No Sweetie Pie: Newly Uncovered Role for PAI-1 in Inflammatory Responses to Ischemia/Reperfusion

In 1983 Loskutoff et al. reported that bovine aortic endothelial cells (ECs) secrete a highly stable inhibitor of fibrinolysis, which eventually was purified, cloned, and termed plasminogen activator inhibitor-1 (PAI-1).^{1, 2} Early scientific investigations of PAI-1, which took place during the heyday of thrombolytic therapy for acute myocardial infarction, centered on its role in stabilizing intravascular fibrin, regulating hemostasis, and predisposing to thrombosis. In 1996 Stefansson et al. reported that PAI-1 regulates smooth muscle cell (SMC) migration by competitively blocking SMC binding interactions with extracellular matrix vitronectin.³ This study and others demonstrated that the functional role of PAI-1 is more complex than simply regulating plasmin formation and fibrinolysis. Over the ensuing two decades, PAI-1 was shown to exert a wide variety of effects that are independent of its anti-fibrinolytic actions. These include regulation of atherosclerosis and intimal hyperplasia,^4–13 myoendothelial junctions,¹⁴ insulin signaling,^{15, 16} obesity,^{17, 18} Alzheimer’s disease,¹⁹ multiple sclerosis,²⁰ cellular senescence,²¹ and cancer.²

In this issue of ATVB, Praetner et al. provide evidence supporting another non-fibrinolytic function of PAI-1 by demonstrating its role in directing neutrophil trafficking during ischemia-reperfusion (I/R) injury.²² This pro-inflammatory effect is accomplished by rapid immobilization of PAI-1 on the surface of microvascular ECs and transmigrating leukocytes early after reperfusion. The immobilized protease inhibitor provokes affinity changes in β2 integrins expressed on rolling neutrophils that allow them to firmly adhere to ECs, transmigrate into tissues, and increase microvascular permeability. The β2 integrin affinity changes are instigated by PAI-1 via a pathway that involves low-density lipoprotein receptor-related protein-1 (LRP1) and mitogen activated protein kinase (MAPK) family members. To characterize the functional relevance of PAI-1 originating from leukocytes vs. other sources, cell transfer studies were conducted. The numbers of adherent and emigrated WT donor leukocytes were reduced in PAI-1-deficient mice, while adherent and transmigrated PAI-1-deficient donor leukocytes also were reduced in WT recipient animals, suggesting that PAI-1 derived both from leukocytes and non-leukocyte sources promoted post-ischemic neutrophil adherence. The work of Praetner et al. is important because increased leukocyte-EC interactions and microvascular permeability, which are amongst the earliest detectable responses to tissue I/R, promote neutrophil transmigration and tissue injury secondary to release of reactive oxygen species (ROS) and hydrolytic enzymes.^{23, 24} Interestingly, genetic deletion of PAI-1 potently down-regulated the inflammatory response to I/R without significantly altering light/dye-induced thrombosis, suggesting that drug targeting of PAI-1 may prove useful in inhibiting I/R tissue injury without disrupting microvascular hemostasis.

While this informative report by Praetner et al. provides strong evidence supporting a role for PAI-1 in I/R injury, it also raises several questions. For example, what factors mediate PAI-1 release and immobilization on ECs after I/R? It is tempting to speculate that I/R-induced production of tumor necrosis factor-α (TNF-α) and platelet-activating factor (PAF) are important precipitating events because these cytokines not only provoke ECs and other cells to release PAI-1,^{25, 26} but also induce post-ischemic neutrophil sequestration.^{23, 24} Mast cell activation may also contribute, since these sentinel cells degranulate after I/R and release cytokines that promote neutrophil infiltration,^{23, 24} as well as exosomes that stimulate PAI-1 secretion by ECs in a thrombin-dependent manner.²⁷ It remains unclear what effect PAI-1 has on pericytes, which surround post-capillary venules, produce PAI-1, and play a critical role in guiding neutrophils that have breached the endothelial barrier into tissues.²⁸ Given that multiple cell types produce PAI-1, it would be interesting to repeat some of the experiments of Praetner et al. in cell-type-specific PAI-1 knockout mice to study the role of PAI-1 derived from ECs, neutrophils, platelets, pericytes, adipocytes, or other cells in I/R injury.

Although the studies performed by Praetner et al. support the hypothesis that endothelium-derived PAI-1 promotes the post-ischemic inflammatory response, it remains unclear precisely how PAI-1 sequestered on ECs during I/R produces this effect. While the authors showed that PAI-1 doesn’t alter adhesion molecule expression level on ECs, it is possible that PAI-1 could trigger spatial or affinity changes in EC adhesion molecules, such as clustering of intercellular adhesion molecule-1 (ICAM-1). Since PAI-1 expression on the surface of ECs and neutrophils increased dramatically after I/R, it is assumed that the cell-bound serpin was responsible for promoting leukocyte adhesion and emigration. This is consistent with the concept that PAI-1 binds to surface glycosaminoglycans, which enhance its biologic activity²⁹ and may position PAI-1 near LRP1 on cells. However, one might wonder if PAI-1 initially immobilized on ECs after I/R is released into the flowing blood during reperfusion and travels to distant organs to produce remote injury. This is an important question, as neutrophil-dependent multi-organ failure can arise secondary to severe myocardial ischemia or if the volume of tissue affected by I/R is large, as can occur following occlusion of limb or major intestinal arteries.

Generation of ROS also plays a major role in post-ischemic inflammation and tissue injury,^{23, 24} but is unclear whether PAI-1 modulates oxidative stress in I/R. While Praetner et al. establish that LRP1-mediated MAPK activation produces the inflammatory effects of PAI-1, how these signaling steps couple to other events that mediate β2-integrin affinity changes, and how PAI-1 regulates microvascular barrier function, remain to be explored. The latter question is especially intriguing in light of reports that PAI-1 enhances endothelial barrier function (i.e. reduces permeability) in in vitro systems devoid of neutrophils,^{30, 31} which was also demonstrated by Praetner et al. Finally, Praetner et al. showed that post-ischemic neutrophil transmigration is markedly reduced in PAI-1-deficient mice and that exogenous administration of latent (inactive) PAI-1 stimulates neutrophil emigration. Taken together, these observations suggest that the ability of PAI-1 to instigate post-ischemic neutrophil sequestration occurs by mechanisms independent of its anti-protease activity. However, other work suggests that PAI-1 inhibits plasmin-induced shedding of interleukin (IL)-8 bound to syndecan-1 ectodomains on ECs, resulting in disinhibition of neutrophil migration.³² This observation suggests that PAI-1 may act to stabilize chemoattractant tethering to the EC surface and promote neutrophil transmigration by inhibiting plasmin-induced cleavage of IL-8/syndecan-1 complexes – i.e. via its anti-protease activity. Whether this mechanism also contributes to PAI-1-dependent neutrophil diapedesis in vivo after I/R has not been tested.

Figure.

PAI-1 promotes neutrophil diapedesis and tissue injury after ischemia-reperfusion. PAI-1 accumulates on microvascular endothelial cells after ischemia-reperfusion and is encountered by rolling neutrophils, which triggers affinity changes in β2 integrins, particularly Mac-1/CD11b, by a process that requires LDL receptor-related protein-1 (LRP1) and mitogen-activated protein kinase (MAPK) family members. Binding interactions between activated β2 integrin and cell adhesion molecules expressed on endothelial cells, including intercellular adhesion molecule-1 (ICAM-1), lead to firm neutrophil adhesion. PAI-1 also triggers an increase in endothelial permeability, which is accompanied by neutrophil transmigration into subendothelial tissue. The post-ischemic, transmigrated neutrophils, which bear PAI-1, release reactive oxygen species (ROS) and hydrolytic enzymes that induce tissue injury.