Matrix Metalloproteinases in Myocardial Infarction and Heart Failure

Abstract

Cardiovascular disease (CVD) is the leading cause of death, accounting for 600,000 deaths each year in the U.S. In addition, heart failure accounts for 37% of healthcare spending. Matrix metalloproteinases (MMPs) increase after myocardial infarction (MI) and correlate with left ventricular dysfunction in heart failure patients. MMPs regulate the remodeling process by facilitating extracellular matrix (ECM) turnover and inflammatory signaling. Due to the critical role MMPs play during cardiac remodeling, there is a need to better understand the pathophysiological mechanism of MMPs, including the biological function of the downstream products of MMP proteolysis. Future studies developing new therapeutic targets that inhibit specific MMP actions to limit the development of heart failure post-MI are warranted. This book chapter focuses on the role of MMPs post-MI, the efficiency of MMPs as biomarkers for MI or heart failure, and the future of MMPs and their cleavage products as targets for prevention of post-MI heart failure.

Introduction

Despite significant advancements in risk prediction, cardiovascular disease (CVD) remains a leading cause of death.¹ Adverse cardiac remodeling that involves excessive extracellular matrix (ECM) turnover contributes to high morbidity and mortality in patients with myocardial infarction (MI).² Elevated matrix metalloproteinase (MMP) levels strongly correlate with left ventricular dysfunction in CVD patients.

MMPs are a family of 25 proteolytic enzymes that regulate ECM turnover and inflammatory signaling. Only about half of the known MMPs have been measured in the post-MI left ventricle (LV), which leaves a significant knowledge gap in the post-MI MMP literature.³ Following MI, MMPs facilitate ECM degradation and recruit inflammatory cells for removal of necrotic cardiomyocytes. The upregulation of pro-inflammatory cytokines initially results in robust MMP activation, however, long-term stimulation increases tissue inhibitor of metalloproteinase (TIMP) levels. This ultimately leads to a decrease in the MMP/TIMP ratio and results in ongoing long-term remodeling.^{4, 5}

The fate of the myocardium post-MI depends on the balance between several competing events that occur during the wound healing response to form the ECM scar (Figure 1). Development of heart failure post-MI can be induced by exaggerated cardiac remodeling leading to impaired cardiac physiology. In response to myocyte injury induced by ischemia and infarction, a series of events occur in three distinct, but temporally overlapping, phases of wound healing: inflammation, proliferation, and maturation.⁶ Each phase contributes to the temporal changes in MMP levels in the post-MI infarct. Multiple cell types in the post-MI myocardium express MMPs, including neutrophils, macrophages, endothelial cells, myocytes, and fibroblasts, making MMPs key regulators in the cardiac remodeling progression.^7–10 This book chapter focuses on the role of MMPs during the post-MI development of heart failure and discusses the future of MMP inhibitors to prevent the development of heart failure.

Figure 1.

Myocardial wound healing is dependent on the balance between extracellular matrix (ECM) breakdown and synthesis. Matrix metalloproteinase (MMPs) are critical during this process as key regulators of inflammation, fibrosis, angiogenesis, and collagen degradation. Optimal scar formation (left) requires (1) appropriate inflammation, (2) fibroblast differentiation, proliferation, and migration to the wound site, (3) suitable angiogenesis, and (4) proper synthesis, cross-linking, and alignment of collagen at the infarct site. Too much or too little of these events will result in insufficient scare formation (right) and can facilitate in the development of heart failure post-myocardial infarction.

MMPs as biomarkers for heart failure

MMPs have been widely studied as possible markers to predict the development of CVD, particularly in post-MI remodeling and heart failure. The use of proteomic techniques over the last 10 years has amplified the discover of candidate biomarkers, due to enhanced sample preparation protocols, improvements in database searching, capabilities of mass spectroscopy, and bioinformatics analytic tools. Combined, these improvements have made identification of biomarkers for heart failure post-MI more attainable. This is especially true when it comes to biomarkers associated with cardiac ECM.^{11, 12}

Discovering novel substrates and the biological functions of peptide fragments generated by MMPs is vital to fully comprehend MMP function post-MI. MMPs have a broad number of substrates that contribute to scar formation and wound healing post-MI. Table 1 lists known MMP substrates and their biological function. MMPs also release ECM fragments called matricryptins or matrikines that are key during the development of heart failure post-MI.^13–17 Multiple studies have suggested matricryptins could be potential therapeutic targets for heart failure patients.^{14, 18, 19} Below, we summarize our currently knowledge on the involvement of MMPs in post-MI remodeling.

Table 1.

Summary of MMP substrates and their post-myocardial infarction (MI) functions. See manuscript text for references.

MMP substrates	MI functional roles	Cleaved by MMP
Angiostatin	Angiogenesis inhibitor, cardiomyocytes death, ↑heart failure	−2, −3, and −9
C-1158/59	increased migration rate of fibroblast cells, ↑ wound healing	−2 and −9
C-terminal telopeptide of collagen I	exaggerated myocardial fibrosis	−1, −2, −8, and −9
CD36	↓ macrophage phagocytosis and neutrophil apoptosis	−9 and −12
Citrate Synthase	↓ mitochondrial function	−9
Endostatin	suppresses proliferation and migration of endothelial cells	−2, −9, and −13
Fibronectin	act as chemoattractant, ↑ inflammation, migration of monocytes	−2, −7, −9, −12, and −13
Galectin-3	↑ collagen deposition, ↓ LVEF	−9
Hyaluronan	↑ inflammation, ↓ neutrophil apoptosis, induce cardiac dysfunction	−9, and −12
Laminin	inhibit migration of macrophages into the inflammatory region	−2
Osteopontin	↑ migration rate of cardiac fibroblast, ↑ wound healing	−2, −3, −7, −9, and −12
Periostin	↑ myocardial fibrosis, ↑ heart failure	−2, −9, −14
SPARC	anti-angiogenic effect, maturation of ECM	−2, −3, −7, −9, and −13
Tenascin-C	unknown	−3, −4, −7 and −9

SPARC-secreted protein acidic and rich in cysteine

MMP-1

Understanding the role of MMP-1 in the post-MI LV has been hindered due to humans only having one isoform of MMP-1 and mice having two: MMP-1a (59% homology with human MMP-1) and MMP-1b (57% homology with human MMP-1).²⁰ MMP-1 is mainly expressed by leukocytes, fibroblasts, and endothelial cells.²¹ In serum of post-MI patients who undergo reperfusion, MMP-1 increases 4 days after admission reaching a peak concentration around day 14. By day 28 MMP-1 levels decrease by 50% compared to day 14.²² In addition, serum MMP-1 levels negatively correlate with the LV end systolic volume index and positively correlate with LV ejection fraction.²²

MMP-1 preferentially degrades collagens I and III; compared to the 25 known MMPs, MMP-1 has the highest affinity for fibrillar collagen. MMP-1 initiates the degradation of collagen fibers within the LV by cleaving collagen into 3/4 and 1/4 fragments. These fragments then become unfolded and degraded by MMP-2, −9, and −3.²³

MMP-2

MMP-2 is expressed by cardiomyocytes, endothelial cells, vascular smooth muscle cells, macrophages, and fibroblasts.^24–26 Due to its high constitutive activity, MMP-2 is considered a MMP housekeeping gene that helps regulate normal tissue turnover.²⁷ Post-MI, MMP-2 levels increase both in plasma and within the infarct due to stimulation of the cardiomyocyte and cardiac fibroblast.^28–30 In patients diagnosed with heart failure there is a 4-fold increase in MMP-2 expression compared to controls.³¹ In rats, MMP-2 mRNA and protein levels elevate within 24 hours post-MI and peak around day 14 post-MI.³² Similar to rats, MMP-2 activity in mice rapidly increases within 4 days post-MI, peaks at day 7, and remains elevated until day 14.³³

Matsumura et al. demonstrated the MMP-2 generated fragments of laminin inhibited migration of macrophages into the inflammatory region and resulted in delayed wound healing after MI.³⁴ MMP-2, in addition to MMP-3, −7, and −9, processes vitronectin into multiple fragments, however, the biological function of these fragments in the post-MI environment has not been evaluated.³⁵ Recently, Sheng Zhao et al. showed periostin increases collagen fibrogenesis in the human failing heart and was associated with elevated MMP-2 levels.³¹ MMP-2 also generates the matricryptin C-1158/59 from collagen.¹⁴ In vitro stimulation with the downstream peptide from C-1158/59 increases fibroblast migration rate and angiogenesis to improve wound healing.

There are multiple MMP-2 polymorphisms, of which only 5 have been associated with MI. Elevations in MMP-2 levels due to the −1575 A/G gene polymorphism increase risk for MI by 4-fold in Hispanic males indicating MMP-2 may act as a strong biomarker for MI incidence in this population.³⁶ Similar to the −1575 A/G polymorphism, a MMP-2 single nucleotide polymorphism, −1306 T/C, displayed a 2-fold increase in promoter activity resulting in increased MMP-2 expression and enzymatic activity. In Hispanics, this polymorphism also associates with increased risk for MI and coronary triple-vessel disease.³⁷ In France, the −1306 T/C showed no association with heart failure related deaths.³⁸

A study in African- and Caucasian-Brazilian patients diagnosed with heart failure of any etiology and reduced ejection fraction (˂45%) implicates the −1575 A/G, −1059 A/G, and −790 T/G MMP-2 polymorphisms with a 2.5-fold increase in heart failure risk compared to non-diseased controls.³⁹ Interestingly, all three MMP-2 polymorphisms associate with heart failure related deaths only in Caucasians.³⁹ Similarly, the −790 T/G MMP-2 polymorphism, in addition to the −735 C/T polymorphism, also associates with chronic heart failure in patients from the Czech Republic.⁴⁰ How these polymorphisms effect heart failure related deaths is still unclear.

MMP-3

MMP-3 is secreted by cardiac fibroblasts and macrophages.²⁸ In post-MI patients, circulating MMP-3 concentrations increase steadily between admission and discharge and are higher at 3 months compared to 48 hours after MI.⁴¹ MMP-3 levels at 72–96 hours post-MI associate with left ventricular dysfunction, adverse left ventricular remodeling, and prognosis of heart failure.⁴¹ Although MMP-3 correlates with MI severity, MMP-3 mechanisms of action are not clear.

MMP-3 breaks down multiple ECM components, including collagen, fibronectin, laminins, proteoglycans, and vitronectin.⁴² MMP-3 also activates a number of MMPs including MMP-1, −7, and −9. As such, MMP-3 is considered an upstream MMP activator. MMP-3 proteolytic action on proMMP-1 is critical for the generation of fully active MMP-1.⁴³ Secreted protein acidic and rich in cysteine (SPARC) can also be cleaved by MMP-3 generating three biologically active peptides (Z-1, Z-2, and Z-3).⁴⁴ Fragment Z-1 increases angiogenesis and vascular growth, whereas fragments Z-2 and Z-3 inhibit cell proliferation. MMP-3 also generates tenascin-C fragments.⁴⁵ The significance of these fragments in cardiac pathology is not yet clear.

Due to their ability to regulate MMP-3 activity, MMP-3 polymorphisms have been implicated as regulators of MI prevalence and heart failure outcomes.^{46, 47} For example, the −1171 6A allele has lower promoter activity compared to the 5A allele and is found significantly less frequent in MI patients than in control subjects.^{46, 48} This polymorphism impacts on cardiac survival in heart failure patients with ischemic and non-ischemic cardiomyopathy differ. MMP-3 6A allele is an independent predictor of cardiac mortality in patients with non-ischemic heart failure. In contrast, there is no evidence for any effect of the MMP-3 genotype on cardiac events in patients with ischemic cardiomyopathy.⁴⁹

MMP-7

MMP-7 is expressed in endothelial cells, cardiomyocytes, and macrophages.^{9, 50, 51} In animal models of MI, MMP-7 increases three-fold in both remote and infarct regions at 7 days post-MI.⁹ MMP-7 activity is linked to increased risk for major adverse cardiac events, including decreased post-MI survival and increased hospitalization for congestive heart failure.^{9, 52–54} Elevated serum MMP-7 levels is associated with LV structural remodeling in 144 patients with LV hypertrophy.⁵³

MMP-7 has an extensive portfolio of substrates, including collagen IV, connexin-43, fibronectin, laminin, peroxiredoxins, tenascin-C, and tumor necrosis factor-α.^{9, 54–56} MMP-7 can also cleave other MMPs, including MMP-1, MMP-2, and MMP-9, leading to their activation and implicating MMP-7 as both a direct and indirect regulator for LV remodeling.⁵⁷ Of these substrates, MMP-7 has major effects on connexin-43 and serves as a predominant mechanism for post-MI arrhythmias.⁹

MMP-8

MMP-8 is expressed by neutrophils and macrophages.^{58, 59} MMP-8 is a major player during the inflammatory response. Serum levels of MMP-8 is a significant predictor of LV remodeling, cardiac rupture, and development of heart failure after MI.^{58, 60} MMP-8 increases six-fold within 6 hours post-MI and peaks at 12 hours due to infiltration of neutrophils.^{59, 61} At day 3 post-MI, MMP-8 spikes again most likely due to macrophage expression during the later stages of remodeling.⁶²

MMP-8 coordinates leukocyte trafficking through cleavage of collagen and chemokine binding proteins.^63–65 MMP-8 processes bioactive molecules, such as LPS-induced CXC chemokine (LIX).⁶⁶ Cleavage of the N-terminus of LIX by MMP-8 enhances neutrophil chemotaxis in response to lipopolysaccharide stimulation. MMP-8 cleavage of interleukin (IL)-8 and CXCL5, the human orthologues of LIX, also increases neutrophil chemotaxis.⁶⁶ How MMP-8 substrates effect the wound healing response post-MI has not been studied.

MMP-9

MMP-9 is secreted by a wide number of cell types, including cardiomyocytes, endothelial cells, neutrophils, macrophages, and fibroblasts.⁶⁷ Circulating MMP-9 increases at day 1 and remains elevated until day 7 post-MI in mice.⁶⁸ Blankenberg and colleagues were the first to implicate MMP-9 as a novel prognostic biomarker for the development of LV dysfunction and late survival in patients with CVD.^{69, 70} MMP-9 correlates with IL-6, C-reactive protein, and fibrinogen concentrations in the plasma indicating MMP-9 can predict cardiovascular outcome independent of an association with inflammatory markers.⁶⁹ Squire et al. demonstrated that increased MMP-9 correlates with larger LV volumes and greater LV dysfunction following MI.⁷¹

MMP-9 regulates tissue remodeling by directly degrading ECM and activating cytokines and chemokines.⁶⁷ Using an ECM-targeted proteomic approach, Zamilpa et al. identified multiple in vivo MMP-9 substrates in the post-MI setting including fibronectin, a known in vitro MMP-9 substrate.⁷² Cleavage of MMP-9 substrates is both detrimental and beneficial for wound healing post-MI. For example, MMP-9 mediated degradation of CD36 post-MI decreases macrophage phagocytosis and prolongs neutrophil inflammation leading to an enlarged LV post-MI.⁷³ In contrast, MMP-9 cleavage of osteopontin generates two biologically active peptides that increased the migration rate of cardiac fibroblasts resulting in enhanced infarct wound healing.⁷⁴ For this reason, targeting substrates downstream of MMP-9 may serve as a feasible alternative for predicting and preventing LV dysfunction post-MI.

There is clinical evidence that MMP genetic polymorphisms can contribute to MMP protein levels and thus influence cardiovascular outcomes.⁷⁵ The MMP-9 −1562 C/T polymorphism associates with increased MI incidence.⁷⁶ Associations vary across ethnic populations. For example, in healthy white subjects MMP-9 genetic polymorphisms did not associate with plasma MMP-9 levels whereas a positive association was discovered in healthy African American subjects.^77–79

In respect to heart failure, the MMP-9 −1562 C/T polymorphisms has had conflicting results.^{49, 76} In a study performed in patients in Brazil, this MMP-9 polymorphism did not associate with heart failure susceptibility or heart failure-related survival.⁴⁹ In a separate study completed in France, increases in the T allele of the MMP-9 polymorphism did not associate with lower ejection fraction or higher end diastolic diameter, but is an independent predictor of cardiac mortality.³⁸ Both studies had similar patient characteristics excluding patients who had an ejection fraction higher than 45%. This suggests the differences observed were most likely due to the differences in genetic ancestrality and that the −1562 C/T polymorphism may not be a strong biomarker across multiple populations. Before this conclusion can be made, the alternative explanation that the two studies measured different end points needs to be ruled out.

MMP-12

Historically macrophages have been regarded as the main cell type to express MMP-12 and hence the earlier term macrophage metalloelastase.^80–82 In recent literature, neutrophils have been identified as an early source of MMP-12 post-MI.⁸¹ MMP-12 is also expressed by endothelial cells, fibroblasts, and vascular smooth muscle cells.^{83, 84} Post-MI, MMP-12 protein levels increase in the infarct at day 1 and remain elevated through day 7 in mice.⁸¹ MMP-12 inhibition in mice led to impaired cardiac function post-MI compared to saline controls, revealing a protective role for MMP-12 in the post-MI setting.⁸¹

MMP-12 has broad substrate specificity, including type IV collagen fibronectin, heparan sulfate, laminin, and vitronectin.^85–87 Cleavage of these ECM proteins play a vital role during cardiac remodeling. For example, MMP-12 cleavage of type IV collagen disrupts the basement membrane and enables fibroblast and macrophages to access to the injured site.⁸⁷ MMP-12 can process pro-TNFα into mature TNFα indicating that MMP-12 has the potential to amplify to TNFα-driven inflammation.⁸⁵ MMP-12 also cleaves CD36 into one major fragment, but whether this peptide has biological activity is not known.⁷³

While little is known about MMP-12 mechanisms in the post-MI LV remodeling process, recent evidence indicates the MMP-12 −82 A/G polymorphism increases MI risk, as the MMP-12 −82 AG and GG genotypes were associated with a 3.7 fold increase in risk of having two or three occluded vessels.³⁶ This polymorphism is located at the activator protein-1 transcription factor binding site for MMP-12. In vitro studies have indicated the G allele of this polymorphism results in lower MMP-12 promoter activity and thus lower transcriptional activity.³⁶

MMP-14

MMP-14 is expressed in cardiomyocytes, macrophages, and fibroblasts.²⁷ MMP-14 increases 20-fold at 3 days post-MI and peaks at 16 weeks post-MI in non-infarcted LV regions, indicating critical roles in both early and late remodeling events and in both infarct and remote regions.^{8, 24, 32, 88} Elevated MMP-14 in both plasma and the LV infarct post-MI correlate with extensive LV remodeling including significant cardiac fibrosis, reduced LV function, and lower survival.^{62, 88}

Post-MI, MMP-14 degrades collagen, fibronectin, and gelatin leading to a loss of ECM structure and support.⁸⁹ In addition to ECM structural proteins, MMP-14 can process pro-fibrotic signaling molecules, such as transforming growth factor-β and periostin, leading to increased fibrillar collagen synthesis and accumulation.⁹⁰ MMP-14 can also activate MMP-2 and −13.⁹¹ Subsequent activation of these MMPs would result in continued ECM degradation and instability thereby contributing to adverse LV remodeling, dilation, and possibly heart failure.⁸⁹

MMP-28

MMP-28 is the newest identified member of the MMP family. MMP-28 is expressed by cardiomyocytes, neutrophils, and macrophages.⁹² In contrast to what is generally observed for MMPs post-MI, total MMP-28 actually decreases post-MI due to the loss of myocytes.⁹³ While myocyte-derived MMP-29 decreases, macrophage derived MMP-29 increase in the post-MI LV.⁹³ In vitro, MMP-28 has been shown to proteolytically processes casein, Nogo-A (a myelin component), and neural cell adhesion molecule-1, however, little is known about the role of MMP-28 post-MI.⁹⁴ Deletion of MMP-28 exaggerates LV dysfunction and cardiac rupture post-MI by reducing the inflammatory and fibrotic response and tilting the balance away from adequate wound healing and high-quality scar formation.⁹³ Inhibition of MMP-28, therefore, would most likely not be a successful target for improving patient outcomes post-MI.

Clinical use of MMP inhibitors post-MI

MMPs can process a variety of substrates, and the same substrate may be processed by a variety of MMPs; combined, this complexity yields a post-MI in vivo environment that is still not fully understood. For this reason, the clinical use of MMP inhibitors (MMPi) in post-MI patients is still under investigation (Figure 2).⁹⁵

Figure 2.

Therapeutics given to post-myocardial infarction patients either directly or indirectly inhibit matrix metalloproteinase (MMP) activity leading to extracellular matrix (ECM) remodeling of the left ventricle (LV).

Direct non-selective inhibition

Broad-spectrum MMPi have been used in clinical trials as an attempt to prevent heart failure post-MI. To date, more than 25 MMP inhibitors have been investigated in post-MI clinical trials. However, the majority of MMP inhibitors developed to date have not proved efficacious. Side effects due to MMPi treatment includes joint pain, stiffness, edema, skin discoloration, and reduced patient mobility.⁹⁶ However, this musculoskeletal syndrome is reversible on cessation of the drug intake.⁹⁷ Insufficient knowledge about complex biological role of MMPs and lack of specificity are main reasons for clinical failure.^{98, 99}

Animals treated with CP-471,474 (a broad-spectrum MMPi) had attenuated end-diastolic and end-systolic dimensions and increased the number of vessels post-MI.¹⁰⁰ Similarly, pigs, mice, and rabbits treated with PD166793 (another broad-spectrum MMPi) at a concentration that can inhibit MMP-2, −3, −9, and −13, but not MMP-1, had lower LV end-diastolic dimension and increased TIMP-1 concentration in infarct region post-MI.^100–102

Pigs treated 3 days pre- or post-MI with an orally available broad-spectrum MMPi (PGE-530742) at a concentration that inhibited MMP-2, −3, −8, −9, and −13, but not MMP-1, and −7 increased post-MI end-systolic volume, and attenuated fibrillar collagen content in the infarct zone.¹⁰³ Pre-MI treatment, however, increased collagen content in the border zone and decreased collagen content in the remote zone in comparison to post-MI treatment, demonstrating that timing plays an important role on MMPi effects on ECM remodeling. In a similar study, post-MI mice were given 2R-2-[5-[4-[ethyl-methylamino]phenyl] thiophene-2-sulfonylamino]-3-methylbutyric acid (TISAM) orally, at a dose that inhibited MMP-2, and −9, but not MMP-1, −3, or −7. The treatment improved survival rate, decreased cardiac rupture rate, and decreased macrophage infiltration, but increased necrotic area at day 7 post-MI.³⁴

PG-116800 is an oral MMPi of the hydroxamic acid class with high affinity for MMP-2, −3, −8, −9, −13, and −14 and low affinity for MMP-1 and −7. In a phase II double-blind, multicenter RCT PREMIER (Prevention of Myocardial Infarction Early Remodeling) trial, post-MI patients did not improve in echocardiography parameters or in clinical outcomes after 90 days treatment with PG-116800.⁹⁷ This may be due in part by the clinical dose used, which was four-fold lower than what was shown effective for achieving a LV anti-remodeling effect in pre-clinical studies.¹⁰³

Doxycycline is an antibiotic used at sub-antimicrobial doses as a broad-spectrum MMPi. Its use is approved by FDA for periodontal disease treatment.¹⁰⁴ Rats treated with doxycycline showed attenuated mRNA transcription and protein expressions of MMP-8, MMP-13, TIMP-1, TIMP-2, and type I collagen content in the remote zone post-MI.¹⁰⁵ In another study, rats treated with doxycycline showed attenuated MMP-2 activity, increased TIMP-1 expression, and improved endothelial dysfunction post-MI.¹⁰⁶ In the phase II TIPTOP (Early Short-term Doxycycline Therapy In Patients with Acute Myocardial Infarction and Left Ventricular Dysfunction to Prevent The Ominous Progression to Adverse Remodeling) trial, treatment with doxycycline (100 mg twice daily) reduced end-diastolic volumes index, infarct size, and infarct severity in comparison to patients who received only the standard treatment.¹⁰⁷

Direct selective inhibition

First generation MMPi targeted a broad range of MMPs and therefore lacked specificity and selectivity, leading to less effective use and inconclusive results.¹⁰⁸ The Fields laboratory developed a potent and highly selective MMP-9 inhibitor (MMP-9i).^{109, 110} A recent study in mice showed that early inhibition of MMP-9 had no effect on infarct size or survival. By day 7 post-MI, MMP-9i enhanced infarct wall thinning and worsened cardiac dysfunction opposing studies utilizing MMP-9 deleted mice.^{73, 108} This was due in party by MMP-9i treatment increasing expression of Mmp8, Mmp12, and Mmp14, and decreasing collagen deposition. MMP-9i treatment also increased neutrophil numbers at day 1 post-MI and macrophage infiltration at day 7 post-MI.

A selective MMP-12 inhibitor (RXP 470.1; MMP-12i) was also found to promote adverse cardiac function post-MI.⁸¹ This is in contrast to an atherosclerotic model, where MMP-12 inhibition was found to reduce atherosclerosis progression in apo-E null mice by inhibiting MMP-12 activity in macrophages.⁹⁸ Post-MI, MMP-12i repressed neutrophil functions leading to impaired cardiac wound healing. This suggests MMP-12i in an acute inflammation model may be harmful while MMP-12i in a chronic inflammation model may be favorable.

These studies reveal that using a selective MMP inhibitor will require complete assessment and complex data interpretation. Studies utilizing MMPi diverge greatly from those seen with global or cell-specific MMP deletion.^{108, 111} In addition, the overall effect of MMPi differs among CVD pathologies and timing of intervention. This is due in party by the fact that MMPs have both detrimental and beneficial mechanisms that are dependent on injury stimuli and on substrate availability.⁸¹

Indirect MMP inhibition

While the use of direct MMP inhibitors is still under examination, MMP inhibition is being achieved through indirect mechanisms. The majority of current medications used for MI and heart failure (e.g., angiotensin converter enzyme (ACE) inhibitors, angiotensin receptor antagonists, beta-blockers, and statins) act as an indirect inhibitors for MMPs.

ACE inhibitors are well documented to improve post-MI outcomes.¹¹² The catalytic domain of ACE is similar to that of MMPs, thus ACE inhibitors simultaneously have an inhibitory effect on MMPs.¹¹³ For example, captopril and lisinopril inhibit MMP-2 activity at concentrations greater than 4 and 1 mmol/L, whereas MMP-9 was inhibited by 87 nmol/L of captopril.^{114, 115} ACE inhibitors bind to S1 and S1′ subsite of MMP-9, which forms a deep hydrophobic pocket similar to the hydrophobic moieties in the ACE active site.¹¹³ Lisinopril is stabilized to the MMP-9 active site by specific hydrogen bonds and hydrophobic interactions, and its hydrophobic group showed greater affinity with the S1 site in comparison to the S1′ site.¹¹³ In addition, imidapril is stabilized to the MMP-9 active site with less molecular distortion, which may explain the greater MMP-9 inhibitory activity of imidapril compared to lisinopril.¹¹⁶ In fact, in vivo experiments showed that hamsters treated with lisinopril or imidapril had decreased MMP-9 activity in the post-MI LV, with MMP-9 activity being lower in the group that received imidapril compared to lisinopril.¹¹⁶ In a separate study, post-MI patients treated with perindopril showed decreased levels of MMP-1 in plasma and attenuated LV dysfunction.¹¹⁷

Similarly to ACE inhibitors, angiotensin II receptor antagonists inhibits MMPs levels and improve ECM remodeling post-MI.¹¹⁸ Rats treated with losartan showed attenuated mRNA transcription and protein expression of MMP-8, MMP-13, TIMP-1, TIMP-2, and type I collagen content in the remote zone post-MI.¹¹⁸ Treatment with valsartan (a selective angiotensin II type 1 receptor antagonist, AT₁) but not PD123319 (a selective angiotensin II type 2 receptor antagonist) decreased levels of MMP-2, −3, and −9 post-MI in rats, indicating that angiotensin II receptor blockage effects on ECM remodeling is due to inhibition of the AT₁ receptor.¹¹⁹ Deletion of the AT₁ receptor in mice increased survival rate and decreased LV remodeling post-MI.¹²⁰ Moreover, patients treated with trandolapril and valsartan showed attenuated MMP-9 levels in plasma and suppressed LV remodeling post-MI.¹²¹ While further investigation is needed to assess the mechanism whereby ACE inhibitors and angiotensin II receptor antagonists decrease MMPs and ameliorate LV dysfunction post-MI, the therapeutical inhibition of MMPs by inhibition of the angiotensin pathway is encouraging.

The National Institutes for Health and Clinical Excellence (NICE) recommends aldosterone antagonist therapy should be given within 3 to 14 days post-MI (preferably after treatment with ACE inhibitor has been initiated) to treat and prevent LV systolic dysfunction and heart failure in post-MI patients.¹²² Aldosterone increases MMP-2 and −9 activity in cultured adult rat ventricular myocytes through modulating mitogen/extracellular signal-regulated kinase and extracellular signal-regulated kinase 1/2 phosphorylation, and increasing ROS production.¹²³ Pretreatment with spironolactone (aldosterone receptor antagonist) abolished the aldosterone-induced increase in MMP activity and decreased collagen deposition post-MI.^{123, 124} Long-term treatment (24 weeks) with spironolactone in heart failure patients improved LV dysfunction, and attenuated plasma MMP-9, TIMP-1, and type I collagen carboxyterminal telopeptide concentrations.¹²⁵

Others vasoactive peptides, such as endothelin-1, can alter MMP expression and ECM remodeling post-MI. The endothelin receptor subtype A (ET_A) is predominate on myocytes and its activation induces myocyte hypertrophy.^{126, 127} Thus, endothelin increases collagen expression and collagenase activity in cardiac fibroblasts.¹²⁸ Rats treated with sitaxsentan (ET_A receptor antagonist) attenuated MMP-1, −2, and −9, increased TIMP-1, and improved post-MI cardiac dilation.¹²⁹

The American Heart Association and American College of Cardiology recommend initiating beta-adrenergic antagonists in all post-MI patients and continuing therapy indefinitely.¹³⁰ Dogs treated with atenolol, at supra-therapeutic doses, showed decreased MMP activity and improved LV stiffness in an experimental pacing-induced heart failure model.¹³¹ Rats treated with metoprolol showed decreased MMP-2 mRNA levels and decreased oxidative stress markers post-MI.¹³² Similar results were observed in post-MI pigs treated with carvedilol or metoprolol, both of which decreased MMP-2 activity, MCP-1 expression, and macrophage infiltration.¹³³ Patients with heart failure treated with standard therapy plus carvedilol showed reduced MMP-9 activity in plasma.¹³⁴

Antiplatelet, anticoagulant, and thrombolytic therapies are administrated after cardiovascular events to improve patient outcome.¹³⁵ Post-MI patients treated with tissue plasminogen activator (tPA) showed increased plasma MMP-9 levels.¹³⁶ Direct thrombin inhibitors and heparin are commonly used in post-MI therapy.^{137, 138} Heparin treatment increases MMP-1 and −2 in cultured fibroblasts, and MMP-1, −2, −3, and −9 in mesangial cells.^{139, 140} The effect of heparin therapy on MMPs in vivo has not been determined.

The use of non-steroidal anti-inflammatory drugs (NSAIDs) are highly associated with risk of MI, however, selective cyclooxygenase-2 inhibitors reduce C-reactive protein levels and improve endothelial function.¹⁴¹ Aspirin is an NSAID administrated as an antiplatelet and is highly recommended for patients after first MI or with heart failure.¹⁴² MMP-2 gene expression was attenuated with NSAID treatment by inhibiting specific protein (SP)-1 transcription factor from binding to the MMP-2 promoter site.¹⁴³ Up to date, no studies have examined the effects of NSAIDs on MMPs post-MI. In vitro experiments in human endothelial cells indicate that aspirin attenuates MMP-1 but not MMP-2 or −9.^{144, 145}

Statins (hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitors) exert a variety of pleiotropic effects, such as inhibiting isoprenoid intermediates production, which decrease the activation of Rho-family small GTP-binding proteins.¹⁴⁶ The inhibitory effect of statins on MMP-9 expression is dependent on nitrite-mediated mechanisms.¹⁴⁷ Interestingly, macrophages treated with statins showed increased MMP-12 expression, which is consistent with the concept that MMP-12 may be beneficial post-MI.⁸⁰ Human cardiac myofibroblasts treated with simvastatin showed reduced TNFα-induced invasion by MMP-9-dependent mechanisms.¹⁴⁸ Patients treated with pravastatin showed decreased MMP-2, MMP-9, c-reactive protein, and CD40L levels post-MI.^149–151 Combined there is ample evidence to support MMPs as off targets of a variety of CVD drugs.

General considerations

Multiple clinical studies have demonstrated the role of direct and indirect MMPi during the development of end-stage heart failure and LV remodeling. However, clinic trials using nonspecific MMPi have been inconclusive.⁹⁷ In addition, animal studies with direct MMPi have given mixed results. To develop a successful therapeutic treatment, additional studies are needed to understand the entire composite of MMP roles. Identifying the downstream effects of MMPs may provide answers and serve as novel and more selective treatment strategies for post-MI patients.⁷³

Future Directions

While the cardiac MMP literature field has exploded over the last 15 years, future research is needed. There is a need to better understand the biological function of MMPs in cardiac maintenance and tissue repair after injury. Out of the 25 MMPs identified to date, about half of the known MMPs have been characterized post-MI with MMP-2 and −9 being the main focus. In addition, the role of the MMPs that have been studied is complicated by to interactions between MMP family members. For example, inhibition of a specific MMP can result in an increase of other MMPs due to compensatory effects. MMPs also compete with each other for the same substrate. Competition assays evaluating the interaction with other MMPs is necessary in order to fully characterize the role of MMPs in the post-MI environment.

Furthermore, each MMP has a broad range of substrates that include chemokines, cytokines, adhesion molecules and growth factors as well as ECM components such as collagen and fibronectin.⁷² Post-MI, small fragments or peptides generated by MMP proteolysis are increased both in plasma and in the infarct area. Evaluating the efficacy of using MMP cleavage products for diagnostic and therapeutic purposes in heart failure is warranted.

Temporal and spatial MMP patterns during LV remodeling need to be determined to gain a more extensive understanding of MMPs. For example, MMP-9 gene deletion has proven beneficial post-MI in mice, yet, macrophage-specific transgenic overexpression of MMP-9 has also shown to be beneficial by attenuating the post-MI inflammatory response.¹¹¹ This suggests that timing and concentration of MMPs may dictate divergent mechanisms of response.^{108, 111}. Combined, the elucidation of these MMP mechanisms will help us to more completely understand how MMPs coordinate the post-MI response.

Conclusion

In conclusion, understanding the role of MMPs during post-MI remodeling remains an important issue.¹⁵² A better understanding of pathophysiological processes, including the biological function of the downstream products of MMPs, may lead to new strategies for the post-MI patient, particularly therapies limiting heart failure progression.