| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print September 7, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* AIDS Pathogenesis Research Program, Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne, Victoria, Australia;
Department of Immunology, Monash University, Melbourne, Victoria, Australia; and
Department of Medicine, Monash University, Melbourne, Victoria, Australia
1 Correspondence: AIDS Pathogenesis and Clinical Research Program, Macfarlane Burnet Institute for Medical Research and Public Health, 85 Commercial Rd., Melbourne 3004, Australia. E-mail: crowe{at}burnet.edu.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
Key Words: pathogenesis • cell migration • HIV-associated dementia • chemokines • cytokines
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
Classification and nomenclature
The metzincin super-family, of which MMPs are a member, has a highly conserved zinc-binding signature motif (HEXXHXXGXXH) [4
]. The three other members of the metzincin super-family are serralysins, astacins, and the ADAMS/adamalysins and are beyond the scope of this review. Table 1
lists the MMPs, their common names, and some of their substrates. There are 23 human MMPs [5
]. MMPs may be further classified in terms of differences in their structure and key substrate preferences. The three main functional groups are interstitial collagenases (MMP-1, -8, -13, and -18), which preferentially cleave collagen type I, II and III; gelatinases (MMP-2 and -9), which also known as type IV collagenases (although some controversy exists over whether they are actually true type IV collagenases able to initiate degradation of full-length or insoluble type IV collagen or just break down products such as pepsin-solubilized or denatured type IV collagen [6
]); and the stromelysins (MMP-3, -10, and -11), which have specificity for laminin. Other groups include membrane-type MMPs [MMP-14 (MT1-MMP), MMP-15 (MT2-MMP), MMP-16 (MT3-MMP), MMP-17 (MT4-MMP), MMP-24 (MT5-MMP) and MMP-25 (MT6-MMP), matrilysins (MMP-7 and-26), the elastase MMP-12, and others (MMP-19,-20,-23, -28)]. There are also endogenous tissue inhibitors of metalloproteinases (TIMPs-1 to -4).
|
|
|
REGULATION OF MMPs |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
|
Normally, the TIMP family of endogenous proteins strictly regulates the activity of MMPs. TIMPs play a major role in the fine balance in cell degradation under routine physiological control and in pathology [15 , 16 ]. Imbalance between TIMPs and MMPs can lead to excessive degradation of matrix components. There are four TIMPs, differentially expressed by cells in various tissues. TIMP-1, -2, and -4 are secreted, whereas TIMP-3 is bound to the ECM and TIMP-4 is mainly found in vascular tissue [17 ]. TIMPs can protect MMPs from cleavage by other MMPs. For example, TIMP-1 binds pro-MMP-9, which protects MMP-9 from MMP-3 cleavage [18 ]. TIMPs usually inhibit MMP activity but can also be required for MMP activation, such as TIMP-2 in the activation of pro-MMP-2 by MMP-14 (MT1-MMP). Inhibition of MMP function is generally achieved through non-covalent binding between residues within the N-terminal consensus sequence (VIRAK) of the TIMP and the MMP catalytic zinc binding site (reviewed in [16 ]). TIMP-1 activity in monocytes is regulated by both prostaglandin-dependent and -independent mechanisms [19 ]. In dendritic cells, prostaglandin-dependent inhibition of TIMP-1 has been reported to reduce MMP-controlled cellular migration through the ECM [20 ].
While much is known regarding TIMPs, autoproteolytic inactivation of active MMPs is poorly understood. The hemopexin domain of MMPs, shed during activation, has been found to inhibit the activity of some MMPs [10
] and alter their affinity for TIMPs [21
]. Other endogenous MMP inhibitors include 
-2-macroglobulin, which is a major inhibitor of MMPs in tissue fluids and irreversibly clears MMPs via scavenger receptor-mediated endocytosis [9
].
Another way that MMPs are regulated is via control of their localization. Proteolytic processing of specific proteins within the cell environment can control extracellular signaling events occurring at or near the cell membrane, thus controlling cell function [22 ]. The localization of MMPs at or near the cell surface is regulated through cellular expression of membrane-bound MMPs [MMP-14, -15, -16, and -24 (MT1-, MT2-, MT3-, and MT5-MMP) and GPI-linked MMP-17 and -25 (MT4- and MT6-MMP)], as well as by the binding of MMPs to cell surface receptors [e.g., integrins, the ECM metalloproteinase inducer (EMMPRIN, CD147)] and by the expression of cell surface receptors for specific MMP activators [23 ].
|
MMP FUNCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
While substrate specificity for MMPs is not fully characterized either in vitro or in vivo, common substrates to which MMPs collectively bind with varying efficacy are the ECM proteins, which include collagens (types I, II, III, IV, V, VI, VII, VIII, IX, X, XIV), laminin, fibronectin, elastin, entactin, vitronectin, myelin basic protein, aggrecan, tenascin, gelatin I, and fibrillin (Table 1)
. MMPs can also cleave a number of non-matrix substrates [26
] including plasminogen, fibrin and fibrinogen, E-cadherin, casein, MCP-3, and certain pro-cytokines such as membrane-bound TNF- and pro-TGFß [27
]. While in many cases the activity of MMPs promotes the activity and availability of specific cytokines and growth factors, they may also play an inhibitory role. For example, a number of MMPs [MMP-1, -2, -3, -13, - 14 (MT1-MMP)] cleave and thus inactivate MCP-3 [28
]; MMP-1, -2, -3, and -9 cleave IL-1ß [29
]; whereas fibrinogen and Factor XII are inactivated by MMPs-8, -12, -13, and -14 (MT1-MMP), thus suppressing normal clotting function [30
].
|
MMP PRODUCTION BY MONOCYTES AND MACROPHAGES |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
and 100 ng/mL LPS leads to up-regulation of MMP-3 expression and down-regulation of MMP-7, -9, -14 (MT1-MMP), -21, and -25 (MT6-MMP) transcripts [32
]. MMP-12, -13, and -20 are difficult to detect ex vivo but may still be biologically important when induced with optimal stimulation. For example, GM-CSF induces MMP-12 in monocytes [34
] and MMP-13 has been detected in rat alveolar macrophages stimulated with, e.g., LPS or PMA [35
]. Circulating blood monocytes highly express EMMPRIN [36
], and EMMPRIN is further up-regulated by GM-CSF-induced monocyte differentiation. TLR recognize pathogen-associated molecular patterns that lead to cytokine production and expression of co-stimulatory molecules, including CD40 to facilitate the innate and adaptive immune responses. CD14 assists the ligand interaction with TLR2 and TLR4, with TLR4 an essential signal transducer for LPS [37 ]. MMPs can be induced through binding of bacteria to TLR2 in monocytes [38 ] and CD14, TLR2, and TLR1 in MDM [39 ].
Signaling through CD40 engagement on monocytes and macrophages can induce MMP production. Binding of T cells high in CD40L, or soluble CD40L, increases MMP-9 production [40 ], and MMP-8 release by monocytes [41 ]. Both MMP-8 and MMP-9 are stored in intracellular pools and can therefore be rapidly released independent of mRNA synthesis [42 ]. MMP-12 is essential in macrophage migration [43 ], as it digests elastin and a number of ECM molecules. Engagement of CD40 on MDM results in enhancement of the cell differentiation and cytokine induced MMP-12 expression [44 ] as well as MMP-1 secretion through activation of NF-kB [45 ]. In tissues, differences in MMP induction by T cell binding to macrophages may be related to the levels of CD40 expressed. For example, activated T cell membranes can also induced production of MMP-1, MMP-9, and TIMP-1 by lung tissue macrophages but not in alveolar macrophages [46 ], which many be reflective of low/negligible CD14 and CD40 expression on alveolar macrophages.
Expression of the endogenous TIMPs is enriched in monocytes compared with T and B lymphocytes [32 ]. Monocytes express TIMP-1 in particular but also TIMP-2, -4, and the insoluble TIMP-3, which is found exclusively bound to ECM. Levels of TIMP expression tend to be reduced upon stimulation of monocytes with the exception of TIMP-3 [32 ]. The major TIMPs produced by macrophages are the inducible TIMP-1 and the constitutively expressed TIMP-2 [47 ].
MMP expression by monocytes and macrophages is difficult to study in vitro, as methods of cell preparation, purity of populations, and culture conditions can have a major influence on the phenotype and function of the cells and thus the expression of MMPs. For example adherent culture activates monocyte/macrophages and thus increases levels of production of some MMPs [47 ]. Ligation of cell receptors with antibody or beads during some purification procedures can also provide activating signals, altering MMP expression profiles. For this reason, methods of cell isolation and culture conditions used to generate data have been identified in this review to qualify observations where they may bias interpretation.
Changes in MMP production with maturation/differentiation
MMP production depends on the stage of maturation/cellular differentiation. In general, MMP expression increases significantly as blood monocytes differentiate into monocyte-derived macrophages. Stimulation of freshly isolated monocytes for 48 h with M-CSF promotes cellular differentiation and is associated with a 5-fold increase in MMP-9 expression [48
]. Monocytes, purified by elutriation and then adhered to plastic (resulting in activation), secrete MMP-9 and low levels of MMP-1, but lack MMP-2 and MMP-3 profile. Con A stimulation of monocytes does not change their MMP expression profile [49
]. We (unpublished data) and others have shown that MMP-7 expression also increases with monocyte differentiation to MDM, reaching a plateau after 5 days in culture [50
]. MMP-7 is also induced in macrophages by glucocorticoids, retinoids, LPS, opsonized zymozan [51
], nitroglycerine [52
], and hypoxia [53
] and inhibited by IFN, IL-4, and IL-10 [51
]. Cellular differentiation of blood monocytes to macrophages during 10 days in culture is associated with up to 9- and 25-fold increases in MMP-2 and -9 mRNA expression, respectively, and increases in activity as assessed by gelatin zymography [47
]. In comparison to blood monocytes, donor-matched alveolar macrophages adhered to plastic produce MMP-9 and low levels of MMP-1 and -2. LPS stimulation significantly up-regulates MMP-9 and MMP-1 production, increases MMP-2, and induces MMP-3 production [54
], indicating that the stromolysin MMP-3 can be secreted by activated, and fully differentiated macrophages [49
]. The ability of differentiated macrophages to produce an increased level of range of MMPs reflects their function within tissue, maintaining homeostasis through ECM remodeling and their ability to respond to the local tissue environment and to immune activation. Augmented MMP expression may be directly related to functional changes that occur during cellular differentiation (e.g., increased migration capacity of mature compared with immature dendritic cells) [17
], as discussed later.
MMP production by microglia
Microglia are the long-lived resident macrophages of the CNS. They express a wide range of MMPs both in vitro and ex vivo. Human fetal microglia constitutively produce high levels of MMPs de novo (including MMP-1,-2,-3, and -9) that do not change with time in culture but can be further elevated by CD40-ligand mediated activation [47
]. Chemokine (MCP-1, MIP-1ß, RANTES, and fractalkine) and cytokine (IL-8) activation can increase MMP-2 as well as TIMP-1 and -2 secretion by CHME3, a human fetal microglia cell line [55
]. Likewise, rat microglial cultures constitutively produce MMPs, particularly MMP-2 and -9, with the levels of MMPs significantly increased through activation with proinflammatory cytokines (TNF [56
], IL-1 [56
, 57
]), mitogens (LPS [56
, 57
], a range of plant lectins, ConA, and PHA [58
]), and chemokines (MCP-1, MIP-1ß, and fractalkine [55
]) with elevated MMP production resulting in increased microglial degradation of fibronectin-gelatin matrix [57
]. Elevation in MMP production can be inhibited by a number of factors including IFN, which has been shown to inhibit MMP-9 production in response to LPS [56
]. Microglia also express membrane-bound MMPs, and expression is dependent on their location within the brain. Microglia in the white matter of human brain tissue express MMP-14 (MT1-MMP) [59
], and MMP-16 (MT3-MMP) is found on microglia both in the white and gray matter when sections are examined by immunohistochemistry [60
]. Distribution of expression in the brain has been confirmed by RT-PCR [60
], suggesting specialized functions of microglia dependant on their location within the brain.
Induction and inhibition of monocyte/macrophage MMP expression with cytokines and other stimuli
A range of factors including cell adhesion to lymphocytes and ECM, stimulation by MCP-1, HIV-1 Tat, lectins, PMA, endotoxin, IL-1ß, and prostaglandin E2 (PGE2) can induce MMP-9 production by monocytes [61
, 62
, 19
], whereas MMP-9 is negatively regulated by IL-4, IL-10, IFNß, IFN, and TGF-ß in macrophages [63
, 64
]. Chemokines, such as RANTES/CCL5 and SDF-1/CXCL12, induce MMP-9 (both mRNA and MMP-9 secretion) during culture for 24 h, while TIMP-1 levels remain unchanged. CCR1 is highly expressed on monocytes, and binding of RANTES increases MMP-9 production by CCR1+ monocytes. The CCR1 antagonist BX471 decreased pro-MMP in culture medium, potentially suggesting a therapeutic role for CCR1 antagonists to decrease MMP-9 production [65
]. Cytokines can have synergistic effects to induce or suppress MMP production. Cytokines such as TNF and GM-CSF also increase MMP-9 production by monocytes and together induce MMP-1 in monocytes [66
] and increase MMP-1, -9, and TIMP-1 levels in macrophages [19
]. IFN, GM-CSF, and TNF enhance MMP-1 and induce TNF by a mechanism involving Caspase 8, independent of apoptosis [66
]. IFN, IL-4 [63
], and IL-10 [63
] inhibit MMP production by cultured monocytes and can also suppress the synergistic effects of TNF or IL-1 and GM-CSF [67
].
Adherent MDMs produce minimal MMP-1 and MMP-3 and produce basal levels of MMP-9. The level of production of these MMPs is increased by TNF and IL-1ß [68
, 54
]. IFN can suppress TNF and IL-1 up-regulation of MMP-9 [69
]. IFN also suppresses LPS-induced MMP-1 and -3 in alveolar macrophages. At high concentrations, IFN can also reduce MMP-9 and TIMP-1 production [69
].
There are many feedback loops in cytokine-MMP interactions. For example, activated monocytes secrete MMP-9, which degrades matrix proteins, disrupts the basal lamina, and activates pro-TNF. Active TNF in turn enhances synthesis of MMP-9 by the activated monocyte [50
]. TGFß plays a role in cell proliferation, chemotaxis, phagocytosis, and cytokine secretion. In addition, TGFß is involved in monocyte invasion through the basement membrane and migration via increasing both integrin and MMP expression levels [67
, 70
]. TGFß has also been shown to suppress TNF-induced MMP-9 mRNA production and MMP-9 secretion by MonoMac6 cell line. The TGFß suppression was reversed with PGE2, indicating that the PGE pathway is involved in this regulation [68
].
Differences in MMP expression between primary cells of macrophage lineage and monocytic cell lines
As highlighted by the various stimuli for MMP induction and suppression, precise isolation and culture conditions used (e.g., media with LPS contamination or the degree of cell-cell contact) markedly affect the results of MMP expression in monocyte/macrophages obtained in different experiments. As in all areas of monocyte biology, results obtained using monocytoid cell lines do not necessarily reflect data obtained using primary cells, further complicating the interpretation of results. For example, U937 is not a good monocyte model for MMP-9 production. Catecholamines potentiate the LPS-induced up-regulation of both MMP-9 and MMP-1 production by U937 cells, but only MMP-1 in primary monocytes and MDM [71
]. Vitamin D3, a potent differentiating agent, strongly inhibits MMP-9 but has no effect on MMP-1 levels in alveolar macrophages or primary monocytes and up-regulates MMP-1 in U937 cells without altering MMP-9 levels [72
].
|
ROLE OF MONOCYTE/MACROPHAGE MMP PRODUCTION IN THE INFLAMMATORY RESPONSE |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
MMP-induced regulation of cytokines, chemokines, and other soluble proteins
MMPs are increasingly being recognized for their role in modulating immune responses via specific cleavage of chemokines [28
]. MMPs can activate cytokines and chemokines through cleavage of the "pro" sequences, which maintain the cytokine in a latent form. However they may also cleave chemokines, with resulting loss of chemokine activity [75
]. This can down-modulate chemokine activity at precise points within the inflammatory response. For example, cleavage of MCP-3 late in the inflammatory response by MMP-2 results in the formation of a new cleaved protein that is an active antagonist of macrophage chemotaxis [76
].
MMP regulated release of soluble proteins can modulate monocyte and macrophage activation in an inflammatory response. MMP-9 and -12 contribute to soluble CD14 shedding, with proteolysis of CD14 playing an anti-inflammatory role by decreasing the responsiveness of monocytes to LPS [77 ]. Studies using broad-range inhibitors of MMPs suggest involvement of MMPs in shedding of CD16 by alveolar macrophages [78 ] and, as we demonstrated also by monocytes and MDM, particularly during phagocytosis [79 ].
Cell-cell contact and contact with matrix proteins up-regulates MMP production in monocytes and macrophages
Regulation of MMP expression is influenced by direct contact of monocytes and macrophages with other cell types in various tissues, as well as cellular interaction with matrix components [80
]. The production and secretion of MMPs by monocytes and macrophages is augmented significantly by contact with matrix proteins. For example MMP-1 production is enhanced by macrophage contact with collagen types I and III [80
]. Monocytes incubated with stimulated T cells display significant increases in MMP-1, MMP-9 [81
], and TIMP-1 secretion, as well as enhanced TNF and IL-1ß production [82
].
Cell migration and role of MMPs
Monocyte migration from blood to tissues is critical for normal immune surveillance and is a key process during the inflammatory response. Following endothelial transmigration, monocytes traverse the sub-endothelial basement membrane and the interstitial matrix of collagens and fibronectin. This is achieved via MMP-induced cleavage of cell surface molecules, which might include CD16 [83
] and L-selectin [84
]; unmasking of cryptic adhesion sites [85
]; reviewed in [23
]; regulation of cellular receptors; and degradation of components of the basement membranes.
The requirement of MMP-14 (MT1-MMP) during human monocyte migration has recently been demonstrated. MMP-14 (MT1-MMP) is up-regulated by monocytes following their attachment to fibronectin and to TNF-activated endothelial cells [23
]. Analysis by confocal microscopy of dual-labeled monocytes demonstrates MMP-14 (MT1-MMP) co-localizes with the leading-edge marker profilin and also clusters along lamellipodia, critical structures for site-directed monocyte migration. The mechanism by which MMP-14 (MT1-MMP) modulates monocyte migration is not known but may involve exposure of cryptic adhesion sites on vascular endothelia or regulation of ß1 and ß2-integrin receptor function [23
]. MMP-9 assists in monocyte migration [86
], and monoclonal antibodies directed against MMP-9 inhibit migration of bone marrow stem cells into the circulation [87
]. It has been suggested that MMP-9/ß2 integrin complexes may be part of the mechanism guiding migration, from experiments examining migration of PMA-stimulated THP-1 cells through human microvascular endothelial cell monolayers in response to chemoattractants, in which migration is blocked by inhibitors of MMP-9 or ß2-integrins [88
].
The role of fractalkine in monocyte migration is still under debate. However, recent data suggest that soluble fractalkine can inhibit MCP-1-induced MMP-2 secretion from monocytes [89 ]. MCP-1 is a key chemokine, which regulates monocyte chemotaxis and induces transendothelial migration of monocytes across endothelial cells [90 , 91 ] including brain microvascular endothelial cells [92 ].
Migration of dendritic cells and the role of MMPs
Induction of MMPs can modulate DC functions including DC migration through endothelial barriers. Cell migration is critical for dendritic cells (DCs) in the initiation of the immune response. DCs migrate from the skin and other tissues to the draining lymph nodes to transport and present immunogenic MHC-peptide complexes to antigen-specific T cells. Immature and mature DCs express, produce, and secrete several MMP, including MMP-1, -2, -3, and -9 and inhibitor TIMP-1 and -2 [93
]. These MMPs influence DC migratory ability and thus modulate innate immunity.
Similar to monocyte/macrophages, in vitro conditions and maturity of the DC can play a large role in MMP production by DCs and their migratory ability. The cellular environment can regulate the balance between MMP and TIMP gene expression. For example hypoxia increases TIMP-1 gene expression and down-regulates MMP-9 and MMP-14 (MT1-MMP), resulting in significantly reduced migratory capacity of DCs [94 , 95 ]. Chemotractants such as CCL3, CCL5, and MIP-3ß increase MMP production, particularly MMP-9 secretion, by monocyte-derived DCs and increase migration in vitro [42 , 96 , 97 ]. Migration of DCs can be significantly reduced by TIMP-1 and antibodies directed against MMP-9 [42 ]. Mature DCs predominantly express both the active form of MMP-9 and low levels of TIMP-1 [96 ] and TIMP-2 [17 ]. Mature DCs are highly migratory cells compared with immature DCs [96 , 97 ], which enables the mature DCs with potent antigen presenting capacity to reach and present antigens to specific T cells in the lymph nodes.
Surface expression of secreted MMPs such as MMP-2 and MMP-9 [98 , 96 ], through their binding to cell surface receptors including CD11b and CD44 [97 ], plays an important role in migration of mature monocyte-derived DCs [96 ] and also dermal DCs in skin explant culture models [98 ]. The migration of both Langerhans cells and dermal DCs from murine and human skin explant models is inhibited by a broad-spectrum inhibitor of MMPs (BB-3103), by antibodies directed against MMP-9 and -2, and by TIMP-1 and -2 [98 ], further demonstrating a role for MMP-2 and MMP-9 in DC migration. The importance of MMP-9 has been emphasized using MMP-9-deficient mice in which Langerhans cell migration from skin explants is strikingly reduced [98 ].
MMPs may influence DC maturation; however, data are conflicting, dependent on the system used. In MMP-9 deficient mice MMP-9 is required only for Langerhans cell migration and not maturation, as epidermal Langerhans cells matured normally with regard to morphology, phenotype, and T cell stimulatory function [98 ]. In another murine model, intradermal injection of purified MMP-9 induced marked increases in cell size, dendrite extension, and enhanced expression of MHC class II in Langerhans cells, strongly suggesting a role for MMP-9 in both migration and morphological and phenotypic maturation [99 ], which may be compensated for by other MMPs in the MMP-9 knockout mice. MMP expression by DCs is regulated by the environmental stimuli, including chemokine regulation of chemotaxis as well as the state of DC maturation, and MMP may in turn influence the maturation process.
MMP in pathogenesis of disease
MMPs play a critical role in the pathogenesis of certain diseases, through an imbalance between MMP and TIMP activity. They have been implicated in the pathogenesis of arthritis; autoimmune disease including multiple sclerosis, malignancy, cardiovascular disease, lung disease (reviewed elsewhere); and more recently in HIV infection, the focus of this review.
|
POSSIBLE ROLES FOR MMPS IN HIV PATHOGENESIS |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
-2-macroglobulin). These MMPs may potentially induce ECM degradation and cause breaches in endothelial barriers, thus facilitating viral dissemination in tissues [105
]. MMP-1 expression has also been reported to increase following in vitro HIV infection of monocyte/macrophages with cell free viral inoculum amplified in MDM [47
]. These virus preparations from culture supernatants may carry cytokines and other soluble factors that can influence both the signals the cells receive and their subsequent MMP production. Preparation of virus can thus influence MMP induction. Often conditions for preparation of viral stocks are not reported. To avoid confounding experimental results, we suggest that high-titer culture supernatants should be clarified and ultra-centrifuged before resuspension in culture medium in order to investigate the direct effect of HIV-1 on MMP production, as described in [106
]. HIV infection of monocyte/macrophages in vitro is generally considered to result in up-regulation of MMP-9, although data from various laboratories differ. Data from the Wahl laboratory using microarray analysis of HIV-1 infected macrophages revealed up-regulation of MMP-9 and MMP-12, by 3.1- and 2.2-fold, respectively, compared with uninfected macrophages [107 ]. Preliminary data from our laboratory suggest no difference in MMP-12 RNA levels between HIVBAL infected and uninfected MDM up to 2 weeks postinfection, even though levels of MMP-1, -2, and -7 were increased with infection of cells from the same donors. Investigators in the Gendelman laboratory show a decrease in MMP-9 both by RT-PCR and zymography, with HIV infection of MDM (investigated during a 48 h window from days 12–14, 5 days postinfection [47 ]) using and a range of CNS- and CSF-derived M-tropic isolates [108 ]. The differences in reported levels of MMP-9 may relate to differing experimental conditions. For example, when pro-MMP-9 levels are measured 12 h after complete media change, an increase in MMP-9 levels can be observed [109 ], which contrasts with the decrease observed in both mRNA and protein levels 48 h after media change either 5 days postinfection [47 ] or 14 days postinfection [108 ] . Studies of the kinetics of MMP production, both monocyte to macrophage maturation and time postinfection, are important to dissect the changes in the levels of both pro- and active-MMP production with time. For example, preliminary results from our laboratory using zymographic analysis of culture supernatants collected daily from MDM infected with HIVBAL 5 days post-isolation and followed for 15 days indicate an earlier peak in pro-MMP-2 production by HIV-infected vs. uninfected MDM, followed by a decline in level of this protein not seen in uninfected cells. These results suggest a complicated alteration in the kinetics of MMP production following HIV-1 infection, which may not have been detected with other study designs.
MMP production and neuropathogenesis of HIV-1-asscociated dementia (HAD)
HAD continues to be a major cause of morbidity in HIV-infected individuals. One of the key features within the brain of individuals with HAD is the marked accumulation of CD16+ cells of macrophage lineage, predominant in perivascular spaces [110
], strongly implicating monocytes and macrophages in the pathogenesis of HAD. HAD is mediated in part by mononuclear phagocyte (MP) secretory products and their interaction with neural cells, resulting in neuronal dysfunction [47
]. Overwhelming evidence implicates MMP involvement in the increase in permeability of the blood brain barrier (BBB) in HAD [111
] and the associated influx of inflammatory cells (Fig 2
).
|
Monocyte activation, not HIV-1 infection per se, may be the central event affecting HIV-associated monocyte migration across the BBB [113
] and altered MMP production. HIV Tat is a potent chemoattractant for monocytes that may directly promote monocyte entry into the brain during HIV-1 infection [114
] and indirectly via stimulation of astrocyte and endothelial cell production of MCP-1 [115
–117
]. Injection of Tat into mouse brain increases the production of MCP-1 and TNF- by macrophages and microglia and stimulates an influx of monocytes from vessels near the site of injection [118
]. HIV Tat protein has been implicated in the pathogenesis of HAD through a number of possible mechanisms, including monocyte recruitment, macrophage, or microglial activation (with subsequent production of other potential neurotoxins), increasing the release and or activation of MMPs [119
] and causing altered permeability of the BBB as shown using brain microvascular endothelial cells [120
]. Soluble HIV Tat up-regulates MMP-9 expression in monocytes (and in the monocytic cell line U937 [121
]) as assessed by gelatin zymography [62
] by inducing production of TNF- and IL-1-ß [62
]. Tat-induced MMP-9 expression can be reportedly blocked by protein tyrosine phosphatase inhibitors, which block NF-kB activation and IkB degradation [121
]. Tat-induced neurotoxic effects can be inhibited by antibodies directed against MMP-2 and -7 in vitro and by the MMP-2, -3, -9, -13, and -14 inhibitor prinomastat in murine studies [122
]. Administration of BB101, a TNF-release inhibitor that also inhibits MMP secretion, has also resulted in reduced inflammatory changes within the brains in a SCID mouse model of HIV-1 encephalitis [123
]. Gp120 activates the p38 and SAPK/JNK MAPK pathways inducing MMP-9 secretion. MMP-9 is increased in the CSF of HIV-infected patients and demyelization associated with dementia. Inhibition of p38 MAPK activation by gp120 in T cells and glioma abolished MMP-9 production [124
], indicating that targeting p38 may be a novel mechanism to control the HIV-induced cytopathic effects of MMP-9.
The influx of monocyte-derived cells into the brain of patients with HAD is considered to be due to the increased permeability of the BBB. MMPs may contribute to ECM degradation and transendothelial migration of HIV-infected cells, including monocytes. MMPs that target critical BBB proteins are elevated in the CSF of patients HIV-infected patients, particularly in those with HAD [125 –127 ], indicating leakage of the blood brain/CSF barrier. Collagen type IV, a primary constituent of the basal membrane of the BBB and substrate of MMP-2 and MMP-9, is reduced in HIV-infected brains [128 ]. Degradation of laminin, another substrate of MMP-2 and -9, can result in neuronal death [129 ]. Macrophages secrete more MMP-2 and MMP-9 following in vitro infection with chimeric HIV clones containing brain-derived envelope fragments from patients with HAD compared with clones from nondemented patients [112 ]. Other MMPs produced by macrophages such as MMP-7 and MMP-12 may also contribute to neurotoxicity [130 ].
Astrocytes are a significant component of the BBB and also function as immune effector cells in the CNS. Interactions between astrocytes and macrophages can result in altered production of inflammatory and neurotoxic factors and /or the promotion of monocyte infiltration, which can potentially amplify the deleterious effects. HIV-infected macrophages secrete pro-MMP-2, which is activated by MMP-14 (MT1-MMP) on neurons. Active MMP-2 processes SDF-1, which is overexpressed by astrocytes during HIV-1 infection, into a highly neurotoxic protein [131 ]. Astrocyte activation has been implicated in the neuropathogenesis of HAD through MMP/TIMP dysregulation [106 , 132 ]. Astrocytes may be a key source of MMP-9 in the inflamed brain. Secretion by astrocytes of the precursors of MMP-2 and MMP-9 is slightly up-regulated following exposure to HIV-1 in vitro [106 ]. MMP-2 and -9 may degrade laminin-surrounding neurons, resulting in altered integrin-related signaling and ultimately neuronal death through lack of survival signals [12 ]. Although acute astrocyte activation with IL-1-ß leads to up-regulation of TIMP-1, prolonged activation in vitro is associated with reduced TIMP-1 levels, which suggests that astrocytes may lose their ability to counteract MMP-induced neurotoxicity over time [132 ]. Supporting the idea that prolonged activation suppresses TIMP-1 secretion is the down-regulation of TIMP-1 expression in brain tissue and CSF samples from individuals with HAD, compared with controls [132 ]. Antiretroviral drugs which prevent HAD decrease MMP-2 and MMP-9 secretion by LPS-stimulated astrocytes [133 ].
Increased MMP-9 production may play a role in facilitating migration of HIV-1 infected monocytes across vascular endothelium [134 ] and this MMP is elevated in CSF of HIV-infected patients those with HAD [125 –127 , 135 ] but absent from controls with noninflammatory neurological disease [127 ]. Elevated CSF MMP-9 levels have in some reports correlated with elevated CSF white cell counts and protein levels in patients with HAD, further suggesting a causal relationship between MMP activity and BBB leakage [126 , 127 ]. Increased levels of other MMPs (such as MMP-2 and -7) in CSF have not been consistently observed [125 , 127 ].
Elevated MMP-9 levels have also been reported in a number of animal models of HIV-1 infection. In rats injected intracisternally with Nef, BBB breakdown parallels Nef-induced MMP-9 expression by peripheral blood mononuclear cells and macrophage and changes in CSF MMP-9 levels. Pretreatment with an MMP inhibitor abrogates vascular permeability [136 ]. Macaque monkeys infected with SIVmac239 that highly express MMP-9 in their microglia undergo more rapid disease progression and develop cognitive and motor deficits when compared with control monkeys with low MMP-9 expression, slow disease progression, and little or no evidence of HAD [137 ].
Roles of MMPs and other HIV-associated diseases
MMPs have been implicated in the pathogenesis of HIV-related dental disease, Kaposi’s sarcoma (KS), and HIV nephropathy and may contribute to the poorer prognosis of HIV and hepatitis C virus co-infected individuals.
HIV-infected patients frequently report gingival inflammation and may develop progressive periodontal tissue breakdown or require dental extraction due to weakened attachment. HIV-related periodontal disease, including gingivitis; periodontitis; and bacterial, viral and fungal infections has been associated with increased levels of MMP-1, -3, and -8 in saliva from HIV-infected individuals, and these MMPs may play a role in the development of HIV-associated periodontitis [138 ].
MMP-9 and TIMP-1 are overexpressed in the glomeruli of patients with HIV-associated nephropathy, as determined by RT-PCR and immunohistochemistry. It is unlikely that macrophages are the major source, as very little infiltrate is observed in kidneys of patients with this condition [139 ].
Plasma from patients co-infected with both HIV-1 and HCV show increased TIMP-1 levels particularly in patients with advanced CD4-depletion. No difference was observed in MMP-9 levels, indicating an imbalance between TIMP-1 and MMP-9. Decreases in ECM breakdown by MMP-9 lead to ECM accumulation and subsequent exacerbation of liver fibrosis observed in patients with HCV/HIV co-infection [140 ].
MMP-7 generates soluble FasL (CD178), which is involved in the apoptotic loss of CD4+ T cells. Polymorphisms in MMP-7 promoter have been studied and found to have no effect CD4+ T cell recovery and plasma viral load of HIV-positive patients in the first six months of HAART, indicating MMP-7 is not a key regulatory gene in response to therapy [141 ]. Antiretroviral therapy may influence MMP levels. A dose-dependent inhibition of MMP-9 mRNA and protein expression has been observed in LPS-stimulated primary rat astrocytes and microglia exposed to zidovudine and indinavir [133 ].
MMPs are involved in angiogenesis and tumor invasion, including the pathogenesis of HIV-related KS. Although HHV8 reactivation is causally associated with the development of KS, Tat has also been linked to the development and disease progression. In vitro Tat can promote the migration, invasion, and growth of KS cells (reviewed in [142 ]). MMPs are also potent inducers of vascular permeability and, with Tat, have been postulated to play a role in the development of edema, a major cause of morbidity in KS patients. An increase in MMP-2 has been found in plasma from KS patients as well as high expression of MMP-2 within KS lesions [142 ]. In vitro studies show that inhibitors of MMP-2 inhibit endothelial invasion induced by AIDS-KS cell supernatants [143 ]. HIV protease inhibitors indinavir and saquinavir block the activation of pro-MMP-2 released by KS and endothelial cells in vitro [144 ]. The chemically modified tetracycline COL-3, which inhibits MMP expression and activity in carcinoma cell lines and cell invasiveness in vitro, provided a partial response in 44% of patients with KS treated in a phase I study [145 ].
Potential uses of MMP inhibitors (MMPI) in therapy for HIV-infected patients
MMPIs may have potential for therapy of HIV-infected patients. Lessons can be learned from benefits of MMPIs in various animal models of disease and trials of MMPIs as therapy for stroke, aortic aneurism, and cancer, among others, for application of therapy for HIV-infected patients. Until recently most clinical trials have been disappointing, which in hindsight is not so surprising given the complexity of the targets, inadequate target validation, lack of identification of the MMP substrates, and poor appreciation their physiological roles [15
, 146
]. For example, in the past, early investigators have used broad-range inhibitors of MMP for therapy of advanced-stage cancer, when most MMPs are already up-regulated and precise knowledge of the MMP families being inhibited was fragmented. Survival benefits have been achieved more recently using the MMPI Marimastat (BB-2516) for treatment of glioblastoma multiform patients [147
] and a subset of gastric cancer patients [148
]. Timing of therapy is important, as demonstrated in stroke therapy, where MMPIs are beneficial in the acute phase but may be detrimental for the recovery phase where remodeling is required [149
]. Likewise, design of MMPIs for cancer therapy is now shifting to specifically targeting particular MMPs early in the course of disease progression to improve their success and possibly avoid side effects [146
].
For therapy in HIV-1 infection, identifying which MMPs are involved in pathology and the regulatory mechanisms involved in their synthesis and activity might allow a targeted therapy for specific HIV-related conditions such as HAD. Therapeutic strategies under development include targeting specific MMPs at the gene transcription level and specific signaling pathways. Other approaches would require new classes of inhibitors to be developed in order to inhibit pro-MMP activation through targeting proteases, which activate the MMPs, or blocking of the active site cleft or the substrate binding sites. Specific targeting and therapy for the different stages in disease progression should avoid deregulation of normal biological processes, demonstrated by the mild phenotype in most MMP knockout mice [146 ], which suggests that individual MMPs are not detrimental to survival. To our knowledge no current clinical trials of MMPIs are being conducted in HIV-infected individuals other than in KS. Treatment or prevention of HAD by these approaches would require more precise knowledge of MMPs and their contribution to loss of BBB integrity.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
Received March 4, 2006; revised June 4, 2006; accepted June 5, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONREGULATION OF MMPsMMP FUNCTIONMMP PRODUCTION BY MONOCYTES...ROLE OF MONOCYTE/MACROPHAGE MMP...POSSIBLE ROLES FOR MMPS...CONCLUSIONSREFERENCES |
|---|
