Attenuation by Gemfibrozil of Expression of Plasminogen Activator Inhibitor Type 1 Induced by Insulin and its Precursors
Background Insulin and its precursors found in increased plasma concentrations in non–insulin-dependent diabetes mellitus (NIDDM) augment synthesis of plasminogen activator inhibitor type 1 (PAI-1) in Hep G2 cells in vitro and in rabbit liver in vivo. Reduced endogenous fibrinolysis secondary to increased PAI-1 activity may exacerbate atherogenesis. Recently, the reduction of the coronary heart disease incidence in the Helsinki Heart Study has implicated favorable modulation of endogenous fibrinolysis by gemfibrozil.
Methods and Results In Hep G2 cells, 500 (700) μmol/L gemfibrozil decreased basal secretion of PAI-1 by 26% (43%) (P=.012 and P=.021, respectively) and attenuated insulin-induced (10 nmol/L) augmentation of PAI-1 in conditioned media by 61% (109%) (P=.010) within 24 hours. Inhibition was dependent on the duration of exposure (0 to 48 hours) and on the concentration of gemfibrozil (0 to 700 μmol/L) but not on the concentration of insulin (0.1 to 100 nmol/L). Gemfibrozil attenuated the augmentation of PAI-1 secretion induced by proinsulin (>100%), by des(31,32)proinsulin (75%), and by des(64,65)proinsulin (77%) as well (10 nmol/L each). The specificity of these effects was confirmed by the unaltered levels of newly synthesized protein (metabolic labeling) and of total protein (both in conditioned media and cell lysates). Secretion of fibrinogen by Hep G2 cells was not affected by gemfibrozil. Changes in PAI-1 protein levels reflected modulation of PAI-1 gene expression as manifested by changes in levels of 3.2-kb PAI-1 mRNA (Northern blots).
Conclusions Gemfibrozil attenuated the augmentation of synthesis and secretion of PAI-1 induced by insulin and its precursors directly and specifically. Accordingly, gemfibrozil may exert favorable therapeutic effects normalizing impaired fibrinolysis in patients with hyperinsulinemia such as NIDDM.
In the Helsinki Heart Study, a primary prevention trial involving more than 4000 asymptomatic middle-aged men with primary dyslipidemia (non-HDL cholesterol ≥200 mg/dL) followed for 5 years, gemfibrozil reduced the incidence of CHD (myocardial infarction and cardiac death as end points) significantly, by 34%, in comparison with results in the placebo control group. The decline in incidence with gemfibrozil became evident in the second year.1 It has been thought that the protective effect was probably mediated via an increase in HDL cholesterol and a decrease in LDL cholesterol.2 However, a recent reanalysis showed that gemfibrozil reduced the incidence of CHD mainly in overweight subjects with additional risk factors known to contribute to or predispose to insulin resistance.3 The observation of fewer fatal cerebrovascular accidents attributable to ischemia in the gemfibrozil-treated group and of fewer but more frequently fatal intracranial hemorrhages than in the placebo group with accentuation of this trend at the 8.5-year follow-up supports the hypothesis that benefit was conferred in part through attenuation of impairment of fibrinolysis.4 Subjects without primary dyslipidemia but with diminished endogenous fibrinolysis, such as those with NIDDM and other hyperinsulinemic conditions, may benefit from treatment with gemfibrozil as well.
In most studies of restenosis associated with coronary angioplasty, diabetes and unstable angina have emerged as the only variables consistently associated with restenosis.5 In diabetic patients, coronary angioplasty is associated with a high rate of infarction within 5 years after the procedure.6 Patients with multivessel CHD and diabetes mellitus who were treated with insulin or oral hypoglycemic agents were found to have a significantly higher 5-year mortality after bypass surgery and angioplasty than nondiabetic subjects. In the diabetic patients, surgery was associated with a much lower (though still elevated) 5-year mortality than that seen with angioplasty.7 Thus, reduced endogenous fibrinolysis secondary to increased plasma activity of PAI-1 typically observed in NIDDM may play an important pathogenetic role in accelerating vasculopathy in such patients,8 9 10 particularly after angioplasty.6
Gemfibrozil reduces plasma activity of PAI-1 (and concentrations of triglycerides) in patients with primary hypertriglyceridemia11 12 13 and in survivors of myocardial infarctions with hypertriglyceridemia.14 If gemfibrozil reduces plasma activity of PAI-1 directly, it could be beneficial in patients with increased plasma PAI-1 even with normal serum triglycerides. Thus, it would be an attractive therapeutic agent for reduction of plasma PAI-1 despite increases induced by insulin and its precursors proinsulin, des(31,32)proinsulin, and des(64,65)insulin.10 15 16 17 The present study was performed to determine whether gemfibrozil induces direct (independent of changes in lipid concentrations) effects on the augmentation of PAI-1 synthesis induced by insulin and its precursors.
Studies were performed with Hep G2 cells, highly differentiated human hepatoma cells, acquired from the American Type Culture Collection (HB 8065). Hep G2 cells were used because they simulate hepatocyte PAI-1 with respect to expression of PAI-1 in vivo in response to agonists and antagonists and because hepatocytes are one of the primary sources of circulating PAI-1. The cells were grown to confluence (final concentration, 7.9×105±1.2×105 cells per well of a 24-well plate) in minimum essential medium with Earle's salts and l-glutamine (Life Technologies) supplemented with 10% FCS, 30 U/mL penicillin, and 30 μg/mL streptomycin (Life Technologies). Monolayers of confluent cells were serum-starved in Dulbecco's modified Eagle's medium with Ham's nutrient mixture F-12, l-glutamine, and HEPES buffer (DMEM, Life Technologies) for 16 to 24 hours, an interval we had found to be sufficient for a decline of PAI-1 synthesis to basal levels.
After serum starvation, cells were exposed to fresh media constituted with selected concentrations of human recombinant insulin, human recombinant proinsulin (Sigma Chemie), human des(31,32)proinsulin, human des(64,65)proinsulin (Eli Lilly), or human synthetic C-peptide (Sigma) dissolved in 0.9% sodium chloride (Braun Melsungen) with 0.5% BSA (fraction V, low endotoxin, cell culture tested, Sigma). Immediately before use, gemfibrozil (Sigma) was first dissolved in DMSO (Roth) and then diluted further in DMEM. The final concentration of DMSO to which the Hep G2 cells were exposed was ≤0.1% (vol/vol), a concentration that affects neither cell morphology nor basal non–insulin-stimulated elaboration of PAI-1. Control experiments were performed with vehicle alone.
PAI-1 Protein and Activity
After selected time intervals, conditioned media from Hep G2 cells were supplemented with Tween 80 (final concentration, 0.01%) and stored at −20°C until assay. The concentration of PAI-1 protein was measured by ELISA as described previously.18 Active, latent, and TPA-complexed forms of PAI-1 were detected with equal sensitivity.
PAI-1 activity was assayed by a solid-phase enzyme immunoassay with TPA that had been immobilized onto plastic microtiter plates with an anti-TPA monoclonal antibody. The active site of TPA remained exposed and could therefore bind free, active PAI-1 contained in the conditioned medium samples. Latent PAI-1 and PAI-1 that had already complexed to TPA in conditioned media could not bind to the immobilized TPA and were washed off the microtiter plate. The bound active PAI-1 was detected by conventional techniques with an enzyme-labeled monoclonal antibody (Technoclone Actibind PAI-1, Immuno). The intra-assay and interassay variations were <5% and <10%, respectively.
The amount of total protein in conditioned media and in cell lysates was measured with a colorimetric microassay procedure after detergent solubilization (DC Protein Assay, BioRad).19
The rate of synthesis of total protein was quantified by metabolic labeling of newly synthesized total protein. After serum starvation of the Hep G2 cells, DMEM was replaced by DMEM devoid of unlabeled methionine and supplemented with [35S]methionine (Tran35S-Label, ICN) at a final concentration of 50 μCi/mL. After 24 hours, aliquots of conditioned media were diluted in PBS with 0.01% Tween 80 and 1% BSA. Proteins were immunoprecipitated with an excess of 10% trichloroacetic acid and incubated at room temperature for 15 minutes. After centrifugation at 13 600g at 4°C for 10 minutes, the pellets were dissolved in 300 μL 0.3 mol/L sodium hydroxide and neutralized with 90 μL 1 mol/L hydrochloric acid. Radioactivity was assayed by scintillation spectrometry with a liquid scintillation counter.
Probes used in Northern blotting experiments were produced as described previously.18 In brief, Esherichia coli were transfected with the vector M13 mp18 carrying ampicillin resistance and a 54-base polylinker including the human PAI-1 cDNA at the recognition site for EcoRI. The plasmids were purified and digested with the restriction enzymes EcoRI and Sal I (Boehringer Mannheim), yielding a 0.9-kb probe for the detection of PAI-1 mRNA. Before hybridization, the cDNA probes were labeled with digoxigenin-dUTP (nonradioactive) by the random oligonucleotide primer method (DIG DNA Labeling Kit, Boehringer Mannheim).
The preparation and blotting of cellular RNA were performed as described previously.18 RNA was fixed to nylon membranes by exposure to UV light. The integrity, equal loading, and equal transfer of RNA were verified by ethidium bromide staining of ribosomal RNA. The prehybridization solution was supplemented with 200 μg/mL denatured calf thymus DNA (Sigma). The hybridization solution was prepared with labeled and denatured PAI-1 cDNA derived from 200 ng native cDNA after random-primed DNA labeling for 16 to 20 hours. After washing, the hybridized probes were detected with anti-digoxigenin Fab fragments conjugated to alkaline phosphatase and visualized with the chemiluminescence substrate CSPD (DIG Luminescent Detection Kit, Boehringer Mannheim). Dephosphorylation of CSPD by alkaline phosphatase led to light emission, which was recorded on x-ray films. The latter were analyzed by densitometric scanning (Scanner JX-325, Sharp, and ImageMaster 1-D Software, Pharmacia Biotech).
Determination of PAI-1 mRNA Half-life
After incubation of serum-starved Hep G2 cells with insulin in the presence or absence of gemfibrozil for 24 hours, the cells were exposed to actinomycin D (Sigma, 5 μg/mL final concentration). At specific times thereafter (0, 2, 4, and 24 hours), cellular RNA was prepared and PAI-1 mRNA was quantified as described above.
For flow cytometry experiments, Hep G2 cells were trypsinized and resuspended in Tyrode's solution containing 0.15 mol/L sodium chloride, 2.5 mmol/L potassium chloride, 1.2 mmol/L sodium hydrogen carbonate, 2 mmol/L calcium chloride, 2 mmol/L magnesium chloride, 0.1% (wt/vol) glucose, and 0.1% (wt/vol) BSA. After incubation of 300 000 cells (in 50 μL) at 4°C for 30 minutes with 10 μg/mL monoclonal mouse anti–human insulin receptor antibody (Biodesign), cells were washed and resuspended in Tyrode's solution. Cells were incubated at 4°C for 30 minutes with 10 μg/mL fluorescein-conjugated goat anti–mouse IgG antibody (Dianova), washed, fixed (CellFix, Becton Dickinson), and assayed for fluorescence with a FACScan (Becton Dickinson) equipped with a 15-mW argon laser. Binding specificity was determined with monoclonal mouse anti–human IgG antibody (Fab specific, clone GG-6, Sigma).
Cell lysates were prepared by the cells' being washed twice in culture with PBS, incubated at −20°C for 20 minutes, and exposed to lysis buffer. The 5× stock of the lysis buffer (Promega) consisted of 125 mmol/L Tris, pH 7.8, 10 mmol/L CDTA, 10 mmol/L DTT, 50% glycerol, and 5% Triton X-100.
Fibrinogen was measured by ELISA (Fibrinostika Intact Fibrinogen, Organon Teknika) with a solid-phase, murine monoclonal anti-fibrinogen antibody specific for the carboxyl-terminus of the Aα-chain of fibrinogen and a horseradish peroxidase–labeled, murine monoclonal anti-fibrinogen antibody specific for the amino-terminus of the Aα-chain.
Data are presented as mean±SEM. The significance of differences between groups was assessed by one-way ANOVA and with Student's t test as appropriate. A value of P<.05 was considered to be significant.
Effects of Gemfibrozil on Insulin-Induced PAI-1 Elaboration
Hep G2 cells were exposed to media containing 0, 500, or 700 μmol/L gemfibrozil for 24 hours in the absence or presence of 10 nmol/L insulin. Fig 1⇓ shows the concentrations of PAI-1 protein in conditioned media of these cells. In the absence of gemfibrozil, 10 nmol/L insulin induced a 1.9-fold increase of PAI-1 elaboration, confirming results obtained previously.18 Gemfibrozil at 500 and 700 μmol/L inhibited basal PAI-1 secretion by 26% and 43%, respectively (P=.012 and .021, respectively); it inhibited the increase of PAI-1 secretion induced by insulin by 61% (500 μmol/L gemfibrozil) and markedly so with 700 μmol/L gemfibrozil (109%, P=.010 compared with values without gemfibrozil). These effects occurred without any detectable changes in cell number or cell morphology.
The inhibitory effect of gemfibrozil on insulin-induced secretion of PAI-1 was more pronounced when the cells were pretreated with gemfibrozil for 2 to 6 hours before exposure to insulin (data not shown). In another set of experiments, cells were incubated with 700 μmol/L gemfibrozil and 10 nmol/L insulin. After 24 hours, the conditioned medium was removed and the cells were washed and incubated with fresh serum-free media without gemfibrozil or insulin. Table 1⇓ shows the amounts of PAI-1 protein secreted within 4-hour intervals thereafter. Results indicate that the inhibiting effect of gemfibrozil disappeared 4 to 8 hours after its removal, whereas the stimulating effect of insulin lasted for ≥12 hours after its removal.
To define concentration-response relationships, Hep G2 cells were exposed to gemfibrozil at concentrations of 0, 100, 200, 300, 400, 500, 600, or 700 μmol/L for 24 hours. Fig 2⇓ illustrates the concentration-dependent decrease of PAI-1 protein in conditioned media seen with gemfibrozil. Both basal and insulin-induced elaborations of PAI-1 were affected by gemfibrozil, with the inhibitory effect of gemfibrozil being more pronounced in the presence of 10 nmol/L insulin (slope of −0.497 with insulin compared with a slope of −0.236 without insulin, r=−.955 and r=−.966, respectively, P=.002 for difference between both slopes).
Cells were incubated with insulin at a concentration of 0.1, 0.3, 1, 3, 10, 30, or 100 nmol/L for 24 hours. Fig 3⇓ demonstrates the concentration-dependent increase of PAI-1 protein with insulin. In accordance with the results above, the increase (slope of 29 without gemfibrozil, r=.984) was suppressed by the presence of 500 μmol/L gemfibrozil (slope of 23, r=.989) and in an even more pronounced manner by the presence of 700 μmol/L gemfibrozil (slope of 16, r=.991, P<.05 versus slope of 29 and slope of 23). In addition, the inhibitory effect of gemfibrozil was not affected by the concentration of insulin to which the cells were exposed. Thus, the inhibitory effect of gemfibrozil on insulin-induced elaboration of PAI-1 was dependent on the concentration of gemfibrozil but not on the concentration of insulin.
The secretion of PAI-1 into the conditioned media of Hep G2 cells as a function of the duration of exposure of the cells to insulin and gemfibrozil was determined. Cells were incubated with 0 or 700 μmol/L gemfibrozil and were exposed to 10 nmol/L insulin or vehicle alone as a control. Samples of conditioned media were removed every 12 hours for up to 48 hours. In the absence of any gemfibrozil, the concentration of PAI-1 protein in conditioned media increased linearly with time (Fig 4⇓). In the presence of 700 μmol/L gemfibrozil, the concentration of PAI-1 exhibited an increase for up to 24 hours. Subsequently, however, there were no further increases in the concentration of PAI-1 protein, which suggested that PAI-1 expression had been altered by gemfibrozil. Throughout the interval of exposure, neither cell number per well nor cell morphology was affected.
Effects of Gemfibrozil on Overall Protein Synthesis
Table 2⇓ shows the results of experiments evaluating the elaboration of both total and active PAI-1 protein in conditioned media 24 hours after exposure to gemfibrozil and insulin. The very small amount of active PAI-1 detected in comparison with that of total PAI-1 protein is consistent with rapid conversion of active PAI-1 to its latent form in conditioned media of cultured cells in the absence of exogenous vitronectin.20
Neither total protein synthesis measured by metabolic labeling nor total protein secretion measured colorimetrically was affected significantly by gemfibrozil (Fig 1⇑), which suggests that the inhibition of insulin-induced elaboration of PAI-1 by gemfibrozil was specific. Total protein (and PAI-1 protein) measured in cell lysates was not altered significantly by 500 or 700 μmol/L gemfibrozil.
Effects of Gemfibrozil on Elaboration of PAI-1 Protein Induced by Proinsulin or des(31,32)Proinsulin and des(64,65)Proinsulin
Because proinsulin and proinsulin-like molecules account for up to 60% of total immunoreactive insulin in plasma of patients with NIDDM,10 the potential modification of their effects by gemfibrozil was examined. Confluent and serum-starved Hep G2 cells were incubated with 0 or 700 μmol/L gemfibrozil and exposed to 0, 1, 10, or 100 nmol/L concentrations of proinsulin, des(31,32)proinsulin, des(64,65)proinsulin, or C-peptide. Fig 5⇓ depicts the results and confirms the stimulatory action of the insulin precursors on PAI-1 concentration in conditioned media. As expected, 10 nmol/L C-peptide induced no significant change in the elaboration of PAI-1 (177±12 in the absence and 87±7 ng/mL in the presence of 700 μmol/L gemfibrozil) compared with control conditions (164±7 and 82±3 ng/mL, respectively). Gemfibrozil 700 μmol/L inhibited the increase in PAI-1 elaboration induced by proinsulin completely, that with des(31,32)proinsulin by 75%, and that with des(64,65)proinsulin by 77% (with 10 nmol/L concentrations of insulin precursors in each case).
Effects of Gemfibrozil on Steady-state Levels of PAI-1 mRNA
PAI-1 mRNA was assayed by Northern blotting 24 hours after exposure of the cells in culture to 0 or 700 μmol/L gemfibrozil and to 0 or 10 nmol/L insulin. Two forms of PAI-1 mRNA were detected: a 3.2-kb and a 2.2-kb form (Fig 6⇓). Insulin increased the concentration of PAI-1 mRNA 1.8-fold, corresponding to the observed 1.9-fold increase in the elaboration of PAI-1 protein. Gemfibrozil did not affect total expression of the PAI-1 gene significantly (total of 10 Northern blot experiments performed in triplicate). However, gemfibrozil reduced the amount of the 3.2-kb species of PAI-1 mRNA in proportion to total PAI-1 mRNA from 0.41±0.03 to 0.28±0.03 (P=.008). These results are in accordance with the previous observation that the rate of transcription of the PAI-1 gene is not affected by insulin and that insulin prolongs the half-life of the unstable 3.2-kb form without affecting the half-life of the relatively more stable 2.2-kb form.21 Thus, gemfibrozil seems to reduce the steady-state level of the 3.2-kb form of PAI-1 mRNA, consistent with the magnitude of reduction in elaboration of PAI-1 protein. The preponderance of the 2.2-kb species may explain the modulation of the response to insulin by gemfibrozil despite a lack of decrease in total PAI-1 mRNA.
To determine whether the decrease in 3.2-kb PAI-1 mRNA induced by gemfibrozil was attributable to increased degradation, cells were exposed to 10 nmol/L insulin in the presence or in the absence of 700 mmol/L gemfibrozil. After 24 hours, the transcription inhibitor actinomycin D was added (time, 0 hours) and PAI-1 mRNA was quantified at 0, 2, 4, and 24 hours. The half-life of the 3.2-kb PAI-1 mRNA species was 24 hours with insulin alone. With gemfibrozil present, the half-life was reduced to 6 hours (two experiments performed in triplicate, r=.78). The relatively long half-lives are in agreement with the several-fold prolongation of 3.2-kb PAI-1 mRNA half-life after incubation of Hep G2 cells with insulin for 16 hours compared with control conditions21 and with the observation that the stimulating effect of insulin lasted for ≥12 hours after its removal (Table 1⇑). Degradation of the 2.2-kb PAI-1 mRNA species was negligible over the intervals observed. In summary, the decrease of the 3.2-kb PAI-1 mRNA levels induced by gemfibrozil appeared to reflect increased degradation of preformed 3.2-kb PAI-1 mRNA, possibly by activation of a specific ribonuclease or relief of insulin-induced ribonuclease inhibition.
Quantification of Insulin Receptors
The number of insulin receptors on Hep G2 cells is thought to be downregulated in the presence of insulin.22 Accordingly, the effect of gemfibrozil on the number of insulin receptors was determined. Flow cytometry experiments were performed after the cells had been incubated with 0, 10, or 100 nmol/L insulin for 24 hours. At 10 nmol/L, insulin reduced the number of insulin receptors by 42% and at 100 nmol/L by 66%, confirming observations by others who assayed binding with radioactively labeled insulin.22 In the presence of 700 μmol/L gemfibrozil, a similar but less marked downregulation was observed (17% at 10 nmol/L and 21% at 100 nmol/L insulin). By contrast, gemfibrozil at concentrations up to 900 μmol/L in the presence or absence of 10 nmol/L insulin did not influence the number of insulin receptors significantly, consistent with the lack of an effect of gemfibrozil on total protein synthesis and suggesting that the action of gemfibrozil on insulin-induced augmentation of PAI-1 is not dependent on transmembrane signal transduction.
Effects of Gemfibrozil on Elaboration of Fibrinogen
Fibric acids can reduce plasma fibrinogen,23 but gemfibrozil has been reported to exert diverse effects. Because fibrinogen is synthesized by hepatocytes, assay of fibrinogen in conditioned media of Hep G2 cells after exposure to gemfibrozil for 24 hours was of interest. In the absence of gemfibrozil, cells under control conditions elaborated 786±37 μg/mL fibrinogen; with 10 nmol/L insulin, they yielded 662±43 μg/mL in conditioned media. The corresponding results in the presence of 700 μmol/L gemfibrozil were 678±52 and 582±58 μg/mL, respectively (from eight experiments performed in triplicate). Statistical analysis revealed no significant effect of insulin or gemfibrozil on the elaboration of fibrinogen, consistent with a lack of influence of gemfibrozil on total protein synthesis and a lack of effect on blood fibrinogen in vivo.
Gemfibrozil directly attenuated the augmentation of PAI-1 synthesis induced by insulin and its precursors proinsulin, des(31,32)proinsulin, and des(64,65)proinsulin. Although several PAI-1 synthesis inhibitors are under development,24 poor bioavailability and possible severe side effects may limit their use in vivo. Nevertheless, inhibition of PAI-1 synthesis is a promising strategy for diminution of increased plasma activity of PAI-1. Gemfibrozil has been studied thoroughly in patients. Its absorption is rapid and virtually complete. Gastrointestinal symptoms and rash are the only side effects seen frequently with its use.25 Mean peak plasma concentrations of gemfibrozil are in the range of 100 μmol/L.25 However, transient concentrations in portal venous blood of compounds given orally markedly exceed those in peripheral blood. Thus, concentrations of gemfibrozil in Hep G2 cells exposed to the agent in our study are probably consistent with concentrations to which hepatocytes are exposed in vivo. The specificity of action of gemfibrozil documented in our study by the lack of an effect on total protein synthesis in cell culture is in accordance with the absence of serious side effects in most patients. The lack of dependence of the effect of gemfibrozil on the concentration of insulin in conditioned media suggests that gemfibrozil taken by patients may be effective in both moderate and severe forms of hyper(pro)insulinemia.
Elaboration of fibrinogen by Hep G2 cells was not affected by gemfibrozil, consistent with at least relative specificity of the effect on PAI-1. In patients with primary hypertriglyceridemia treated with gemfibrozil for 12 weeks, fibrinogen levels either before or after venous occlusion were significantly lower than those seen with placebo.13 A similar reduction was observed in patients with type IV hyperlipoproteinemia after treatment with gemfibrozil for 90 days.11 In contrast, concentrations of fibrinogen in plasma increased in survivors of myocardial infarction with hypertriglyceridemia after treatment with gemfibrozil for 8 weeks14 as well as in patients with primary hypercholesterolemia treated with gemfibrozil for 1 to 4 months.26 These conflicting observations suggest that other mechanisms in addition to effects on PAI-1 influence the concentrations of fibrinogen.
The expression of the PAI-1 gene after stimulation of cells with insulin was affected by gemfibrozil. Gemfibrozil did not affect steady-state levels of total PAI-1 mRNA; conversely, it reduced steady-state levels of the 3.2-kb species of PAI-1 mRNA. This observation is in contrast with the reduction of levels of both PAI-1 mRNA species by gemfibrozil seen in Hep G2 cells exposed to platelet-associated epidermal growth factor and transforming growth factor-β.27 The discrepancy may be explained by the presence of at least two pathways for PAI-1 mRNA processing and degradation with differential susceptibility to modification by specific mediators, as has been demonstrated previously with insulin stabilizing only the 3.2-kb species and insulin-like growth factor 1 stabilizing both species of PAI-1 mRNA.21 The short half-life of processing enzymes or enzyme complexes suggests that transformation of the 3.2-kb form to the 2.2-kb form21 may depend on a liver nuclear receptor known to interact with fatty acids and possibly with fibric acids such as gemfibrozil as anomalous fatty acids.23 28 Modulation of the translation of PAI-1 mRNA to PAI-1 protein by gemfibrozil may be involved as well, as has been postulated recently in the attenuation of PAI-1 synthesis by niacin, another lipid-lowering agent.29 Gemfibrozil, as a pharmacological agent that inhibits PAI-1 synthesis directly, may provide insight for design of congeners that may inhibit PAI-1 synthesis without altering lipid metabolism. A challenge for the future remains the elucidation of intracellular mechanisms affected by gemfibrozil that alter transcription of the PAI-1 gene, the stability of both PAI-1 mRNA species, and the translation of PAI-1 protein to permit development of more specific approaches for therapeutic modulation of endogenous fibrinolysis.
Gemfibrozil is a promising agent for normalizing impaired endogenous fibrinolysis not only in patients with hypertriglyceridemia but also in patients with hyper-(pro)insulinemia, ie, in patients with insulin-resistance syndromes, including obesity, essential hypertension, and NIDDM (as suggested by the reduced coronary heart disease incidence in the participants of the Helsinki Heart Study suffering from NIDDM).30 By normalizing endogenous fibrinolysis, gemfibrozil may reduce the propensity for accelerated atherosclerosis and restenosis after angioplasty in patients with hyperinsulinemia such as NIDDM.
Selected Abbreviations and Acronyms
|CHD||=||coronary heart disease|
|NIDDM||=||non–insulin-dependent diabetes mellitus|
|PAI-1||=||plasminogen activator inhibitor type 1|
|TPA||=||tissue plasminogen activator|
This article is dedicated to Prof Dr Dr hc Hans Erhard Bock, Department of Medicine, University of Tu¨bingen, Germany, on his 93rd birthday. This work was supported in part by a grant from the Deutsche Forschungsgemeinschaft (DFG No. 214/2-1) and by a grant-in-aid from the Medizinische Fakulta¨t Heidelberg (148/95). The authors thank the Lilly Research Laboratory, Eli Lilly Co, for human des(31,32)proinsulin and des(64,65)proinsulin; and Stephanie Weckesser for excellent technical assistance.
Presented in part at the 68th Scientific Sessions of the American Heart Association, Anaheim, Calif, November 13-16, 1995, and published in abstract form (Circulation. 1995;92[suppl I]:I-555).
- Received May 13, 1996.
- Revision received September 11, 1996.
- Accepted September 24, 1996.
- Copyright © 1997 by American Heart Association
Frick MH, Elo O, Haapa K, Heinonen OP, Heinsalmi P, Helo P, Huttunen JK, Kaitaniemi P, Koskinen P, Manninen V, Ma¨enpa¨a¨ H, Ma¨lko¨nen M, Ma¨ntta¨ri M, Norola S, Pasternack A, Pikkarainen J, Romo M, Sjo¨blom T, Nikkila¨ EA. Helsinki Heart Study: primary prevention trial with gemfibrozil in middle-aged men with dyslipidemia: safety of treatment, changes in risk factors, and incidence of coronary heart disease. N Engl J Med. 1987;317:1237-1245.
Manninen V, Elo MO, Frick MH, Haapa K, Heinonen OP, Heinsalmi P, Helo P, Huttunen JK, Kaitaniemi P, Koskinen P, Ma¨enpa¨a¨ H, Ma¨lko¨nen M, Ma¨ntta¨ri M, Norola S, Pasternack A, Pikkarainen J, Romo M, Sjo¨blom T, Nikkila¨ EA. Lipid alterations and decline in the incidence of coronary heart disease in the Helsinki Heart Study. JAMA. 1988;260:641-651.
Tenkanen L, Ma¨ntta¨ri M, Manninen V. Some coronary risk factors related to the insulin resistance syndrome and treatment with gemfibrozil: experience from the Helsinki Heart Study. Circulation. 1995;92:1779-1785.
Califf RM, Fortin DF, Frid DJ, Harlan WR III, Ohman EM, Bengtson JR, Nelson CL, Tcheng JE, Mark DB, Stack RS. Restenosis after coronary angioplasty: an overview. J Am Coll Cardiol. 1991;17:2B-13B.
Sobel BE. Potentiation of vasculopathy by insulin: implications from an NHLBI clinical alert. Circulation. 1996;93:1613-1615.
Ferguson JJ. NHLBI BARI clinical alert on diabetics treated with angioplasty. Circulation. 1995;92:3371.
Nordt TK, Sawa H, Fujii S, Sobel BE. Induction of plasminogen activator inhibitor type-1 (PAI-1) by proinsulin and insulin in vivo. Circulation. 1995;91:764-770.
Schneider DJ, Sobel BE. Augmentation of synthesis of plasminogen activator inhibitor type 1 by insulin and insulin-like growth factor type I: implications for vascular disease in hyperinsulinemic states. Proc Natl Acad Sci U S A. 1991;88:9959-9963.
Nordt TK, Schneider DJ, Sobel BE. Augmentation of the synthesis of plasminogen activator inhibitor type-1 by precursors of insulin: a potential risk factor for vascular disease. Circulation. 1994;89:321-330.
Fattal PG, Schneider DJ, Sobel BE, Billadello JJ. Post-transcriptional regulation of expression of plasminogen activator inhibitor type 1 mRNA by insulin and insulin-like growth factor 1. J Biol Chem. 1992;267:12412-12415.
Anfosso F, Chomiki N, Alessi MC, Vague P, Juhan-Vague I. Plasminogen activator inhibitor-1 synthesis in the human hepatoma cell line Hep G2: metformin inhibits the stimulating effect of insulin. J Clin Invest. 1993;91:2185-2193.
Charlton PA, Barnes CS, Bent FM, Bryans JS, Faint RW, Folkes AJ, Machin SJ, Mackie I, Bevan P. Evaluation of a low molecular weight inhibitor of plasminogen activator inhibitor-1. Circulation. 1995;92(suppl I):I-486. Abstract.
Fujii S, Sobel BE. Direct effects of gemfibrozil on the fibrinolytic system: diminution of synthesis of plasminogen activator inhibitor type 1. Circulation. 1992;85:1888-1893.
Brown SL, Sobel BE, Fujii S. Attenuation of the synthesis of plasminogen activator inhibitor type 1 by niacin: a potential link between lipid lowering and fibrinolysis. Circulation. 1995;92:767-772.
Koskinen P, Ma¨ntta¨ri M, Manninen V, Huttunen JK, Heinonen OP, Frick MH. Coronary heart disease incidence in NIDDM patients in the Helsinki Heart Study. Diabetes Care. 1992;15:820-825.