Metabolic Stress: Mitochondrial Function in Neointimal Formation

Abstract

Mitochondria regulate major aspects of cell function by producing ATP, contributing to Ca²⁺ signaling, influencing redox potential, and controlling levels of reactive oxygen species. In this review, we will discuss recent findings that illustrate how mitochondrial respiration, Ca²⁺ handling, and production of reactive oxygen species (ROS) affect vascular smooth muscle cell function during neointima formation. We will review mitochondrial fission/fusion as fundamental mechanisms for smooth muscle proliferation, migration, and metabolism, and examine the role of mitochondrial mobility in cell migration. In addition, we will summarize novel aspects by which mitochondria regulate apoptosis.

Graphical Abstract

INTRODUCTION

Approximately 500,000 percutaneous-coronary and 50,000 peripheral-balloon angioplasties are performed annually in the U.S.¹. Following balloon angioplasty, neointimal formation leads to clinically significant restenosis of a target vessel in about 3-20% of patients.² In addition, neointimal formation is the leading cause of vein graft failure after coronary artery bypass surgery ³ and occurs after surgical endarterectomy ⁴.

The de-differentiation and proliferation of vascular smooth muscle cells (VSMCs) and their migration from the media to the neointima are major contributors to neointimal formation ^5,6. Lessons from animal models have demonstrated that proliferation in the intima increases within 96 hours after injury and persists for up to 8 weeks ^6,7. The effects of proliferation and migration are counterbalanced by apoptosis of VSMCs that occurs during the acute phase following injury and at later time points; however, this process is insufficient to abolish neointimal formation ^8-10.

The proliferation and migration of VSMCs is likely affected by mitochondrial function, as mitochondria provide energy and regulate reactive oxygen species (ROS) and Ca²⁺ levels. Cytosolic signaling triggers mitochondrial activity through multiple pathways, including uptake of Ca²⁺ through the mitochondrial Ca²⁺ uniporter (MCU) ¹¹ or activation of the mitochondrial K_ATP channel that triggers mitochondrial ROS production ^12,13. Notably, mitochondrial ROS and energy production are promoted by Ca²⁺-dependent activation of dehydrogenases of the TCA cycle (Figure 1) ¹⁴. In addition, changes in mitochondrial shape affect mitochondrial respiration, Ca²⁺ and ROS levels ^15,16, adding further complexity to their effect on smooth muscle phenotypes. Thus, there are multiple mechanisms by which mitochondria can affect VSMC function, which are discussed in detail below.

Figure 1: Overview of mechanisms by which mitochondria modulate neointima formation.

Ca²⁺ influx via the mitochondrial Ca²⁺ uniporter (MCU) complex increases activity of the tricarboxylic acid (TCA) cycle by upregulating the activity of pyruvate dehydrogenase (PDH), isocitrate dehydrogenase (IDH), alpha-ketoglutarate dehydrogenase (αKDGH), and succinate dehydrogenase (SDH). MCU activity is increased by phosphorylation by the Ca2+/calmodulin-dependent kinase II (CaMKII). Activity of the electron transport chain (ETC) via oxidative phosphorylation generates ROS (O2•−) and ATP, promoting vascular smooth muscle cell (VSMC) proliferation and migration. Superoxide dismutase 2 (SOD2) reduces mitochondrial O2•− production, proliferation, and migration. Uncoupling protein-2 (UCP2) decreases the mitochondrial membrane potential, O2•− production and neointima formation. VSMC apoptosis that counteracts neointima formation occurs upon opening of the mitochondrial permeability transition pore (mPTP) or by formation of BAX/BAK complexes.

MITOCHONDRIAL ROS PRODUCTION

Smooth muscle cells respond to growth factor stimulation by increasing intracellular production of ROS ¹⁷. Mitochondria are the major source of intracellular oxygen radicals under physiological conditions. They contribute to O2•− production as electrons are moved through the electron transport chain (ETC), in particular in complex I and II, where two-electron carriers donate to one-electron carriers ¹⁸. Mitochondrial O2•− is readily converted to H₂O₂ by superoxide dismutase 2 (SOD2), which is degraded by mitochondrial peroxiredoxin 3 ¹⁹ or freely diffuses into the cytoplasm where it induces cytoplasmic ROS, e.g. via activation of PI3 kinase and RAC1 ²⁰. Other sources of mitochondrial O2•− include alpha-ketoglutarate dehydrogenase, pyruvate dehydrogenase, glycerol 3-phosphate dehydrogenase, and fatty acid beta-oxidation ^21,22. Whether the NADPH oxidase NOX4 contributes to O2•− production in the mitochondrial matrix of vascular cells remains controversial ^23,24.

The suppression of ROS production reduces VSMC proliferation, migration and neointima formation ²⁵. The contribution of mitochondrial ROS production to VSMC proliferation, migration and neointima formation has been corroborated in loss- and gain-of-function studies of SOD2 ^26,27. The regulation of SOD2 expression or activity has been studied as an approach to reduce neointima formation. Overexpression of the peroxisome proliferator-activated receptor-gamma coactivator-1alpha, a major regulator of mitochondrial biogenesis, increased SOD2 expression and inhibited VSMC migration and neointima formation in the carotid balloon injury model ²⁸. The enzymatic activity of SOD2 is reduced after acetylation of lysines 68 and 122 at its catalytic center ²⁹, which is reversed by the mitochondrial deacetylase sirtuin-3 ³⁰. A direct effect of SOD2 acetylation has, to date, not been shown in neointima formation. However, this may be the mechanism by which the increased enzymatic activity of sirtuin-3 by stimulation of a nicotinic Ach receptor subtype reduced VSMC migration ³¹.

Negative mitochondrial matrix membrane potential is a result of the constant transport of protons from the mitochondrial matrix to the intermembrane space via the ETC. In contrast, uncoupling proteins (UCPs) are mitochondrial anion carriers that allow reentry of protons in the absence of ATP synthesis, leading to dissipation of the membrane potential. The O2•− leakage from the ETC depends on the mitochondrial proton gradient and the mitochondrial potential.³² Thus, overexpression or activation of uncoupling protein-2 (UCP2) is expected to limit mitochondrial ROS production, although this should also decrease maximal ATP synthesis as protons are diverted from the ATP synthase. In support of a role of UCP2 in neointima formation, it was downregulated from days 7 to 21 after injury in vivo, when neointimal formation was maximal ³³. Moreover, UCP2 knock down promoted VSMC proliferation and migration ³³. As a downstream mechanism, it was proposed that UCP2 decreased PDGF-dependent NFκB activation by reducing cytosolic ROS production. This mechanism links the antioxidant function of a specific mitochondrial protein to a canonical signaling pathway involved in neointima formation. Collectively, these studies provide evidence that interventions to lower ROS production, selectively in mitochondria may provide novel approaches to limit neointima formation.

MITOCHONDRIAL FISSION AND FUSION

Mitochondrial shape is highly synchronized during the cell cycle in order to accommodate the energy demands of the proliferating cells ³⁴. It is tightly regulated by GTPases that are responsible for mitochondrial fission (Dynamin-Related Protein-1 (DRP1)) and fusion (Optic Atrophy-1 (OPA1) and Mitofusin- (MFN)1 and 2, Figure 2) ³⁵. Interference with mitochondrial fission and fusion impairs cell migration and proliferation, for example in rat kidney ³⁴ and human lung cancer cells ³⁶, providing a rationale for exploring fission/fusion in neointima formation. Indeed, disrupting mitochondrial fission, by application of the pharmacologic DRP1 inhibitor Mdivi-1 ³⁷ or by transgenic expression of the dominant-negative DRP1 mutant K38A that does not hydrolyze GTP ¹⁵, reduced neointima formation in vivo.

Figure 2: Schematic representation of mitochondrial dynamics.

Mitochondrial shape is regulated by fission and fusion, which affect mitochondrial respiration, mobility, Ca²⁺ uptake, ROS production, and cell division. Mitochondrial fission is regulated by Dynamin-Related Protein 1 (DRP1), fusion by Optic Atrophy 1 (OPA1) at the inner and Mitofusin (MFN)1 and 2 at the outer mitochondrial membrane. Interfering with fission reduces VSMC proliferation, migration, and neointima formation. Mitochondria in dedifferentiated VSMCs are mobile, in part by MIRO1-dependent transport along microtubules. Blocking mitochondrial mobility interferes with focal adhesion turnover and VSMC migration.

DRP1 resides in the cytosol and, upon activation, associates with the outer mitochondrial membrane ³⁸. Here, it oligomerizes under GTP hydrolysis, leading to mechanical constriction of the membrane and consequently, mitochondrial fission ³⁹. In proliferating VSMCs, treatment with PDGF activated DRP1 by phosphorylation at Ser616 and induced its association with mitochondria, which promoted mitochondrial fragmentation ¹⁶. As expected, Mdivi-1 decreased VSMC proliferation and migration ⁴⁰. Mdivi-1 prevented agonist-induced mitochondrial fragmentation but also reduced PKCδ and MAP kinase activation, implying that inhibition of proliferation by DRP1 was in part mediated by cytosolic signaling. However, though considered as specific DRP1 inhibitor, Mdivi-1 potently blocks activity of the ETC complex I, and other non-specific effects cannot be excluded ³⁷. The expression of DRP1 K38A inhibited VSMC cell migration ¹⁵. This was attributed to decreased mitochondrial metabolism and ROS production, yet direct evidence for the effects of DRP1 K38A on metabolism was not provided. In summary, while mitochondrial fission and its regulation by DRP1 has been studied in some detail, the evidence that DRP1 exclusively controls neointima formation via its actions at the outer mitochondrial membrane is incomplete, in part because of the limitations associated with the frequently used inhibitor Mdivi-1. To date, direct evidence that regulation of mitochondrial fusion plays a role in neointima formation is missing.

MITOCHONDRIAL MOBILITY AND Ca²⁺ HANDLING

Calcium enters the VSMCs through ORAI1 Ca²⁺ channels that are regulated by the sensor STIM ⁴¹. Increases in cytosolic Ca²⁺ levels promote neointima formation and VSMC proliferation ^42-44. Coordinated cytosolic Ca²⁺ waves are necessary for VSMC migration ⁴⁵. Mitochondria sequester Ca²⁺ from cytoplasmic microdomains and alter the spatio-temporal pattern of Ca²⁺ gradients following Ca²⁺-influx into the cytosol or its release from the ER ^46,47. This is achieved by the mitochondrial Ca²⁺ uniporter (MCU) complex ^48,49. In VSMCs, inhibition of MCU activity by blocking the MCU activator mitochondrial Ca2+/calmodulin-dependent kinase II (CaMKII), delayed cytosolic Ca²⁺ clearance and prolonged cytosolic Ca²⁺ transients ¹¹. Unpublished data from our laboratory demonstrate that deletion of the MCU blocks PDGF-induced VSMC proliferation by inhibiting cell-cycle dependent increases in mitochondrial fusion and respiration in G1/S phase.

Mitochondrial Ca²⁺ uptake regulates its trafficking within the cell. The Rho-GTPase MIRO1 at the outer mitochondrial membrane controls mitochondrial motility along microtubules (Figure 2) ⁵⁰. In subcellular domains with increased Ca²⁺ concentrations, Ca²⁺ binds to two EF hand domains of MIRO1, leading to dissociation of mitochondria from microtubules ^51,52. In migrating VSMCs, mitochondria localize to focal adhesions, likely to provide ATP for focal adhesion turnover ^11,53. Consequently, loss of MIRO1 or MCU or inhibition of MCU by blocking mitochondrial CaMKII arrest mitochondrial mobility and VSMC migration ¹¹. Mice that express an inhibitor of CaMKII selectively in mitochondria of VSMCs displayed significant decreases in neointima formation after endothelial denudation. Independently, Chalmers demonstrated that inhibiting mitochondrial trafficking reduced VSMC proliferation ⁴⁰.

Mitochondrial Ca²⁺ concentrations may affect neointima formation by regulating metabolism and ROS production. The influx of mitochondrial Ca²⁺ activates tricarboxylic acid (TCA) cycle dehydrogenases, thereby promoting respiration and ROS production. Inhibiting MCU in proliferating VSMCs reduced mitochondrial ROS production ¹¹, glycolysis and mitochondrial respiration (Koval et al, unpublished observation). Moreover, excessive mitochondrial Ca²⁺ uptake is responsible for cytochrome C release and initiates apoptosis ⁵⁴. Together, these data indicate that manipulation of mitochondrial Ca²⁺ uptake is sufficient to affect neointima formation and associated VSMC phenotypes.

MITOCHONDRIAL METABOLISM

Cell division requires the generation of ATP and building blocks for the biosynthesis of new cells. Perez and colleagues hypothesized that increased bioenergetic capacity through glycolysis and mitochondrial respiration is required for VSMC proliferation ⁵⁵. As anticipated, PDGF treatment induced glycolytic flux and increased the mitochondrial reserve capacity of the ETC ⁵⁵. These findings suggest that the metabolic response during VSMC proliferation relies on increased aerobic glycolysis and mitochondrial respiration. Inhibitors of glycolysis blocked cell proliferation, lactate production and mitochondrial oxygen consumption, suggesting that mitochondrial respiration was fueled by pyruvate production generated from lactate. Together, the findings underscore the importance of glycolysis for mitochondrial respiration in VSMCs. Further, in proliferating human coronary smooth muscle, glycolysis and oxidative phosphorylation contributed equally to ATP production ⁵⁶.

PDGF treatment enhanced mitochondrial respiration in VSMCs in some studies ^15,55 but reduced it in others ^16,57. The conflicting results may be explained by the differences in PDGF treatment and growth media used. Salabei reported that PDGF treatment reduced the oxygen consumption rate by glucose oxidation by 20%, and increased fatty acid oxidation and attributed these findings to decreased pyruvate availability under growth conditions ¹⁶. However, no direct evidence for this was provided.

A few studies have linked changes in metabolism to neointima formation. They also revealed an unexpected role for regulation by extracellular matrix proteins ^57,58. For instance, fragments of the atypical FAT1 cadherin accumulated in the mitochondrial matrix of VSMCs, although it is unknown what triggers its processing and the translocation of its C-terminal fragments to the mitochondria. FAT1 fragments interacted with the mitochondrial proteins NDUFS3, an ETC complex I-specific subunit, and prohibitin, an important regulator of ETC complex I stability ⁵⁸. This association decreased complex I activity and altered formation of ETC supercomplexes. As a result, VSMCs lacking FAT1 grew faster, consumed more oxygen for ATP production, and contained more aspartate, a limiting metabolite in rapidly proliferating cells ⁵⁹. In addition, targeted expression of the FAT1 intracellular domain to mitochondria reduced mitochondrial respiration in FAT1^−/− VSMCs, however, it was not reported if this also inhibited proliferation. Expression of FAT1 protein within the neointima was strongly increased after injury; and mice with smooth muscle-specific deletion of FAT1 developed a significantly larger neointima. However, since FAT1^−/− mice were used, it cannot be ruled out that the effect was mediated by the extramitochondrial functions of FAT1, such as its regulation of cyclin D levels or its effect on cell-cell contacts ^60,61.

Additionally, Jia and colleagues reported that the extracellular glycoprotein cartilage oligomeric matrix protein COMP/TSP-5 localized to the mitochondrial matrix via a putative N-terminal mitochondrial localization domain ⁵⁷. There, it interacted with the scaffolding protein prohibitin-2, a regulator of mitochondrial protein synthesis and DNA stability. Overexpression of a prohibitin-2 mutant that does not bind COMP lowered baseline rates of mitochondrial respiration and reserve capacity, and increased VSMC dedifferentiation and neointima formation. However, it remains unclear whether the reported effect on neointimal size can solely be attributed to the mitochondrial actions of COMP given its presence in the cytoplasm and in the extracellular matrix, and its role in cell migration ^62,63.

While the studies of FAT1 correlated neointima formation with increased mitochondrial respiration, those of COMP/TSP-5 suggested that lower respiration rates promote dedifferentiation of VSMCs and neointima formation. These contrasting findings highlight the need for further exploration. VSMC proliferation was not significantly affected by inhibiting the ETC complex I by rotenone or upon knock down of the ETC complex I subunit NDUF3 ⁵⁸, thus challenging the concept that ETC-dependent respiration is critically important for VSMC proliferation. The seemingly contrasting findings described above may be reconciled by the emerging concept that the essential metabolic function provided by mitochondrial respiration is to generate building blocks for the biosynthesis of new cells. This is in contrast to the view that ATP synthesis via oxidative phosphorylation is the critical output of respiration during cell proliferation ^59,64. In particular, the biosynthesis of aspartate from the TCA cycle intermediate oxaloacetate has been defined as a limiting factor ^59,64. While the regulation of neointima formation by TCA cycle metabolites has not been explored, it is inhibited upon knockdown of pyruvate dehydrogenase kinase, which inactivates pyruvate dehydrogenase ⁹.

In summary, mitochondrial respiration is altered in the context of VSMC proliferation. While it is tempting to speculate that interfering with “metabolic remodeling” could provide a novel, upstream approach to inhibit VSMC dedifferentiation, a complete picture that defines the role of mitochondrial respiration in phenotypic switching of VSMCs, proliferation, and neointima formation has not yet emerged. Thus, additional research is needed to understand how neointimal formation is affected when mitochondrial metabolism is altered.

MITOCHONDRIA-DEPENDENT APOPTOSIS

Mitochondria are the nodal point in the intrinsic pathway of apoptosis. Following a variety of stress signals, the BCL2 effector protein, BAX, translocates from the cytosol to insert into the outer mitochondrial membrane, where it associates with BAK. BAX and BAK change conformation and assemble into oligomeric pore complexes in the outer mitochondrial membrane (Figure 1) ⁶⁵. Proteins from the mitochondrial intermembrane space then empty into the cytosol to activate proteases that dismantle the cell. Increased expression of BAX, or knockdown of apoptosis inhibitors such as BCL2, prevent neointima formation ^66,67. Moreover, BAX and BAK bind to the mitochondrial permeability transition pore (mPTP, Figure 1) ⁶⁸, a large non-selective channel in the inner mitochondrial membrane. Opening of the mPTP induces release of cytochrome C, resulting in mitochondrial depolarization and dissipation of the mitochondrial membrane potential. The ensuing swelling of the mitochondrial matrix leads to rupture of the outer membrane and further release of proteins from the intermembrane space.

While the composition of the mPTP is not entirely understood ⁶⁹, it is believed to contain the F1/F0 ATP synthase ⁷⁰ and cyclophilin D ^71,72, a peptidyl-prolyl isomerase in the matrix. Opening of the mPTP was modulated by hexokinase-II ⁹, which binds to the voltage-dependent anion channel (VDAC) in the outer mitochondrial membrane and reduces channel conductance ⁷³. The association of hexokinase-II with the mitochondria coincided with hyperpolarization and resistance to apoptosis ⁹. The association was reduced by dichloroacetate, a rapid-acting small molecule, thus increasing apoptosis and abrogating neointima formation in a pig coronary artery angioplasty model ⁹.

In the mitochondrial matrix, the mitochondrial chaperone HSP90 forms a physical complex with cyclophilin D ⁷⁴. After injury, HSP90 associated with the apoptosis inhibitor, Survivin, which blocks access to the ATP-dependent binding pocket of HSP90 ⁷⁵. Disrupting this interaction attenuated neointimal lesion formation in vivo and increased apoptotic events ⁷⁵. This effect was abrogated with the mPTP inhibitor Cyclosporin-A, suggesting that the decrease in neointima formation was driven by the actions of HSP90 within mitochondria. Thus, induction of apoptosis by forcing open the mPTP or by inhibiting Survivin may be a pathway to reduce neointima formation. These two studies used pharmacologic approaches (a small molecule and a peptidomimetic) to abrogate neointima formation in in vivo studies ^9,75 that may be translated to human therapy.

Lastly, the overexpression of MFN2 that drives fusion of the outer mitochondrial membrane induced apoptosis and attenuated neointimal formation in the carotid artery balloon angioplasty model. In this early study, apoptosis was attributed to decreased cytosolic Akt activation independent of mitochondrial fusion ⁷⁶. However, alternative mechanisms have been described recently; MFN2 is tethered to mitochondria in the ER ⁷⁷, which potentiates Ca²⁺ transfer between the two organelles and the sensitivity to cell-death stimuli. These data illustrate that many mitochondrial regulators have multiple functions: here, a GTPase that promotes mitochondrial fusion putative, has some putative cytosolic actions and affects mitochondrial Ca²⁺ uptake and apoptosis. The overlapping functions of mitochondrial regulators must be better understood in order to fully delineate their role in neointima formation.

CONCLUSIONS

Over the last decade, a plethora of studies have provided insight into how mitochondrial form and function is controlled and how mitochondrial functions impinge on health and disease, yet, their role in vascular physiology and neointima formation has remained relatively understudied. Frequently, studies have addressed particular aspects of mitochondrial function and their effect on proliferation, migration, or apoptosis in isolation, without integrating their interrelated functions, e.g., a change in mitochondrial shape may affect metabolism or manipulating a fission/fusion regulator could induce apoptosis. In addition, there is often little consideration for the extramitochondrial function of candidate regulators of mitochondria. Further, few studies have provided direct in vivo evidence that mitochondrial function drives neointima formation. Thus, continued investigation into these areas is necessary to address the current knowledge gaps.