PAI-1: cardiac friend or foe?

In this issue of Blood, Xu and colleagues provide evidence that suggests PAI-1 is cardioprotective as demonstrated by the observation that spontaneous cardiac fibrosis occurs in aging PAI-1–deficient (PAI-1^−/−) mice, preceded by early changes in cardiac vascular barrier integrity, enhanced inflammation, dysregulated extracellular matrix remodeling, and, eventually, pervasive fibrosis and compromised cardiac function.¹ 

Plasminogen activator inhibitor-1 (PAI-1), a serine protease inhibitor (SERPIN), is the major physiological inhibitor of plasminogen activation through inactivation of the plasminogen activators, urokinase (uPA) and tissue plasminogen activator (tPA). (Plasminogen activation results in generation of plasmin, the major fibrinolytic enzyme.) In addition, functions of PAI-1 unrelated to regulation of fibrinolysis continue to be elucidated. The interaction of PAI-1 with the matrix protein vitronectin and regulation of cell-surface signaling through uPA/uPA receptor internalization in complex with low-density lipoprotein receptor–related protein-1 (LRP-1) have profound effects on a variety of cellular processes, including cell adhesion, migration, and proliferation. These activities are dependent on distinct domains within the PAI-1 molecule. The conservation of these functional domains within PAI-1 has been demonstrated in vitro with human and murine recombinant PAI-1 protein homologs.

Whether PAI-1 is a mediator or inhibitor of cardiac fibrosis is still controversial. Other studies have demonstrated that PAI-1 contributes to cardiac fibrosis potentially through blocking plasmin activation of pro–matrix metalloproteases. However, these studies were performed in cardiac injury challenged young mice, that is, after coronary occlusion/myocardial infarction.^2-4  Interestingly, in one of these studies, myocardial hemorrhage and inflammation were enhanced in PAI-1^−/− mice, in association with greater infarct size compared with wild-type mice, but fibrosis was attenuated.³  In the current study by Xu et al, investigating spontaneous cardiac changes with aging, increases in spontaneous myocardial hemorrhage and inflammation were observed in PAI-1^−/− mice. However, cardiac fibrosis was enhanced in PAI-1^−/− mice relative to age-matched wild-type mice. Increased fibrosis in 1-year-old PAI-1^−/− mouse hearts, in the absence of injury, has also been observed.⁵  The Xu study characterizes the temporal development of spontaneous cardiac fibrosis in PAI-1^−/− mice. Evidence for spontaneous hemorrhage and up-regulation of proinflammatory cytokines in cardiac tissue occurred in young mice (at 12 weeks), preceding the development of fibrosis observed in aged mice (at 36-48 weeks). Taken together, these results may suggest differential effects of PAI-1 on myocardial fibrosis leading to a more profibrotic effect in response to cardiac injury and leading to a more antifibrotic effect in response to aging. These contrasting effects may, in turn, be mediated by differences in the relative contribution of protease inhibitor-dependent (eg, hemorrhage and breakdown of the cell-extracellular matrix interface) and independent (cell signaling) regulatory functions of PAI-1. Also of interest is whether the age-dependent increases in plasma PAI-1 concentrations observed in mice and humans may play an important role in the maintenance of normal cardiac architecture and cardiac functional homeostasis with aging.

The development of cardiac fibrosis has been observed in other mice with altered expression of coagulation and fibrinolysis proteins. Mice expressing low levels of tissue factor (TF) or its ligand, factor VII (FVII), and mice overexpressing uPA also develop cardiac fibrosis,^5-7  whereas the absence of uPA in mice prevents cardiac fibrosis in response to myocardial infarction.^8,9  In low-TF mice, altered regulation of uPA expression was also observed¹⁰  and uPA levels were increased in PAI-1^−/− mice in the Xu study, consistent with increased matrix remodeling.¹  Therefore, uPA may be a driving force regulating the cardiac fibrosis phenotype in these models. PAI-1 may regulate fibrosis by inhibiting proteolytic damage at the cell-extracellular matrix interface caused by uPA proteolytic activation of plasmin and/or by altering uPA/uPA receptor signaling. In addition, the absence of PAI-1 may lead to increased plasmin activation of the profibrotic cytokine, TGF-β, concomitantly with increased TGF-β synthesis (as shown in the Xu et al study). In vivo studies that selectively alter functional properties of PAI-1 and uPA will further contribute to an understanding of how these proteins regulate events leading to cardiac fibrosis.