MMP-9 signaling in the left ventricle following myocardial infarction

Abstract

Following myocardial infarction (MI), the left ventricle (LV) undergoes a series of cardiac wound healing responses that involve both the stimulation of robust inflammation to clear necrotic myocytes and tissue debris and the induction of extracellular matrix (ECM) protein synthesis to generate an infarct scar. The collective changes in myocardial structure and function are termed LV remodeling, and matrix metalloproteinase-9 (MMP-9) is a key instigator of post-MI LV remodeling. Through direct molecular effects on ECM and inflammatory protein turnover as well as indirect effects on major cell types that coordinate cardiac wound healing, namely the infiltrating leukocytes and the cardiac fibroblasts, MMP-9 coordinates multiple aspects of LV remodeling. In this review, we will discuss recent research that has expanded our understanding of post-MI LV remodeling, including recent proteomic advances focused on the ECM compartment to provide novel functional and translational insights. This overview will summarize how our understanding of MMP-9 has evolved over the last decade and will provide insight into future directions that will drive our understanding of MMP-9-directed cardiac ECM turnover in the post-MI LV.

this article is part of a collection on Annual Meeting of the International Academy of Cardiovascular Sciences: North American Section. Other articles appearing in this collection, as well as a full archive of all collections, can be found online at http://ajpheart.physiology.org/.

The matrix metalloproteinase (MMP) family is comprised of zinc-dependent endopeptidase enzymes that share similar sequence homology and structure, and all MMPs by definition can cleave at least one extracellular matrix (ECM) protein. Based on domain structures and substrate preferences, MMPs can loosely be catalogued into one of six classes: 1) collagenases MMP-1, -8, and -13; 2) stromelysins MMP-3 and MMP-10; 3) gelatinases MMP-2 and MMP-9; 4) matrilysins MMP-7 and MMP-26; 5) six membrane-type MMPs; and 6) other types (e.g., MMP-28). Of note, this classification is an imprecise catalogue, as MT1-MMP (MMP-14), for example, fits into both the collagenase and MT-MMP groups (72). MMPs are expressed by a broad array of cell types, including the cardiac-relevant neutrophils, macrophages, fibroblasts, cardiomyocytes, endothelial cells, and lymphocytes (6, 95). Of the MMPs that have been evaluated in the setting of cardiovascular disease, MMP-9 plays a pivotal role in atherosclerosis, hypertension, myocardial infarction (MI), and heart failure. Under physiological conditions, MMP-9 expression at the gene and protein levels is low; MMP-9 is robustly elevated under pathological conditions including multiple cardiovascular diseases.

This review article will provide an overview summary of MMP-9 signaling at the molecular, cellular, and tissue levels to provide insight into MMP-9 pathogenic roles in the post-MI left ventricle (LV; Fig. 1). Delineating modulation at each level provides a foundation on which to form the global network of MMP-9 actions and can serve as a template for the examination of other MMPs in the system.

Fig. 1.

Simplified matrix metalloproteinase-9 (MMP-9) signaling network, with molecular inputs in layers 1, 2, and 3 and cellular and tissue level phenotype outputs in layers 4 and 5. ECM, extracellular matrix; LV, Left ventricular; miRNA, microRNAs.

Molecular Level MMP-9 Signaling

Figure 2 depicts the signaling cascade from stimulation of MMP-9 synthesis to its activation and proteolytic processing of a wide range of MI-relevant substrates. By in vitro cleavage assay or in silico evaluation, the MMP-9 substrate list includes ECM proteins (e.g., collagen, fibronectin, laminin, thrombospondin and tenascin-C), non-ECM substrates including a variety of cytokines and chemokines [e.g., interleukin (IL)-1β, CXC motif ligand (CXCL)1, CXCL4, CXCL5, CXCL7 and CXCL12], and novel substrates (e.g., CD36 and citrate synthase) (24, 25, 43, 47, 61, 62, 74, 106).

Fig. 2.

MMP-9 signaling begins at the extracellular-cell membrane interface, where cytokines, chemokines, growth factors, or ECM proteins engage cell surface receptors to transmit signaling cascades that culminate in MMP-9 release, activation, and proteolysis of substrates. The green arrows depict the signaling pathway, and the orange arrows depict modifiers of the pathway.

Stimuli that upregulate MMP-9 synthesis.

There are a number of well-known stimuli that elevate MMP-9 gene levels, including cytokines, chemokines, growth factors, and ECM proteins (10, 32, 33, 91, 103, 107). Below, we provide one example for a cytokine and one for a growth factor that demonstrate bidirectional signaling with MMP-9.

Tumor necrosis factor-α (TNF-α) is a classical proinflammatory cytokine synthesized as a membrane-bound protein that is subsequently cleaved to release a soluble form. Soluble TNF-α circulates as a stable homotrimer (10). Soluble TNF-α is an endogenous mediator of inflammatory processes leading to multiple cellular responses, including activation of genes involved in inflammatory and immunoregulatory responses, cell proliferation, and cell death (9). Elevated TNF-α levels in the LV leads to the local induction of myocardial MMP-9 and MMP-13 but not MMP-2 (10). Loss of TNF-α reduces post-MI leukocyte infiltration, a major MMP-9 source, and attenuates collagen degradation to decrease the incidence of LV rupture in mice (73). Partial inactivation of phosphatase and tensin homolog (PTEN) drives down post-MI TNF-α and MMP-9 expression, indicating that PTEN serves as a positive upstream modifier (73).

Transforming growth factor (TGF)-β₁ is a pleiotropic factor with divergent roles in the post-MI setting (13). TGF-β₁ has a bidirectional regulatory loop with MMP-9, such that their functional interplay is due to a heterotypic reciprocal interaction between the two factors. TGF-β₁ needs to be proteolytically triggered by MMPs to exert its cellular functions, and activated TGF-β₁ modulates the balance of ECM remodeling by controlling the expression of MMPs and their tissue inhibitors of metalloproteinases (TIMPs), particularly MMP-9 and TIMP-1 (50). MMP-9 initiates TGF-β₁ signaling to maintain ECM integrity and stability (13). MMP-2 and MT1-MMP are also capable of activating TGF-β₁ to maintain matrix integrity and stability (39, 48). In the mdx mouse model of muscular dystrophy, inhibition of MMP-9 by administration of a nuclear factor-κB (NF-κB) inhibitory peptide lowers the levels of active TGF-β₁ concentrations and decreases cardiac fibrosis (55). Vaday and colleagues (32, 91, 107) showed that TGF-β₁ acts as a potent suppressor of TNF-α-induced monocyte MMP-9 synthesis via a prostaglandin E₂- and cAMP-dependent mechanism, indicating that TNF-α and TGF-β₁ intersignal and that MMP-9 is both upstream and downstream of these signaling mechanisms. To delineate the precise role of MMP-9, it will be important to decipher the overall signaling network and the cross talk among the individual components.

Intracellular signaling cascades that stimulate MMP-9 expression.

Regulatory elements that are present on the MMP-9 gene include binding sites for NF-κB, activator protein-1 (AP-1), specificity protein 1 (Sp-1), serum amyloid A-activating factor (SAF)-1, E-twenty six (Ets) transcription factors, and polyomavirus enhancer A-binding protein-3 (PEA3) transcription factors (101). While we know these transcription factors control MMP-9 expression, whether these transcription factors have overlapping or distinct temporal and cellular expression profiles in the post-MI setting has not been entirely elucidated. For example, NF-κB and AP-1 have relevance in driving MMP-9 expression in neutrophils, macrophages, fibroblasts, and cardiomyocytes (17, 68, 87, 93); Sp-1 has been shown in neutrophils, macrophages, and myocytes (11, 53, 90), and SAF-1 is expressed in macrophages while Ets is found in fibroblasts (79, 85). Understanding the entire complement of regulatory elements of the MMP-9 gene within the different post-MI cell types over the time continuum of LV remodeling is still needed.

MMP promoter regions can be categorized into one of three groups based on the cis-element composition: 1) MMPs that contain a TATA box, an AP-1-binding site and an upstream PEA3-binding site to increase transcription by cytokines (e.g., MMP-9); 2) MMPs with a TATA box but no AP-1-binding site (e.g., MMP-8, -11, and -21); and 3) MMPs with multiple other transcription sites such as Sp-1 that binds to the GC box (e.g., MMP-2, -9, -14, and -28). (63) Note that MMP-9 falls in both the first and third groups due to binding sites for AP-1 and Sp-1 promoters (63).

The transcription factor NF-κB is a dimer protein that consists of five members: p50, p52, RelA/p65, c-Rel, and RelB (71). NF-κB activation and recruitment into the nucleus can be initiated by proinflammatory cytokines and growth factors. In vascular smooth muscle cells, IL-1α triggers NF-κB to strongly enhance MMP-9 expression (7, 21). The global absence of the NF-κB subunit p50 improves post-MI remodeling, in part by decreasing MMP-9 expression and lowering collagen content to improve survival and attenuate LV dilation (30).

In human carotid atherosclerotic plaques, CD147, a member of the immunoglobulin superfamily, mediates MMP-9 induction through extracellular signal-regulated kinase (ERK) and the nuclear translocation of NF-κB (46). In primary mouse cardiac fibroblasts, MMP-9 induction is mediated through multiple different transcription factors such as IκB kinase (IKK)/NF-κB, jun kinase (JNK)/AP-1, and specificity protein-1 (Sp1)-mediated reversion-inducing-cysteine-rich protein with kazal motifs (RECK) suppression (82). These studies provide strong evidence that NF-κB is a crucial transcription factor controlling MMP-9 synthesis. Depending on the cell type and stimulus used, the downstream mechanisms that follow NF-κB activation to enhance MMP-9 synthesis vary.

Overexpression of the ETS-related transcription factors EIA-F and PEA3 induce the MMP-9 promoter to increase MMP-9 synthesis (98). The fine tuning of MMP-9 expression is modified by the functional interplay among NF-κB, AP-1, and ETS factors (94), and Sp-1 inhibition can downregulate MMP-9 expression (70). SAF-1 is an inflammatory transcription factor that induces MMP-9 transcription through AP-1, and mutation of either SAF-1 or AP-1 compromises the transactivation potential and weakens cytokine responsiveness of the MMP-9 promoter (78).

MMP-9 can also be transcriptionally tuned by epigenetic mechanisms, including DNA methylation, histone acetylation, or chromatin remodeling (19). DNA methylation of cytosines within CpG islands in the promoter region inhibits MMP-9 gene expression, whereas chromatin remodeling promotes MMP-9 expression (19). Since LV remodeling involves both acute and chronic inflammation, the possibility that epigenetic mechanisms modify MMP-9 concentrations cannot be ruled out. Further studies are needed to understand the underlying epigenetic pathways that control MMP-9 gene expression.

MMP-9 expression can by posttranslationally fine-tuned by miRNAs. The ectopic expression of miR-885-5p or miR-491-5p reduces MMP-9 and inhibits cellular invasion in glioma cells (102). In multiple different types of cancer cells, miR-125, miR-21, miR-143, and miR-181 augment MMP-9 (56). Each of these miRNAs is also expressed in cardiac myocytes (18, 64, 77). This may provide another layer for MMP-9 release to be coordinated in a cell and time-dependent manner.

While we know that promoters among the structurally and functionally related MMP-2 and MMP-9 are distinct, MMP-9 transcription is not a well explored area in the setting of MI. This is due, in part, to the importance that has been placed on the activation step as the major point of regulation. Understanding how signal transduction cascades allow strict tuning of amplification, feedback, cross talk, and branching of the initial signals triggered among the different cell types to result in precise MMP-9 localization needs to be evaluated. Further studies are required to understand the full transcriptional control of MMP- 9 at the genomic level.

MMP-9 activation.

MMP-9 can be activated by a variety of endogenous and exogenous factors, including plasmin, heparin, chymase, and other MMPs such as MMP-3 (43). Historically, priority has been placed experimentally on measuring MMP activity as a marker of function. We now know that this may not be the perfect indicator, as zymograms evaluating activity in tissue samples do not take into consideration the presence of tissue inhibitors of metalloproteinases (TIMPs) and other endogenous inhibitors within the sample, and pro-MMP-9 can have activity when presented with substrate through a mechanism that does not require removal of the prodomain (43). While measuring pro- and active-MMP-9 amounts can give you an idea of recruitability and current activity potential, monitoring the actual generation of substrate cleavage products is a final confirmation of MMP activity.

MMP-9 inhibition.

Specific and selective MMP-9 inhibition has been difficult to achieve, for a variety of reasons that are well-reviewed elsewhere and briefly include the following: 1) promoter polymorphisms; 2) involvement of multiple MMPs in cardiovascular diseases; and 3) lack of knowledge of complete temporal and spatial profiles for each MMP and members of other protease families (26, 34, 43, 101, 105). Fields laboratory (52) has recently constructed an inhibitor with a K_i of 0.98 ± 0.09 nM for MMP-9; at low dose, this inhibitor is selective for MMP-9 over MMP-2, which has a K_i of 2.24 ± 0.11 nM. Studies evaluating the effects of this MMP-9 inhibitor on post-MI remodeling are warranted.

Cellular Sources of MMP-9

In the post-MI LV, neutrophils and macrophages are predominant early sources of MMP-9. In addition, fibroblasts, cardiomyocytes, and endothelial cells are relevant sources. Signaling initiated during the post-MI inflammatory response strongly stimulates MMP-9 in a variety of cell types (23, 67).

Neutrophils.

The post-MI inflammatory response includes robust leukocyte infiltration, and the neutrophils are an early leukocyte type to home to the infarct. Within 15 min of reperfusion, neutrophils are found in the infarcted LV and provide a first source of MMP-9 (58). Activated neutrophils degranulate to quickly release preformed MMP-9 (16, 20). The inactive pro-form of MMP-9 is stored in neutrophil gelatinase granules, indicating that MMP-9 derived from neutrophils is primarily posttranslationally controlled. MMP-9 release from neutrophils is stimulated by the phorbol ester formyl-Met-Leu-Phe, TNF-α, and IL-8 (15). In the extracellular space, pro-MMP-9 is activated by serine elastase and other proteinases (57, 58). Neutrophils isolated from human blood and stimulated with IL-8 showed robust MMP-9 release that was mediated through CXCR2 induction of two downstream signaling pathways, the first involving protein kinase C and ERK1/2 and the second involving Src-family kinases (15).

In addition to being released from neutrophils to degrade collagen, fibronectin, and other ECM components necessary for clearance of necrotic myocardium, MMP-9 can degrade intracellular proteins, including actin, tubulin, annexin 1, and high mobility group box 1 (HMGB1) protein, indicating that active MMP-9 may serve a protection feature to limit the injury induced by damage-associated molecular patterns (DAMPs) released from ischemic cells (14, 49). At the same time, MMP-9 signals back to the neutrophil to prevent apoptosis, which prolongs inflammation (25). The in vitro stimulation of isolated neutrophils with MMP-9 lowers caspase-9 expression, indicating that MMP-9 can directly control neutrophil apoptosis (25). While the above studies provide evidence that MMP-9 is involved in multiple neutrophil functions, the precise signaling pathways are not yet known.

Monocytes and macrophages.

Macrophages in the post-MI LV originate from circulating blood monocytes, and the conversion of monocytes to macrophages is accompanied by the induction of several cytokines, chemokines, and proteases. For example TNF-α, TGF-β, IL-1, MMPs -1, -2, -3, -7, -8, -9, -12, -14, and -28 as well as TIMPs -1, -2, -3, and -4 are markers of differentiated macrophages (100). A principal function of the macrophage post-MI is to enable wound healing and scar formation by phagocytosis of necrotic or apoptotic cells and by secretion of angiogenic molecules and growth factors (100).

While Mac-3-positive macrophages infiltrate into the infarct beginning at day 3 post-MI and peak at day 5 post-MI, Nahrendorf and colleagues (27, 31, 36, 38, 54, 59, 92) have reported an earlier pool that infiltrates beginning at the onset of ischemia. This indicates that the marker(s) used to identify the macrophage is important. In mice, MMPs in general and MMP-9 specifically are markers of proinflammatory M1 macrophages (8). Peripheral blood mononuclear cells from acute MI patients that were differentiated in vitro to macrophages produce high levels of MMP-9 compared with cells from healthy controls, indicating that macrophages are a significant cellular source of MMP-9 in humans as well (34).

Macrophage-specific transgenic overexpression of MMP-9 has been shown to improve post-MI cardiac function by blunting the inflammatory response and limiting ECM synthesis (105). On the other hand, targeted deletion of the MMP-9 gene also improves cardiac remodeling and survival post-MI albeit through different mechanisms. MMP-9 deletion stimulates the resolution of inflammation by increasing neutrophil apoptosis and macrophage phagocytosis and promoting neovascularization (25, 26, 37, 59). The fact that deletion and overexpression generate overall beneficial cardiac phenotypes indicates that LV remodeling can proceed through different pathways to end up at the same place.

Cardiac fibroblasts.

In addition to directly modulating macrophage function through multiple mechanisms, MMP-9 indirectly regulates fibroblast functions. MMP-9 secreted by M1 proinflammatory macrophages activates TGF-β to stimulate fibroblast proliferation (51). While fibroblasts are not the major source of MMP-9 in the post-MI LV, fibroblasts can elevate their MMP-9 in vitro expression in response to hypoxia or ischemia and oxidative stress (33, 83). The increase in MMP-9 expression in cardiac fibroblasts associates with a decrease in collagen synthesis rates. MMP-9 is also actively involved in cardiac fibroblast migration (97). MMP-9 levels increase in cardiac fibroblasts after stimulation with IL-1β or TNF-α through ERK1/2 and NF-κB signaling pathways (12). A common signaling pathway for MMP-9 synthesis in cardiac fibroblasts involves Notch 3 upregulation (107). Zhang et al. (107) showed that in post-MI mice, Notch3 siRNA augments MMP-9 expression levels to reduce cardiac fibrosis post-MI. Furthermore, Notch3 siRNA-dependent MMP-9 upregulation inhibited TGF-β₁-induced fibroblast-myofibroblast transition, indicating that MMP-9 and downstream TGF-β₁/Smad3 signaling are controlled by Notch 3. Notch 3 is activated by G protein-coupled receptors and the extracellular signal-regulated kinase pathway, which suggests that these signaling pathways may cross talk to trigger the downstream effects of MMP-9 (88).

Cardiac myocytes.

MI results in the dramatic loss of cardiac myocytes (74), and prolonged ischemia can cause myocyte vacuolization, often termed myocytolysis. Myocytolysis is characterized by cell swelling, myofibril lysis, and phagocytosis of necrotic myocytes (74). Ischemic cardiac myocytes are smaller in size with concomitant elevations in intracellular edema (74). Since myocardial ECM provides the structural integrity surrounding the myocyte, limiting MMP activity to the site of necrotic myocyte removal and keeping it at sufficient amounts are important (74). An imbalance between MMPs and TIMPs provides a major mechanism for the extension of infarction into the remote area (74). MMP-9 is expressed in infarcted myocytes post-MI (80). MMP-9 expression is induced in cultured rat cardiac myocytes exposed to a hypoxic environment (80). Aldosterone induces CaMKII, which can also increase the expression of MMP-9 in cardiomyocytes, indicating that MMP-9 production by myocytes occurs via several different mechanisms (35). Likewise, PPARβ/δ activation in cardiac myocytes inhibits reactive oxygen/nitrogen species generation and drives down MMP-9 gene expression (2). The PPARβ/δ agonist GW0742 markedly suppresses IL-1β-induced MMP-9 expression in vascular smooth muscle cells. PPAR signaling controls both MMP-2 and MMP-9, whereas TNF-α signaling controls only MMP-9 and not MMP-2 (2, 10). This stimulus-dependent selective induction of MMP-9 illustrates the fine tuning of a highly orchestrated process of MMP synthesis across a myriad of cellular sources.

Other cell types.

MMP-9 is also expressed in endothelial cells and lymphocytes (1). Endothelial cells are involved in the complex process of angiogenesis to form new blood vessels (81). Angiogensis includes multiple interactions between endothelial cells, surrounding smooth muscle cells, ECM, and angiogenic cytokines and growth factors (81). MMPs in general and MMP-9 specifically regulate the clearance of the ECM surrounding endothelial cells post-MI, which allows for cell proliferation and migration (4). VEGF initiates the release of pro-MMP-9 and induces migration of human CD34 endothelial progenitor and stem cells (33). This suggests that VEGF controls downstream MMP-9 signaling under hypoxic conditions. MMP-9 accelerates VEGF interaction with its receptor to trigger the angiogenic switch (4, 99).

In a chronic heart failure model, MMP-9 was shown to induce endothelial cell apoptosis and endothelial-myocyte uncoupling (74). At the same time, MMP-9 is required for adequate angiogenic revascularization of ischemic tissues and macrophage secreted MMP-9 is involved in capillary branching (45). In an indirect role, MMP-9 cleaves type IV collagen in the basement membrane to expose a cryptic regulatory peptide sequence that induces endothelial cell growth and migration (4.) MMP-9, therefore, is prominently involved in both inducing and limiting angiogenesis under pathophysiological conditions. This demonstrates the dual negative and positive functions of MMP-9.

Ras/mitogen-activated protein kinase (MAPK) signaling pathways may act as inhibitory signals for MMP-9 expression in T lymphocytes (29). Inhibition of phosphatidylinositide 3-kinases and MAPKs increase fibronectin-induced MMP-9 expression, revealing that inhibitory signals are transduced through Ras/Raf-1/MAPK pathways. Furthermore, upstream inhibition of the MAPK cascade by the G-protein inhibitor pertussis toxin enhances MMP-9 production (29).

Overall, a number of cell types contribute to the release of MMP-9 both in the LV infarct and remote regions. The cell source of MMP-9 is relevant, because MMP-9 function is dictated by the presence of substrates available for proteolytic processing. MMP, cell, and substrate presence are all managed in time and location. During the temporal evolution of remodeling, the orchestration among cells that both augment MMP-9 and provide a shift in substrates available merges to provide the net overall effect. For example, MMP-9 released from the neutrophil is high at day 1 post-MI, while MMP-9 from the macrophage is high at day 5 post-MI (58, 59). Likewise, IL-1β and other cytokines are elevated at day 1 post-MI, while TGF-β₁ and other growth factors are elevated at day 5 post-MI (26, 75). Stimulation of neutrophil-derived MMP-9, therefore, would exert a greater effect on substrates present at day 1 and generate a different cell and tissue level phenotype compared with stimulation of macrophage-derived MMP-9.

MMP-9 inhibitors have been used to test effects at different stages of MI (42). The administration of the broad-spectrum MMP inhibitor CP-471,474 immediately post-MI decreased LV dilation 4 days post-MI in mice (74). CP-471,474 inhibits MMP-1, -2, -3, -9, and - 13. Administration of PD166793 in pigs 5 days after MI led to decrease in MI size and expansion rate by 2 wk post-MI (69). PD166793 inhibits MMP-1, -2, -3, -7, -9, -13, and -14. The MMP inhibitor PGE-530742 blocks MMP-2, -3, -9 and -13 reduces dilation after MI (104). The MMP inhibitor 2R-2-{5-[4-(ethyl-methylamino)phenyl] thiophene-2-sulfonylamino}-3-methylbutyric acid (TISAM) also showed beneficial changes to remodeling (66). In addition to MMP-2, TISAM also inhibits MMP-9, and -14. Doxycycline regulates coronary artery disease, in part by inhibiting MMP-2, -8, -9, and -13 (3, 42). The cannabinoid receptor antagonist rimonabant decreased MMP-9 activity and TGF-β₁ expression in rats, leading to reduced collagen content and attenuation of fibrosis at 6 wk post-MI (84). Rimonabant regulated cardiac remodeling by indirect modulation of MMP expression and activity (84). Salvianolic acid A, a competitive inhibitor of MMP-9 prevented LV remodeling post-MI by preventing fibroblast proliferation and myofibroblast transdifferentiation (44). Even though these broad spectrum inhibitors showed a beneficial effect post-MI, clinical trials to date have not shown efficacy. This may be attributed to the lack of complete knowledge of protease roles in specific cell types and at specific times point post-MI. Cell-based targeting studies are needed to examine MMP-9 levels under a variety of conditions to provide information regarding the best time to promote or inhibit MMP-9.

Tissue Level Effects of MMP-9 on LV Function Post-MI

MI is an acute event that develops after prolonged disruption of blood supply leading to irreversible myocardial tissue injury (101). Acute MI is followed by a progressive wound healing process that can evolve to prolonged pathological remodeling of the LV. MMP-9 directs many aspects of the inflammatory and proliferative stages of acute MI (26). MMP-9 levels increase very early post-MI and remain high for the first week in both animal models of MI and in human patients with MI (25). The early increases in MMP-9 levels post-MI correlate with elevated leukocyte numbers and LV dimensions, as well as with impaired LV function (25, 101). Targeted deletion of MMP-9 reduces the number of macrophages post-MI leading to attenuated enlargement of the LV, prevents collagen accumulation, promotes neovascularization, and improves LV remodeling (34). Interestingly, transgenic overexpression of MMP-9 only in macrophages unexpectedly also showed improved cardiac function by stimulating inflammation resolution (105). These studies reveal that timing and cellular source of MMP-9 determine net function at the tissue level (Table 1).

Table 1.

A summary of MMP-9 effects in the left ventricle following myocardial infarction

Model	Time of Post-MI Evaluation, days	Cellular Source	MMP-9 Expression	Post-MI Effects
MMP-9 null mice (25, 59, 76)	Up to 28	Global	↓	Deletion:
				↑Neutrophil apoptosis, macrophage phagocytic potential, and neovascularization
				↓Macrophage infiltration, collagen accumulation and LV enlargement and dysfunction
Transgenic overexpression of MMP-9 in macrophages (105)	5	Macrophage	↑	Overexpression in macrophages:
				↓Inflammation, macrophage polarization, ECM synthesis, and LV dysfunction
NF-κB p50 null mice (30)	56	Global deletion	↓	Deletion:
				↑Early survival
				↓LV dilation
Diabetic ischemia/reperfusion (5)	30-min ischemia/2-h reperfusion	Global	↑	Diabetes:
				↑Vascular remodeling and cardiovascular complications

MMP-9, matrix metalloproteinase-9; MI, myocardial infarction; LV, left ventricular; ECM, extracellular matrix.

Future Directions and Concluding Remarks

MMP-9 has both negative and positive effects in the post-MI LV. Thus any kind of therapeutic intervention targeting MMP-9 must be carefully evaluated. There are several areas where additional in vivo and in vitro studies are required to provide a more complete understanding of MMP-9 functions. The interplay among temporal and spatial effects on MMP-9 expression remains incompletely understood. MMP-9 interactions with other MMPs remains to be examined in terms of MMP competition for a particular substrate. Likewise, whether MMP-9 exhibits substrate preference when a variety of substrates are present has not been evaluated by competition assay. A more complete understanding of MMP-9 substrates is required to know which substrates are preferentially cleaved in the post-MI setting and what is the biological function of the MMP-9 generated substrate peptides (61).

There is a close association between MMP-9 expression and diabetes (41). Diabetes enhances vascular MMP-9 activity, presumably by increasing local oxidative stress (89). In animal MI models, MMP-9 expression is linked with diabetic microvascular complications. At 6 wk after induction of diabetes, increases in MMP-2 and MMP-9 were enhanced in the infarct region of diabetic rats following 30 min of ischemia and 2 h of reperfusion (5). Increased MMP-9 levels in ischemia/reperfusion-injured rats lead to vascular remodeling and cardiovascular complications (5). In a porcine model with streptozotocin-induced diabetes, MMP-2 and -9 levels decreased in serum and activities decreased in the LV myocardium compared with the control group (65). Diabetic patients have increased risk of developing heart failure post-MI (28). Human diabetic plaques showed increased MMP-9 levels and decreased collagen content, which could lead to vulnerable plaque formation (22). Hyperglycemia associates with impaired microvascular function and cardiac wound healing after MI (40). While the association between MMP-9 and diabetic complications is strong, whether MMP-9 could serve as a diagnostic marker for the progression of diabetic complications remains to be tested.

MMP-9 signaling is another area open for future investigation. The majority of studies to date use MMP-9 as an output measurement rather than an input signal (60). Experiments where MMP-9 itself is used as the input stimuli are needed to then evaluate downstream mechanisms. The MEROPS database assimilates information of MMPs and their substrates, including known cleavage sites (see http://merops.sanger.ac.uk) (105). By evaluating known cleavage sites in a variety of substrates, an MMP-9 consensus sequence has been generated (Fig. 3). A more complete understanding of MMP-9 and its substrate interactions will likely identify specific targets that could be inhibited or overexpressed to provide therapeutic means to improve LV remodeling or to serve as diagnostic early indicators of MI.

Fig. 3.

MMP-9 Substrate Consensus Sequences derived from analysis of known cleavage sites. Abbreviations for the amino acids are given, with X indicating any amino acid. Derived from http://merops.sanger.ac.uk.

In conclusion, MMP-9 is a complex protein that plays a key role in several molecular mechanisms operational across multiple cell types during the post-MI LV remodeling continuum (Fig. 4). MMP-9 targeted studies for selective inhibition (preferably within a specific cellular source) are warranted at both the in vitro and in vivo levels to elucidate the signaling mechanisms and provide new insights to improve outcomes for the post-MI patient.

Fig. 4.

Diagram showing the temporal rise of MMP-9 post-myocardial infarction (MI) as it relates to the influx of neutrophils and macrophages and the activation of cardiac fibroblasts (34, 80, 86, 96).

GRANTS

This work was supported by American Heart Association Postdoctoral Grant 14POST18770012; National Institutes of Health Grants HL-075360, HL-129823, GM-114833, HL-051971, and GM-104357; and the Biomedical Laboratory Research and Development Service of the Veterans Affairs Office of Research and Development Award 5I01BX000505.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

R.P.I., M.J., and M.L.L. edited and revised manuscript; R.P.I., M.J., and M.L.L. approved final version of manuscript; M.L.L. prepared figures; M.L.L. drafted manuscript.