LEADING ARTICLE
Drugs & Aging 2001; 18 (2): 87-99 1170-229X/01/0002-0087/$22.00/0 © Adis International Limited. All rights reserved.
The Clinical Potential of Matrix Metalloproteinase Inhibitors in the Rheumatic Disorders Sarah Elliott and Tim Cawston Department of Rheumatology, University of Newcastle, Newcastle-upon-Tyne, England
Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Structure and Function of Cartilage . . . . . . . . . . . 2. Proteinases Involved in Cartilage Turnover . . . . . . . 3. Proteoglycan Degradation . . . . . . . . . . . . . . . . 4. Inhibition of Matrix Metalloproteinase Activity . . . . . 5. Synthetic Matrix Metalloproteinase Inhibitors (MMPIs) . 6. Strategic Approaches for the Use of MMPIs in Arthritis 7. Preclinical and Clinical Studies of MMPIs in Arthritis . . 8. Safety of MMPIs . . . . . . . . . . . . . . . . . . . . . . . 9. Future Prospects . . . . . . . . . . . . . . . . . . . . . .
Abstract
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
87 88 89 91 92 92 94 95 97 98
Rheumatoid arthritis (RA) and osteoarthritis are chronic diseases that result in cartilage degradation and loss of joint function. Currently available drugs are predominantly directed towards the control of pain and/or the inflammation associated with joint synovitis but they do little to reduce joint destruction. In the future, it will be important to have drugs that prevent the structural damage caused by bone and cartilage breakdown. In this review, we will outline the structure and function of cartilage and the key features of matrix metalloproteinases (MMPs), enzymes involved in joint destruction. We will present evidence for the role of MMPs in RA and osteoarthritis, and describe the potential of synthetic inhibitors to control MMP activity and so prevent joint destruction. MMPs are able to cleave all components of the cartilage matrix. Regulation of MMPs is aberrant in osteoarthritis and RA, and MMPs have been implicated in the collagen breakdown that contributes to joint destruction in these diseases. Synthetic MMP inhibitors have been developed. In animal models of osteoarthritis and/or RA, these agents have shown chondroprotective effects. However, results from clinical trials in RA have been equivocal, with some studies being terminated because of lack of efficacy or safety concerns. Nevertheless, this approach remains promising. Increased understanding of the structure, regulation and function of individual MMPs may lead to more effective strategies, and approaches aimed at multiple steps of the pathogenesis of arthritis may be needed to break the chronic cycle of joint destruction.
88
Elliott & Cawston
Rheumatoid arthritis (RA) and osteoarthritis are chronic diseases that result in cartilage degradation and loss of joint function. Currently available drugs are predominantly directed towards the control of pain and/or the inflammation associated with joint synovitis but they do little to reduce joint destruction.[1] In the future, it will be important to have drugs that prevent the structural damage caused by bone and cartilage breakdown. In this review, we will outline the structure and function of cartilage and the key features of matrix metalloproteinases (MMPs), enzymes involved in joint destruction. We will present evidence for the role of MMPs in RA and osteoarthritis, and describe the potential of synthetic inhibitors to control MMP activity and so prevent joint destruction. 1. Structure and Function of Cartilage Cartilage is an aneural and avascular tissue which covers and protects the articular surface of bones. It is maintained by one cell type: the chondrocyte, which synthesises the cartilage matrix components. These cells respond to their environment, carefully regulating the balance between synthesis of new matrix components and their
breakdown. Cartilage is a matrix composed primarily of collagen and proteoglycan. Proteoglycan is synthesised as a protein core, termed aggrecan, to which negatively charged carbohydrate side chains are attached, attracting cations and water (fig. 1). As pressure is placed on cartilage, for instance during joint movement, the aggrecan is compressed and fluid is expelled.[2,3] As pressure is released, the carbohydrate side chains move apart by charge repulsion and fluid flows back into the cartilage. Proteoglycans are readily lost from cartilage but are rapidly replaced.[4-6] Collagen molecules consist of 3 polypeptide chains that bind together in a right-handed triple helix (fig. 2).[7] The helices are aligned in a staggered array to produce collagen fibrils and this structure provides a scaffold of high tensile strength that supports other components within cartilage. Principally, it is type II collagen that is found in adult articular cartilage but collagen types VI, IX, X and XI are also present. Collagen loss is difficult to induce but, once lost, collagen is rarely replaced.[8] The prevention of collagen degradation is thought to be pivotal to the effective treatment of rheumatoid arthritis (RA) and osteoarthritis.
G2 G3
G1
Link protein Keratan sulphate rich region
Chondroitin sulphate rich region
Hyaluronan
Fig. 1. Structure of aggrecan. Aggrecan, the main proteoglycan in human articular cartilage, is composed of a central protein chain
which has 3 globular domains (G1, G2 and G3). In the interglobular domain between G2 and G3 are attached keratan sulphate and chondroitin sulphate chains which attract water and cations into the cartilage. Aggrecan binds to hyaluronan via the G1 domain, and link protein enhances this interaction. The region between G1 and G2 is particularly susceptible to cleavage by a number of metalloproteinases.
© Adis International Limited. All rights reserved.
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
89
64nm
Microfibrils
Collagen fibril 64mm periodicity Overlap zone Staggered collagen molecules packed and crosslinked
Collagen molecule
α2 Triple helix Gly at every 3rd residue α1 α1
Hole zone
280nm
1.5nm
Fig. 2. Structure and organisation of collagen. Collagen is a protein composed of α helices (α1, α2) that wind together to form a triple
helix. Triple helices are aligned in a regular staggered array such that the final fibril has a striated appearance. The structure of collagen is such that it provides cartilage with its tensile strength. Gly = glycine.
Cartilage also contains small amounts of the proteoglycans biglycan, decorin and fibromodulin, and other minor constituents [hyaluronan, link protein, cartilage oligomeric protein (COMP) and anchorin]. These molecules bind and link different matrix components (fig. 3). There is an ordered arrangement of chondrocytes and matrix in cartilage from the superficial layer adjacent to the joint space to a calcium-rich layer adjacent to the underlying bone (fig. 4). In osteoarthritis and RA, the cartilage matrix is degraded and the chondrocytes are unable to fully repair the tissue, resulting in tissue loss, eventually revealing the subchondral bone. 2. Proteinases Involved in Cartilage Turnover Endoproteases are classified as cysteine, aspartate, serine and metallo-proteinases according to the chemical group found at the active site which is responsible for peptide bond cleavage (fig. 5).[9] Cysteine and aspartate proteinases act intracellularly at low pH whilst serine and metallo-proteinases act extracellularly at neutral pH. It is likely © Adis International Limited. All rights reserved.
that both intracellular and extracellular routes are involved in cartilage matrix turnover, and that the resorptive conditions dictate which proteinases have prominence. Within the metalloproteinase class, the matrix metalloproteinases (MMPs) as a group are able to cleave all components of the cartilage matrix, as described in table I.[10] All MMPs share a number of common characteristics: 1. They are synthesised as inactive pre-pro-peptides: the pre-sequence targets them for secretion. 2. The pro domain blocks the active site of the enzyme until removed by the action of other MMPs, plasmin or kallikrein to expose the active site of the enzyme.[11] 3. The amino-terminal domain has catalytic activity, and requires zinc and calcium for activity. 4. There is a conserved amino acid sequence which forms the active site and to which the zinc is coordinated. 5. A hinge region links the amino-terminal and carboxy-terminal domains. 6. The carboxy-terminal domain contains repeated amino acid sequences termed ‘haemopexin-like’ and has various functions including substrate and Drugs & Aging 2001; 18 (2)
90
inhibitor binding or involvement in activation processes. 7. All are inhibited by a family of inhibitors, the tissue inhibitor of metalloproteinases (TIMPs) (table II). TIMPs comprise a 4-member family that block the action of MMPs by forming irreversible, noncovalent, 1 : 1 complexes.[12] The basic arrangement of the pro, catalytic, hinge and haemopexin domains in MMPs is illustrated in figure 6, with alterations to this basic form also shown. For example, MMPs 2 and 9 have additional sequences within their catalytic domain that bind the substrate gelatin.[13-15] The membranetype MMPs (MT-MMPs) have a membrane-spanning domain at the carboxy-terminus that anchors them to the cell membrane.[16] MT-MMPs and MMP-11 have a sequence in their pro domain that
Elliott & Cawston
the furin family of serine proteinases cleave in the Golgi to activate the enzyme intracellularly.[17,18] The enzyme is then secreted in an active form. Under normal physiological conditions, MMPs do have important roles to play in the body, e.g. in growth and development.[19] However, in pathological states such as cancer, periodontal disease, RA and osteoarthritis, MMPs are over-produced and/or inadequately inhibited.[10,20] They have been proposed as the principal enzyme group responsible for collagen turnover.[21] MMPs are classified according to their preferred substrate into the collagenases, gelatinases, stromelysins and MT-MMPs (table I). Tissue distribution and substrate specificity often differs between members of the same group. For example, although collagenases all cleave native fibrillar
Fig. 3. Macromolecules in the extracellular matrix of articular cartilage. The principal collagen is type II with which type IX and XI are associated. Decorin and fibromodulin may also bind to type II collagen. Aggrecan, the main proteoglycan of cartilage, is anchored in the cell matrix by its interaction with link protein and hyaluronan. There are minor cartilage constituents such as decorin, fibromodulin, biglycan, anchorin and cartilage oligomeric protein (COMP).
© Adis International Limited. All rights reserved.
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
91
Zone 1
Zone 2
Zone 3 Tidemark Zone 4 Osteochondral junction
Subchondral bone
Fig. 4. The zones of articular cartilage. Zone 1 consists of densely packed collagen fibrils that are parallel to the articular surface, and flattened discoidal chondrocytes. Zone 2 has an increased proteoglycan content, an oblique collagen fibre network and the chondrocytes are more rounded. Zone 3 has a high proteoglycan content and the collagen fibres form a radial network. The chondrocytes are rounded and arranged in columns. The tidemark separates the noncalcified layers from Zone 4 which has a high concentration of calcium salts, and no proteoglycans. Collagen fibres are anchored within the calcified matrix, and the chondrocytes are held within a chondron. Below the calcified layer is the subchondral bone.
collagen into three-quarter and one-quarter fragments, MMP-1 preferentially cleaves type III collagen, MMP-8 type I and MMP-13 type II.[22-24] The role of each collagenase in particular physiological and pathological processes remains controversial.[25] Determining the key enzymes in disease may be important in identifying pharmacological targets. These MMPs and the collagen fragments produced by them are found in stimulated cartilage explant cultures, in the synovium and serum of patients with rheumatoid disorders, and in animal models of RA and osteoarthritis.[26-30] These findings provide evidence of MMP involvement in collagen breakdown in vivo. 3. Proteoglycan Degradation Although a number of MMPs can cleave aggrecan at a site known to be revealed in diseased tissue, recent evidence implicates a different fam© Adis International Limited. All rights reserved.
ily of metalloproteinases in this process.[31] Like MMPs, members of this family have conserved domains, namely pro, catalytic, dis-integrin-like, epidermal growth factor-like, cysteine-rich, transmembrane, and cytosolic domains. Hence, they have been named ADAM: a dis-integrin and metalloproteinase. The ‘aggrecanase’ enzymes (ADAMts 4 and 5) are from an ADAM sub-family which lack a transmembrane domain and possess thrombospondin-like sequences found in their carboxy terminus.[32] It is not known if these enzymes, purified from cartilage stimulated in vitro for long periods, are the same enzymes as those responsible for the initial cleavage of aggrecan in vivo. We have identified a membrane-associated activity that can cleave aggrecan,[33] and are currently characterising this activity to determine if it shares homology with the ADAMts molecules. In general, TIMPs do not inhibit ADAMs, although TIMP-3 Drugs & Aging 2001; 18 (2)
92
Elliott & Cawston
is a weak inhibitor of ADAM-17.[34] It is unclear at present which TIMPs, if any, inhibit ADAMts 4 and 5. 4. Inhibition of Matrix Metalloproteinase Activity The MMPs are required in normal physiological processes and are tightly regulated. However, these regulatory processes are aberrant in osteoarthritis and RA. With knowledge of the control points in MMP expression and activity, potential pharmacological targets can be identified (fig. 7). Available arthritis therapies do target a range of sites. For example, anti-tumour necrosis factor-α (antiTNFα) therapy blocks TNFα activity, reduces inflammatory effects, and decreases MMP levels and joint destruction in patients with RA.[35,36] A number of inhibitors are proposed to function via inhibition of intracellular signalling mechanisms, e.g. esculetin and tenidap.[37,38] Considerable efforts have been directed toward the design of synthetic MMP inhibitors which will substitute for TIMPs
Intracellular
Aspartate proteinases
Cysteine proteinases
Low pH
Example Cathepsin D
Examples Cathepsin B Cathepsin H Cathepsin K Cathepsin L Cathepsin S
and directly block active MMPs, as described in section 5. 5. Synthetic Matrix Metalloproteinase Inhibitors (MMPIs) Highly specific MMP inhibitors have been developed and the prospect for prevention of cartilage breakdown using synthetic MMP inhibitors is promising. Initial attempts to produce synthetic MMP inhibitors were based on the collagen cleavage site, with the scissile bond replaced by a chelating group, such as hydroxamate, to co-ordinate the catalytic zinc ion (fig. 8). This zinc-binding group was attached to a peptide that mimicked the cleavage site. Inhibitors have also been synthesised with carboxylic acid, thiol and phosphorous ligands as zinc-binding groups. These ligands co-ordinate with the zinc in the active site, blocking the action of the MMP on its substrate. These inhibitors initially had problems as they were modified or destroyed when adsorbed by the gut but now compounds have been made with good oral bioavaila-
Extracellular
Metalloproteinases
Serine proteinases
Neutral pH
Examples Collagenases MMP -1, -8, -13 Stromelysins MMP -3, -10, -11 Gelatinases MMP -2, -9 Membrane-type MMP MMP -14, -15, -16, -17 ADAMs
Examples Elastase Cathepsin G Plasmin Plasminogen activator
Fig. 5. Proteinases that degrade connective tissue matrix. Endoproteinases cleave proteins at peptide bonds and are classified according to the chemical group present at the active site that is crucial for this function. The aspartate and cysteine proteinases are predominantly responsible for intracellular turnover at low pH, and the metallo- and serine proteinases for extracellular turnover at neutral pH. ADAM = a dis-integrin and metalloproteinase; MMP = matrix metalloproteinase.
© Adis International Limited. All rights reserved.
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
93
Table I. Matrix metalloproteinases (MMPs) and their substrates. MMPs are classified according to their preferred substrate, as shown. There is variability with respect to tissue distribution and substrate specificity between members of the same groupa MMP
Alternative names
Group
MMP-1
Interstitial collagenase; vertebrate collagenase; fibroblast collagenase Gelatinase A; 72 kDa Gelatinase; 72 kDa type IV collagenase Stromelysin-1, transin; proteoglycanase; procollagenase activator
MMP-2
MMP-3
Collagenase
Latent MW (kDa) 55
Active MW (kDa) 45
Gelatinase
72
66
Stromelysin
57
45
MMP-7
Matrilysin, PUMP-1; Other uterine metalloproteinase
28
19
MMP-8
Neutrophil collagenase
Collagenase
75
58
MMP-9
Gelatinase A; 92 kDa gelatinase; 92 kDa type IV collagenase Stromelysin-2; transin-2
Gelatinase
92
86
Stromelysin
57
44
Stromelysin-3 Metalloelastase; macrophage elastase Collagenase-3
Stromelysin Other
51 54
44 45/22
Collagenase
60
48
Membrane type-1 MMP (MT1-MMP) MT2-MMP MT3-MMP MT4-MMP Xenopus collagenase
MT-MMP
66
56
MMP-10
MMP-11 MMP-12 MMP-13 MMP-14
Substrates Collagens I, II, III, VII, VIII, X; gelatin; aggrecan; link protein; versican; casein; α1-PI; α1-ACT; α2M Collagens I(?), IV, V, VII, X, XI, XIV; gelatin; aggrecan; link protein; versican; elastin; fibronectin; laminin Collagens III, IV, IX, X; gelatin; casein; type I collagen telopeptides; aggrecan; link protein; versican; elastin; α1-PI; α1-ACT; α1-M; fibronectin; laminin; activates MMP-1, MMP-8, MMP-13 Collagen IV, X; gelatin; casein; elastin; aggrecan; link protein; fibronectin; entactin; vitronectin; laminin; α1-PI; activates MMP-1 Collagens I, II, III, V, VII, VIII, X; gelatin; aggrecan; fibronectin; α1-PI Collagens IV, V, VII, X, XIV; gelatin; aggrecan; link protein; versican; elastin; α1-PI Collagens III, IV, V; gelatin; casein; aggrecan; link protein; fibronectin; elastin; activates MMP-1, MMP-7, MMP-8, MMP-9 α1-PI Collagen IV; gelatin; fibronectin; laminin; vitronectin Collagens I, II, III, IV; gelatin; aggrecan; tenascin; PAI-2 Collagens I, II, III; gelatin; casein; aggrecan; fibronectin; activates MMP-2, MMP-13 Gelatin; fibronectin; activates MMP-2 Activates MMP-2
MMP-15 MT-MMP 72 60 MMP-16 MT-MMP 64 52 MMP-17 MT-MMP 57 53 MMP-18 Collagenase 55 42 MMP-19 Other 54 45 Aggrecan MMP-20 Enamelysin Other 54 22 Amelogenin MMP-21 XMMP Other 70 53 MMP-22 CMMP Other 52 43 Gelatin; casein MMP-23 Other Unknown Unknown MMP-24 MT5-MMP MT-MMP 63 45 Activates MMP-2 MMP-25 MT6-MMP; leukolysin MT-MMP Gelatin MMP-26 Endometase Other Gelatin; α1-PI a MMP-4, -5 and -6 are missing from the list because these names were assigned for use but subsequently withdrawn. α1-ACT = α1-antichymotrypsin; α1M = α1-macroglobulin; α1-PI = α1-proteinase inhibitor; α2M = α2-macroglobulin; CMMP = chicken MMP; kDa = kilodaltons; MT-MMP = membrane-type MMPs; MW = molecular weight; PAI-2 = plasminogen activator inhibitor 2; XMMP = xenopus MMP.
bility. Nonpeptide inhibitors have also been designed. The solution of crystal structures for MMPs, in particular of the catalytic domains of MMP-1, -2, -3, -7, -8, -13, -14, and -16, has allowed inhibitors with increased specificity to be made. At least © Adis International Limited. All rights reserved.
some of the variation in substrate specificity amongst the MMPs can be explained by differences in the 6 specificity subsites in the active site cleft and surrounding sequences (fig. 8). It is here that selectivity can be designed into inhibitors. The first specificity subsite on the carboxy-terminal Drugs & Aging 2001; 18 (2)
94
Elliott & Cawston
side of the substrate scissile bond, the so-called S1’ pocket, is particularly important, e.g. being deeper in MMP-3 and larger in MMP-8 than in MMP-1. This, and differences out to the S4’ position, offer target sites for engineering specificity into synthetic MMP inhibitors. The detailed chemical design of MMP inhibitors is thoroughly reviewed in Beckett et al.,[39] and Bottomley et al.[40] 6. Strategic Approaches for the Use of MMPIs in Arthritis Firstly, it is not clear if it is preferable to inhibit release of proteoglycan, collagen or both as a therapeutic strategy for arthritis. Proteoglycan release precedes that of collagen, so blocking this release could prevent the loss of both proteoglycan and collagen from joints.[4,5] However, some researchers believe that release of proteoglycan is a normal physiological response of cartilage that is designed to protect the tissue. Once the stimulus, either mechanical or chemical, is removed proteoglycan then is rapidly replaced. However, if collagen release is targeted by therapeutic approaches, a problem arises in that it is not yet known which collagenase is responsible for collagen release.[41] Some researchers argue that this supports the use of broad spectrum inhibitors as these will test
whether MMP inhibition is therapeutically effective before testing specific inhibitors targeted to individual enzymes. The danger of using broadspectrum inhibitors is that undiscovered enzymes involved in unrecognised but essential pathways may be very important physiologically, and undesirable or unexpected adverse effects could result (see section 8). For these reasons, it is important to screen potential new compounds in in vivo animal models prior to initiating trials in humans. It could be argued that specificity must be considered when designing a synthetic metalloproteinase inhibitor for use in the treatment of arthritis in order to avoid these problems.[42] However, as far as the collagenases are concerned, insufficient data are available to pinpoint one enzyme to a specific disease. Nevertheless, as irreversible cartilage damage appears to occur only after the collagen network is destroyed, it is important to ensure that the collagenases are inhibited. In addition, the results of some studies in mice genetically engineered to lack MMPs suggest that loss of a single MMP family member is not sufficient to prevent arthritis; presumably other enzymes that degrade the same substrates can compensate.[43] One argument in support of the use of broadspectrum MMP inhibitors is that they could inhibit
Table II. Properties of tissue inhibitors of metalloproteinases (TIMPs) Characteristic
TIMP family member TIMP-1
TIMP-2
TIMP-3
TIMP-4
MMPs inhibited
All (poor inhibition of MT1-MMP)
All
All
MMP-1, -2, -3, -7, -9
Mature protein size (kDa)
20.243
21.729
21.676
22.609
Glycosylated?
Yes
No
Yes
No
Localisation
Diffusible
Diffusible
ECM bound
?
Chromosomal location of gene
Xp11.23-11.4
17q2.3-2.5
22q12.1-13.2
3p 25
Size of RNA transcripts (kb)
0.9
3.5, 1.0
4.5-5.0, (2.8, 2.4)
1.2-1.4
Regulation of expression
Inducible
Constitutive
Inducible
?
Major tissue sites of expression
Bone, ovary
Lung, ovaries, brain, testes, heart, placenta
Kidney, brain, lung, heart, ovary
Kidney, placenta, colon, testes, brain, heart, ovary, skeletal muscle
Binds to pro-enzyme of which MMP?
MMP-9
MMP-2
MMP-2, MMP-9
MMP-2
ECM = extracellular matrix; kb = kilobases; kDa = kilodaltons; MMP = matrix metalloproteinase; MT-MMP = membrane-type MMP.
© Adis International Limited. All rights reserved.
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
Function
95
Maintains MMP in inactive form
Zn2+ is involved in cleavage of matrix protein
Linking peptide
Binds to substrate other proteins
Propeptide
Catalytic domain
Hinge
Carboxy terminal domain
QPRC92GVP
H218ELGHSLGLSH
C278
C466
Zn2+
Additional domains
RXKR
Function
Sequence Gelatin-binding domains recognised by furin
MMP
MMP-11, MT-MMPs
MMP-2, -9
Collagen-like homology
Transmembrane domain and cytoplastic tail
MMP-9
MT-MMPs
Fig. 6. Domain structure of matrix metalloproteinases (MMPs). MMPs are composed of a common domain pattern, each domain contributing a particular function to the enzyme. There are additional domains in some MMPs that confer particular functions and examples of these are illistrated in the diagram. C = cysteine; E = glutamic acid; G = glycine; H = histidine; K = lysine; L = leucine; MT = membrane type; P = proline; Q = glutamine; R = arginine; S = serine; V = valine; X = any amino acid.
metalloproteinases in general. It may be sensible to use one broad-spectrum MMP inhibitor to reduce the activity of both MMPs and ADAMs. Because similar amino acid sequences are found in the catalytic domains of the ADAMs and MMPs, a single MMP inhibitor may be active against both families. However, this approach has resulted in safety concerns (see section 8). An interesting potential target for inhibition would be a novel member of the ADAM family, ADAM-17, also known as TNF-α convertase (TACE).[44] ADAM-17 ‘sheds’ membrane-associated TNF-α from the cell surface, increasing serum TNFα levels and promoting inflammation. Thus, inhibition of TACE would decrease serum TNFα levels and reduce inflammation. 7. Preclinical and Clinical Studies of MMPIs in Arthritis Many of the synthetic MMP inhibitors currently under development were designed prior to discoveries such as the 3-dimensional structure of the active sites of MMPs and the collagenolytic activity © Adis International Limited. All rights reserved.
of MMP-2 and -14.[45,46] Early studies were primarily concerned with the treatment of cancer. The first synthetic MMP inhibitor was marimastat (BB-2516), an orally administered, hydroxamatebased inhibitor. Agouron developed prinomastat (AG3340), an MMP inhibitor with an inhibition constant (Ki) in the picomolar range for the inhibition of MMP-2, -9, -13 and -14, but with much lower inhibitory activity against MMP-1 and -7.[20] Chiroscience developed D2163, now adopted by Bristol-Myers Squibb as BMS-275291.[20] This compound causes dose-dependant inhibition of endothelial cell migration in a mouse model of angiogenesis (unpublished observations). It also prevents tendinitis in a marmoset model of joint pain (unpublished observations). Phase I studies in healthy volunteers have been completed and the data indicate that good plasma levels are achieved and the compound is well tolerated (unpublished observations). Novartis has described an orally active, hydroxamate-based MMP inhibitor, CGS 27023A, for arthritis. This broad-spectrum inhibitor has a Ki in the nanomolar range against MMP-1, -2, -3, -9, -12 Drugs & Aging 2001; 18 (2)
96
Elliott & Cawston
and -13, and was chondroprotective in both the rabbit menisectomy model of osteoarthritis and a guinea pig model of spontaneous osteoarthritis.[20] Another synthetic MMP inhibitor for the treatment of osteoarthritis, tanomastat (BAY 12-9566), targets MMP-3 and with little inhibition of MMP1, although the biological basis for this selectivity was not clear. Tanomastat also inhibits MMP-2, -8, -9 and -13. It is effective in guinea pig and canine models of osteoarthritis, and no musculoskeletal adverse effects were reported in trials in which BAY 12-9566 was given to 300 patients with osteoarthritis for 3 months. The drug was detectable in human cartilage of treated patients undergoing joint replacement. However, tanomastat was withdrawn from a phase III trial in 1800 patients with arthritis following reports of adverse effects in cancer trials of the drug.[20] Cipemastat (Ro-323555), a selective collagenase inhibitor, was in phase III trials for the treatTranscriptional control of MMPs and TIMPs
Activation
ment of RA. It has a Ki in the low nanomolar range against MMP-1, -8 and -13, with approximately 10- to 100-fold lower activity against MMP-2, -3 and -9. It blocks cartilage degradation in a rat granuloma model, a Propionibacterium acnes-induced rat arthritis model and a mouse osteoarthritis model.[47] In these animal models, clear evidence of protection of bone and cartilage from degradation was apparent with cipemastat treatment even where active inflammation was present. Cipemastat had no effect on acute inflammation in rodent models so presumably it does not inhibit TACE at the dosages administered. However, large-scale trials of cipemastat in patients with RA were terminated presumably because of a lack of efficacy. It is hoped that future studies will determine if adequate amounts of this compound within the joint can be achieved with an appropriate dosing schedule. This was the first large-scale trial of a collagenase inhibitor in patients with RA and the failure Inhibition
Tissue destruction
TIMPs Cytokines Interleukin-1 Tumour necrosis factor-α Oncostatin M Interleukin-17
Stromelysin
TIMPs Active MMPs degrade connective tissue protein but no turnover can occur if TIMPs are present in excess
Collagenase
TIMPs Growth factors Transforming growth factor-β Retinoic acid Fibroblast
Gelatinase
TIMPs
Fig. 7. Control of matrix metalloproteinase (MMP) activity. MMP activity is controlled at 3 main points: stimulation of synthesis of pro-enzyme (under control of cytokines, growth factors and transcription factors), activation of the pro-enzyme (by furin within cell or by other MMPs and/or serine proteinases extracellularly) and inhibition by tissue inhibitor of metalloproteinases (TIMPs).
© Adis International Limited. All rights reserved.
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
S3
S2
97
S1
S1′
S2′
S3′
Enzyme subsites
Zn O H3N
P3
P2
P1
CN
P1′
P2′
P3′
Pro
Gln
Gly
H
IIe
Ala
Gly
COO
−
Substrate Collagen I
ZnBG
RHS inhibitor
ZnBG
LHS inhibitor
ZnBG
Combined inhibitor
Fig. 8. Matrix metalloproteinase (MMP) inhibitor interactions. (a) Subsites around the catalytic zinc (Zn) bind amino acids in the substrate on either side of the cleavage site. Synthetic MMP inhibitors use a zinc binding group (ZnBG) attached to modified peptides that can bind tightly to these subsites. Various ZnBGs can be used (Beckett et al.[39]). LHS = left hand side; P = positions of residues in the peptide; RHS = right hand side; S = subsites of active site.
to prove efficacy leaves the future of therapies targeted at the collagenases in question.[20] Several studies have shown that the antibacterial tetracycline and its derivatives inhibit MMPs. Micromolar concentrations of tetracycline are sufficient to inhibit collagenase activity by 50%; greater activity is seen with some modified tetracyclines. Doxycycline hyclate (Periostat®), a subantimicrobial dose of doxycycline, is the only MMP inhibitor approved by the US Food and Drug Administration (FDA) and it is indicated as an adjunct therapy in adult periodontitis.[20] The results of studies of tetracyclines in the treatment of patients with rheumatic disease are equivocal.[48] There is evidence that in combination with nonsteroidal anti-inflammatory drugs, these compounds can be effective and further studies are planned with the more potent derivatives.[49] Clinical evaluation of MMP inhibitors in arthritis is difficult as long trials have to be conducted with radiographs being the most reliable measure of joint damage. 8. Safety of MMPIs When any new class of drugs is used for the first time it raises issues concerning safety. MMPs are involved in a range of physiological processes, including wound healing, growth and fetal development.[50] Therefore, these processes could all be © Adis International Limited. All rights reserved.
affected by administration of MMP inhibitors. Normally, there is a balance between matrix synthesis and degradation in connective tissues. Inhibition of MMPs could lead to an excess deposition of matrix, resulting in fibrosis. However, it should be possible to adjust the dosage of MMP inhibitors such that excess degradation of matrix is blocked without leading to excess synthesis. The most advanced safety data available is for marimastat. Musculoskeletal pain and tendonitis are identified as adverse effects in patients treated with this drug.[42] These effects commence in the small joints of the hand and spread to the arms, shoulders and other joints if the treatment is continued. The symptoms are time- and dose-dependent and could be reversible. These symptoms are also seen in patients who receive the compound Ro-319790 and led to termination of its development as an arthritis treatment.[20] Some researchers suggest that the musculoskeletal pain and tendonitis are caused by inhibition of MMP-1, but the effect is reproducible in rodents (which have no MMP-1). Also, no such adverse events were noted in patients who received cipemastat, which inhibits MMP-1 very effectively.[51] It is likely that inhibition of an uncharacterised "sheddase" enzyme, similar to TACE, contributes to the adverse events, possibly via effects on inflammation. All new compounds can be very effectively screened in rodent Drugs & Aging 2001; 18 (2)
98
Elliott & Cawston
models for these musculoskeletal events so that unsuitable compounds can be discarded at an early stage of development.[20] Administration of tanomastat appeared to be associated with increased tumour growth and poor survival times in patients with small cell lung cancer. However, it is not logical to assume that an effect seen with one member of this class of compounds will automatically be seen with all, since there are significant differences in chemical structure and metabolism between individual MMP inhibitors.[25] 9. Future Prospects As more detailed information about the structure of MMPs and their interaction with substrates becomes available, it may be possible to design inhibitors which target areas of the enzyme other than the active site. For example, the carboxyterminal, haemopexin-like domain of collagenases has long been known to be required for collagenolysis presumably because of interactions with substrate. The process of activation of the proenzyme is also a valid target; again, this approach will require a detailed knowledge of the underlying biology. An understanding of the regulation of expression of both MMPs and TIMPs at the molecular level may allow us to modulate the levels of both enzymes and inhibitors expressed by cells during disease. This approach will require detailed analysis of the signalling pathways involved in transforming a cytokine signal at the cell surface to induce the expression of proteinase or inhibitor. Finally, there is interest in the synthesis of modified TIMPs that specifically aim to inhibit specific enzymes.[52] It will be interesting to see whether the blocking of one enzyme in the MMP family, with some inhibition of other MMP family members, is sufficient to halt the progressive and chronic destruction of connective tissue seen in arthritides. Furthermore, as noted above, MMPs are not alone in being implicated in joint disease. Serine proteinases are believed to be involved in MMP activation and cysteine proteinases have been shown to de© Adis International Limited. All rights reserved.
grade collagen. It may be necessary to combine proteinase inhibitors, either sequentially or with other agents that hit other specific steps in the pathogenesis, before the chronic cycle of joint destruction found in these diseases can be broken. References 1. Conaghan PG, Brooks P. Disease-modifying antirheumatic drugs, including methotrexate, gold, antimalarials, and Dpenicillamine. Curr Opin Rheumatol 1995; 7 (3): 167-73 2. Hardingham TE, Fosang AJ. Proteoglycans: many forms and many functions. FASEB J 1992; 6: 861-70 3. Muir H. The chondrocyte, architect of cartilage: biomechanics, structure, function and molecular biology of cartilage matrix macromolecules. Bioessays 1995; 17: 1039-48 4. Dingle JT, Page TDP, King B, et al. In vivo studies of articular tissue damage mediated by catabolin/interleukin 1. Ann Rheum Dis 1987; 46: 527-33 5. Page-Thomas DP, King B, Stephens T, et al. In vivo studies of cartilage regeneration after damage induced by catabolin/ interleukin-1. Ann Rheum Dis 1991; 50: 75-80 6. Fell HB, Barratt ME, Welland H, et al. The capacity of pig articular cartilage in organ culture to regenerate after breakdown induced by complement-sufficient antiserum to pig erythrocytes. Calcif Tissue Res 1976; 20 (1): 3-21 7. van der Rest M, Garrone R. Collagen family of proteins. FASEB J 1991; 5 (13): 2814-23 8. Jubb RW, Fell HB. The breakdown of collagen by chondrocytes. J Pathol 1980; 130: 159-62 9. Cawston TE. Proteinases and inhibitors. Br Med Bull 1995; 51: 385-401 10. Birkedal-Hansen H, Moore WGI, Bodden MK, et al. Matrix metalloproteinases: a review. Crit Rev Oral Biol Med 1993; 4: 197-250 11. Van Wart HE, Birkedal-Hansen H. The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family. Proc Natl Acad Sci U S A 1990; 87: 5578-82 12. Henriet P, Blavier L, DeClerck YA. Tissue inhibitors of metalloproteinases (TIMP) in invasion and proliferation. APMIS 1999; 107: 111-9 13. Collier IE, Wilhelm SM, Eisen AZ, et al. H-ras oncogene-transformed human bronchial epithelial cells (TBE-1) secrete a single metalloprotease capable of degrading basement membrane collagen. J Biol Chem 1988; 263 (14): 6579-87 14. Wilhelm SM, Collier IE, Marmer BL, et al. SV40-transformed human lung fibroblasts secrete a 92-kDa type IV collagenase which is identical to that secreted by normal human macrophages. J Biol Chem 1989; 264 (29): 17213-21 15. Allan JA, Docherty AJP, Barker PJ, et al. Binding of gelatinases A and B to type-I collagen and other matrix components. Biochem J 1995; 309: 299-306 16. Sato H, Takino T, Okada Y, et al. A matrix metalloproteinase expressed on the surface of invasive tumour cells. Nature 1994; 370: 61-5 17. Sato H, Kinoshita T, Takino T, et al. Activation of a recombinant membrane type 1-matrix metalloproteinase (MT1-MMP) by furin and its interaction with tissue inhibitor of metalloproteinases (TIMP)-2. FEBS Lett 1996; 393 (1): 101-4 18. Pei D, Weiss SJ. Furin-dependent intracellular activation of the human stromelysin-3 zymogen. Nature 1995; 375 (6528): 244-7
Drugs & Aging 2001; 18 (2)
MMP Inhibitors in Rheumatic Diseases
19. Matrisian LM. The matrix-degrading metalloproteinases. Bioessays 1992; 14: 455-63 20. Clark IM, Rowan A, Cawston TE. Matrix metalloproteinase inhibitors in the treatment of arthritis. Curr Opin Anti-inflamm Immunomod Invest Drugs 2000; 2 (1): 16-25 21. Shingleton WD, Hodges DJ, Brick P, et al. Collagenase: a key enzyme in collagen turnover. Biochem Cell Biol 1996; 74 (6): 759-75 22. Welgus HG, Jeffrey JJ, Stricklin GP, et al. Characteristics of the action of human skin fibroblast collagenase on fibrillar collagen. J Biol Chem 1980; 255 (14): 6806-13 23. Hasty KA, Jeffrey JJ, Hibbs MS, et al. The collagen substrate specificity of human neutrophil collagenase. J Biol Chem 1987; 262 (21): 10048-52 24. Knauper V, Lopez-Otin C, Smith B, et al. Biochemical characterisation of human collagenase-3. J Biol Chem 1996; 271: 1544-50 25. Cawston TE, Rowan AD. Prevention of cartilage breakdown by matrix metalloproteinase inhibition: a realistic therapeutic target? Br J Rheumatol 1998; 37: 353-6 26. Clark IM, Powell LK, Ramsey S, et al. The measurement of collagenase, tissue inhibitor of metalloproteinases (TIMP), and collagenase-TIMP complex in synovial fluids from patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum 1993; 36: 372-9 27. Mort JS, Dodge GR, Roughley PJ, et al. Direct evidence for active metalloproteinases mediating matrix degradation in interleukin 1-stimulated human articular cartilage. Matrix 1993; 13 (2): 95-102 28. Dodge GR, Pidoux I, Poole AR. The degradation of type II collagen in rheumatoid arthritis: an immunoelectron microscopic study. Matrix 1991; 11 (5): 330-8 29. McCachren SS. Expression of metalloproteinases and metalloproteinase inhibitor in human arthritic synovium. Arthritis Rheum 1991; 34 (9): 1085-93 30. Brinckerhoff CE. Joint destruction in arthritis: metalloproteinases in the spotlight. Arthritis Rheum 1991; 34 (9): 1073-5 31. Arner EC, Pratta MA, Trzaskos JM, et al. Generation and characterization of aggrecanase – a soluble, cartilage-derived aggrecan-degrading activity. J Biol Chem 1999; 274 (10): 6594-601 32. Abbaszade I, Liu RQ, Yang F, et al. Cloning and characterization of ADAMTS11, an aggrecanase from the ADAMTS family. J Biol Chem 1999; 274 (33): 23443-50 33. Billington CJ, Clark IM, Cawston TE. An aggrecan-degrading activity associated with chondrocyte membranes. Biochem J 1998; 336 (1): 207-12 34. Amour A, Slocombe PM, Webster A, et al. TNF-α converting enzyme (TACE) is inhibited by TIMP-3. FEBS Lett 1998; 435 (1): 39-44 35. Brennan FM, Browne KA, Green PA, et al. Reduction of serum matrix metalloproteinase 1 and matrix metalloproteinase 3 in rheumatoid arthritis patients following anti-tumour necrosis factor-a (cA2) therapy. Br J Rheumatol 1997; 36 (6): 643-50 36. Lipsky P, St. Clair W, Furst D, et al. 54 week clinical and radiographic results from the ATTRACT trial: a phase III study of infliximab (Remicade™ ) in patients with active RA despite methotrexate [abstract]. Arthritis Rheum 1999; 42 Suppl. 11: S401 37. Elliott S, Rowan A, Carrère S, et al. Esculetin inhibits cartilage resorption induced by IL-1α in combination with OSM. Ann Rheum Dis 2001; 60: 158-65 38. Martel-Pelletier J, Mineau F, Tardif G, et al. Tenidap reduces the level of interleukin 1 receptors and collagenase expression
© Adis International Limited. All rights reserved.
99
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
in human arthritic synovial fibroblasts. J Rheumatol 1996; 23: 24-31 Beckett RP, Davidson AH, Drummond AH, et al. Recent advances in matrix metalloproteinase inhibitor research. Drug Disc Today 1995; 1: 16-26 Bottomley KM, Johnson WH, Walter DS. Matrix metalloproteinase inhibitors in arthritis. J Enzyme Inhib 1998; 13 (2): 79-101 Martel-Pelletier, Pelletier JP. Wanted: the collagenase responsible for the destruction of the collage network in human cartilage. Br J Rheumatol 1996; 35 (9): 818-20 Nemunaitis J, Poole C, Primrose J, et al. Combined analysis of studies of the effects of the matrix metalloproteinase inhibitor marimastat on serum tumor markers in advanced cancer: selection of a biologically active and tolerable dose for longerterm studies. Clin Cancer Res 1998; 4 (5): 1101-9 Shapiro SD. Mighty mice: transgenic technology ‘knocks out’ questions of metalloproteinase function. Matrix Biol 1997; 15: 527-33 Black RA, Rauch CT, Kozlosky CJ, et al. A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 1997; 385 (6618): 729-33 Konttinen YT, Ceponis A, Takagi M, et al. New collagenolytic enzymes cascade identified at the pannus-hard tissue junction in rheumatoid arthritis: destruction from above. Matrix Biol 1998; 17 (8-9): 585-601 Knauper V, Will H, Lopez-Otin C, et al. Cellular mechanisms for human procollagenase-3 (MMP-13) activation: evidence that MT1-MMP (MMP-14) and gelatinase a (MMP-2) are able to generate active enzyme. J Biol Chem 1996; 271 (29): 17124-31 Brewster M, Lewis EJ, Wilson KL, et al. Ro 32-3555, an orally active collagenase selective inhibitor, prevents structural damage in the STR/ORT mouse model of osteoarthritis. Arthritis Rheum 1998; 41 (9): 1639-44 Greenwald RA. Treatment of destructive arthritic disorders with MMP inhibitors: potential role of tetracyclines. Ann NY Acad Sci 1994; 732: 181-98 Greenwald RA, Golub LM, Ramamurthy NS, et al. In vitro sensitivity of the three mammalian collagenases to tetracycline inhibition: relationship to bone and cartilage degradation. Bone 1998; 22 (1): 33-8 Cawston TE. Metalloproteinase inhibitors and the prevention of connective tissue breakdown. Pharmacol Ther 1996; 70 (3): 163-82 Shaw, T. Inhibition of matrix metalloproteinases in rheumatoid arthritis. In: Firestein GS, Panayi GS, Wollheim F, editors. Rheumatoid arthritis. Oxford: Oxford University Press, 2000: 551-61 Butler GS, Hutton M, Wattam BA, et al. The specificity of TIMP-2 for matrix metalloproteinases can be modified by single amino acid mutations. J Biol Chem 1992; 274 (29): 20391-6
Correspondence and offprints: Professor Tim Cawston, Department of Rheumatology, University of Newcastle, Floor 4 Catherine Cookson Building, The Medical School, Framlington Place, Newcastle-upon-Tyne, NE2 4HH, England.
Drugs & Aging 2001; 18 (2)