Mitochondrial mitophagic mechanisms of myocardial matrix metabolism and remodelling

Abstract

High levels of homocysteine (Hcy), known as hyperhomocysteinmia (HHcy), are correlated with an increase in extracellular matrix remodelling (ECM) via the matrix metalloproteinases (MMPs) and plasminogen/plasmin system. This results in an increase deposition of collagen that leads to endothelial-myocyte (EM) and myocyte-myocyte (MM) uncoupling; the physiological consequences are a plethora of cardiovascular pathologies. Homocysteine-induced increase in intracellular and mitochondrial Ca²⁺ plays an important role in increasing reactive oxygen species (ROS) within mitochondria and instigating mitophagy within the cell. This occurs via several Hcy-mitigated processes: agonizing N-methyl-d-aspartate receptor-1 (NMDA-R1), decreasing expression of peroxisome proliferator activator receptor (PPAR) [thereby increasing oxidation], impairing Ca²⁺ handling via Na⁺/Ca²⁺ exchanger (NCX1) and Sarco endoplasmic reticulum Ca²⁺ ATPase (SERCA-2a). The end result is an increase in ROS that directly or indirectly lead to MMP activation within mitochondria or the cytoplasm. Hcy induces a mitochondrial permeability transition that allows MMPs to be released from mitochondria thereby metabolizing matrix and impairing cardiac function. Further work remains to be elucidated concerning the specific mitochondrial mitophagic mechanisms under which matrix metabolism and remodelling occurs. Moreover, the therapeutic implications of NMDA and PPAR ligands are some promise to patient.

Introduction

A search for other risk factors for heart disease instigated the discovery of homocysteine (Hcy). Hcy has been shown to be an independent risk factor in several cardiovascular pathologies (Wilcken & Wilcken, 1976). Moreover, elevated levels of Hcy have been found in diabetics, which could explain the association between diabetes and heart disease (Hofmann et al., 1997; Robillon et al., 1994). Vitamin treatment has been shown to be successful in significantly lowering Hcy levels; however, it still remains a puzzle for why clinical trials have yielded mixed results in lowering the risk for cardiovascular-related death.

The heart’s conduction system is of utmost importance for proper cardiac function. The mechanism of transmission of membrane depolarization is via gap junctions; among proteins that make up these gap junctions are the following: connexin 40 (Cx40), connexin 43 (Cx43), connexin 45 (Cx45) (Gros & Jongsma, 1996; Miquerol et al., 2004). Deposition of collagen in extracellular matrix (ECM) can impair cardiovascular function via endothelial-myocyte uncoupling (EM uncoupling) and myocyte-myocyte (MM uncoupling) whereby the signal is not properly transmitted (Givvimani et al., 2011; Moshal et al., 2008b). Matrix metalloproteinase’s (MMPs) and the plasminogen/plasmin system are responsible for modifying the ECM; activation of MMPs is considered pathological, whereas activation of plasmin is thought to be beneficial in most cases (Iimoto et al., 1988; Jalil et al., 1989; Mukherjee & Sen, 1990; Norton et al., 1997; Takeshita et al., 2004). Most MMPs are generally considered to be either secretory or membrane-anchored (Hadler-Olsen et al., 2011; Pei et al., 2000). However, MMPs have also been found to have roles within the nucleus and mitochondria (McCawley & Matrisian, 2001).

Activation of NMDA-R1 will increase intracellular Ca²⁺ levels, and mitochondrial Ca²⁺ levels, resulting in oxidative stress (Gao et al., 2007). Mitochondria are involved in translocating and activating several proteins (Hansson Petersen et al., 2008). Importing proteins involves translocation through outer (OMM) and inner mitochondrial membranes (IMM) via complex protein machinery (Ow et al., 2008; Rassow et al., 1994). Another function of mitochondria aside from the role of ATP generation is sequestration of Ca²⁺, and the generation and detoxification of cellular ROS; the electron transport chain is a key connector for these roles (Brand et al., 2004; Turrens, 2003; Xi et al., 2005). It is concluded that an increase in Ca²⁺ will disrupt the membrane potential of mitochondria, decrease oxygen utilization (since oxygen is the final electron acceptor to generate water) and produce greater amounts of superoxide (Archer, 2010; Dalton et al., 1999). One other mechanism for MMP-9 activation involving mitochondria is via the calpain system. It was shown that Hcy activates and translocates calpain-1 from the cytosol to mitochondria; this results in intra-mitochondrial oxidative stress, resulting in MMP-9 activation within mitochondria (Tyagi et al., 2010).

The activation of the PPAR receptor by Hcy has been shown to promote a reducing environment (antioxidant) (Brude et al. 1999; Hunt & Tyagi, 2002; Inoue et al., 1998). However, an increased concentration of Hcy correlates with a decreased expression of the PPAR receptor and its antioxidant effects (Brude et al., 1999; Gillespie et al., 2011; Inoue et al., 1998).

Homocysteine as a marker for cardiovascular pathology

Homocysteine (Hcy) is a sulphur-containing amino acid that is derived from the essential amino acid, methionine (HofMann et al., 2001; Zhou et al., 2001). Moreover, Hcy is shown to be metabolized from two pathways: trans-sulphuration by cystathionine-β-synthase (CBS) in hepatic cells or via re-methylation to methionine in non-hepatic cells (Loscalzo 2006). Hyperhomocysteinemia (HHcy) is considered an independent risk factor for cardiovascular disease (CVD) (Zhou and Austin 2009). Elevated Hcy levels are caused by two factors: genetic defects in enzymes involved in Hcy metabolism or nutritional deficiencies in vitamin cofactors (folate, vitamin B12, vitamin B6 (Milani and Lavie 2008). Other factors involved are the following: chronic kidney disease, hypothyroidism, psoriasis, cancers, and several drugs (Milani & Lavie, 2008).

Folic acid, Vitamin B12, and Vitamin B6 combinations are able to reduce Hcy concentrations without lowering the risk of further cardiovascular events; it has been suggested that some of these vitamin treatments may counteract the beneficial effects of lowering Hcy, and cause damage themselves (Bonaa et al. 2006). For instance, treatment with B vitamins did not lower the risk of recurrent CV disease after MI. Pathogenesis of coronary artery disease show reduced ability to metabolize Hcy in premature coronary artery disease (Nasir et al., 2007; Wilcken & Wilcken 1976). One study indicates that the greater the increase in Hcy induced by fenofibrate, the smaller the increase in HDL-c and apoA-I, two proteins that are helpful in cholesterol uptake; hence cholesterol levels would remain higher (Taskinen et al., 2009). Similar results were noted in a study of ischemic stroke patients: a higher level of Hcy was independently associated with ischemic stroke (Dhamija et al., 2009).

Vitamin treatment did lower Hcy levels by 27% among patients given folic acid plus Vitamin B12 (Bonaa et al.,2006). One study found that a Hcy level of ≥20 μmol/L is associated with a high mortality risk (odd ratio 2.57) (Maurer et al., 2010). Another study found that increased Hcy levels cause abnormalities in Na⁺ currents in human atrial cells via the following mechanism: slowing inactivation and promoting recovery of Na⁺ channels (Cai et al.,2009). Another study found that folic acid supplementation resulted in a significant intima-media thickness reduction after 18 months in patients with at least one CV risk (Ntaios et al., 2010). Finally, Hcy was shown to act as an independent risk factor for an increase in arterial stiffness (Ruan et al., 2009).

Homocysteine and diabetes act in synergy

Elevated levels of Hcy have been found in diabetics, which could explain the association between diabetes and heart disease (Hofmann et al., 1997; Robillon et al. 1994; Wijekoon et al. 2007). Another study showed an association between Hcy and silent myocardial infarction (SMI) in diabetic patients (Tarkun et al., 2004). However, HHcy was not detected in adolescent patients with type 1 diabetes (Pavia et al., 2000). In cases where there were no renal complications in both Type 1 and Type 2 diabetes, Hcy levels were even lower than controls (Wijekoon et al., 2007). It was concluded that in Type 1 diabetes, increased activity of the trans-sulphuration enzymes were the major cause of reduction in plasma Hcy. In Type 2 diabetes, BHMT(betaine:homocysteinemethyltransfer ase) was considered to be responsible in increased Hcy catabolism (Wijekoon et al., 2007). Plasma levels of Hcy are usually normal in diabetes; however, both high and low values have been reported. They have been modulated by hyperfiltration and renal dysfunction, and low folate status, while insulin resistance does not seem to play a role in HHcy (Huijberts et al., 2005). It appears that the presence of diabetes contributes to worsening HHcy determined cardiovascular risk, and may even act in synergy in evoking their vascular effects (Becker et al., 2003; Soinio et al., 2004). One review indicates that, in patients with Type 2 diabetes, elevated Hcy levels have an independent risk associated with higher rates of CHD events and CHD mortality (Soinio et al., 2004).

Contraction and conduction – gap junctions

The integrity of the heart’s conduction system is paramount to maintaining proper cardiac function. This specialized pacemaker system consists of modified cardiomyocytes and includes the following: sinoatrial node (SA), atrioventricular node (AV), His Bundle with two branches, and Purkinje fibres (PFs). The molecular mechanism of transmission of membrane depolarization is via gap junctions; among proteins that make up these gap junctions are the following: Connexin 40 (Cx40), Connexin 45 (Cx45), and Connexin 43 (Cx43) (Gros & Jongsma, 1996; Miquerol et al., 2004). A single gap junction, for instance, is composed of 12 Cx43 units (Schulz and Heusch, 2006). Cx40 is highly expressed in active atrial myocytes, the central part of the AV node, the His bundle, as well as the Purkinje fibres. One study found that a null mutation in Cx40 results in impaired conduction and conduction block, suggesting the pivotal role that this connexin plays in transmitting signal from atria to ventricles (Gros et al., 2004; Tamaddon et al., 2000; van Rijen et al., 2001). Another connexin involved in gap junction machinery is Connexin 30.2 (Cx30.2); this protein was found in both SA and AV nodes; (Kreuzberg et al. 2006). Cx30.2 was actually shown to dampen the rate of impulse propagation in AV node in control mice versus mutant, determined by the PQ interval (Kreuzberg et al., 2006). This provides further evidence of the importance of different types of connexins as structural components of gap junctions in the cardiac pacemaker.

Furthermore, Cx43 is the primary connexin component of active myocytes of the atria and ventricles; ablation causes reduced conduction velocity, increased dispersion of conduction, and enhanced electrical sensitivity on the ventricle (Beyer et al., 1989; Reaume et al., 1995; van Rijen et al., 2004). Connexin 45 is located in SA and AV nodes (Coppen et al., 1998; Kruger et al., 2000; Verheijck et al., 2001), and its ablation is lethal in mice due largely to defects in vascularization (Kruger et al., 2000). Connexins are expressed in several tissues including heart, blood vessels, and neural tissue (Rackauskas et al., 2007). In fact, basilar artery SMCs are coupled in vivo with Cx43, Cx40, and other conductance channels; many of the channels involve non-homotypic components [not entirely comprised of the same connexin protein] (Li and Simard, 1999).

Endothelial-myocyte uncoupling and myocytemyocyte uncoupling

Remodelling is a process whereby there is synthesis and degradation of the ECM involving a very precise balance of proteinase/antiproteinase activity; an increase in this ratio has been shown to result in systolic and diastolic heart failure with uncoupling cardiomyocytes (Hunt et al., 2002). This would result in impaired depolarization of the signal between endothelial cells and cardiomyocytes [Endothelial-Myocyte uncoupling, E-M uncoupling], or between cardiomyocytes [Myocyte-Myocyte uncoupling, M-M uncoupling]. In fact, one study suggests a direct cause-and-effect relationship between MMP-9 activation and EM uncoupling in LV myocardium after chronic volume overload (Moshal et al., 2008b).

During heart disease, including hypertension, tissue inhibitors of matrix metalloproteinase’s (TIMPs) are oxidized and inactivated, thereby allowing matrix metalloproteinase’s (MMPs) to be activated (Rucklidge et al. 1992). A normal heart expresses four TIMP species: TIMP1, TIMP2, TIMP3, TIMP4; these TIMPs are altered during progression of human heart failure (Mann & Taegtmeyer, 2001). A reduction in TIMP3, thereby allowing MMP activation, results in adverse remodelling affects (Fedak et al. 2003). MMPs act as collagenases and elastases; however, their primary function is as collagenases. Hence, collagen deposition is greater than elastin deposition when MMPs are active. The accumulation of collagen disrupts the aforementioned connexin proteins, interfering with depolarization of cardiomyocytes, and impairing heart function. One study found that congenic transfer of TIMP ameliorated LV hypertrophy and cardiac dysfunction by inactivating MMP-9 involved in remodelling (Rodriguez et al., 2008). It was found that there are chamber-specific alterations in myocardial collagen content and MMP and TIMP levels that may provide for diagnostic and other insight into pathogensis of atrial fibrillation and chronic heart failure (CHF) (Mukherjee et al., 2006).

Moreover, nitric oxide (NO) generation from endocardial endothelium has a role in myocyte contraction, relaxation and heart rate (Brady et al., 1993; Pinsky et al., 1997). Hence, an increase of collagen between the endothelium and myocyte will lead to a longer contracting period since endothelial-mediated relaxation via NO is impaired (Moshal et al., 2005). In accordance with other results, another study found that an increased MMP activation contributes to the LV dilation and increased wall stress with pacing CHF (McElmurray III et al., 1999).

Remodelling via plasmin and MMPs

The dynamic nature of the cardiovascular architecture requires such remodelling systems to be in place and work in harmony to accommodate various pressures and stresses. However, when these systems are not in proper balance, or when the stresses exceed the accommodating capacity, physiological problems ensue (Heistad et al., 1991; Wilson et al., 1993). It was found that myocardial fibrosis and ECM remodelling are characteristics of the failing heart (Weber et al. 1992).

The purpose of collagen fibres is to provide an elastic force in the myocardium that allows Starling Forces to operate properly (Weber et al., 1989). The following are characteristic of ECM remodelling that can contribute to stiffness and systolic/diastolic failure: total collagen content, collagen subtypes, collagen protein stability, collagen cross-linking (Iimoto et al., 1988; Jalil et al., 1989; Mukherjee & Sen 1990; Norton et al., 1997). The activation of MMPs involves collagen degradation with replacement of fibrotic tissue (Dollery et al., 1995; Li et al., 2000b; Maquart et al., 1988).

Remodelling of the vessel wall will determine lumen diameter after vascular injury or hemodynamic forces (de Smet et al., 2000; Mintz et al., 1996; Mondy et al., 1997; Tyagi 1999). The smooth muscle cells (SMCs) remodel existing extracellular matrix, as well as contributing to further matrix deposition/removal, thereby altering the phenotype (Tummalapalli & Tyagi 1999). One aspect of atherosclerosis involves the migration of SMCs from media to intima of vasculature (Ross, 1993; Schwartz, 1997). In order for this deleterious migration to occur, a degradation of matrix is necessary via MMPs and fibrinolytic plasminogen/plasmin (Cho and Reidy, 2002; Clowes et al., 1990; Kenagy et al., 1996).

This destructive cascade is first initated from the conversion of plasminogen to plasmin by two plasminogen activators: tissue plasminogen activator (tPA) and urokinase plasminogen activator (uPA) (Carmeliet et al., 1995; Lijnen, 2001). Since MMPs are secreted as inactive zymogens (pro-MMPs), they require proteolytic activation by t-PA (Malemud, 2006). Plasminogen activator inhibitor-1 (PAI-1) is responsible for inhibition of uPA and tPA, thereby mediating the destructive process whereby matrix is metabolized and migration of SMCs occurs (Carmeliet et al., 1997; Hasenstab et al., 2000). Further evidence for higher levels of PAI-1 becoming detrimental was cited in a review of cancer and plasmin activators (Andreasen et al., 1997). One study showed the effects of mutations in the following: u-PA, t-PA, MMP-9 (Heymans et al., 2005). In t-PA deficient mice, cardiomyocyte hypertrophy was discovered in conjunction with myocardial fibrosis, LV dilation, dysfunction after 7 weeks (Heymans et al., 2005); this is logical since plasmin levels would decrease, thereby reducing fibrinolysis. One study showed evidence that reduced inactivation (moderate activation) will result in a decrease in plasmin levels, thereby decreasing remodelling after myocardial infarction (Askari et al., 2003). In conjunction with this study, another study showed that an increase of PAI will serve the opposite role: decrease plasmin, and increase in myocardial fibrosis after infarction (Takeshita et al., 2004). One study indicated that the use of MMP-inhibitors would preserve cardiac pump function in LV overloading (Heymans et al., 2005).

Injury of a vessel can lead to leakage of proteins into the interstitial space, which activates the coagulation cascade with deposits of fibrin, the major substrate for plasmin (Loskutoff & Quigley, 2000). However, it is generally accepted that fibrinolysis is a good thing since fibrotic disease is detrimental in all major tissues. Hence, activation of PAI-1 results in greater fibrosis, whereas the inactivation of PAI-1 results in increased fibrinolysis. This was demonstrated in PAI-1 deficient mice compared with control mice: fibrinolysis was enhanced, collagen build-up was reduced, and survival was dramatically prolonged in bleomycin-treated mice (Hattori et al., 2000). One in vivo study found that in SMCs PAI-1 plays a role in limiting flow-induced SMC migration, thereby playing a pivotal role in controlling vascular remodelling (Cullen et al., 2004).

MMPs: intracellular, extracellular, intranuclear, intramitochondrial

A second system of remodelling includes MMPs; MMPs are a class of zinc endopeptidases that initially exist in a pro-form that is further activated upon cleavage. Like most biological mediators, the two systems [MMPs, plasminogen/plasmin] that play a role in remodelling are not mutually exclusive. Plasmin, for instance, can also cleave the inactive zymogen MMP to the active form (Dollery et al., 1995; Lijnen 2001). Both of these systems are known to be active in plaque formation as part of the atherosclerotic process; all of the following are increased in expression/activity: MMPs (MMP-2, MMP-9), tPA, uPA (Dollery et al., 1995; Lijnen 2001). This role was further confirmed by using MMP inhibitors and in vivo models of mice lacking uPA or uPA and tPA; in such cases, SMC migration and intimal thickening were reduced (Bendeck et al., 1996; Dollery et al., 1995).

Most MMPs have two methods for completing their role of digesting substrate: secretory and membrane-anchored roles [via a type 1 transmembrane domain or glycosylphosphatidylinositol linkage] (Pei et al., 2000). However, MMPs have also been found to have roles within the nucleus and mitochondria. There has been a recent understanding that MMPs are not only located in the matrix (McCawley & Matrisian 2001), but also act intracellularly. In fact, it has been shown that MMP-2 is expressed by fibroblasts and cardiomyocytes, and can be found with contractile proteins such as troponin-I and sarcomeres (Schulz, 2007; Wang et al., 2002). Moreover, MMP-2 activation has been shown to reduce performance of contraction after ischemia-reperfusion injury (Singh et al., 2000). The mitochondria have also been shown to contain MMPs: mtMMPs. One study has shown that ROS, possibly generated from mitochondria, can increase MMP-2 expression as well as activation (Nelson and Melendez, 2004).

Another study has found that MMP-1 was not only found in interstitial space, but also intracellularly, intranuclear, and within mitochondria (Limb et al., 2005). Inhibition of the enzyme with interference RNA (RNAi) or broad MMP inhibitor resulted in faster degradation of lamin A, activation of caspases, and fragmentation of DNA compared to controls. This suggests that MMP-1 expression allows the cell to resist apoptosis, thereby explaining a known mechanism whereby tumour cells may survive for a longer period of time (Limb et al., 2005).

One study has found that MMP-3 is located in the nucleus and is involved in apoptosis (Si-Tayeb et al., 2006). A mutation of MMP-3 resulted in decreased apoptosis; hence, MMP-3 activation is involved in apoptosis (Si-Tayeb et al., 2006). This is in stark contrast to MMP-1 expression in cells that allow cells to resist apoptosis (Limb et al., 2005). Another study found that MMP-2 is present in the nucleus of cardiac myocytes with the role of cleaving poly (ADP-ribose) polymerase (PARP) in vitro (Kwan et al., 2004).

Hcy and NMDA-R1 activation leading to an increase in Ca²⁺

Chronic heart failure (CHF) includes propensity of arrhythmias, systolic failure and diastolic failure; moreover, an overactive sympathetic system can contribute to the abnormality, and decrease in normal function (Colucci et al., 1981; Singh et al., 2000; Sood et al., 2002). CHF has been shown to correlate with an increase in glutamatergic activity that mediates sympathetic regulation. An upregulation of N-methyl-D-aspartate receptor-1 subunits (NMDA-R1) in the hypothalamus during CHF has been demonstrated (Li et al., 2003). Moreover, ischemia and reperfusion-induced arrhythmias are sensitive to NMDA-R1 blockade (D’Amico et al., 1999). One study has shown that the antagonist to NMDA-R1, MK-801, protects against Hcy-induced oxidative damage in neurons (Folbergrova, 1994), and an increase in heart rate by an analogue of NMDA (D’Amico et al., 1999). Another study found that the activation of NMDA-R by Hcy increases oxidative stress and Ca²⁺ load in mitochondria, leading to cardiomyocyte death in neonatal rats (Gao et al., 2007).

NMDA-R1 is well-known as being expressed in neural tissue; however, NMDA-R1 is now known to be expressed in cardiomyocytes and endothelial cells (Huang & Su, 1999; Krainc et al., 1998; Qureshi et al., 2005). Activation of NMDA-R1 will increase intracellular Ca²⁺ levels, and mitochondrial Ca²⁺ levels, resulting in oxidative stress (Gao et al., 2007). HHcy was shown to decrease myocyte contractile performance by agonizing the NMDA-R1 receptor. An increase in Hcy decreased the contraction amplitude with an increase in Ca²⁺ concentration; recent studies suggest that HHcy condition increased mitochondrial NO levels and mitochondrial permeability transition (MPT), leading to the poor cardiac performance (Moshal et al., 2009). This cascade of events is illustrated in Figure 1.

Figure 1.

Hcy increases intracellular Ca²⁺ by: agonizing NMDA-R1 receptor, impairing ability of NCX-1 protein to extrude Ca²⁺ from the cell in exchange for Na⁺, impairs SERCA-2a uptake of ER Ca²⁺. This increases Ca²⁺ in mitochondria, disrupting electron transport chain, and increasing presence of ROS. An increase in ROS will activate MMPs; Calpain-1 will activate mitochondrial pore transition, resulting MMPs exiting mitochondria and causing contractile dysfunction.

Mitochondrial mechanism of ECM metabolism

Hcy-induced increase of Ca²⁺ leads to ROS production. The mitochondria are involved in several cellular processes aside from its well-known role of providing energy through oxidative phosphorylation. In fact, mitochondria have a well-known role in cellular death that includes the release of many pro-apoptotic intermembrane space proteins: cytochrome c, apoptosis inducing factor, endonuclease G, and DIABLO/Smac (Du et al., 2000; Kroemer & Reed 2000; Li et al., 2001; Liu et al., 1996; Spiess et al., 1999; Susin et al., 1999; Van et al., 2001; Verhagen et al., 2000). Another protein found to be released is Omi, a homologue to the bacterial HtrA gene product: a chaperone and active protease (Van et al., 2002a). HtrA2/Omi has a role in degrading improperly folded proteins in times of cellular and endoplasmic reticulum stress, heat-shock, and even ischemia/reperfusion (Faccio et al., 2000; Gray et al., 2000). In fact, many studies demonstrate a role for serine proteases in apoptotic cell death (Kagaya et al., 1997; Wright et al., 1997). One study also found that the alteration of mitochondrial membrane potential contributes to apoptosis. A decrease in mitochondrial membrane potential leads to matrix condensation with an exposure of cytochrome c into IMM space; this facilitates cytochrome c release and cell death (Gottlieb et al., 2003).

Mitochondria are involved in translocating and activating several proteins. For example, amyloid β-peptide is imported into mitochondria via translocase of outer membrane (TOM) import machinery, and is localized into the mitochondrial cristae (Hansson Petersen et al., 2008). Import of proteins involves translocation through the outer (OMM) and inner mitochondrial membranes (IMM) involving complex protein machinery (Rassow et al., 1994). One such example of this is the MIM44 and mt-hsp70 cooperation in translocation of pro-proteins (Rassow et al., 1994). For instance, it was found that the translocation of the protein, Bax, and its activation to mitochondria will alter the mitochondrial transmembrane potential; the consequences could be cell death (Tikhomirov and Carpenter, 2005).

Another function of mitochondria aside from the role of ATP generation is sequestration of Ca²⁺, and the generation and detoxification of cellular ROS; the electron transport chain is a key connector for these roles (Brand et al., 2005). One study found that endothelin-1 (ET-1) promoted oxidative stress through mitochondrial ROS in vascular smooth muscle cells (Touyz et al., 2004). Rate of superoxide formation within mitochondria is greatly affected by the coupling state of mitochondria; Ca²⁺ plays a great role in this state (Dalton et al., 1999). Furthermore, the redox state can determine the oxidation state of thiols and pyrimidine nucleotides (Dalton et al., 1999). An increase in intracellular Ca²⁺ will also increase mitochondrial Ca²⁺ via a simple mechanism linked to the proton gradient that is established via the electron transport chain (Dalton et al., 1999). Hence, it is concluded that an increase in Ca²⁺ will disrupt the membrane potential of mitochondria, decrease oxygen utilization (since oxygen is the final electron acceptor to generate water) and produce greater amounts of superoxide (Dalton et al., 1999).

A mitochondrial manganese-containing superoxide dismutase (Sod2) is one of several enzymatic defences to reduce injury from oxidation (Van et al., 2002b). Sod2, for instance, is responsible for catalysing the reaction of superoxide to hydrogen peroxide. One study also indicates that a Sod2-dependent production of hydrogen peroxide leads to MMP-1 expression. Mice that have this gene removed will develop fibrosis and collagen deposition. Under basal conditions, mitochondria have a buffering capacity that is largely determined by glutathionine redox system (Ranganathan et al., 2001). Sod2 expression results when such systems are overwhelmed (Ranganathan et al., 2001). Another study showed that over expression of Sod stimulated activation of MMP-2 with an increase of ROS (Zhang et al., 2002). One possible mechanism for MMP activation in mitochondria is generation of hypochlorous acid (HOCl) from H2O2 by the enzyme, myeloperoxidase (Fu et al., 2001). One study indicated that HOCl regulates the activity of MMP-7 in vitro. In addition, HOCl activated pro-MMP-7 to MMP-7 in vivo via converting thiol residue of cysteine switch to sulphinic acid (Fu et al., 2001). Figure 2 illustrates how ROS, produced from mitochondria, can cleave pro-MMP to active MMPs; this results in contractile dysfunction. This is a distinctly different mechanism from proteolytic cleavage of MMP (Mukherjee et al., 2006).

Figure 2.

Hcy increases intracellular Ca²⁺ by: agonizing NMDA-R1 receptor, impairing ability of NCX-1 protein to extrude Ca²⁺ from the cell in exchange for Na⁺ impairs SERCA-2a uptake of ER Ca²⁺. This increases Ca²⁺ in mitochondria, disrupting electron transport chain, and increasing presence of ROS. An increase in ROS will activate MMPs. ROS is also generated via decreased expression of PPAR receptor, allowing greater presence of ROS that will activate MMPs within the cell, and result in contractile dysfunction.

One other mechanism for MMP-9 activation involving mitochondria is via the calpain system. It was shown that Hcy activates and translocates calpain-1 from the cytosol to mitochondria; this results in intra-mitochondrial oxidative stress, resulting in MMP-9 activation within mitochondria. There has also been a link to ERK ½ pathway in activating calpain-1 (Moshal et al., 2006). Moreover, this is a Ca²⁺-dependent mechanism whereby calpain-1 is activated via induction of dissociation of calpain subunits (Moshal et al., 2006). The requirement for Ca²⁺ activated calpain-1 in MMP-2 and MMP-9 expression was also demonstrated via the calpain inhibitor, CP1B which reduces expression of MMP-2 and MMP-9 (Popp et al., 2003). Use of ERK ½ blocker also resulted in a decrease in MMP-9 expression (Moshal et al. 2006). It has also been suggested that there is a negative feedback mechanism involved whereby an increase in ROS would impair mitochondrial membrane potential, thereby disrupting function (Zhou et al., 2007).

A mechanism whereby Hcy controls this process begins with calpain protease activation; upon activation, there is an induction of mitochondrial permeability transition (MPT) (Moshal et al., 2008a). Treatment with MK-801, a blocker of NMDA-R1 will attenuate the induction of MPT in the presence of Hcy (Moshal et al., 2008a). The mechanism by which this is proposed to occur is Hcy-induced ROS and Ca²⁺ load in the mitochondria. One study reports that HHcy increases MMP-9 expression by agonizing the NMDA-R1 receptor, the consequences of which are an increase in ROS and Ca²⁺ load in mitochondria (Moshal et al., 2008c). One study found that in conditions of HHcy, the Ca²⁺ clearance rate declined from a decrease in the expression of SERCA-2a and NCX – Ca²⁺ handling proteins. An increase in Ca²⁺ may have induced MPT, thereby reducing the ability of the mitochondria to generate ATP resulting in a decline of myocyte contractility (Moshal et al., 2008c). The mechanism of Hcy-induced activation of calpain and MPT is emphasized in Figure 1, whereby activated MMPs translocate and can cause contractile dysfunction.

Another study analysed the effects of reactive oxygen species on neutrophil and fibroblast collegnases; it was found that pro-MMP-8 was preferentially activated by ROS such as hydrogen peroxide and hypochlorous acid versus the traditional serine proteinases: trypsin, chymotrypsin (Saari et al., 1992). Again, figure 2 illustrates this kind of interaction whereby ROS of mitochondrial origin cleaves pro-MMP to MMP. Studies have shown that the plasminogen/plasmin system can be activated by oxidative stress (Tanaka et al., 1997), and also by xanthine/xanthineoxidase generation (Liu & Gaston Pravia, 2010). Another study indicated that reactive oxygen species are able to activate NF-κB. NF-κB is involved in many cytokine genes, such as TNF that stimulates plasminogen activators (Mariappan et al., 2010).

Other mechanisms of ROS generation and extracellular metabolism

There are several nuclear transcription factors (NF) receptors that control oxidation/reduction balance of the cell. NF-κB has been shown to be induced by Hcy (Ferlazzo et al., 2008; Wang et al., 2002). Activation of the PPAR receptor by Hcy has been shown to promote a reducing environment (antioxidant) (Brude et al., 1999). Moreover, an increased concentration of Hcy correlates with a decreased expression of the PPAR receptor and its antioxidant effects (Inoue et al., 1998). An activated PPAR receptor will increase the expression of superoxide dismutase (SOD) and catalase while decreasing NAD/NADPH oxidase (Inoue et al., 2000; Inoue et al., 2001; Poynter and Daynes, 1998; Takenouchi et al., 2010). This can serve as one potential mechanism by which high levels of Hcy result in the increased production of reactive oxygen species (ROS) and cell injury (Berman & Martin, 1993; Jiang et al., 2011; Mujumdar et al., 2001; Zhang et al., 2000). Moreover, it was found that agonists of PPAR will decrease oxidative stress and MMP activity in macrophages (Lee et al., 2011; Marx et al., 1998; McGregor et al., 2000). Another method whereby matrix remodelling is decreased by PPAR activity is via a decrease in mRNA expression of the plasminogen activators and increase of plasminogen activator inhibitors (Xin et al., 1999). One study suggests Hcy may enhance vascular constrictive remodelling by inactivating PPARα and PPARγ in ECs and PPARγ in SMCs (Mujumdar et al., 2002). Some data indicates members of the plasminogen activator system, in addition to MMP-2/9, increase with growing potential of ovarian tumours; hence there has been some interest in using MMP-inhibitors to treat certain types of cancer (Schmalfeldt et al., 2001).

The transcription factor, TNFα, was also found to increase MMP expression; over expression of TNFα, and subsequent MMP expression, can cause heart failure phenotype (Kubota et al., 1997; Li et al., 2002). One study found that ECM remodelling in transgenic mice that over express TNFα can be modulated using an anti-TNFα treatment (Li et al., 2000a).

Conclusions

The role of mitochondria in myocardial matrix metabolism and remodelling is still not clear. This review briefly mentions some mechanisms that could activate the MMP system and modulate plasminogen/plasmin that involves Hcy-induced production of oxidative stress during cardiovascular remodelling.