Increased expression of PAI-1 in diabetes is undoubtedly multifactorial. A direct effect of insulin on the expression of PAI-1 has been suggested by a positive correlation between the concentration of insulin and PAI-1 in vivo (93,94,96,100-103,106). Triglycerides and their constituents (fatty acids) appear to contribute to the overexpression of PAI-1 in view of the fact that both insulin and triglycerides independently increase expression of PAI-1 by human hepatoma cells in vitro (105,107-109). Liver steatosis is another determinant of elevated concentrations of PAI-1, perhaps indicative of the response of both to derangements in the tumor necrosis factor signaling pathway (110). Insulin and triglycerides exert a synergistic increase in accumulation of PAI-1 in conditioned media when both are present in pathophysiological concentrations (105). Analogous results are obtained with insulin in combination with very low-density lipoprotein-triglyceride, emulsified triglycerides, or albumin-bound free fatty acids (FFAs) (nonestrified). Thus, the combination ofhyperinsulinemia and hypertriglyceridemia increases expression of PAI-1 consistent with the possibility that the combination is a determinant of the increased PAI-1 in people with diabetes in blood in vivo. Furthermore, because elevated concentrations of glucose increase expression of PAI-1 by endothelial cells and vascular smooth muscle cells in vitro (111,112), the metabolic state typical of diabetes may elevate concentrations of PAI-1 in blood-emanating release of PAI-1 from vessel wall cells.
A combination ofhyperinsulinemia, hypertriglyceridemia, and hyperglycemia increases the concentration of PAI-1 in blood in normal subjects (99). Although neither the infusion of insulin with euglycemia maintained by euglycemic clamping nor the infusion of triglycerides without induction of hyperinsulinemia in normal subjects increases the concentration of PAI-1 in blood, the induction of hyperglycemia, hypertriglyceridemia, and hyperinsulinemia by infusion of glucose plus emulsified triglycerides plus heparin (to elevate blood FFAs) does increase concentrations of PAI-1 in blood. Of note, the infusion of insulin under euglycemic clamp conditions results in a marked decrease in the concentration of blood triglycerides and FFAs. Thus, results of the infusion studies demonstrate that the combination of hyperinsulinemia, hyperglycemia, and hypertriglyceridemia is sufficient to increase expression of PAI-1 in healthy subjects. However, results in these studies do not answer the question of whether, as in the case in vitro, insulin increases expression of PAI-1 when concentrations of glucose, triglycerides, and FFAs are all maintained within normal ranges. What is clear is that a combination of hormonal (hyperinsulinemia) and metabolic (particularly hypertriglyceridemia) derangements typical of type 2 DM elevate the concentration of PAI-1 in blood. The elevations of PAI-1 may subject people with diabetes to double jeopardy because the ratio of PAI-1 activity to the concentration of PAI-1 protein increases when the latter is high. This appears to reflect a slower rate of loss of PAI-1 activity associated with higher concentrations of PAI-1 protein (113).
Adipose tissue is another potential source of the increased blood PAI-1 in subjects with type 2 DM. Studies performed on genetically obese mice demonstrated that PAI-1 mRNA expression was increased four- to fivefold in mature adipocytes (114). The injection of insulin into lean mice increased expression of PAI-1 in adipocytes, an effect seen also with 3T3-L1 adipocytes in vitro. We have found that elaboration ofPAI-1 from adipocytes is increased by transforming growth factor (TGF)- P, known to be released from activated platelets (115) secondary to increased transcription and furthermore, that caloric restriction per se lowers elevated PAI-1 in blood in obese, nondiabetic human subjects (116). Thus, the elevated concentrations of PAI-1 in blood seen in subjects with type 2 diabetes appear to be secondary to effects of hyperinsulinemia, particularly in combination with hypertriglyceridemia, and to effects of other mediators implicated in the prothrombotic state seen with diabetes on expression of PAI-1 by hepatic, arterial, and adipose tissue.
In addition to elevated PAI-1 in blood, expression of PAI-1 in vessel walls with subsequent elaboration into blood is increased by insulin (117). Pathophysiological concentrations of insulin increase the expression of PAI-1 by human arteries in vitro (117), an effect seen in both arterial segments that appear to be grossly normal and those that exhibit atherosclerotic changes. The increased PAI-1 expression is seen in arterial segments from subjects with or without insulin-resistant states. Augmented expression of PAI-1 is seen in response to insulin with VSMCs in culture (118) and with co-cultured endothelial cells and SMCs (117). Insulin increases expression of PAI-1 by vascular tissue in vivo. Local elaboration of PAI-1 follows perfusion with insulin in forearm vascular beds of healthy human subjects (119).
With the use of a co-culture system one mechanism by which insulin increases arterial wall expression of PAI-1 has been characterized (117). In vivo, insulin present in the luminal blood is known to be transported from the luminal to the abluminal surface of endothelial cells. In vitro, SMCs exposed to insulin have been shown to release a soluble factor(s) that increases endothelial cell expression of PAI-1. Thus, it appears likely that insulin in vivo alters expression of PAI-1 in arterial walls through a direct effect on VSMCs that, in turn, increases endothelial cell expression of PAI-1 in a paracrine fashion.
Therapy designed to reduce insulin resistance, the resultant hyperinsulinemia, or both have been shown to reduce PAI-1 in blood as well. Thus, treatment of women with the polycystic ovarian syndrome with metformin or troglitazone decreased concentrations in blood of insulin and of PAI-1 (100,103). Changes in the concentrations of PAI-1 in blood correlated significantly with those of insulin (100). The concordance supports the view that insulin contributes to the increased PAI-1 expression seen in vivo.
Human subjects who participate in relatively large amounts of leisure time physical activity have low levels of PAI-1 activity in blood (120). After adjustment for variables indicative of syndromes of insulin resistance such as high body mass index and waist to hip ratio in addition to advanced age and elevated concentrations of triglycerides, the association of PAI-1 activity with physical activity was no longer significant. This observation, particularly in combination with the results seen after therapy with troglitazone and metformin in women with the polycystic ovarian syndrome, demonstrates that interventions designed to attenuate insulin resistance will lower concentrations of PAI-1 in blood and increase fibrinolytic system capacity.
The exposure of human hepatoma cells to gemfibrozil decreases basal and insulin-stimulated secretion of PAI-1 (121). This inhibitory effect has been observed in vitro but not in vivo (122,123) despite reductions in vivo in the concentration of triglycerides in blood by 50% to 60%. No changes in insulin sensitivity or concentrations of insulin in the blood were seen after treatment of patients with gemfibrozil. Thus, unlike therapy with agents that reduce insulin resistance and lower concentrations of insulin, therapy with gemfibrozil that reduces triglycerides without affecting concentrations of insulin does not lower PAI-1 in vivo. These observations support the likelihood that insulin is the critical determinant of altered expression of PAI-1 in subjects with insulin resistance such as those with type 2 DM. As judged from results in studies in which human hepatoma cells were exposed to insulin and triglycerides in vitro, modest elevations in the concentrations of triglycerides and FFAs in the setting of hyperinsulinemia may be sufficient to augment expression of PAI-1. Thus, although the concentration of triglycerides in patients treated with gemfibrozil was decreased by 50%, the prevailing concentration of triglycerides may have been sufficient to lead to persistent elevation of PAI-1 in blood in the setting of hyperinsulinemia. Recent results in studies with several statins including atorvastatin fail to show concordant changes in PAI-1 in blood, consistent with this possibility (124).
Was this article helpful?