.
L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique.
L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.
MTl-MMP in Neutrophils: Potential Mechanism for Collagenase Activation Jim Yuan Lai Faculty of Dentistry, University of Toronto M.%. 2000
ABSTRACT
Matrix metalloproteinases (MMPs) are important enzymes in the destruction of
extracellulz natriccs in perioilontai diseases. One of these enzymes, MMP-8, is derived
largely from neutrophils. The active f o m of MMP-8 is found in the gingival crevicular
fluid of progressive periodontitis lesions but the rnechanism by which the latent enzyme
is converted to active forms in vivo is poorly understood. As activation of MMP-2 cm be
mediated by a membrane-bound MMP, membrane-type-1 MMP (MT1-MW), 1 tested
the hypothesis that MT-MMPs are expressed in neutrophils and can activate latent MMP-
8. My objectives were: 1) to assess the presence and location of MT-MMP in peripheral
blood neutrophils; 2) to detemine if MT-MMPs c m activate MMP-8. By RT-PCR, 1
found that human peripheral blood neutrophils expressed MTI-MMP mRNA.* The
plasma membrane and specific granule fractions of neutrophils were isolated by
discontinuous Percoll gradients and from these fractions 1 found MT1-MMP in the
plasma membrane fraction. Soluble biotinylated collagen assays showed some
collagenase activity (-28% collagen digestion) in the membrane fraction. The specific
granule fraction contained latent MMP-8 that could be activated by APMA. The
collagenase activity of a plasma membrane fraction combined with specific granules of
neutrophils (54% digestion) was higher than the collagenase activity of the individual
fractions (28% for plasma membrane, 1% for specific granules). These results are
consistent with the notion that MTI-MMP may contribute to the activation of MMPd on
the ce11 surface of peripherai blood neutrophils .
ACKNOWLEDGEMENTS
1 would like to express my gratitude and appreciation for the tremendous support and
assistance offered by the members of CIHR group in Periodontal Physiology.
My deepest thanks and admiration goes to Chns McCulIoch. He was not only a
supervisor, but a mentor who offered vaiuable advice and guidance on many things. His
expenence, patience and support guided me to the completion of rny thesis.
I would also like to acknowledge the help 1 received in the lab fiom Maki MacGillivray,
Jiaxu Wang, Pam Arora, Wilson Lee, Cheung Lo, and Kevin Ko. I especially would like
to thank Carol Laschinger. Her thoroughness, patience and guidance were extremely
valuable and much appreciated.
Finaily, I thank my parents for their unconditional love and support. Their trust and
encouragement have helped me achieve my goals and find happiness in life.
TABLE OF CONTENTS
. . Abstract ............................................................................................... 11 ... Ac knowledgements ................................................................................ .HL
Table of Contents ................................................................................... iv List of Figures ....................................................................................... vi .. Abbreviations ..................................................................................... ..vil
Review of the Literature .......................................................................... 1
Periodontd Diseases ........................................................................ 1 A . Classification, Naturai History and Clinical Course ............................... 1
7 B . Infiammation and Destruction of Extracellular Matrices ...................... ..., C . Role of Neutrophils in Lnflarnmation and Connective Tissue Destruction ...... 4
II . Destruction of Extracellular Matrices .................................................... 6 . A General Overview of the MMP family .............................................. 7
B . Domain Structure and Function of Collagenase .................................... 8 .................................................. . C Substmte Specificity of Collagenase 9
D . Activation Mechanisms of Collagenase ............................................ I O
III . Membrane-type Matrix Metalloproteinases ............................................ 12 . A Domain Structure and Function ...................................................... 12 . B Regulation & Activation of MT-MMP ............................................. 15 . C Substrate Specificity of MT-MMP .................................................. 17
D . Relationship to TIMP-2 .............................................................. 18 E . Membrane Fixation .................................................................. -20
[V . Introduction and Statement of the Problem ............................................. 21
Materials and Methods ......................................................................... - 2 2 33 Reagents ......................................................................................
Isolation of Polymorphonuclear Neutrophils 33 by Human Plasma/Percoll Blood Ce11 Separation ..............................
Isolation of PMN Plasma Membrane (y-fraction) and Specific Granules (p-fraction) By Discontinuous Percoll Gradients ........................... - 2 3
....................... Identification of MT 1 -MMP mRNA expression by RTPCR 24 hmuniocaiization of MT 1 MMP ..................................................... -25 Western BIot Analysis ................................................................... 2 6 Soluble Biotinylated Collagen Assay (SBA) ........................................ - 2 6 Preparation of samples for SBA ...................................................... 2 7
ResuIts ............................................................................................ -29 ............................................ A . MT1-MMP mRNA expression in PMN -29
8 . Immunolocalization of MT 1 -MW in HGF and PMN ............................ 29 ........................... C . Identification of MT1-MMP by Western Blot Andysis 33
..... D . Verification of Specific Granule @-fraction) as a source of latent MMP-8 33 .............. E . Collagenase Activity in NeutrophiI Plasma Membrane (yfraction) 35
F . Increased Collagenase Activity of Specific Granules @-fraction) Latent MMP-8 by Neutrophil Plasma Membrane .......................... 38
........................................................................................ Discussion -39 ..................... A . Expression of MT 1 - M W in Peripheral B lood Neutrophils -39
.................................................................. B . Foms of MT I MMP -40 ................................................... C . MMP-8 activation by MT I .MMP -42
D . Summary and Suggestions for Future S ~ d i e s ..................................... -45
...................................................................................... Conclusions -47
........................................................................................ References A8
LIST OF FIGURES
Figure 1 : MT1 -MMP mRNA Expression in HGF and PMN ............................... 30
Figure 2: Immunolocdization of MT1 -MMP in KGF ...................................... 31
Figure 3: immunolocalization of MT1 -MMP in PMN ...................................... 32
Figure 4: Identification of MT1 -MMP ........................................................ 34
Figure 5: Identification of MMP-8
Figure 6: Collagenase Activity in Plasma Membrane (y-Fraction) And Specific Granules (p.fraction) ................................................ 37
MMP:
MT-MMP:
TIMP:
Pm:
LPS:
ECM:
SBA:
RT-PCR:
PMA:
Dl-r:
APMA:
DAPI:
NHS-LC-Biotin:
matrix metalloproteinase
membrane-type matnx metalloproteinase
tissue inhibitor of matrix metalloproteinase
polynorphonuclear leukocyte
Iipopolysacc haride
extracellular matrices
soluble biotinylated collagen assay
reverse transcription-polymerase chah reaction
phorbol myristate acetate
dithiothreitol
P-aminophenylmercuric acetate
4,6-Diamidino 2-p henylindole
N-hydroxysul fosuccinimide- long chain- B iotin
vii
REVIEW OF THE LITERATURE
1. Periodontat Diseases
A. Classification, Natural History and Clinical Course
There are two major categories of inflammatory periodontai diseases, gingivitis and
periodontitis, the division of which is based on destruction of alveolar bone. Gingivitis is
a reversible inflamrnatory disease of only the gingiva while periodontitis involves the
destniction of the marginal supporting structures of the tooth including the alveolar bone.
The hailmark of gingivitis is gingival inflammation which is caused by adherent
subgingival bacterial plaques (Loe et al., 1965). Based on a sequence of retrospectively
analyzed histopathologic events, idammatory processes of the periodontium have been
separated into
and Schroeder
three hypothetical stages: the initial, earl y and established lesions
The initial lesion is an acute inHammatory reaction that occurs 2 to
4 days after the onset of subgingival plaque formation. This lesion is characterized by the
loss of perivascular collagen and increased migration of neutrophils through high
endothelial venules, across the junctiond epithelium and into the gingival sulcus. From 4
to 7 days, the initial lesion progresses to the early lesion in which 60-70% of the collagen
in the marginal gingival comective tissue may be destroyed and the inflammatory ce11
infiltrate is dominated by lymphocytes and macrophages. Afier 2 to 3 weeks, the
established lesion is formed which is synonymous with chronic aduit gingivitis. The key
histological feature of the established lesion is the presence of plasma cells in the
idammatory infiltrate. This lesion cm persist for months or years without progression to
periodontitis.
In a small proportion of individuals with gingivitis, some of the affected sites progress to
periodontitis (Loe et al. 1986). Several classifications of periodontitis have been
suggested but a recent review by Armitage (1 999) suggests that there are four major types
of periodontitis: i) chronic periodontitis; ii) aggressive periodontitis; iii) periodontitis as a
manifestation of systemic diseases; and iv) necrotizing ulcerative periodontitis. It has
been suggested that each disease type has a different etiology, a different rate of disease
progression and response to treatment, but a comrnon pathway of extracellular matrix
(ECM) destruction. Indeed the degradation of ECM rnacromolecules including collagen
fibers, elastic fibers, proteoglycans and glycoproteins is the structural basis for the
destruction of the periodontal ligament, the disruption of its attachent to cementum, and
ultimately the resorption of alveolar bone (Page and Schroeder 1976, Page et al. 1997).
B. Inflammation and Destruction of ECM
In periodontitis, there are several different mechanisms that rnay lead to degradation of
ECM rnacromolecules (Birkedal-Hansen 1993). One proposed mechanism is that
proteolytic enzymes from putative pathogenic bactena can directly degrade the ECM. For
example, Porphyrornonas gingivalis and dctinobacilhs actinomycetemcomitans produce
hydrolytic enzymes that cleave native type I and III collagen fibrils at physiologicai pH
and temperature (Birkedal-Hansen et al. 1988; Robertson et al. 1984). However. the
relative abundance of these bacterial enzymes in the lamina propria of the gingiva and the
periodontal ligament may not be sufficient to account for the arnount of collagen
degradation seen in periodontitis.
A second complementary mechanism is the release of bacterial virulence factors that
may act directly on host cells to perturb homeostasis. For example, in response to the
vinilence factor lipopolysaccharide (LPS) that is present in the envelope of several gram
negative anaembic pathogens, macrophages undergo major shifts in their expressed - gene
repertoire which can result in the induction, for exarnple, of matrix metailoproteinases
(Wahl et ai. 1974; Welgus et al. 1990). Other virulence factors such as proteinases from
the pathogen P. gingivalis have been demonstrated to induce the expression and mediate
the activation of MMPs from mucosal keratinocytes and fibroblasts (Birkedal-Hansen et.
a1 1984).
The third and most studied mechanism is a cytokine-dependent, host immune.
inflammatory response to the antigenic challenge of subgingival pathogens such as
Porphyromonus gingivalis, Bacteroides forsyrhus, and ndctinobacilltrs
actinomycetemcornitans. These pathogens colonize teeth and grow to form a biofilm
(Darveau et al. 1997) which eventudly extends subgingivally if left undisturbed
(Haffajee and Socransky 1994). The biofilm and its metabolites disrupt the anachment of
the junctional epitheliurn to the 100th. Subsequently, in a susceptible host, the presence of
viruience factors from perîodontal pathogens can tngger the synthesis and release of
proinflammatory cytokines such as IL4 P, TNF-a and IFN-y. These cytokines, in turn.
induce and enhance host ce11 pathways for the degradation of the ECM (Page 1998).
C. Role of Neutrophiis in Infiammation and Connective Tissue Destruction
One of the key inflammatory cells involved in the degradation of the ECM is the
neutrophil. Neutrophils comprise 4040% of the circulating leukocytes in the peripheral
blood and are a first line of defense against microbial attack. Neutrophils are produced in
the bone marrow and after maturation, are released into the bloodstream with a
circulatory half Iife of 6 to 7 hours (Gordon 1994).
In the blood flow through capillaries, there are continuous reversible rolling contacts
between neutrophils and uninilamed endotheliurn. These transient, adhesive rolling
contacts are mediated by L-selectin, a ce11 surface protein that is constitutively expressed
on the surface of the neutrophil (Smith et al. 199 1). However, in the presence of
proinfiammatory cytokines such as Il- 1 P and TNF-a. endothelial cells rapidly translocate
P-selectins from Weibel-Palade bodies (McEver et al. 1989) to the ce11 surface and
induce biosynthesis of E-selectins (Bevilacqua et al. 1987). The carbohydrate-bearing
moieties on the neutrophil surface, gpl50-Lewis x and siaiyl Lewis-X, will respectively
bind to P- and E-selectins (Fukuda et al. 1984: Symington et of. 1985). This
consequently strengthens the adhesive interactions between the neutrophils and
endothelial cells and increases the number of neutrophils rolling and attaching to
inflamed postcapillary high endothelial vendes. Subsequently, interleukin-8 (a product of
endothelial cells) induces neutrophils to shed L-selectin and express p2-integins (LFA-1'
Mac-1, CR4) that are stored in the specific granules (Miyasaki 1996). LFA- 1 and Mac-1
bind tightiy to the endotheliai adhesion molecules (ICAM-I and -2) and this binding
initiates the tramendothelid migration of the neutrophil through the endothelid ce11
junctions into the extravascular cornpartment, a process known as diapedesis.
In periodontal tissues, neutrophils infiltrate the perivascdar connective tissue, migrate
dong a gradient of chemoattractants generated by subgingival bactena and pass through
the junctional epithelium to fom a defensive barrier between the subgingival plaque and
the gingival tissues (Theilade et al. 1985). Upon encountenng bacteria the neutrophil
utilizes various bactenal killing mechanisms which can be broadly classified into
oxidative and nonoxidative system. Oxidative killing involves the production of toxic
superoxide anions (023, hyarogen peroxide (H202), and hydroxyl radicals (OH).
However, no studies have conclusively s h o w that these metabolites alone are capable of
producing toxic effects under pathophysiologicai conditions. Nonoxidative mechanisms
involve secretion of antimicrobial components stored in granules. For exarnple, the
primary or anirophilic granules contain myeloperoxidase. defensins, and lysozyme while
secondary or specific granules contain lactofemn and lysozyme. (Van Dyke 1994).
In the presence of bactenal factors or complement factors (e.g. C k ) , the granule contents
are released extraceilulariy either by secretion or when the neutrophil undergoes
apoptosis or cytolysis (Miyasaki 1991). Although the primary function of the neutrophil
is host protection, the extracellular release of the granule contents may lead to excessive
tissue breakdown because proteolytic enzymes such as elastase. collagenase and
gelatinase are also found within these granules and are able to attack key components of
the extracellular matrix (Weiss 1 989). Notabl y, the neutrop hil collagenase (MW-8)
found in the specific granules appears to have a direct role in the tissue destruction of
periodontitis. In periodontitis, much of the collagenase activity is denved from
neutrophils and not from bacteria or other host cells, as based on the pattern of collagen
substrate degradation (Gangbar et al. 1990, Lee et al. 199 1 & 1995, O v e d l et al. 1991).
A longitudinal cohort study that examined gingival crevicular fluid in patients
demonstrated that active collagenase activity was 5 - 6 fold higher in groups with active
periodontitis than groups with gingivitis (Lee et cil. 1995). Large increases of active
collagenase were detected concurent with connective tissue anachment loss. In contrat,
latent collagenase activity was 2-fold higher in patients with inflammation but no
destruction.
II. Destruction of Extracellular Matrices
Birkedal-Hansen (1993) identified five distinct pathways that lead to degradation of the
extracellular matrices of the penodontium. The first pathway involves the conversion of
plasminogen into plasmin, a senne protease that cleaves fibrin and fibronectin (Dano et
al. 1985). The second pathway is the neutrophil-serine proteinase pathway. When
released, the neutrophil serine proteinases, elastase and cathepsin G, cleave a variety of
ECM molecules such as type N collagen, laminin, fibronectin and proteoglycans (Weiss
1989). The third pathway involves intracellular degradation by fibroblasts and
macrophages that phagocytize collagen fibrils and degrade them within phagolysosomes.
This intracellular pathway is particularly important in sites of rapid collagen turnover
such as the healthy gingiva or periodontai Ligament (Melcher and Chan 1981). The
fourth, osteoclastic bone resorption pathway is unique in that removal of minerai (i.e.
hydroxyapatite) is required before degradation of matrix proteins cm be initiated (Vaes
1988). The f i f i pathway is the matrix rnetalloproteinase pathway. This pathway will be
discussed in detail in the following section as it is centrai to the objectives of this thesis.
A. General Ovewiew of the MMP family
The MMP family is a group of metal-dependent endopeptidases which are capable of
degrading most extracellular matrix macromolecules. MMPs are involved in comective
tissue rernodeling and degradation, embryonic growth and development, and diseases
such as rheumatoid arthntis, periodontitis, tumour growth and metastasis (Birkedai-
Hansen et al. 1993). The MMPs share sequence homology but differ in terms of substrate
specificity and transcriptional regulation.
Based on substrate specificity, the MMPs are classified into four broad categones: i)
collagenases ii) gelatinases iii) stromelysins including matrilysin and metalloelastase and
iv) membrane-type MMPs (Polette et al. 1998). The collagenases (MMP-1, MMP-8.
MMP-13) cleave type I, II, III, VI1 and X coliagen. Neutrophil collagenase (MMP-8) is
produced largely by neutrophils whereas interstitial collagenase (MMP-1) is produced by
many ce11 types such as fibroblasts, keratinocytes, endothelid cells, macrophages.
chondrocytes and osteoblasts (Birkedaf-Hansen rf al. 1993). A recent paper (Hanernaaijer
et al. 1997) has s h o w that other cells can produce MMP-8 but the pathophysiologicd
significance of this is unclear. MMP-13 is produced by both epithelial and mesenchymd
cells in inflarned and remodelling connecting tissues. Its ability to cleave type II collagen,
type X collagen and cartilage aggrecan suggests that MMP-13 plays a significant role in
cartilage collagen degradation (Freije et al. L994, Mitchell et al. 1996, Knauper et al.
1976). MMP- 13 ha also been identified in gingival crevicular Buid (Mancini et al. 1999)
but the degradative impact of MMP-13 on periodontal connective tissues has not been
assessed. Gelatinases (MMP-2, MMP-9) are also synthesized by a wide variety of cells.
In addition to gelatin, gelatinases also cleave type IV collagen found in basement
membranes. The expression of gelatinases is often associated with invasive and
metastatic tumours (Liotta 1980). Stromelysins degrade a wide variety of collagenous
and noncollagenous ECM substrates which include collagen. gelatin, laminin and
proteoglycans. The final group, membrane-type MMPs, often contain a transmembrane
domain. This hydrophobic stretch of amino acids anchors the protein to the plasma
membrane and leaves the catalytic domain exposed extracellularly. In contrast, the other
MMPs are released extracellularly as soluble proteinases.
B. Domain Structure and Function of Collagenase
Collagenases have a five-domain modular structure that is shared by the other MMPs
(Fig. A) (Birkedal-Hansen ef al. 1993). The hydrophobic signal sequence of about 1 7-29
residues is followed by a 77-87 residue propeptide domain that constitutes, for example,
the NH2-terminal domain of the secreted MMP-8 precursor. This propeptide contains a
highly conserved sequence (PRCGVPD) which maintains the latency of the enzyme
extracellularly until it is removed in the process of enzyme activation (Springman et al.
1990). The caialytic domain of about 160 residues contains a highly conserved 2nZ'
binding active site (Lovejoy er al. 1994). A 5-50 residue proline-nch hinge region
c o ~ e c t s the catalytic domain with a 200 residue pexin-like COOH domain which plays a
role in substrate specificity.
Figure A: Domain structure of Collagenase a) hydrophobic signal sequence; b) NH2-terminal propeptide; c) catalytic domain with zinc molecule; d) proline-rich hinge region; e) hemopexin-like COOH terminal domain
While the MMPs share certain structural motifs, there are sorne notable differences
between closely related enzymes such as the neutrophil collagenase (MMP-8) and the
fibroblast collagenase (MMP-1). MMP-8 has a higher molecular mass than MMP-1
(80kDd75kDa vs. 57 kDa/SZkDa) because of the greater glycosylation of MMP-8. It is
speculated that die carbohydrate moieties of MMP-8 encode targeting signals that direct
the enzyme to specific granule storage sites (Birkedal-Hansen et al. 1993). Another
significant difference between the two enzymes is the mode of transcriptional regulation.
MMP-8 is synthesized during neutrophil maturation and is rapidly released fiom the
specific granules after neutrophil activation whereas MMP-1 is not stored in granules but
instead is synthesized in response to a wide variety of stimuli including IL-1 and TNF-a.
C. Substrate Specificity of Collagenase
Collagenases can cleave the native triple helix of type I1 II, and III collagens at a single
site in each polypeptide chah at physiological temperature and pH. In type 1 collagen, the
cleavage occurs at the glycine775-isoleucine776 bond in the alpha 1 chah and at the
glycine775-leucine bond776 in the alpha 2 c h a h This single cut produces a N-terminal K
and a C-terminal % collagen fragment. Subsequent degradation of the collagen may be
mediated by the gelatinases. Wu et al. (1990) demonstrated by site-directed mutagenesis
that the primary structure of the collagenase-sensitive site in collagen is an important
factor in determining the susceptibility to collagenase. However, the susceptibility of this
site cannot be accounted entirely by the amino acid sequence of collagen alone. The
collagenase-sensitive region rnay also be a locus that at 37OC more readily unfolds and
relaxes its triple helicai structure than other regions (Birkedal-Hansen 1987). In
cornparison to the fibroblast collagenase, the neutrophil collagenase exhibits 10 to 30 fold
higher catalytic efficiency on al1 substrates except for type III collagen (Netzell-Arnett et
al. 1991).
D. Activation Mechanisms of Collagenase
The activation of the latent collagenase is a critical regdatory step especially with MMP-
8. As mentioned above. latent forms of MMP-8 are stored in specific granules of
neutrophils and upon neutrophil activation, the granule contents are released
extracellularly (Weiss et al. 1985, Desrochers et ul. 1992). However, collagen
degradation cannot occur until MMP-8 is activated. The latency of the collagenase is
rnaintained by the formation of a cysteine9n" bond that links the unpaired propeptide
cysteine residue to the active site ~ n " (Van Wart el al. 1990). In enzyrne activation.
dissociation of the bond between the cysteine thiolate moiety and zinc atom is assumed to
be a crucial step. This process is described by Van Wart (1990) as the cysteine switch
activation mechanism. In vitro studies have demonstrated that there are different ways in
how t h i s activation c m be achieved.
Organrnercurials, metal ions, thiol reagents and oxidants can directly interact with the
cysteine residue to disrupt the cpteine-~n" bond in vitro. This interaction results in a
confirmation shift that triggen a senes of autolytic cleavages to occur. The first cleavage
leads to a reduction in molecular mass but not activation. The second cleavage. which
occurs at ~ s ~ ~ - ~ e t ~ ' , results in 40% increase of the mêuimum enzymatic activity. The
final cleavage occurs at ~ h e ' ~ - ~ e t ' ~ or ~ e t ' * - ~ e u ~ ' after prolonged incubation with the
activator (Knauper el al. 1 990, Blaser et al. 199 1).
Another activation system is based on Ni vitro studies using chaotropic agents (KI,
NaSCN) or detergents (SDS) to disnipt the cysteine-~n" bond, a procedure which
induced conformational changes in the polypeptide backbone. This conformational shift
also leads to several autolytic cleavages in which a fully processed active form of the
enzyme is generated (Nagase 1997).
A third activation process involves proteolytic enzymes (trypsin, plasmin, chymotrypsin,
neutrophii elastase, cathepsin B, plasma kallikrein). These enzymes excise a portion of
the propeptide which causes the cysteine switch to open. Activation requires the cleavage
of peptide bonds probably between residues 70 and 82. specifically, ~ h e ' ~ . et" and
eu" (Nagase 1990).
However, despite these advances based on MMP-8 activation systems studied in vitro,
the biologicai mechanism by which MMP-8 activation actually occurs in vivo is poorly
understood. The pathway involvhg proteolytic enzymes may be the closest surrogate to
the in vivo situation. Activation in vivo is frequentiy associated with a decrease in
molecular mass due to the removal of the propeptide. For example. latent MMP-8 in
gingival crevicular fluid is 78 kDa while the active form is 60 kDa (RomaneIli et al.
1999). Since activation of collagenase is a cntical regulatory step, there have been
continuing searches for an activation mechanism that truly reflects the in vivo situation.
The recently discovered membrane-type MMP was identified as the first physiological
activator of gelatinase A (Sato 1994) but little is known about the role of membrane-type
MMPs in the activation of M W - 8 .
III. Membrane-Wpe Matrix Metallo~roteinase
A. Domain structure and Function of MT-MMP
The majority of the MMPs are in a soluble form. However, by the use of RT-PCR and
screening a human placenta cDNA library, Sato (1 994) discovered a protein of 582 amino
acids that had the cornmon MMP five domain structure with three unique insertions.
First, there is an insertion of I l amino acids between the propeptide and the catalytic
domain. This insertion contains a stromeiysin-3-like RXKR furin cleavage motif. The
second insertion is an 8 arnino acid residue within the cataiytic domain whose h c t i o n
remains undehed while the third insertion of 24 hydrophobic arnino acids in the C-
terminus represents the trammembrane domain. The trammembrane domain allows the
pmtein to be anchored to the plasma membrane with its catalytic domain exposed to the
extracellular space. Hence, this protein is cailed membrane-type MMP. Currently, the
family of MT-MMP has been expanded to include MTI-, MT& MT3-, MT4-, MT5-,
MT6-MMPs (Sato 1994, Will 1995, Takino 1995. Puente 1996, Pei 1999 & 1999a).
Many tissues and cells express MTI-MMP. For example, during mouse embryogenesis,
MTl-MMP is expressed at hi& levels in developing blood vessels, kidney,
osteocartilaginous and musculotendinous tissues (Kinoh et al. 1 996, Apte et ai. 1997,
Sato & Seiki 1996). MTI-MMP is also found in lung tissue, kidneys, microglia in the
brain, ocular tissues, enamel and the pulp organ of teeth (Takino et al. 1995a & b, Sato et
al. 1994, Caron et al. 1998). At sites of vascular injury during wound healing, demal
fibroblasts express MT1 -W. Other cells that express MT I -MMP include smooth
muscle cells, endothelid cells, ameloblasts, odontoblasts, chondrocytes. and osteoclasts
(ha i et al. 1997, Sato & Seiki 1996, Sato ef al. 1997)
In various tumours of lung, stomach, colon, breast, ovary, cervix. urethra, bladder and
pancreas, stroma1 and cancer cells have elevated MTI-MMP expression (Sato & Seiki
1996). Notably, MTl-MMP expression levels correlate with gelatinase A activation and
with the malignant potential and invasiveness of the turnour. In an experimental
metastasis assay, MTI-MMP expression has been shown to enhance metastatic activity
(Tsunezuka et al. 1996). In the context of cancer biology, gelatinase A plays a cntical
role in the invasion of tumour cells through the basement membrane (Stetler-Stevenson et
aL 1993). Because progelaûnase A is produced constitutively in high concentration by
many ce11 types, its activity is regulated mainly at the level of proenzyme activation
whereas the expression of other MMPs such as fibroblast collagenase, strornelysin and
gelatinase B is enhanced by various growth factors (Birkedal-Hansen et al. 1993). By
gelatin zymography, MTI-MMP has been demonstrated to induce progelatinase A
activation (Sato et al. 1994). This rnechanism will be discussed below.
There is growing evidence that MT-MMP rnay be a key regulator of ECM turnover.
MTI-MMP deficient mice exhibit inadequate collagen turnover which leads to dwarfism,
osteopenia, arthritis, and connective tissue disease (Holmbeck er al. 1999). MTI-MMP
deficiency greatly affects the skeleton, seen in reduced longitudinal growth, cranial
dysmorphism, osteoclasia and osteopenia. MT1 -MMP is also essential for growth plate
function and secondary ossification. As bone formation requires remodeling of
unmineralized comective tissue, when MTI-MMP is deficient, the remodeling of the
periskeletal soft tissues is impaired. The block of remodelling leads to increased bone
resorption and osteoclastic activity. Collectively. these finding demonstrate the
importance of MTI-MMP since this enzyme cannot be adequately compensated by other
MMPs or by other collagen degrading mechanisms (Holmbeck et al. 1999). In
corroboration of these fmdings, Zhou et al. (2000) have also demonstrated that MT1-
MMP deficient mice have impaired endochondral ossification in skeletal development
and impaired angiogenesis. However. neither of these studies examined perturbations of
inflammation, so the relative importance of MTI-MMP in the idammatory response is
unknown*
B. Replation & Activation of MT-MMP
The processes that regulate the synthesis and activation of MT1-MMP are not clearly
understood. In some papers (Lehti et al. 1998, Stanton et al. 1988, Hernandez-Barrantes
et al. 2000), the latent form of MT1-MMP is reported to be 60 kDa and the active form is
57 kDa However, the active MTl-MMP can be M e r processed to a functionaily
inactive form (44 kDa) that lacks the entire catalytic domain but maintains the
hemopexin-like domain and hinge region (Hemandez-Barrantes et al. 2000). MTI-MMP
expression is regulated in part by cytoskeleton-ECM interactions. It is induced in
fibroblasts by culturing in three dimensional collagen matrices, by mechanical stretching
and by treatment of cells with the cytoskeleton disrupting agent cytochalasin D (Gilles et
al. 1997, Tyagi et al., 1998, Ailenberge Br Silveman 1996).
MTl-MMP is unlike most other MMP genes. The majority of MMP genes are either not
expressed in unstimulated cells or are expressed at low levels; their gene expression is
increased in the presence of proinflammatory cytokines. On the other hand, in vitro,
MTl-MMP is constitutively expressed by different ce11 types including fibroblasts,
endothelid cells and smooth muscle cells. Its expression is not dfected by TGF-P but is
decreased by dexamethasone and only modestly enhanced by PMA, TNF-a, and the
lectin concanavalin A (Lohi et al. 1996). Thus. other aspects of MT-MMP regulation
need to be considered such as its activation mechanism.
Like the other W s , MTI-MMP contains a propeptide domain that maintains latency.
As mentioned before, an 11 amino acid sequence that contains a stromelysin-3-like
RXKR furin cleavage
facilitate constitutive
motif precedes the catalytic domain. This motif has been shown to
intracellular processing and activation of stomelysin-3 (Pei and
Weiss 1995) and MTI-MMP itself. The sequence is recognized by furin, a proprotein
convertase present in the Golgi apparatus that is able to activate recombinant MTl-MMP
(Sato et al. 1996). When fin specifically cleaves MTI-MMP in vitro between Arg"'-
T ~ T " ~ , MTl-MMP is activated, which in tum enables MTI-MMP to activate
progelatinase A. This contention is supported by Pei and Weiss (1996) who reported
activation of MTI-MMP using a mutant protein lacking the trammembrane domain
purified From MDCK cells. However, Cao et al. (1 996) demonstrated that hirin-induced
activation of MTI-MMP is not a prerequisite for progelatinase activation. Activation of
gelatinase continued despite the presence of aiPIpirr, a M n inhibitor, and within COS4
cells cotransfected with furin and mutant forms of MT1 -MMP in the RRKR"' site. The
discrepancy between these studies is explained by Yana & Weiss (2000). First, they
reported that alPIpm h i n inhibitor is unable to efficiently inhibit the proprotein
convertase-dependent pathways. Second, it was discovered that the MTl-MMP contains
a secondary M n cleavage site of mg9, which was only used when RRKR"' was
mutated. Thus in Cao's study, îürin converted mutant MTl-MMP to an active form
through cleavage at the secondary site. The activated MTI-MMP was then able to active
progelatinase A.
Another possible activation mechanism suggests that plasmin can activate the pro-MT1 -
MMP by cleavage of Arglo8 and ~ r ~ " ' (Okumura 1997). These authors suggested that
the pro-MT1 -MMP is activated by plasmin extraceIIularIy.
C. Substrate Specifieity of MT-MMP
MT1 -MMPY s substrates include a variety of ECM proteins such a s type 1, II and III
collagens, gelatin, fibronectin, vitronectin, tenascin, entactin and laminin-1 (Ochuchi et
al. 1997, d'Orth0 et al. 1997). In cornparison to MMP- 1, MT1-MMP is 5-7.1 fold less
efficient at cleaving type 1 collagen while its gelatinolytic activity is 8-fold higher. MTI-
MMP is also an efficient fibrinolytic proteinase: overexpression of MT1 -MMP enables
non-fibrinolytic cells to invade fibrin gels (Hiraoka er al. 1998). However, more
significantly, other MMPs are substrates for MT-MMPs. MTI-MMP has been implicated
in the activation of pro-MMP2 and pro MMP 13 (Sato et al. 1994, Murphy et al. 1999.
Knauper et a[. 1996a).
nie fint MMP substrate discovered for a MT-MMP was progelatinase A. Sato et al.
( 1994) transfected MT I -MMP plasmid into human fibrosarcoma UT 1080 cells which
constitutively secrete pro-gelatainase A into the culture medium. MT1 -MMP generated
two new gelatinolytic bands which represented the intermediate and active forrns of
gelatinase A. MT1 -MW recognizes the ~ s n ~ ~ - ~ e u bond (Kinoshita et al. 1996). After
cleavage, the progelatinase A is subsequently processed to an intennediate form. Then. an
autoproteolytic reaction occurs to produce the fully active fom which is dependent on the
presence of gelatinase A at the ce11 surface (Sato et al. 1996a). This enzymatic activity is
inhibited by TIMP-2 and TIMP3, but not by n M P l (Will et al. 1996). TIMP-2 has a
complex and critical regulatory role in the activity of MTI-MW. In vitro experiments
revealed that both MT1-MMP and TIMP-2 are required for the binding of gelatinase A to
the MTI-MMP at the cell surface and for subsequent membrane activation of gelatinase
A (Strongin et al. 1995, Atkinson et al. 1995).
D. Relationship to TIMP-2
TIMP-2, a 21 kDa non-glycosylated protein. belongs to a family of endogenous
inhibitors, the tissue inhibitors of metalloproteinases. These proteins fonn non-covalent
stoichiometric complexes with both latent and active MMPs. TIMPs regulate rnatrix
degradation mainly by bloc kage of autolytic MMP activation (DeClerck et al. 1 99 1 ).
Previous studies have s hown that at low concentrations, TIMP -2 stimulates progelatinase
A activation whereas at high concentrations. it inhibits activation (Strongin et al. 1995.
Kinoshita et al. 1998). Strongin (1 995) also demonstrated in cross-linking experiments
that TIMP-2 binds to MT1-MMP and to the hemopexin-like domain of progelatinase A.
Based on these observations. a mode1 for activation of progelatinase A was proposed in
which the catalytic domain of MTLMMP binds to the N-terminus of TMP-2 while the
C-terminus of TIMP- binds the C-terminus of progelatinase A (Strongin et al. 1995,
Butler et al. 1998). This complex enables progelatinase A to cluster at the ce11 surface
near a second active MTI-MMP molecule (Fig. B). If not bound by TIMP-2, this second
MTI-MMP wi11 then cleave the propeptide of progelatinase A to initiate activation of
geiatinaseA.
Figure 8: Activation of MMP-2 (progelatinase A) by MT1 -MMP
Figure C: Excess of TIMP-2, no activation of MMP-2
Because TIMP-2 is a potent and specific inhibitor of MTI-MMP and gelatinase A, the
optimal concentration of TIMP-2 to enable gelatinase A activation is quite low and is
within a narrow range. This mechanism permits active MTI-MMP to activate the
progelatinase A bound ont0 the complex. An excess of TIMP-2 will lead to the inhibition
of the activation reaction (Fig. C).
In addition to regulation of progelatinase A activation, TIMP- has an unique interaction
with MTI-MMP in tems of the activity of the enzyme on the ceIl surface. With
recombinant vaccinia viruses encoding Full length MT1-MMP or TIMP-2. Hernandez-
Barrantes (2000) was able to control various expression levels of TIMP-2 in mammalian
cells. The resdts demonstrated that TIMP-2 directly and positively regulated the
concentration of active M T 1 - W . In the absence of TIMP-2, there is a significant
decrease in the amount of active MT1-MMP on the cell surface due to the generation of
membrane-bound inactive 44 kDa species. However, in cells that CO-express MT 1 -MMP
and TIMP-2, cleavage at the ~ l ~ - ~ l # ~ ~ and the subsequent generation of the inactive
f o m is significantly inhibited by TIMP-2. The authors of this study concluded that
because TIMP-2 and not TIMP-1 inhibits the generation of 44 kDa species, the MTI-
MMP processing is an autocatalytic event. Thus, by binding and inhibiting a fraction of
active MTI-MMP, TIMP-2 reduces the extent of autocatalytic processing of free active
57 kDa species and therefore promotes its accumulation on the ce11 surface. This in tum
enhances progelatirme A activation.
E. Membrane Fixation
The complex of the MTI-MhPTIMP-2 with progelatinase A illustrates the concept of
pericellular proteolysis (Werb 199'7). As mentioned before, MMPs are the major enzymes
that degrade the ECM. The majority of these enzymes are soluble and are released
extracellularly. However, previous studies have shown that ECM degradation in vivo is
confined to the immediate pericellular environment of the ce11 (Andreasen et al. 1997.
Nakahara et al. 1997). This Ieads to the question of how the proteinases that degrade the
ECM can operate in a spatially confined manner. One possible mechanism is the
involvement of membrane-bound enzymes such as MT-MMP. Not only is MTI-MMP
capable of degrading ECM molecules, it is also capable of localizing and activating other
MMPs at the cell surface such as gelatinase A. This activation and Iocalization
mechanism may be applied to other MMPs and other disease processes but this has not
yet been investigated, most notably for MMP-8.
IV. Introduction and Statement of the riroblem
Periodontal disrases are a group of chronic infiammatory disorders which involve the
destruction of extracellular matrices (ECM). One of the destructive enzymes that
degrades collagen in the ECM is the neutrophil collagenase or matrix metalloproteinase-8
(MW-8). MMP-8 is released as an inactive zyrnogen which is subsequently activated
extracellularly by cleavage of the propeptide. Activation of MMP-2 (gelatinase) by a
membrane bound MMP, membrane-type- 1 MMP (MT 1 -MMP) has been demonstrated in
fibroblasts, chondrocytes, and various tumour tissue ce11 lines (Sato et ai. 1994). There is
growing evidence that MT-MMPs may be key regulators of ECM turnover. For exarnple,
MTI-MMP deficient mice exhibits inadequate collagen turnover which leads to
dwarfism, osteopenia, arthntis, and connective tissue disease (Holrnbeck et al. 1999).
However, there is limited information on activation of MMP-8 by MT-MMPs in
neutrophils or indeed whether MT-MMPs are expressed on the neutrophil plasma
membrane at ail. My hypothesis is that MT-MMPs are expressed in neutrophils and can
activate latent MMP-8. To address this hypothesis. rny objectives were to:
1) Determine if MTI-MMP mRNA is expressed in peripheral blood neutrophils by RT-
PCR
2) Determine if MTl-MMP protein is found in the plasma membrane or in granules of
penpheral blood neutrophils
3) Compare by SBA the amount of MMP-8 activation by neutrophil membrane Fractions
that may contain MT-MMPs.
MATERIALS AND METHODS
A. Reagents
h u n o P u r e @ NHS-LC-Biotin for collagen Iabeling was purchased korn Pierce
(Rockford, Illinois). Anti-MT1-MMP rabbit polyclonal antibody used for
irnmunofluorescence and Westem blots was from Sigma (St. Louis, Missouri). Anti-
nebulin monoclonai antibody used for immunofluorescence was from Sigma. Rhodamine
(TEüTC)-conjugated affinipure F(ab')? fragment donkey anti-rabbit IgG (WL) used for
irnmunofluorescence was fiom Jackson Irnmunoresearch Lab, Inc. (West Grove,
Pennsylvania). Anti-MMP-8 mouse monoclonal antibody (IgG) used for Westem blots
was from Calbiochem (San Diego, California). ECL reagents were fiom Amersham
International (Buckinghamshire, UK). Anti-mouse IgGi HRP conjugates (Caltag
Laboratories, Burlingame, CA) and anti-rabbit IgG HRP conjugates used for Western
blots were from Arnersharn Life Science (Buckinghamshire, UK.) Puified MMP-8 was
from Calbiochem (San Diego, California). P-aminophenylmercuric acetate (APMA) was
fiom Sigma. Pefabloc was from Boehringer Mannheim (Laval, Quebec, Canada). MTI-
MMP PCR primers were kindly donated by Dr. C. Overall (UBC, Vancouver, BC).
B. Isolation of PoIymorphonuclear Neutrophils by Human Plasma/Percoll Blood
Ce11 Separation
Blood (80 ml) was drawn from male donors with a 19 gauge butterfly neede and
collected into 3.8% sodium citrate. The blood was centrifuged at room temperature for 20
minutes at 175 x g (1080 rpm, Beckman GPR centrifuge) to separate the platelet nch
plasma (PRP) layer from red ce11 fraction. The PRP layer was centrifuged at 1000 x g
(3400 rpm) at room temperature for 15 minutes to obtain the platelet poor plasma (PPP)
layer. Dextran sedimentation was performed on the red ce11 fraction for 30 minutes to
facilitate sedimentation of erythrocytes. and the leukocyte-rich supematant was aspirated.
Percoll in PPP (2 ml of 42%; v/v) was layered under the leukocyte-rich supematant with
a baked pasteur pipette and then, 2 ml of 51% (vlv) Percoll in PPP was layered under the
42% layer to form a plasma/Percoll gradient. Centrifugation was camed out at 180 x g
(1 l8Orpm) at room temperature for 10 minutes to separate the rnononuclear, neutrophil
and erythrocyte layers. The neutrophil layer was collected and resuspended in Kreb's
Ringer Phosphate with Dextrose (KRPD) buffer at a concentration of 8x10~ cellsfml.
Under light microscopy, the number and purity of neutrophils were determined. The yield
was approximately 1 . 5 ~ 1 o8 cells containhg 90-95% neutrophils, 1 2 % erythrocytes, 3-
5% eosinophils and < 0.5% rnononuclear cells.
C. Isolation of PMN Plasma Membranes (y-fraction) and Specific Granules (B-
fraction) by Discontinuous Percoll Gradients
The protocol used to isolate the plasma membrane and specific granules was based on
Borregaard (1 983) with some modifications. Isolated neutrophils were resuspended in 1X
PIPES with Pefabloc at 5x10' cellslml. Nitrogen cavitation was performed at 4°C for 8
minutes at 700 psi to disrupt the cells. The cavitate was collected in IOX EGTA and
centnfuged at 4°C for 10 minutes at 1600 rpm to remove the nuclei and whole cells. For
discontinuous Percoll gradients, 4.5 ml of Percoll with a density of 1.120 g/ml? was
Iayered under 4.5 ml of Percoll (density of 1.030 g/ml). The sample (1 52 .0 ml) was
applied on top and centrifugation was carried out at 12,000rpm at 4OC for 45 minutes.
The density of the gradient was estimated fiom the migration of caiibration beads of
known density (Pharmacia Fine Chemicals) in gradients run in parallel. The mean density
of the a, p, and y fractions were 1.135 glml. 1 .O84 g M , and 1 .O26 g/ml respectively. The
three fractions were removed and ultracentrifuged at 4OC for 90 minutes at 35.000 T m to
remove the Percoll.
D. Identification of MT1-MMP mRNA Expression by RT-PCR
Total RNA was isolated fiom cells by the QIAGEN RNAeasy Total RNA kit according
to the manufacturer's instructions and quantified by spectrophotometry (Ultrospec 3000;
Pharmacia Biotech; Montreal, Quebec). cDNA was produced fIom 5 pg of total RNA per
sarnple, using Superscript II reverse transcriptase (GIBCO) primed with random hexamer
prirners. RNA (5 pg) was incubated with random hexamers (5.0 pl of 100 @pl in
DEPC-water) to a final volume of 27 pl at 70°C for 10 minutes. The buffer contained 6
pl of 0.1 M DTT, 8 pi of 5 m M mixed dNTP, 12 pl of 5X fint strand buffer and 3 pl of
RNA guard and the reaction was initiated at 25OC for 5 minutes. Superscript II (3 pl) was
added and incubated at 2S°C for an additional I O minutes and then heated to 42°C for 1
hour. The terminai reaction was conducted at 70°C for 15 minutes. As a negative control.
a separate reaction was performed without reverse transcriptase to ensure no genomic
DNA were present PCR primers were as follows:
MT1 -MMP: 5'-GGGCCCAACATCTGTGAC-3' and 5'-CCCATCCAGTCC-3'
GADPH: 5 '-GGCATGGACTGTGGTCATGA-3 'and5'-TCACCACCATGGAGAAGGC-3'
PCR reactions were started at 94°C for 1 minutes, continued for 35 cycles at 94OC for 30
seconds, 50°C for 30 seconds, and 7 2 T for 30 seconds followed by a 10 minute
extension at 72OC. PCR reactions were carried out in separate reaction vessels in a
volume of 50 @ with 0.5 pbf primes. 1.5 mM MCJCI~: 3 111 cDNA, 200 p.!! ~NTPs a d
2.5 U Taq DNA polymerase. Products were run on 1.5% agarose gels at 120 V for 1 hour
and visuaiized under UV light.
E. Imrnunolocalization of MT1-MMP
In 8-chamber well glass slides (LaboTeka), neutrophils (5 x 10' per well) were incubated
at 37°C for 5 minutes to allow neutrophil adherence to the slides. The cells were fixed
with ice-cold methanol for 6 minutes at -ZO°C, washed twice with PBS and non-specific
binding sites were blocked with 0.2% BSA at room temperature for 10 minutes. After
washing two times with PBS, the cells were incubated with anti-MT1-MMP antibody (10
pg/ml) in a humidor at 4°C overnight. The cells were washed twice with PBS and
blocked with 0.2% BSA at room temperature for 10 minutes. TRITC goat anti-rabbit
F(ab')2 fragment (1.5 mg/ml at 1: 100) was added for 2 hours at 4°C. Cells were washed
twice with PBS, stained with DAPI (1 pg/ml in 0.01% Nonidet and PBS) for 2 minutes
and washed again with PBS. The cells were covered with irnmunofloure (ICN
Biomedicais. Aurom, OH) rnounted and examined irnmediately by epifluorescence
microscopy. Fluorescence micrognphs were recorded on Kodak High-Speed TMX
P3200 film. The positive controls included human gingival fibroblasts that were grown to
subconfluence in wells coated with fibronectin (10 pg/ml). Negative controls were
neutrophils treated without primary antibody or neutrophils treated with anti-nebulin
antibody (76 pg/ml) as the primary antibody.
F. Western Blot Analysis
Human gingival fibmblast lysates, neutrophil lysates and neutrophil cellular fractions (a,
p, y) were incubated in electrophoresis sample buffer (0.05 M Tris-HCI pH 6.8, 8 M urea,
2% (w/v) SDS, 8% (vh) Bromphenol blue) and 0.15 M of dithiothreitol ( D m
(SBDTT), boiled for 5 minutes and separated by SDS-PAGE on 10% cross-linked
minigels at 120 V for 1 % hours and transferred for 2 hours (64 &gel) onto
nitrocellulose membranes. M e r blocking with 5% ( w h ) Carnation milk, membranes
were incubated with anti-MT1-MMP rabbit antibody (2 pg/ml) or with anti-MMP-8
mouse antibody (5 pg/rnl) for 3 hours at room temperature, waçhed and incubated with
either anti-rabbit IgG HRP (1:3000 dilution) or anti-mouse IgGi HRP (1 2000 dilution)
for I h o u at room temperature. The bands were visualized with ECL reagents and
molecular weights of the fragments were detemined by cornparison with prestained
molecular weight standards. Treatment of blots with secondary antibody alone sho wed no
reactivity when the secondary antibody was incubated in 5% (wh) Carnation milk in
TBS-Tween.
G. Soluble Biotinylated Collagen Assay (SBA)
Aliquots of samples were incubated with biotinylated collagen at ratios of 10 ng
biotinylated collagedpl sample, for 21 hours at room temperature in the presence of
collagenase assay buf5er (CAB - 0.05M Tris, 0.2 M NaCl containing 5 mM CaCl?, 0.5
pVml BRU 35 and 0.2 @ml NaN3). Biotinylated collagen was prepared as described in
Mancini et al. (1999). Reactions were terminated by the addition of SBiDTT and boiled
for 5 minutes. Collagen fragments were separated on 7.5% cross-linked SDS-PAGE gels
and transferred to a nitrocellulose membrane using the PHAST system (Pharmacia). After
blocking with 5% (wh) Carnation milk in TBS-Tween, the membranes were rinsed with
TBS (0.02 M Tris-HC1. 0.14 M NaCl pH 7.6 containing 0.1% Tween 20) and incubated
with HRP-labeled streptavidin (Amersham) diluted Ill 500 in Tris-HCI b a e r pH 7.6.
Afier a second wash, ECL reagents were used for detection of biotinylated collagen
fragments by cherniluminescence. Autoradiographs were scanned and full Iength and %
a-chahs were quantified by cornputer analysis using the IP Lab Gel Scientific Image
Processing prograrn (Signal Analytics- V i e ~ a , Virginia USA). The estimation of
collagenase activity was based on the densitometric data in terms of percentage
biotinylated collagen degradation into % a-chains.
% collagen degraded = densitv of Yt a1 (1) chahs X IO0 densities of % al (1) chains + a l (1) chains
H. Preparation of samples for SBA
The total protein yield of y-fractions and p-fractions fiom the discontinuous Percoll
gradients was dependent on the donor and the initial amount of neutrophils isolated.
From 1.5~10' cells, the total protein obtained was 96 pg and 287 pg for y-fractions and P-
fractions respectively. In preparation for SBA andysis, various combination of these
fractions were rnixed and incubated for 4 hours rotating at 37OC. The samptes included
150 p1 of y, 150 pl y + 50 pl P, 150 pl of P with a total protein concentration of O. 130
pg/pI for the y-fraction and 0.379 pg/$ for the p-fraction. An aliquot (100 pl) of the P
fraction was activated with the addition of 50 p1 1 m M APMA and incubated at room
temperature for 45 minutes. Then, 50 pl from samples were removed and cornbined with
6 pl 10X CAB and 2 pl biotinylated collagen for SBA analysis.
A. MT1-MMP mRNA expression in PMN
Expression of MTI-MMP mRNA was exarnined in whole ce11 lysates of peripheral blood
neutrophils and human gingival fibroblasts by RT-PCR. PCR primers for a 236 bp
GAPDH product were used as a positive control to ensure the presence of cellular mRNA
(Fig. 1). The MT1-MMP primes identified a 580 bp MTI-MMP product in both the
neutrophil and HGF samples. No gene pmduct was identified in a separate reaction
without reverse transcriptase to ensure the absence of genomic DNA.
B. Immunolocalization of MT1-MMP in HGF and PMN
Immunolocalization studies were done to detemine the presence of MTl-MMP. As a
positive control, human gingival fibroblasts (HGF) exhibited strong staining for MT1 - MMP (Fig. 2A) and within the same field, the DAPl stained nuclei were evident (Fig.
2B). No significant immunoreactivity was observed in HGF preparations stained with an
irrelevant antibody, anti-nebulin, (Fig. 2C) or with no primary antibody (Fig. 2E).
Peripheral blood neutrophils stained with ah-MT1-MMP also showed bright staining
(Fig. 3A). The neutrophils were identified based on morphology under light microscopy
and DAPI staining demonstrated the multilobuiar appearance of the neutrophil nuclei.
Neutrophil preparations stained with an irrelevant antibody, anti-nebulin (Fig. 3C) or with
no primary antibody (Fig. 3E) showed no significant immunoreactivity. 1 used a nebulin
antibody as a control because nebulin is a hi& molecular weight protein that is
specifically localized in skeletal muscle rnyofibrils. This immunolocdization study was
HGF PMN
236 bp GAPDH
Figure 1: MTI-MMP mRNA expression in HGF and PMN. MTI-MMP mRNA expression was tested by RT-PCR on samples of whole ce11 lysates of human gingival fibroblasts (HGF) and peripheral blood neutrophils (PMN). Rimers 142 and 90 identified a 580 bp o f MTI-MMP in PMN and HGF. The primer for GAPDH positive control also appears in both samples as 236 bp. -RT lanes were reactions with no reverse transcriptase produced no gene products.
Figure 2: Immunolocaihation of MT1-MMP. A) Preparation of human gingival fibroblasts (HGF) stained wi th anti-MT 1 -MMP primary antibody and TRITC-conjugated secondary antibody. B) same field shown in panel A stained with DAPI. C) PMN stained with ad-nebulin primary antibody and TNTC-conjugated secondary antibody. D) same field shown in panel C stained with DAPI. E) PMN s h e d with no phary antibody and TRITC- conjugated secondary antibody. F) same field shown in panel E s h e d with
I DAPI,
- --
Figure 3: Immunolocaluation of MT1-MMP. A) Reparation of peripheral blood neutrophils (PMN) stained with anti-MT 1 -MMP primary antibody and TRITC-conjugated secondary antibody. B) same field shown in panel A stained with DAPI. C) PMN stained with anti-nebulin primary antibody and TMTC-conjugated secondary antibody. D) same field shown in panel C stained with DAPI. E) PMN stained with no primary antibody and TRITC-conjugated secondary antibody. F) same field shown in panel E stained with DAPI.
completed on neutrophil samples from three different donors with similar results. In ail
samples, the neutrophils incubated with anti-MT1-MMP antibody as the primary
antibody exhibited strong fluorescence with minimal background signais.
C. Identification of MTl-MMP by Western Blot Analysis
Sarnples of human gingival fibroblasts, penpheral blood neutrophils and fractions thereof
were separated by SDS-PAGE and transferred onto nitrocellulose membranes. M e r
incubation with anti-MT1-MMP, three bands were identified at 86 kDa, 80 ma, and 62
kDa for the human gingival fibroblast sample (HGF) (Fig. 4). The same banding pattern
appeared for the plasma membrane fraction (y) fkom neutrophils. For the whole ce11
neutrophil lysate sample (PMN), the 80 kDa and 62 kDa bands were present with
additional bands at 69 kDa and 66 kDa. No bands were detected in the a and P fractions
(Fig. 4). This Western Blot anaiysis was repeated with neutrophil samples from three
other human donon and each blot produced the same banding pattern. The 62 kDa band
is the latent forrn of MTI-MMP and was present in the human gingival fibroblast (HGF),
whole ce11 neutrophil lysates (PMN) and the plasma membrane fraction (y). MT1-MMP
was not present in the specific (p-fraction) or in azurophilic (a-fraction) granules.
D. Verification of specific granules (p-fraction) as a source of latent MMP-8
Samples of purified human MMP-8 (Calbiochem). neutrophil plasma membrane (y-
fraction) and specific granules @-fraction) were separated by SDS-PAGE and transferred
onIo nitrocellulose membrane. M e r incubation with anti-MMP-8, the purified human
MMP-8 sample exhibited two bands at 65 kDa and 55 kDa which represented the latent
PMN HGF a B Y PMN HGF
+ + no primary Ab anti-MT1-MMP primary Ab
Figure 4: Identification of MT1-MMP. Samples of whole ce11 lysates o f peripheral blood neutrophils (PMN) and postnuclear cavitate fractions of y (plasma membrane), f3 (specific granules), a (azurophil granules) were tested by Western blot for immunoreactive MTI-MMP. The proform of MTI-MMP of 62 kDa is predominant in PMN and y (plasma membrane) fhctions. The positive control of human gingival fibroblast (HGF) also demonstrates the presence of MTI-MMP. The samples incubated with no primary antibody demonstrate minimal cross-reactivity.
and active form of MMP-8 respectively. The specific granules sample @-fraction)
exhibited o d y the 65 kDA latent form of MMP-8 whereas in the neutrophil plasma
membrane sample (y-fraction), MMP-8 was not detected (Fig. 5).
Collagenase activity of the specific granules sample Ip-fraction) was tested hy SBA (Fip.
6). Analysis of an diquot (56.9 pg total protein) representing 6.5% of the p-fraction
reveded negligible (1%) digestion of the biotinylated collagen. An equivalent aliquot
(6.5% or 58.1 pg total protein) of specific granules from another donor digested 3% of
the collagen. However, when an aliquot representing 4.3% or 37.3 pg total protein of P-
fraction was treated with APMA, the collagenase activity of the specific granules sample
generated 86% collagen digestion. Thus, the specific granules (P-hction) contain latent
MMP-8 that could be activated. Both specific granule sarnples (P-Fraction) that were
obtained fiom two difTerent human donors produced similar results from the anti-MMP-8
Westem Blots analysis and SBA.
E. Collagenase activity in neutrophil plasma membrane (y-fraction)
The neutrophil plasma membrane sample (y-fraction) contained MT1-MMP as
demonstrated by the ad-MU-MMP immunofluorescence (Fig. 3A) and Westem blot
analysis (Fig. J), but the sample did not contain any forms of MMP-8 as demonstrated by
the anti-blMP-8 Westem Blot anal ysis (Fig. 5).
However, the neutrophil plasma membrane sample (y-fraction) did exhibit some
collagenase activity. Analysis by SBA showed that an aliquot (6.7% or 19.5 pg total
Figure 5: Identification of MMP-8 Western blot analysis with anti- MMP-8 monoclonal antibody demonstrates the presence of latent 65kDa MMP-8 in the P fractions and absence of MMP-8 in the y hctions. Rire human MMP-8 serves as a positive control dernonstrating the presence of both the latent and active form of MMP-8.
APMA
1 % % Digestion
Figure 6: Collagenase activity in plasma membrane (y-fraction) and opecific granules (p-fraction). y, P, APMA activated p, and
samples were tested by SBA for collagenase activity. y alone had some intrinsic collagenase activity while P alone had no activity. B can be activated by the addition of APMA. y + P demonstrated an increase in digestion when combined.
protein) of plasma membrane fraction generated 28% collagen digestion (Fig. 6) . Another
equivalent aliquot (6.7% or 20.4 pg total protein) of plasma membrane fraction from
another donor digested 24% of the collagen.
F. Increased collagenase activity of specific granules @-fraction) latent MMP-8 by
neutrophil plasma membrane
The incubation of an aliquot (19 pg total protein) representing 5% of neutrophil plasma
membrane fraction with an diquot (1 -6% or 19 pg total protein) of specific granules
sample @-fraction) digested 54% of the collagen (Fig. 6). Aliquot mixtures of similar
proportions fiom another donor generated 40% collagen digestion.
DISCUSSION
The main findings of this study are that MTI-MMP is expressed in peripheral blood
neutrophils and is e ~ c h e d in the plasma membrane fraction of these celis. In vitro
experiments showed that cornponents of the plasma membrane fraction can activate the
latent collagenase From the specific granules of neutrophils.
A. Expression of MTI-MMP in Penp heral BIood Neutrophils
RT-PCR demonstrated that peripheral blood neutrophils expressed MTI-MMP mRNA
while immunofluorescence and Western blots showed the presence of the MTI-MMP
protein. I used human gingival fibroblasts as positive controls for MTI-MMP and, as
expected, these cells showed constitutive expression of MTI-MW mRNA and protein as
previously described (Atkinson et al. 1995, Gilles et al. 1997). Many other ce11 types
such as smooth muscle cells, endothelid cells, arneloblasts, odontoblasts, chondrocytes.
and osteoclasts express MTI-MMP (Imai et al. 1997, Sato & Seiki 1996, Sato et al.
1997) however to the best of my knowledge, this is the first report of MTl-MMP
expression by neutrophils.
In experiments using nitrogen-cavitated cells separated by density centrifugation in
Percoll gradients, f localized MTl-MMP to the plasma membrane fractions as
demonstrated by Western blots of the various neutrophil fractions. This finding is
consistent with previous studies showing that MT1-MMP is found in the plasma
membrane preparation of transfected COS4 cells, human skin fibroblasts and osteoclasts
(Sato et al. 1994, Atkinson et al. 1995, Sato et al. 1997). In view of the structure of MT1-
h4MP which contains a 24 amino acid hydrophobie h?uis-membrane domain, it would be
expected that MTI-MMP wouid be tethered to the neutrophil membrane by the tram-
membrane region.
B. Forms of MTI-MMP
By Westem blotting 1 only found the 62 kDa (latent) form of MTI-MMP in neutrophils.
MTl-MMP exists in three forms. The latent fom that contains a propeptide domain is 60
D a form. Cleavage of this propeptide domain results in a 57 kDa form which c m
activate progelatinase A. However, in the absence of TIMP-2, this 57 kDa fom
undergoes autocatalytic conversion to a functionally inactive form (44 kDa) that lacks the
entire cataiytic domain but maintains the hemopexin-like domain and hinge region
(Hemandez-Barrantes et al. 2000). The absence of the active and the functionally
inactive MTl-MMP foms in the Westem blots may be amibuted in part to the neutrophil
preparation method since isolation procedures can exert profound effects on the
activation systems that regulate the proteolytic machinery of these cells (Pabst 1994).
Cognimt of the tendency of neutrophils to be readily activated by inappropriate isolation
procedures, special precautions were taken to prevent ce11 activation following
venipuncture in healthy adult human volunteers. For example, as neutrophils can be
activated by trace amounts of lipopolysaccharides (LPS), 1 used LPS-fiee solutions.
disposable plastics and glassware baked at 180°C for several hours. It is thought that if
these special precautions are taken, neutrophils isolated From circulating biood are largly
in a "resting state" (Haslett et al. 1985). Thus in resting neutrophils, the annamentarium
used for bacterial defence is inactive. Notably, the components of the NADPH oxidase
system are stored in granules and therefore in resting neutrophils the granule contents are
not assembled and are not prepared for destruction of phagocytosed organisms (Weiss
1989). Similarly, the neutrophil collagenase (MMP-8) is stored in the latent, pro-enzyme
form in the specific granules pnor to release and extracellular activation (Doherty et al.
1994). By analogy 1 suggest that in resting neutrophils the MT1 -MMP is in a latent fnm
because it has not been activated to engage in proteolysis. Thus the 57 D a active form
and the 44 kDa functionally inactive form of MTl-MMP were not detected in the
Western blot analysis. Currently, it is not known what mechanism rnay activate MTI-
MMP on the surface of neutrophils. However, 1 speculate that the 57 kDa active form is
present in activated neutrophils. It is likely that this activation mechanism would exert an
important effect on the abiiity of MTI-MMP to activate MMP-8 and thereby initiate
degradation of rnatrix proteins. Thus, if MT1 -MMP cm cleave the N-terminal propeptide
of MW-8, it may be able to regulate the collagenolytic activity of MMP-8.
In vitro, a number of activation mechanisms have been studied which produce different
forms of MMP-8 and with varying levels of collagenolytic activity. For example,
stromelysin-2 processes MMP-8 by a single-step activation mechanism by cleavage of
the ~ l ~ ~ ~ - ~ h e ~ ~ peptide bond in the N-terminal propeptide domain. This active W - 8
displays very high specific collagenolytic activity (Knauper ef al. 1996b). On the other
hand. trypsin requires a two-step activation mechanism in which the fmt cleavage occurs
at Ar&~he' '~ to generate an intermediate latent forrn and then a second cleavage at
~ r ~ ' ~ - C ~ s ' ' to produce an active MMP-8 (Knauper et al. 1990). MMP-8 activation by
HgClz follows a three-step mechanism where the first cleavage is at ~ s n ' ~ - ~ a l ~ ~ . Then,
autoproteolytic cleavage of ~ s ~ ~ ~ - ~ e t ~ ~ produces an intermediate form ( ~ e t ~ j N-
terminus) which displays only about 40% of the maximum collagenolytic activity. Final
activation occurs d e r autoproteolytic cleavage of either P he7'-~et80 or ~ e t ' ' - ~ e u ~ ' (Blaser et al. 199 1).
C. MMP-8 activation by MT1-MMP
Currently, the system(s) by which MMP-8 is activated in vivo is not known although, as
discussed above, a large nurnber of in vitro studies have implicated stromelysin (Knauper
et al. 1996b), oxidants (Nagase 1997), as well as other proteases such as trypsin and
chymotrp ysin (Knauper et al. 1 990).
Initiai pilot experiments to determine if MTI-MMP can activate MMP-8 involved the
incubation of recombinant MTI-MMP with latent rat MMP-8. Based on the SBA, this
combination resulted in complete digestion of the biotinylated collagen. However, the
latent rat MMP-8 which was supposed to be latent produced > 50% digestion.
Unfortunately, this experiment was not reproduced due to the lack of more recombinant
MTl-MMP. Analysis of MTl-MMP mediated activation of MW-8 was also
cornplicated by the dificulty in obtaining a source of pure latent MW-8. The latent rat
MMP-8 most likely was zctivated during shipment or during storage, so I decided to use
the specific grandes from neutrophils as a source of latent MMP-8. As dernonstrated by
the Western biots (Fig. 5) and SBA (Fig. 6), the specific granules were a viable source of
latent MMP-8 than can be activated by APMA.
Since only a limited amount of recombinant MTI-MMP was available, 1 tried to generate
recombinant MTl-MMP by transfections of CHO cells with MT1-MMP plasmids. While
the transfection was successful, the predominant form of MTI-MMP generated was the
inactive form that lacked the catalytic domain. Consequently, the neutrophil plasma
membrane was used as a source of MTI-MMP, a source that required large number of
cells and arduous procedures.
As I was able to show the presence of MTl-MMP in association with the ce11 surface of
neutrophils, 1 conducted expenments in which I combined neutrophil plasma membrane
fractions with the contents of specific granules that contained latent MMP-8. In these
expenments, the membrane preparation exhibited abundant MT I -MMP and moderate
collagenase activity while the specific granules contained abundant latent MMP-8 but
minimal collagenase activity. Afier combuiing the two fractions, collagenase activity was
doubled. This finding indicates that components of the neutrophil membrane including
MTI-MMP have the potential to activate latent MMP-8. While MTI-MMP exhibits
collagenase activity, it is 5-7.1 fold less eficient than MMP-8 (Ohuchi el al. 1997).
However, if MTI-MMP can process the latent form of MMP-8 to an active form, then
this would provide an efficient and spatially segregated mechanisrn for activating MMP-8
on the surface of neutrophils. In this context, previous studies have demonstrated that
MTI-MMP activates other MMPs such as latent gelatinase A and collagenase-3 (Sato et
al. 1994, Knauper et al. 1996). Sato (1 994) transfected MTI-MMP plasmid into hurnan
fibrosarcoma HT1080. These ce11 lines also secreted progelatinase A (66 kDa) into the
culture supernatant. Plasma membrane fiactions of the transfected HT1080 (20 pg
protein) were incubated with the conditioned medium fiom HT1080 that contained
progelatinase A for two hours at 37°C. Gelatin zymography demonstrated that the plasma
membrane generated the processed 64 kDa and 62 kDa gelatinase A. Similarly, Knauper
(1996) incubated fibroblast-derived plasma membranes that contained MT1 -MMP (1 5p1;
Img/ml protein) with 50 ng of procollagenase-3. The plasma membranes processed the
procollagenase-3 to a 48 kDa active enzyme as demonstrated by Western blot analysis.
Taken together, these studies indicate that MT1-MMP activation of MMP-8 is feasible.
However, that MTI-MMP c m directly activate MMP-8 still needs to be established.
As mentioned above, in gingivitis, MMP-8 is predominantly latent, whereas, in
periodontitis, it is conceivable that the combination of a susceptible host and the presence
of periodontal pathogens provides a situation in which an activation cascade can lead to
active MMP-8. 1 suggest that vinilence factors and inflammatory cytokines may trigger
specific enzymes which lead to activation of enzymes such as funn or plasmin. These
enzymes in turn can activate latent MT1 -MMP. The active MT1 -MMP will then activate
MMP-8 which leads to destruction of the periodontium. As many cells types express
MTI-MMP, it is possible that within the penodontiurn or gingival crevice. other ce11
types (e.g. fibrblasts) that express MT1 -MW may be responsible for activation of MMP-
8. However, because neutrophils are so abundant in acute inflammation and are the
primary producers of MMP-8, it seems unlikely that other cells such as fibroblasts play a
significant role in M W - 8 activation. That extensive but revenible destruction of
gingival tissues occur in gingivitis suggests that there may be fundamental differences in
the rnechanism of tissue degradation between periodontitis and gingivitis.
D. Summary and Suggestions for Future Studies
In summary, my results demonstrate that neutrophils express MTl-MMP and the protein
is located on the plasma membrane. The combination of plasma membrane fractions with
latent MMP-8 indicates that components of the neutrophil membrane may be able to
activate MMP-8. I suggest that one of these activating molecules may be M T l - W .
However, the activation of MMP-8 by MT1-MMP still needs to be confmed by more
definitive experiments. For example, a potential future study could involve the use of an
antibody that blocks the enzymatic activity of MT1-MMP or an assay that
immunodepletes MTI-MMP From the plasma membrane fraction prior to incubation with
latent MMP-8. Under these conditions, selective removal of the MTl-MMP by antibody
should reduce the activation of M W - 8 and lead to lower collagenase activity as detected
by the SBA assay.
In considenng activation systems, 1 should also point out the involvement of TIMP-2 in
the hypotheticd activation of MMP-8 by MTl-MMP as there is increasing evidence that
TIMP-2 plays an important role in regulating the activity of MTI-MMP. Strongin et al.
(1995) suggested that TIMP-2 is required for progelatinase A (MMP-2) activation and
Hemandez-Barrantes (2000) demonstrated that TIMP-2 regulated the effective cell-
surface "concentration" of active MTl-MMP by adjusting the autocatalysis of the
enzyme and consequently its availability for interacting with gelatinase A. Notably, in the
absence of TIMP-2, there is uncontrolled autocatalysis of MTl-MMP that leads to
production of the inactive 44 kDa form. However, when there is an excess of TIMP-2, the
enzymatic action of MTI-MMP is also inhibited. In effect, a fiiture mode1 of MTI-MMP
activation of MMP-8 may involve controlling the level of TIMP-2 at the ce11 surface for
appropriate regulation of MTI-MMP and its subsequent activation of MMP-8.
The ability to regulate the activity of MMP-8 at the ce11 surface confines collagenase
activity close to the cell. This illustrates the concept of pericellular proteolysis (Werb
1997). Even though MMP-8 is released extracellularly, ECM degradation in vivo is
confined to the irnmediate pencellular environment of the ce11 (Ancireasen el al. 1997,
Nakahara et al. 1997). Because MT1-MMP is bound on the neutrophil plasma membrane
surface, it provides the oppominity to concentrate the various components involved with
ECM degradation close to the neutrophil. Thus, if MTI-MMP can activate MMP-8, then
MTl-MMP may provide a mechanisrn for controlled localized degradation of the
collagen matrix in periodontitis and other diseases in which there is a marked neutrophil
inflammatory infiltration.
CONCLUSIONS
MT I -MMP is expressed in periphed blood neutrophils.
MTl-MMP is found in the plasma membrane and not in the specific and azurophilic granules.
Specific granules fiom the P-fraction of neutrophils are a source of latent MMP-8 t h t
can be activated by APMA.
The plasma membrane hction has collagenase activity.
The collagenase activity of a plasma membrane fraction cornbined with specific granules of neutrophils is higher than the collagenase activity of the individual fractions.
REFERENCES
Ailenberg, M., Silverman, M. (1996). Ce1lda.r activation of mesangial gelatinase A by cytochalasin D is accompanied by enhanced mRNA expression of both gelatinase A and its membrane-associated gelatinase A activator (MT-MW). Biochemica! Journal 3 13, 879-884.
Ancireasen, P.A., Kjoller, L., Christensen, L, Dum, M.J. (1997). The urokinase-type plasminogen activator system in cancer metastasis: a review. Inrernntional Jozlrna! of Cancer 72, 1-22.
Apte, S.S., Fukai, N., Beier, D.R., Olsen, B.R. (1997). The matrix metalloproteinase-14 (MMP-14) gene is stmcturally distinct from other MMP genes and is CO-expressed with the TIMP-2 gene during mouse embryogenesis. JO urnal of Biological Chemistry 272, 2551 1-25517.
Armitage, G.C. (1999). Development of a Classification System for Periodontai Diseases and Conditions. Annals of Periodontology 4, 1-6
Atkinson, S.J., Crabbe, T., Cowell, S., Ward, R.V., Butler, M.J., Sato, H. Seiki, M., Reynolds, J.J., Murphy, G. (1995) Intermolecular autolpic cleavage can contribute to the activation of progelatinase A by ce11 membranes. Journal of'diological Chemistry 2 70, 30479-30285.
Bevilacqua, M.P., Pober, J.S., Mendrick, D.L., Cotran, R.S., Gimbrone, M.A. (1 987). Identification of an inducible endothelial-leukocyte adhesion molecuie, ELAM-1. Proceedkgs of Nafional Academy of Science, USA 84,9238-9242.
Birkedai-Hansen, H., Wells, B.R., Lin, H-Y., Caufield, P.W., Taylor, R.E. (1984). Activation of keratinocyte-mediated collagen (type 1) breakdown by suspected hurnan periodontopathogen: Evidence of a novel mechanism of connective tissue breakdown. Journal of Periodontal Research 19,645-650.
Birkedal-Hansen, H. (1 987). catabolism and turnover of collagens: collagenases. ltlethods in Enzymology 144: 140- 1 7 1
Birkedal-Hansen, H., Taylo, R.E., Zarnbon, J.J., Barwa, P.K., Neiden, M. (1988). Characterization of colIagenoIytic activity from strains of Bacteroides gingivalis. JO urnd of P eriodonial Research 23,258-264.
Birkedd-Hansen, H. (1993). Role of cytokines and inflammatory mediators in tissue destruction. Journal of Periodonfal Research 28,500-5 10.
Birkedal-Hansen, H., Moore, W.G.I., Bodden, M.K., Windsor, L.J., Birkedal-Hansen, B., Decarlo, A., Engler, LA. (1993). Matrix metailoproteinases: a review. Criticai Revie~vs in Oral Bioiogy and Medicine 4, 197-250.
Blaser, J., Knauper, V., Osthues, A., Reinke, H., Tschesche, H. (1991). Mercurial activation of human polymorphonuclear leucocyte procollagenase. European Journal of Biochemistry 202, 12234230.
Borregaard, N., Heiple, J.M., Skons, E.R., Clark, R.A. (1 983). Subcellular localization of the b-cytochrome component of the human neutrophil microbicidal oxidase: translocation during activation. Journal of Cell Biology 97,5206 1.
Butler, G.S., Butler, M.J., Atkinson, S.J., Will, H., Tamura, T., van Westrum, S.S., Crabbe, T., Clements, J., d'Orth, M.P., Murphy, G. (1998). Journal of Biological Chemistry 273,871-880.
Cao, J., Rehmtulla, A., Bahou, W., Zucker, S. (1996) Membrane type matrix metalloproteinase 1 activates pro-gelatinase A without f i n cleavage of the N-temiinal domain. Journal ofBiological Chemistry 27 1,30 174-30 1 80.
Cao, J., Drew, M., Lee, H.M., Corner, C., Bahou, W.F., Zucker, S. (1998). The propeptide domain of membrane type 1 matrix metalloproteinase is required for binding of tissue inhibitor of metalloproteinases and for activation of pro-geiatinase A. Journal of Biological Chemistry 273,34745-34752.
Caron, C. Xue, J., Bartlett, J.D. (1998) Expression and localization of'membrane type 1 matrix metalloproteinase in tooth tissues. Matrtr Biology 17,jO 1 -5 1 1
Dano, K., Andreasen, P.A., Grondahl-Hansen, J., Knstensen, P., Nielsen, L.S., Skriver, L. (1985) Plasrninogen Activators, Tissue Degradation and Cancer. ildvances in Cancer Research 44, 1 3 9-266.
Danteau, R.P., Tanner, A.. Page, R.C. (1997). The microbial challenge in periodontitis. Periodontology 2000 14, 1 2-3 2.
DeClerck, Y.A., Yean, T.D., Lu, H.S., Ting, J. Langley, K.E. (1991). Inhibition of autoproteolytic activation of interstitial procollagenase by recombinant metalloproteinase inhibitor MVT'MP-2. Journal of Biological Chemistïy. 266,3893-3839
Desroches, P.E., Mookhtiar, K., Van Wart, H.E., Hasty, KA., Weiss, S. J. (1992) Proteolytic inactivation of alpha 1-proteinase inhibitor and alpha 1 -antichymorrypsin by oxidatively activated human neutrophil metalloproteinases. Jotirnal of Biological Chemistry 267,5005-50 12.
Docherty, N.S., Janusz M.J. (1994). 4. Neutrophil protease: their physiological and pathologicd roles. Immunopharrnacology of Neutmphils. eds. Hellewell, P.G, Williams. T. J. Academic Press Limited.
d'ortho, M.P., Will, H., Atkinson, S., Butler, G.' Messent, A., Gavrilovic, J., Smith, B., Timpl, R., Zardi, L., Murphy, G. (1997). Membrane-type matrix metalloproteinases 1 and 2 exhibit broad-spechum proteolytic capacities comparable to many matr;x metalloproteinases. European Journal of Biochemistry 250, 75 1-757.
Freije, J.M., Diez-Itza, I., Balbin, M., Sanchez, L.M., Blasco, R., Tolivia, J., Lopez-Otin, C., (1 994) Molecular cloning and expression of collagenase-3, a novel human matrix metalloproteinase produced by breast carcinomas. Journal of Biologicol Chernistry 269, 16766- 16773-
Fukuda, M., Spooncer, E.S., Oates, LE., Dell, A., Klock, J.C. (1984). Stmcuture of sialyated fucosyl lactoaminoglycan isolated from human granulocytes. Journal of Biologieal Chemistry 259, 10925- 10935.
Gangbar, S., Overall, C.M., McCdloch, C.A., Sodek, J. (1990). Identification of polymorphonuclear collagenase and gelatinase activities in mouthrinse samples: correlation with periodontal disease activity in adult and juvenile periodontitis. Journal of Periodontal Research 25,257-267.
Gilles, C., Polette, M., Seiki, M., Birembaut, P., Thompson, E.W. (1997). Implication of collagen type 1-induced membrane-type 1 -matrix metalloproteinase expression and matrix metalloproteinase-2 activation in the metastatic progression of breast carcinoma. Laboratory Imestigution 76,65 1-660.
Gordon, M.Y. (1994). 2. Origin and Development of Neutrophils. Irnrnunopharmacology of Neutrophils. eds. Hellewell, P.G, Williams, T. I. Acadernic Press Limited
HafTajee. A.D., Socransky, S.S. (1994) Microbial etiological agents of destructive periodontal diseases. Periodontology 2000 5. 78-1 1 1.
Hanemaaijer, R. Sorsa, T, Konttinen, Y.T.? Ding, Y., Sutinen, M., Visser, H., van Hinsbergh, V.W.M., Helaakoski, T., Kainulainen, T., Ronka. H., Tschesche, H., Salo, T. (1997). Matrix metalloproteinase-8 is expressed in rheurnatoid synovial fibroblasts and endothelial cells. Journal of BioZogical Chemisny 272,3 1504-3 1509.
Haslett, C., Guthrie, L.A., Kapaniak, M.M., Johnston, R.B., Henson, P.M. (1985). Modulation of multiple neutrophil functions by preparative methods or trace concentrations of bactenal lipopolysaccharide. American Jozirnal of Pathology 1 19, 1 O 1 - 109.
Hemandez-Barrantes, S., Toth, M., Bemardo, MM., Yurkova, M.. Gervasi, D.C., Raz, Y. Sang, Q.A, Fndman, R. (2000). Binding of active (57kDa) membrane type 1-matrix metalloproteinase (MT I - M W ) to tissue inhibitor of metailoproteinase (TIMP-2) regdates MTI-MMP processing and pro-MMP-2 activation. Journal of Biological Chernistry 275,12080- 12089.
Hiraoka, N., Allen, E., Apel, I.J., Gyetko, M.R., Weiss, S.J., (1998). Matrix metalloproteinases regulate neovascularization by acting as pericellular fibrinolysins. Cell 95,365-377.
Holmbeck, K., Bianco, P., Caterina, J., Yamada, S., Kromer, M., Kuznetsov, S.A., Mankani, M., Robey, P.G., Poole, A. R., Pidow, I., Ward, J.M. Birkedd-Hansen, H. (1 999) MT1-MW-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen turnover. Cell99, 8 1-92.
Imai, K., Ohta, S., Matsumoto, T., Fujimoto, N., Sato, H., Motoham, S., Okada, Y. (1997) Expression of membrane-type 1 matrix metalloproteinase and activation of progelatinase A in human osteoarthntic cartilage. American Journal of Pathology 151 , 245-256
Kinoh, H., Sato, H., Tsunemka, Y., Takino, T., Kawashima, A., Okada, Y., Seiki, M. (1996). MT-MW, the cell surface activator of proMMP-2 (pro-gelatinase A), is expressed with its substrate in mouse tissue during embryogenesis. Journal of Cell Science 109,953-959.
Kinoshita, T., Sato, H., Takino, T., Itoh, M., Akizawa, T., Seiki, M. (1996). Processing of a precursor of 72-kilodalton type IV collgenase/gelatinase A by a recombinant membrane-type 1 matrix metalloproteinase. Cancer Research 56,2535-2538.
Kinoshita, T., Sato, H., Okada, A., Ohuchi. E., Imai, K., Okada, Y., Seiki, M. (1998). TIMP-2 promotes activation of progelatinase A by membrane-type 1 matrix metalloproteinase immobilized on agarose beads. Journal of Biological Chemistry 273, 16098-16103.
Kleiner, D.E., Stetler-Stevenson, W.G. (1999) Matrix metailoproteinases and metastasis. Cancer Chemotherapy and Phurmacology 43(suppl), S42-S5 1 .
Knauper, V., Krarner, S., Reinke, H., Tschesche (1990). Characterization and activation of procollagenase from human ppolymorphonuclear leucocytes. N-terminai sequence determination of the proenzyme and various proteolytically activated forms. Ewopean Journal of Biochemistry 1 89,295-300.
Knauper, V., Will, H., Lopez-Otin, C. Smith, B., Atkinson, S.J., Stanton, H.. Hembry, R.M., Murphy, G. (1 996a) Cellular rnechanisms for human procollagenase-3 (MMP- 13) activation. Evidence that MTI-MMP (MME'-14) and gelatinase a (MMP-2) are able to generate active enzyme. Journal of Biological Chemistry 27 1 . 1 7 1 24- 17 13 1.
Knauper, V., Mwphy, G., Tschesche, H. (1996b). Activation of human neutrophil procollagenase by strornelysin 2. . European Journal of Biochemistry 23 5 , 1 87- 19 1.
Knauper, V., Lopez-Otin, C., Smith, B., Knight, G., characterization of human collagenase-3. Journal of 1550.
Murphy, G. (1996). Biochemical Biological Chemistry 27 1 , 1 544-
Lamster, I.B., Oshrain, R.L., Harper, D.S., Celenti, R.S., Hodiaras, C.A., Gordon, J.M., Enxyrne activity in crevicular fluid for dtection and prediction of clinical attachment loss in patients with chronic adult periodontitis; six month results. Journal of Periodontology, 59,5 16-523.
Lee, W., Aitken, S., Kulkarni, Ci., Birek, P., Overall, CM., Sodek, J., McCulloch, C.A. (1 99 1). Collagenase activity in recment periodontitis: relationship to disease progression and doxycylcine therapy. Journal of Periodontal Research 26,479-485
Lee, W., Aitken, S., Sodek, J., McCulloch, C.A. (1 995). Evidence of a direct relationship between neutrophil collagenase activity and penodontal tissue destruction in vivo: role of active enzyme in in human periodontitis. Journal of Periodontal Research 30,23-33.
Lehti, K., Lohi, J., Valtanen, H., Keski-Oja, J. (1998) Proteolytic processing of membrane-type4 matrix metalloproteinase is associated with gelatinase A activation at the ce11 surface. Biochemical Journal 334,345-353.
Liotta, L.A., Tryggvason, K., Garbisa, S., Hart, I., Foltz, CM., Shafie, S. (i980). Metastatic potential correlates with enzymic degradation of basiment membrane collagen. Nature 284,67-68.
Lohi, J., Lehti, K., Westemarck, J., Kahari, V-M., Keski-Oja, J. (1996). Regulation of membrane-type matrix metalloproteinase-1 (MT-MMP-1) expression by growth factors and the phorbol ester PMN. European Journal of Biochernistry 239,239-247.
Loe, H., Theilade E., and Jensen, S.B. (1965). Experirnental gingivitis in man. Journal of Periodontology 36, 1 77- 1 87.
Loe, H., Anerud, A.? Boysen, H.. and Momson, E. (1986). Natural history of periodontal disease in man. Rapid, moderate and no loss of attachent in S n Lankan laborers 14 to 46 years of age. Journal of Clinical Periodontology 1 3,43 1 -445.
Lovejoy, B. Cleasby, A. Hassell, A.M., Longley, K., Luther, MA., Wegl, D., McGeehan, G., McElroy. A.B., Drewry, D., Lambert, M.H., Jordan, J.R. (1994). Structure of the catalytic domain of fibroblast collagenase complexed with an inhibitor. Science 263. 375
Massova, I., Kotra, L.P., Fridrnan? R.? Mobashery, S. (1998). Matnx metalloproteinases: structures, evohtion, and diversification. FASEB Journal 12, 1075- 1 095.
McEver, R.P., Beckstead, J.H.. Moore, K.L.. Marshall-Carlson, L., Bainton, D.F. (1 989). GMP-140, a platelet alpha-granule membrane protein is also synthesised by vascular
endothelid cells and is locdised in the Weibel-Palade bodies. Journal of Clinical Investigation 84,92-99.
Melcher, A.H., Chan, J. (1981). Phagocytosis and digestion of collagen by gingival fibroblasts in vivo: a study of senai sections. Journal of Ultrastructural Research 77, 1- 36.
Mitchell, P.G., Magna, H.A., Reeves, LM., Lopresti-Morrow, L.L., Yocum, S.A., Rosner, P.J., Geoghegan, K.F., Hambor, J.E. (1996). Cloning, expression. and type 11 collagenoiytic activiw of ma& metailoproteinase- 1 3 fiom hurnan osteoarthntic cartilage. Journal of Clinical Investigation 97,76 1-768.
Miyasaki, K.T. (1 99 1). The Neutrophil: Mechanism of controlling periodontal bacteria. Journal of Periodontology 62,76 1-774.
Miyasaki, K.T. (1996). Cha ter 10: Altered Leukocyte Function and Periodontal Disease. R Clinical Periodontology 8' Ed Carraiua F.A., Newman, M.G. eds.. W.B. Saunders Company, Toronto.
Murphy, G., Stanton, H., Cowel, S., Butler, G., Knauper, V., Atkinson, S., Gavrilovic, J. (1 999) Mechanisms for pro ma& metailoproteinase activation. Acta pathologica et micro biologica Scandinavica 1 07,3 8-44.
Nagase, H., Enghld, J.J., Suzuki, K., Salvesen, G. (1990) Stepwise activation mechanisms of precursor of matrix metalloproteinase 3 (stomelysin) by proteinases and (4-arnino-phenyl) mercurîc acetate. Biochemishy 29,57834789.
Nagase, H. (1997). Activation mechanisms of matnc rnetalloproteinases. Biological Chemistry 378, 15 1-1 60.
Nakahara, H. Howard, L, Thompson, E.W., Sato, H., Seiki, M., Yeh, Y., Chen, W-T. (1 997) Transmembrane/cytoplasmic domain-mediated membrane type 1 -matrix metailoprotease docking to invadopodia is required for ce11 invasion. Proceedings of National Academy of Science USA 94,7959-7964.
Netzel-Arnett, S., Fields, G., Birkedal-Hansen, H., Van Wart, H.E. (1 99 1). Sequence specificities of human fibroblast and neutrophil collagenases. Journal of Biological Chemistry 266,6747-6755.
Ochuchi, E., Imai, K.? Fujii, Y., Sato, H., Seiki, M., Okada Y. (1 997JMembrane type 1 matrix metalloproteinase digests interstitial collagens and other extracellular matrix macromolecules. Journal of Biological Chemistry 272,2446-245 1 .
Okumura, Y., Sato, H.. Seiki, M., Edo, H. (1997). Proteoiytic activation of the precursor of membrane type 1 ma& metailoproteinase by huma. plasmin. A possible ce11 surface activator. Federaiion of European Biochemical Socieiies Le [fers 402, 1 8 1 - 1 84.
Overall, C.M., Sodek, I. McCulloch, C.A., Birek, P. (1991). Evidence for polymorphonuclear leukocyte collagenase and 92-kilodalton gelatinase in gingival crevicular fluid. infection & Immunity 59,4687-4892.
Pabst, M.J. (1 994). 9. Priming of Neutrophils. Immunopharmacology of Neutrophils. rds. HeIlewelI, P.G, Williams, T. J. Academic Press Limited.
Page, R.C., and Schroeder, H.E. (1976). Pathogenesis of inflammatory periodontai disease. A summary of curent work. Lnboratory Investigation 34,235-249.
Page, R.C., Offenbacher, S., Schroeder, H.E., Seymour, G.J., Komman, K.S. (1997). Advances in the pathogenesis of penodontitis: summary of developrnents. clinical implications and future directions. P eriodontology 2000 14,2 16-248.
Page, R.C., (1998). The pathobiology of periodontal diseases may dfiect systemic diseases: inversion of a paradigm. Annals of Periodontology. 3(1), 108- 120
Pei., D ., Weiss, S. (1 99 5). Furin-dependent intracellular activation of the human stromelysin-3 zyrnogen. Nature 375,244-247.
Pei, D., Weiss. S. (1 996). Transmembrane-deletion mutants of the membrane-type matrix rnetalloproteinase-l process progelatinase A and express intrinsic matnx-degrading activity. Journal ofBiologicu1 Chemistry 27 1,9 135-9 140.
Pei, D. (1999). Identification and characterization of the fifth membrane-type matrix metalloproteinase MTS-MMP. Journal of Biologicaf Chemistry 274,8925-8932.
Pei, D. (1999a). Leukolysin/MMP25/MT6-W: a novel matrix rnetalloproteinase specificail y expressed in the leukocyte lineage. Cell Research. 9,29 1-303.
Polette, M. Birembaut, P. (1998) Membrane-type metalloproteinases in Nmor invasion. The International Journal Of Biochemistry & Cell Biology 20, 1 195-1202.
Puente, X.S., Penda, A.M., Liano, E., Vaiasco, G., Lopez-Otin, C. (1996). Molecular cloning of a novel membrane-type matrix metailoproteinase from a hurnan breast carcinoma. Cancer Research 56,944-949.
Roberston, P.B., Lantz, M., Manrcha, P.T., Kornman, K.S., Tnrmmel, CL., Holt. S.C. (1982). Collagenolytic activity associated with bacteroides species and Actinobacillus actinomycetemcomitans. Journal of Periodontu[ Research 17,275-283.
RomaneIli. R., Mancini, S., Laschinger, C., Overall, CM., Sodek, J., McCulloch, C.A.G. (1999). Activation of neutrophil collagenase in penodontitis. Infection and lmmunity 67. 23 19-2326.
Sato, H., Takino, T., Okada, Y. Cao, J., Shinagawa, A., Yamaoto, E., Seiki. M. (1 994). A matrix metalloproteinase expressed on the surface of invasive tumour celis. Nature 370, 61-65.
Sato, H., Seiki, M. (1996). Membrane-type rnatrix metalloproteinase ( M T - W s ) in tumour metastasis. Journal of Biochemistry 1 19,209-2 15.
Sato, H., Kinoshita, T., Takino, T., Nakayama, K., Seiki, M. (1996). Activation of a recombinant membrane-type lrnatrix rnetalloproteinase (MTl-MMP) by funn and its interaction with tissue inhibitor of metalloproteinase (TIMP-2). FEBS Leu. 393, 10 1 - 104.
Sato, H., Takino, T., Kinoshita. K., Imai, K., Okada, Y., Stetler-Stevenson. W.G., Seiki, M. (1996). Ce11 surface binding and activation of gelatinase A induced by expression of membrane-type- 1 -matrix metalloproteinase (MT1 -MMP). FEBS Lett. 3 85.23 8-240.
Sato, T., del Carmen Ovejero, M., Hou, P., Heegaard, A., Kumegawa, M., Foged, N.T.. Delaisse, J. Identification of the membrane-type matrix metalloproteinase MT1 -MMP in osteoclasts. Journal of Ce11 Science 1 10, 589-596.
Smith, C.W., Kishimoto, T.K., Abbass, O., Hughes, R.J., Rothiein, R.. Mclntyre, L.V., Butcher, E. Anderson, D.C. (1991). Chernotactic factors reguiate lectin adhesion molecule- 1 (LECAM- 1 )-dependent neutrophil adhesion to cytokine-stimuiated endotheliai cells in vitro. Journal of Clinical Investigation 87,609-6 18.
Springman, E.B., Angleton, E.L., Birkedal-Hansen, H., Van Wart, H.E. (1990). Multiple modes of activation of latent human fibroblast collagenase: evidence for the role of a Cys73 active-site zinc complex in latency and a cysteine switch mechanism for activation. Proceedings ofNuiiona1 Acaderny of Science USA 87, 364.
Stanton, H., Gavrilovic, J. Atkinson, S. J., d70rtho, M.P., Yamada. K.M., Zardi, L. Murphy, G. (1 998) The activation of ProMMP-2 (gelatinase A) by HT 1 O80 fibrosarcoma cells is promoted by culture on a fibronectin substrate and is concomitant with an increase in processing of MTI-MMP (MMP-14) to a 45 kDa fom. Joztrnal of Cell Science 1 1 1,2789-2798
Stetler-Stevenson, W .G.. Amavoorian, S., Liotta, L. A. (1 993). Tumour cell interaction with the extracellular matrix during invasion and metastasis. Annztal Revieiv of Cell Biology 9,54 1 -5 73.
Strongin, A.Y., Collier, I., Bannikov, G.. Marner, B.L.. Grant. G.A.. Goldberg, G.I. (1995). Mechanism of ce11 surface activation of 72 kDa type IV collagenase. Isolation of the activated fo rm of the membrane metalloprotease. JO urnal of Biologicol C h e m i s v 270,533 1-5338.
Symington, F.W., Hedges, DL, Hakornori, S.I. (1985). Glycolipid antigens of human polymorphonuclear neutrophiis and the inducible HL-60 myeloid leukaemia line. Journal of Imrnunology 134,2498-2506.
Takino, T., Sato, H., Yamamoto, E.9 Seiki, M. (1995a). Cloning of a human gene potentially encoding a novel matnx metalloproteinase having a C-teminal transmembrane domain. Gene 155,293-298.
Takino, T. Sato, H., Shinagawa, A., Seiki, M. (1995b). Identification of the second membrane-rype matrix rnetalloproteinase (MT-MMP-2) gene from a human placenta cDNA library: MT-MMP form a unique membrane-type subclass in the MMP family. Journal of Biological Chemishy 270,23 0 1 3 -23020.
Theilade, J., Attstrom, R. (1985). Distribution and ultrastructure of subgingival plaque in beagle dogs with gingival inflammation. Journal of Periodontal Research 10. 13 1 - 145.
Tsunezuka, Y. Kinoh, H., Takino, T., Watanabe, Y., Okada, Y., Shinagawa, A., Sato, H., Seiki, M. (1 996) Expression of membrane-type matrix metalloproteinase 1 (MT 1 -MMP in tumour cells enhances pulmonary metastasis in an experirnental metastasis assay. Cancer Research 56,5678-5683.
Tyagi, S.C., Lewis, K., Pikes, D., Marcello, A., Mujumdar, V.S., Smilcy, LM., Moore, C.K. (1 998). Stretch-induced membrane type matrix metalloproteinase and tissue plasminogen activator in cardiac fibroblast cells. Journal of Ce12 Physiology 176, 374- 382.
Vaes, G. (1988). Cellular biology and biochemical mechanism of bone resorption Clin. Orrhop. Rel. Res. 23 1,239-27 1.
Van D yke, T.E., Vaikuntam, J. (1 994). Neutrophil function and dys function in peridontal disease. Ctirrent Opinion in Periodontology 19-27
Van Wart, H., Birkedd-Hansen, H. (1 990). The cysteine switch: a rpinciple of regulation of metalloproteinase activity with potentiai applicability to the entire matrix me tallo pro teinase gene famil y. Proceedings of l e iVational A cademy of Sciences USA 87, 5578-5582.
Wahl. L., Wahl? S.M., Mergenhagen, S.E., Martin, GR. (1974). Collagenase production by endotoxin-activated macrophages. Proceedings of the National Academy of Sciences USA 71,3598-3601.
Weiss, S.J., Peppin. G., Ortiz, X., Ragsdale, C., Test, S.T. (1985) Oxidative autoactivation of latent collagenase by human neutrophils. Science 227,747-749.
Weiss, S.J. (1989). Tissue destruction by neutrophils. New EngZund Joirrnal of Medicine 320,365-376.
Welgus, H.G., Campbell, E.J., Cury, J.D. (1990). Neutra1 metalloproteinases produced by human mononuclear phagocytes. Enzyme profile. regulation, and expression during cellular development. Journal of Clinical investigation 86, 1496- 1 502.
Werb, 2. (1997). ECM and ceil surface proteolysis: regulating cellular ecoiogy. Cell91, 43 9-442. Will, H., Hinzmann, B. (1995). cDNA sequence and mRNA tissue distribution of a novel hurnan matrix metalloproteinase with a potential trammembrane segment. European Journal ~~'Biochemistry 23 1,602-608.
Will, H., Atkinson, S.J., Butler, G.S., Smith, B., Murphy, G. (1996). The soluble catalytic domain of membrane type 1 matrix metalloproteinase cleave the propeptide of progelatinase A and initiates autoproteolytic activation. Journal of Biological Chemistry 271, 171 19-17123.
Wu. H., Byrne, M.H, Stacey, A., Goldring, M.B.. Birkhead, J.R., Jaenisch, R., Krane. S.M. (1990). Generation of collagenase-resistant collagen by site-directed mutagenesis of murine proal(I) collagen gene. Proceedings of the National Acodemy of Science USA 87. 5888-5892.
Yana, I., Weiss, S.J. (2000). Regdation o f membrane type4 matrix metalloprofeinase activation by proprotein convertases. Molecular Biology of the Cell 1 1. 2387-240 1.
Zhou, Z., Apte, S.S., Soininen, R., Cao, R., Baaklini. G.Y., Rauser. R.W.. Wang, J., Cao. Y., Tryggvason, K. (2000) lmpaired endochondnl ossification and angiogensis in mice deficient in membrane-type matrix rnetalloproteinase 1. Proceedings of National Acaderny of Science USA 97,4052-4057.