Activator Inhibitor Type-1 by Precursors of Insulin

321

Augmentation of the Synthesis of PlasminogenActivator Inhibitor Type-1 by

Precursors of InsulinA Potential Risk Factor for Vascular Disease

Thomas K. Nordt, MD; David J. Schneider, MD; Burton E. Sobel, MD

Background Both vascular disease and elevated concentra-tions in plasma of plasminogen activator inhibitor type-1(PA-1) are prominent in patients with non-insulin-dependentdiabetes mellitus (NIDDM). We and others have hypothesizedthat the increased PAI-1 may contribute to acceleration ofatherosclerosis in this condition and in other states character-ized by insulin resistance as well. Surprisingly, however, ele-vations of PAI-1 decrease when type II diabetic patients aretreated with exogenous insulin, as do circulating concentra-tions of the precursor of insulin, proinsulin, in plasma. Accord-ingly, the increased PAI-1 in patients with NIDDM may reflecteffects of precursors of insulin rather than or in addition tothose of insulin itself. To assess this possibility directly, thisstudy was performed to identify potential direct effects ofproinsulin and proinsulin split products on synthesis of PAI-1in liver cells, thought to be the major source of circulatingPAI-1 in vivo.Methods and Results Hep G2 cells (highly differentiated

human hepatoma cells) were exposed to human proinsulin,des(31,32)proinsulin and des(64,65)proinsulin (split products of

Ni 'on-insulin-dependent (type II) diabetes melli-tus (NIDDM) and other conditions associatedwith insulin resistance, such as hypertension

and obesity, are characterized by accelerated vasculardisease.1 We and others have hypothesized that onefactor that may be responsible is reduced plasma fibri-nolytic activity,2'3 which may shift the balance betweenthrombosis and fibrinolysis toward thrombosis4 andconsequently increase exposure of luminal surfaces ofvessel walls to clot-associated mitogens. Attenuatedfibrinolysis attributable to increased concentrations ofplasminogen activator inhibitor type-1 (PAT-1) inplasma, the primary physiological inhibitor of endoge-nous fibrinolysis, is a risk factor for deep vein thrombo-sis,S premature coronary artery disease,6-9 acute myo-cardial infarction,10-'2 and restenosis after percutaneoustransluminal coronary angioplasty.13The "hyperinsulinemia" typically seen in patients with

NIDDM reflects contributions to the assayed material,immunoreactive insulin (IRI), not only of insulin but also

Received May 6, 1993; revision accepted July 12, 1993.From the Cardiovascular Division and Center for Cardiovascu-

lar Research, Washington University School of Medicine, StLouis, Mo.Correspondence to Dr. Thomas K. Nordt, Cardiovascular Divi-

sion and Center for Cardiovascular Research, Washington Uni-versity School of Medicine, 660 S. Euclid Ave, Campus Box 8086,St Louis, MO 63110.

proinsulin), or C-peptide. Accumulation of PAI-1 in conditionedmedia increased in a time- and concentration-dependent fashionin response to the two des-intermediates [3.3-fold withdes(31,32)proinsulin and 4.5-fold with des(64,65)proinsulin].C-peptide elicited no increase. Stimulation was transduced atleast in part by the insulin receptor as shown by inhibition ofstimulation by insulin receptor antibodies, mediated at the levelof PAI-1 gene expression as shown by the 2.2- to 2.9-foldincreases in steady-state concentrations of PAI-1 mRNA, andindicative of newly synthesized protein as shown by results inmetabolic labeling experiments.

Conclusions Our results are consistent with the hypothesisthat precursors of insulin (proinsulin and proinsulin splitproducts), known to be present in relatively high concentra-tions in plasma in patients with NIDDM and conditionscharacterized by insulin resistance, may directly stimulatePAI-1 synthesis, thereby attenuating fibrinolysis and acceler-ating atherogenesis. (Circulation. 1994;89:321-330.)Key Words * atherogenesis * diabetes mellitus .

fibrinolysis * proinsulin split products * atherosclerosis

of proinsulin split product precursors of insulin that aredetected in the conventional assays used.14 Split productssuch as des(31,32)proinsulin and des(64,65)proinsulinconstitute a relatively large fraction of IRI in plasma ofpatients with type II diabetes who are not treated withexogenous insulin.14"5 Administration of proinsulin topatients with NIDDM has been associated with delete-rious cardiovascular effects, including acute myocardialinfarction,16"17 and elevated concentrations of both pro-insulin and des(31,32)proinsulin have been identified asmarkers of increased cardiovascular risk.'5We and others have reported that PAI-1 activity is

elevated in plasma of patients with NIDDM.2'3 In suchpatients, increases in IRI, one factor implicated in thegenesis of the PAI-1 elevations, are attenuated whenpatients are given exogenous insulin. Furthermore, con-centrations in plasma precursors of insulin decline.'8 Incontrast, concentrations in plasma of precursors ofinsulin are elevated in patients with NIDDM who arenot being treated with exogenous insulin, in part be-cause of the increased stimulation of pancreatic p-cellsthat results from hyperglycemia secondary to insulinresistance and in part because of the impaired process-ing of proinsulin to insulin that is typical of the condi-tion. The decline in concentration of proinsulin inplasma induced by exogenous insulin is accompanied bya decline in plasma PAI-1.18 Thus, the increased plasmaPAI-1 seen in patients with NIDDM may result fromstimulation of PAI-1 synthesis by proinsulin and proin-

by guest on March 21, 2018

http://circ.ahajournals.org/D

ownloaded from

http://circ.ahajournals.org/

322 Circulation Vol 89, No 1 January 1994

sulin split product precursors of insulin rather than byinsulin itself.

Concentrations of PAI-1 in plasma are thought to bedetermined primarily by hepatic synthesis, althoughdiverse cell types can express the PAI-1 gene. Inpreliminary experiments (see "Results"), proinsulinappeared to increase PAI-1 synthesis in liver cells invitro. Accordingly, we performed the present study todetermine whether liver cell PAI-1 synthesis was influ-enced directly by diverse precursors of insulin, includingproinsulin and proinsulin split products, all of which areknown to contribute to conventionally measured IRI inplasma of patients with NIDDM.

MethodsCell Cultures

Studies were performed with Hep G2 cells, highly differen-tiated human hepatoma cells, because the liver appears to bethe primary site for synthesis of circulating PAI-1 and becausechanges in hepatocyte function are likely to be critical inmodulating prevailing concentrations of multiple constituentsresponsible for net fibrinolytic activity in blood. The Hep G2cells were acquired from the American Type Culture Collec-tion ([ATCC] Rockville, Md), seeded at a concentration of3.0x 105 cells per well (24-well plate) and grown to confluencein minimum essential medium with Earle's salts supplementedwith L-glutamine (2 mmol/L) (Life Technologies, Grand Is-land, NY) and 10% NuSerum (Collaborative Biomedical,Bedford, Mass), 30 U/mL penicillin, and 30 ,ug/mL streptomy-cin (Life Technologies). Monolayers of confluent cells wereserum-starved in Dulbecco's modified Eagle's medium withHam's nutrient mixture F-12 supplemented with HEPESbuffer (DME medium, Washington University Medical SchoolTissue Culture Support Center, St Louis, Mo) for at least 16hours, an interval we found to be sufficient for a decline ofPAI-1 synthesis to basal levels. In independent control exper-iments, we showed that the relatively high concentration ofinsulin in NuSerum did not affect the elaboration of PAI-1 inthese cells after serum starvation.

After serum starvation, cells were exposed to fresh mediaconstituted with selected agents, including human recombi-nant insulin (Sigma Chemical Co, St Louis, Mo); humanproinsulin, human des(31,32)proinsulin, or humandes(64,65)proinsulin (Eli Lilly, Indianapolis, Ind); or humansynthetic C-peptide (Sigma) dissolved in 0.9% sodium chloride(Abbott, North Chicago, Ill) with 0.5% bovine albumin (frac-tion V, low endotoxin, cell culture tested, Sigma). Siliconizedtubes and pipette tips were used to minimize the adherence ofthese agents to storage or transfer devices. Control experi-ments were performed with vehicle alone.

Contamination with endotoxin was excluded in all mediaand reagents by testing with Limulus amebocyte lysate (Pyrotellassay; Associates of Cape Cod, Woods Hole, Mass). Thesensitivity for detection of endotoxin was 0.0125 ng/mL. Stocksolutions, reagents, and vehicle contained no detectable endo-toxin. DME medium contained as much as 0.025 ng/mLendotoxin (calibrated with lipopolysaccharide [Sigma] fromEscherichia coli 0111:B4). However, lipopolysaccharide(0.0001 to 1000 ng/mL) exerted no discernible effects on thesecretion of PAI-1 by Hep G2 cells as opposed to effects inendothelial cells.19 Infection of the cells with mycoplasma wasexcluded with the Gen-Probe Mycoplasma Rapid DetectionSystem (Fisher, Fair Lawn, NJ).Human umbilical vein endothelial cells were acquired from

Endotech (Indianapolis, Ind) and grown to confluence inM199 medium (Sigma) supplemented with 10% NuSerum, 50,ug/mL endothelial cell growth supplement (ECGS), 10 ng/mLepithelial growth factor (EGF), 1 ,tg/mL hydrocortisone (allCollaborative Biomedical), 10 U/mL heparin (Sigma), 100U/mL penicillin, and 100 ,ug streptomycin. Serum starvationwas performed in Endothelial-SFM medium (Life Technolo-

gies) supplemented with ECGS, hydrocortisone, and heparinfor 24 hours. The endothelial cells exhibited typical cobble-stone morphology and were used in passage two.Human aortic smooth muscle cells were prepared from

segments of fresh human thoracic aorta obtained through theTissue Division of Mid-America Eye and Tissue Bank(St Louis, Mo) using explants. The cells were grown toconfluence in RPMI-1640 medium (Sigma), supplementedwith 10% fetal bovine serum and 1% antibiotic antimycoticsolution (100x, Sigma) and serum-starved in RPMI-1640medium for 24 hours. They exhibited typical hill-and-valleymorphology, stained homogeneously with mouse monoclonalanti-a-actin antibody and goat fluorescein conjugated anti-mouse immunoglobulin (Ig)G antibody, and were used inpassages three to eight.

Quantification of PAI-1Conditioned media from Hep G2 cells were supplemented

with Tween 80 (final concentration, 0.01%). Cellular debris wasremoved by centrifugation at 13 600g at 4°C for 1 minute.Samples of conditioned media were stored at -20°C until assay.The concentration of PAI-1 antigen in conditioned media wasmeasured by ELISA with a monoclonal goat antibody, MA-7D4B7, and a polyclonal conjugated antibody (TintElize PAI-1,Biopool, Umea, Sweden). Active, latent, and TPA-complexedforms of PAI-1 were detected with equal sensitivity.PAI-1 activity was assayed spectrophotometrically. Samples

were incubated with exogenous TPA (KabiVitrum, Stockholm,Sweden) at room temperature for 10 minutes under conditionsin which PAI-1 but not other low-affinity inhibitors could bindto the TPA. Subsequently, they were acidified and snap-frozento eliminate residual a2-antiplasmin activity. Residual TPAactivity was assayed spectrophotometrically by incubation withplasminogen (Kabi, Molndal, Sweden) and a chromogenicsubstrate S-2251 (KabiVitrum).20

Metabolic Labeling and Immunoprecipitationof Newly Synthesized PAI-1

After serum starvation and stimulation of the cells inculture, DME was replaced by DME devoid of methionine.After 30 minutes, [35S]methionine (Tran35S-Label; ICN, Irvine,Calif) was added at a final concentration of 100 ,uCi/mL. Afteran additional 60 minutes, the cells were transferred to freshDME without radioactive methionine. One hour later, condi-tioned media, supplemented with 0.01% Tween 80, wereharvested and stored at -20°C until immunoprecipitation wasperformed. The cells were washed with phosphate-bufferedsaline (PBS) and lysed with ice-cold buffer (10 mmol/L Tris[pH 7.4], 150 mmol/L sodium chloride, 1% Nonidet P-40,0.5% sodium deoxycholate, 0.1% SDS, 1 mmol/L phenylmeth-ylsulfonyl fluoride, and 1 mmol/L iodoacetate). Cell lysateswere scraped, disrupted by sonification on ice for 1 minute,and centrifuged at 10 00g for 10 minutes. The supernatantfractions were stored at -20C until immunoprecipitation wasperformed.For immunoprecipitation,21 bovine serum albumin (final

concentration, 0.1%), used to reduce nonspecific binding ofproteins to antibodies, and mouse monoclonal anti-humanPAI-1 antibody (American Diagnostica, Greenwich, Conn) inmolar excess (which binds to active, latent, and TPA com-plexed PAI-1) were added to the samples. After incubationand gentle agitation at 4°C for 16 hours and subsequentincubation with agarose-linked goat anti-mouse IgG antibody(HyClone, Logan, Utah) at room temperature for 2 hours,antibody-bound PAI-1 was pelleted by centrifugation of thesamples in a microcentrifuge. The pellets were washed in PBSwith 0.1% SDS, 0.5% Nonidet P-40, and 0.1% deoxycholicacid twice; in PBS alone once; resuspended in reducing buffer;and heated at 100°C for 3 minutes. The supernatant fractions,which contained PAI-1, were subjected to SDS-polyacryl-amide gel electrophoresis (SDS-PAGE) on stacking gels of4%and separating gels of 10% acrylamide (BioRad, Richmond,



ownloaded from


Nordt et al Proinsulin Split Products and PAI-1 323

Calif; Protogel, National Diagnostics, Manville, NJ). Afterelectrophoresis, the gels were stained with rapid Coomassieblue (Diversified Biotech, Boston, Mass), dried (Drygel Jr SE540; Hoefer, San Francisco, Calif), and subjected to autorad-iography at -70°C for selected intervals. Intensity of individ-ual bands was quantified by laser densitometry (Ultroscan XL;Pharmacia LKB, Piscataway, NJ) and by radioisotopicscanning.

Quantification of PAI-1 mRNATo obtain a probe for Northern blotting, plasmids containing

human PAI-1 cDNA were purified from transfected E. coli byalkaline lysis and with Qiagen resins (Qiagen Plasmid Maxi Kit;Diagen, Hilden, Germany) and digested with the restrictionenzymes, EcoRI and Sal I (Promega, Madison, Wis). The DNAfragments were separated on low melting temperature agarosegels (2% NuSieve GTG agarose; FMC BioProducts, Rockland,Me) in TAE buffer (40 mmol/L Tris acetate, 2 mmol/L EDTA[pH 8.5]). A 0.9-kb fragment containing PAI-1 mRNA wasrecovered with elutip-d minicolumns (Schleicher & Schuell,Keene, NH), precipitated with ethanol (Quantum, Tuscola, Ill)and glycogen (Boehringer Mannheim, Indianapolis, Ind) as acarrier, rinsed in 70% ethanol, and dissolved in TE buffer (10mmol/L TrisCl, 1 mmol/L EDTA [pH 7.4]). Before hybridiza-tion, the PAI-1 cDNA was labeled with deoxycytidine 5'-[a-32P]triphosphate (Amersham, Arlington Heights, Ill) by therandom oligonucleotide primer method (Random Primed DNALabeling Kit; Boehringer Mannheim).22For assay of PAI-1 mRNA, total cellular RNA was extracted

with RNAzol B (Tel-Test, Friendswood, Tex) and chloroform(Fisher), precipitated with isopropanol (Fisher), washed with75% ethanol, dissolved in 1 mmol/L EDTA (pH 7.0), and sizefractionated (10 to 20 jig) on 1.5% formaldehyde agarose gelscontaining 0.53 ,g/mL ethidium bromide in MOPS buffer (20mmol/L MOPS, 5 mmol/L sodium acetate, 1 mmol/L EDTA[pH 7.0]). Northern blotting was performed by capillary transferof RNA to nylon membranes (Biodyne B; Pall Filter, Vienna,Austria) with 20x SSC (3 mol/L sodium chloride, 0.3 mol/Lsodium citrate [pH 7.0]). RNA was fixed by baking at 80°C for1 hour in a vacuum. The integrity, equal loading, and transfer ofRNA were verified by ethidium bromide staining of ribosomalRNA.

Prehybridization was performed at 42°C for at least 2 hours,and hybridization was performed at 42°C for 20 to 24 hours.Membranes were prehybridized in a solution of 50% deionizedformamide, 10x Denhardt's solution, 50 mmol/L Tris-HCl(pH 7.5), 1 mol/L sodium chloride, 0.1% sodium pyrophos-phate, 1% SDS, 10% dextran sulfate, and 100 ,ug/mL dena-tured calf thymus DNA. For hybridization, a labeled PAI-1probe (500 000 cpm/mL) was added.23 Membranes werewashed three times at room temperature for 5 minutes each ina solution of 1% SDS and 2x SSC followed by washing at 60°Cfor 20 minutes in fresh washing solution. Radioactivity ofhybridized bands was quantified by radioisotopic scanning(AMBIS; Radioanalytic Imaging System, San Diego, Calif)with results confirmed and documented by autoradiography.

In additional internal control experiments, membranes wereboiled in Northern strip buffer (10 mmol/L Tris [pH 8.0], 1mmol/L EDTA [pH 8.0], 1% SDS) for 10 minutes to strip offprobes and were rehybridized as above with a labeled glycer-aldehyde-3-phosphate probe (GAP probe, 0.6-kb fragmentafter Xba I/HindIII dehydrogenase digestion, 200 000cpm/mL).

Additional ProceduresProtein in conditioned media supplemented with 0.01%

Tween 80 and in cell lysates was quantified with bicinchoninicacid (Pierce, Rockford, Ill) conventionally. To prepare lysates,cells were washed with PBS three times, incubated in filtered0.5% Triton X-100 at room temperature for 1 hour, andsonified on ice for 1 minute. The supernatant fractions ob-

tained after centrifugation at 12 00(g at 4°C for 10 minuteswere stored at -20°C until assay.

Purified mouse monoclonal anti-human insulin receptorantibody (clone MA 20) was purchased from Amersham. Theantibody binds to an epitope on the a-subunit of the insulinreceptor, inhibits the binding of insulin to its receptor, andinduces no significant autophosphorylation of the insulinreceptor. Rabbit polyclonal anti-human IGF-1 receptor anti-body to an epitope on the a-subunit of the insulin-like growthfactor-1 (IGF-1) receptor was obtained from Upstate Biotech-nology (UBI, Lake Placid, NY). Both antibodies were addedto media bathing the cells at least 1 hour before the additionof an agonist to be tested. IGF-1 was purchased from Collab-orative Biomedical. Actinomycin D (from Streptomyces sp),cycloheximide (from Streptomyces griseus), and genistein wereobtained from Sigmna and dissolved immediately before use.

Statistical AnalysisResults are mean±SEM values. The significance of differ-

ences between groups of cells was assessed by one-wayANOVA and with Student's t tests. Significance was defined asP<.05.

ResultsEffects of Proinsulin, des(31,32)Proinsulin,des(64,65)Proinsulin, and C-Peptide on theConcentration of PAI-1 in ConditionedMedia of Hep G2 CellsHep G2 cells were exposed to media containing 10

nmol/L proinsulin, 10 nmol/L des(31,32)proinsulin, 10nmol/L des(64,65)proinsulin, 10 nmol/L C-peptide, orvehicle alone (as a control) (n=3 for each condition).Samples of conditioned media (less than 2.5% of thevolume of conditioned media each) were removed every12 hours for up to 5 days. Stimulation with proinsulininduced a prompt increase in PAI-1 protein in the first36 hours followed by a plateau for an additional 36hours (with 25 ng/mL PAI-1 being the average) and alate increase between 72 and 96 hours (with 46 ng/mL asan average) (Fig 1). After 24 hours and at all subsequentintervals, PAI-1 was significantly increased with expo-sure of the cells to proinsulin or to des(31,32)proinsulinand des(64,65)proinsulin. The two proinsulin split prod-ucts elicited an increase in PAI-1 that persisted for up to60 hours with a "rapid" increase occurring between 12and 36 hours (from 4 to 45 and from 3 to 50 ng/mL,respectively) and a more modest increase seen between36 and 60 hours (from 45 to 60 and from 50 to 70 ng/mL,

E0,

c

V-,.C

[L

60

40

20

0

+* Control-+-lOnM Proinsulin--- lOnM des(31,32)P.-&- lOnM des(84'65)1P.-*- lOnM C-Peptide

-w - L

0 12 24 36 48 60 72Time (h)

1 1 184 96 108 120

FIG 1. Plot of concentration of plasminogen activator inhibitortype-1 (PAl-i) protein in conditioned media of highly differentiatedhuman hepatoma cells (Hep G2) exposed to 10 nmol/L concen-trations of proinsulin, des(31,32)proinsulin, des(64,65)proinsulin,C-peptide, or vehicle alone as a function of the duration ofexposure.

80 r

- -W -1- 1 JL



ownloaded from



100 r

80 h60 [40

0200zMy0

100F

0.1 1 10Proinsulin (nM)

**

100

c 80Em 60

_ 40

X< 20

1000 AM 0.1 I A 1111111 1 0Mill 101111

1 10 100des(31,32) Proinsulin (nM)

100 r

'U F

I0 *$0

0 Ill]11 I 1111111 1 il ifi I 111111l

0 0.1 1 10 10(des(64,65) Proinsulin (nM)

CEm

rc_s

[L

80

60

40

20

)

"'[I*f~ ~ - -*

11111111 1111111 1 1111111 A 11111110.1 1 10 100

C-Peptide (nM)

**

E0)

CL

II 1 1111 it 1 1111 if 1 1111 1 1 111111

0.1 1 10 100

300

200

100

O

Proinsulin (nM)

9~~~~

1-111111 1111111 1 111111 1111111

0.1 1 10 100des(31,32) Proinsulin (nM)

300 FE

c1-_.

200 F

100 0 0 -

d "1A """AI I I 11illu IA 11111111 ,, 1111111

0 0.1 1 10 10des(64,65) Proinsulin (nM)

00L 11 1 111111 1 1 1111111 1 1 111111

0.1 1 10 100C-Peptide (nM)

-7, ''"""''" -11-Ii .'.1 1 1 1 11 11

FIG 2. Plots of concentration of plasminogen activator inhibitor type-1 (PAl-1) protein in conditioned media of highly differentiatedhuman hepatoma cells (Hep G2) exposed to selected concentrations (0.1, 1, 10, and 100 nmol/L) of proinsulin, des(31,32)proinsulin,des(64,65)proinsulin, or C-peptide for 24 hours (A) and for 48 hours (B) (*P<.05, **P<.005, ***P<.0005).

respectively). After 60 hours, no further increase oc-curred, consistent with a general attenuation of proteinsynthesis over time in Hep G2 cells in culture undermost conditions. Overall, the accumulation of PAI-1induced by des(64,65)proinsulin appeared to be greaterthan that with des(31,32)proinsulin. Between 36 and 84hours, both agonists induced significantly greater accu-mulation of PAI-1 compared with the increases inducedby proinsulin itself. In contrast, Hep G2 cells exposed toC-peptide exhibited no increase in PAI-1.To define concentration-response relations, Hep G2

cells were exposed to each agonist at concentrations of 0.1,

1, 10, and 100 nmol/L for 24 and 48 hours. Fig 2 shows theconcentration-dependent increase of PAI-1 protein inconditioned media seen with proinsulin, des(31,32)-proinsulin, and des(64,65)proinsulin after 24 and 48 hours(n=3 for each). After 48 hours, 100 nmol/L proinsulin hadelicited a 2.0-fold increase. des(31,32)Proinsulin (100nmol/L) and 100 nmol/L des[64,65]proinsulin had eliciteda 3.3-fold and a 4.5-fold increase, respectively. The thresh-old concentration of agonist was 10 nmol/L for proinsulinbut only 1 nmol/L for the other two precursors of insulin.In contrast, C-peptide in concentrations as high as 100nmol/L had no effect.

A

c

v-

1

Gs

C

a-cl

Rn-L

300 -

200 F

B

c

a-a-

100

,,CL

300 j

? 2001-

< 100

9 allil ssli al E asllo A 1M

_f, _.

11 U1 ff

c

E

4

2

) )1

1

, ,-



ownloaded from



E

E008

0.4

0-024 48 24 48 24 48 24 48 24 48 h

des(31,32) deS(84,85)Control Proinsuin Proinsuin Proinsuin C-Peptide (10 nM)

FIG 3. Bar graph of concentration of total protein in conditionedmedia of highly differentiated human hepatoma cells (Hep G2)exposed to 10 nmol/L concentrations of proinsulin, des(31,32)proinsulin, des(64,65)proinsulin, C-peptide, or vehicle alone for24 hours and for 48 hours.

Effects of Proinsulin, des(31,32)Proinsulin,des(64,65)Proinsulin, and C-Peptide on

Overall Protein SynthesisTo determine whether nonspecific effects on protein

synthesis had been induced by the agonists studied, HepG2 cells were exposed to 10 nmol/L proinsulin,des(31,32)proinsulin, des(64,65)proinsulin, C-peptide,or vehicle alone (n=3 each) for 24 and 48 hours. Asshown in Fig 3, no significant increase in total protein inconditioned media was seen compared with values incontrol cells that had not been exposed to the agonists.Total protein increased in a generally linear fashionbetween 24 and 48 hours (from 0.26±0.02 to 0.55+±0.03mg/mL, P<.0005). Total protein in cell lysates did notincrease in response to any of the agonists tested(1.48±0.05 mg/mL at 24 hours, 1.68+0.06 mg/mL at 48hours). Thus, effects of the insulin precursors on PAI-1synthesis were not simply a reflection of a general effecton overall protein synthesis.

TAaLE 1. PAI-1 Activity in Conditioned Media of HepG2 Cells 24 or 48 Hours After the Addition of 10 nmol/LProinsulin, 10 nmol/L des(31,32)Prolnsulln, 10 nmol/Ldes(64,65)Proinsulin, 10 nmol/L C-Peptide, orVehicle Alone

PAI-1 Activity, arbitraryunts/mL

Agonist 24 hours 48 hours

None (control) 1.3±0.4 1.1±0.1

Proinsulin 1.5±0.1 1.0±0.2

des(31,32)Proinsulin 3.9±0.1t 2.0±0.2*des(64,65)Proinsulin 3.0±1.1 1.7±0.2*

C-Peptide 1.6±0.2 1.3±0.2

PAI-1 indicates plasminogen activator inhibitor type-i; HepG2, highly differentiated human hepatoma cells.n=3 for each, mean±SEM.*P<.05, tP<.005.Activities are minuscule with respect to total PAI-1 protein

(which has a specffic activity of approximately 0.5 arbitraryunits/ng when fully active) because of the rapid conversion ofactive PAI-1 to latent PAI-1 and to inactive (cleaved) PAI-1 inconditioned media in the absence of addition of an exogenousbinding protein such as vitronectin.42

Effects of Proinsulin, des(31,32)Proinsulin,des(64,65)Proinsulin, and C-Peptide onPAI-1 Activity

After cells had been incubated with 10 nmol/L pro-insulin, 10 nmol/L des(31,32)proinsulin, 10 nmol/Ldes(64,65)proinsulin, 10 nmol/L C-peptide, or vehiclealone for 24 and 48 hours, respectively, conditionedmedia were harvested and assayed for PAI-1 activity(n=3 for each condition). PAI-1 activity was low in eachcase (Table 1), although values were somewhat higherin cells exposed to the proinsulin split products. Thedichotomy between activity and concentration of PAI-1protein in conditioned media is consistent with rapidconversion of PAI-1 to a latent form and to inactive(cleaved) forms in conditioned media of cultured cellsin the absence of exogenous vitronectin (t/2 <2 hours at370C).24

Effects of Proinsulin, des(31,32)Proinsulin, anddes(64,65)Proinsulin on PAI-1 Synthesis

Metabolic labeling of Hep G2 cells was performed todetermine whether the precursors of insulin stimulatedsynthesis as well as secretion of PAI-1. The midpoint ofthe interval of incubation with labeled methioninecorresponded to the 24- and 48-hour interval after theaddition of proinsulin, des(31,32)proinsulin, or des(64,65)proinsulin (10 nmol/L each, n=6 each for conditionedmedia, n=3 each for cell lysates).As shown in Fig 4, labeled PAI-1 was detectable in a

band of approximately 50 kd.25 In conditioned mediaharvested 24 hours after addition of agonists, newly syn-thesized PAI-1 increased from 107±6 (control) by 1.6-foldto 174±15 (P=.002) with proinsulin, 1.5-fold to 165±12(P=.002) with des(31,32)proinsulin, and 2.0-fold to216±22 cpm (P<.001) with des(64,65)proinsulin. Theseincreases were consistent with results in the concentra-tion-response experiments [1.3-fold increase of PAI-1protein with proinsulin, 1.7-fold with des(31,32)proinsulin,and 1.8-fold with des(64,65)proinsulin] and with increasedplasma concentrations of PAI-1 in type II diabetic pa-tients in comparison with those in nondiabetic subjects.2After 48 hours, differences compared with values in con-trols were still present, consistent with the persistentelevation of PAI-1 mRNA (see later).

In lysates of cells incubated for 24 hours, the in-creases in newly synthesized PAI-1 seen after exposureof the cells to proinsulin and to proinsulin split productsparalleled increases seen in conditioned media after 24hours. Increases seen with des(64,65)proinsulin weremost prominent (1.7-fold increase from 0.44±0.06 to0.75±0.05 arbitrary units, P=.014). Thus, changes inaccumulation of PAI-1 protein in conditioned mediaappeared to reflect changes in the rate of PAI-1 synthe-sis. This interpretation is supported by results of controlexperiments with cycloheximide, an inhibitor of proteinsynthesis, in which Hep G2 cells exposed to the agonistsexhibited no increase in PAI-1 in conditioned mediaafter 24 hours (Table 2).

Effects of Proinsulin, des(31,32)Proinsulin,des(64,65)Proinsulin, and C-Peptide onSteady-State Concentrations of PAI-1 mRNAThe increases in PAT-1 synthesis could reflect

changes in gene expression (transcription or degrada-tion of mRNA), changes in translation, or both. To



ownloaded from



FIG 4. Concentration of 35S-methionine-labeled plasminogen activator inhibitor

- 180 type-1 (PAI-1) in conditioned media ofhighly differentiated human hepatomacells (Hep G2) exposed to 10 nmol/Lconcentrations of proinsulin, des(31,32)

- 1 25 proinsulin, des(64,65)proinsulin, or vehi-cle alone for 24 hours and for 48 hours(**P<.005, ***P<.0005). Labeled PAI-1was assayed by radioisotopic scanning(results are shown in lower B). A repre-sentative autoradiogram is shown in A

75 (obtained after an exposure of 24 hours).The bands at 50 kd represent PAI-1, andaccording to the molecular weight, thebands at 75-kd vitronectin and at 125-kdand 1 80-kd PAI-1 -vitronectin complexes,respectively. In the left lane are [14C]me-

50 thylated proteins as molecular weightmarkers (Amersham).

kD

h

(10 nM)

assess potential effects on gene expression, Hep G2 cellswere exposed to 10 nmol/L concentrations of the ago-nists tested (n=4 for each condition). After 24 and 48hours, total cellular RNA was extracted, and PAL-1mRNA was assayed. Two forms of PAI-1 mRNA were

detected: a 3.2-kb and a 2.2-kb form. The ratio of the3.2- to the 2.2-kb form was constant (0.9) under basalconditions and with stimulation (Fig 5). Steady-statelevels of PAI-1 mRNA were increased 2.3-fold in 24hours with exposure of cells to proinsulin (P-.005),2.2-fold to des(31,32)proinsulin (P=.004), and 2.9-foldto des(64,65)proinsulin (P<.0005). C-peptide did notelicit any increase. Steady-state levels of GAP mRNA

(used as an internal control) did not change signifi-cantly. Increases in PAI-1 mRNA persisted for 48hours. Accordingly, the effects of the insulin precursorsappeared to be mediated in part at the level of geneexpression. Control experiments with actinomycin D, a

specific inhibitor of transcription, were confirmatory inthat this inhibitor completely attenuated the increasesin PAI-1 protein in conditioned media over 24 hours(Table 2).

Mediation of Observed Effects by Insulin ReceptorsTo determine whether the effects observed were

mediated in whole or in part by insulin receptors, a

A

01.

0c)

c

.5Co

Ca.5

0

CM

cm-

a)

0r-

0~%

CD

COa)

01-

200 -

97 -

69-

46 -

E.........

.ER

.........E

i.i... ........ ............

kDB

0

.'00 -oii= -1

CO4)ra

250

200

150

100

50

o illi24 48 24 48 24 48 24 48

des(31,32) des(64,65)Control Proinsulin Proinsulin Proinsulin



ownloaded from


Nordt et al Proinsulin Split Products and PAI1 327

TABLE 2. PAI-1 Protein in Conditioned Media of Hep G2 Cells 24.5 Hours After the Addition of 2.5,mol/L Cycloheximide or 0.2 ,umol/L Actinomycin D and 24 Hours After the Addition of 10 nmol/LProinsulin, 10 nmol/L des(31 ,32)Proinsulin, 10 nmol/L des(64,65)Proinsulin, or Vehicle Alone

PAI-1 Protein, ng/mL

Agonist Control Cycloheximide, 2.5 jumol/L Actinomycin D, 0.2 umol/LNone (control) 37±4 10±2 16±1

Proinsulin 59±2* 11±1 20±1

des(31,32)Proinsulin 60±3* 12±1 18±1

des(64,65)Proinsulin 70±2t 12±1 20±2

PAI-1, plasminogen activator inhibitor type-1.n=3 for each, mean±SEM.*P<.05, tP<.005.Cycloheximide, an inhibitor of protein synthesis, suppressed the stimulation by the agonists used, supporting the

concept that changes in the accumulation of PAI-1 protein in conditioned media reflect changes in the rate of PAI-1synthesis. Actinomycin D, an inhibitor of transcription, suppressed the stimulation as well supporting the conceptthat the effects of insulin precursors are mediated in part at the level of gene expression.

monoclonal antibody against an epitope on the a-sub-unit of the insulin receptor capable of inhibiting thebinding of insulin to its receptor yet not acting as apartial agonist or autophosphorylating the receptor wasused. Hep G2 cells were exposed to 10 nmol/L concen-trations of each agonist for 24 hours with 0, 1, 5, or 10,ug/mL of the insulin receptor antibody (n=3 for eachcondition). Insulin (10 nmol/L) was used as a control.The insulin receptor antibody inhibited the increased

accumulation of PAI-1 protein in conditioned mediacompletely when proinsulin was the agonist and almostcompletely (86%) when des(31,32)proinsulin was theagonist (Fig 6). With des(64,65)proinsulin, the inhibi-tion was 68%. Inhibition was even greater than thatseen with insulin (47%). Thus, effects of the precursorsof insulin appeared to be mediated, largely or exclu-sively, through the insulin receptor. This interpretationis corroborated by our finding that stimulation by theprecursors was not additive to that induced by insulin.In fact, insulin attenuated the effects. In control exper-iments with a polyclonal receptor antibody to an epitopeon the a-subunit of the IGF-1 receptor, no inhibitionwas seen. Effects of genistein,26 an inhibitor of tyrosineprotein kinase that mediates diverse effects of both theinsulin and the IGF-1 receptor, were not characterizedbecause the conditions required were found to damagethe Hep G2 cells.

Quantitative differences in the effects on synthesis ofPAI-1 by the different precursors of insulin seen werenot likely to be attributable to different affinities of theprecursors proinsulin, des(31,32)proinsulin, anddes(64,65)proinsulin to the insulin receptor as judgedfrom the comparable effects of equimolar concentra-tions of each on the time course of increase andsteady-state levels of PAI-1 mRNA. They do not appearto be attributable to conversion of any of the precursorsto insulin as judged from the lack of appearance ofinsulin in the conditioned media based on SDS-PAGEstudies and with autoradiography after addition oflabeled proinsulin to the conditioned medium incu-bated over 48 hours (data not shown).

Vascular Endothelial and Smooth Muscle CellsHuman umbilical vein endothelial cells were exposed

to 1 and 10 nmol/L concentrations of proinsulin,des(31,32)proinsulin, des(64,65)proinsulin, C-peptide,

or vehicle alone (as a control) (n=3 each) for 24 hours.None of the agonists increased PAI-1 protein in condi-tioned media in comparison with results under controlconditions. Human aortic smooth muscle cells wereincubated with 0.1, 1, 10, and 100 nmol/L concentra-tions of the agonists used above for 24 and 48 hours (twodifferent donors, n=3 each). Again, none of the agonistsincreased PAI-1 protein in conditioned media.

DiscussionHep G2 cells were used in this study because hepa-

tocytes appear to be the primary source of circulatingPAI-1 and because Hep G2 cells are highly differenti-ated human cells with gross morphological and func-tional27 properties similar to those in hepatocytes inprimary culture, including interactions of cell surfacereceptors with insulin27 and insulin-induced secretion ofPAI-1.28-30

Conversion of proinsulin to insulin in vivo occursprimarily in immature secretory granules of pancreatic3-cells in reactions catalyzed by two types of calcium-dependent endopeptidases (trypsin-like and carboxy-peptidase B-like endopeptidase) through a branchedpathway. This type of proteolytic cleavage, involved inthe processing of many eukaryotic secretory proteins,occurs on the carboxylic side of the two dibasic sectionsof the proinsulin molecule at the Arg31-Arg32 and atLys&I-Arg65 sites. Proinsulin is processed either to(32,33)split proinsulin and des(31,32)proinsulin or to(65,66)split proinsulin and des(64,65)proinsulin. Subse-quently, each split product can be converted to insulin.The first pathway [formation of des(31,32)proinsulin]dominates. Because of the high activity of carboxypep-tidase B in the pancreatic 13-cell secretory granules,neither of the split proinsulins (in contrast to the des-split products) is readily detected within the pancreaticcells themselves in situ.31,32

Proinsulin has been given intravenously to human sub-jects. Over intervals as long as 24 hours, it is not convertedappreciably to intermediates or to insulin, with 99% of thematerial remaining intact in plasma as proinsulin.31 Innormal human subjects, the typical concentration of pro-insulin in plasma is 0.002 nmol/L under fasting conditionsand 0.010 nmol/L after an oral glucose load. Correspond-ing values for des(31,32)proinsulin are 0.002 nmol/L and0.020 nmol/L. The two moieties together account for only



ownloaded from



FIG 5. The concentration of plasmino-gen activator inhibitor type-1 (PAI-1)mRNA in highly differentiated human

- 3.2 kb hepatoma cells (Hep G2) exposed to10 nmol/L concentrations of proinsulin,des(31 ,32)proinsulin, des(64,65) pro-insulin, C-peptide, or vehicle alone for24 hours and for 48 hours (*P<.05,

- 2.2 kb **P<.005, ***P<.0005). Hybridizedbands were quantified by radioisotopicscanning. A, Representative autoradio-grams and the same membrane afterethidium bromide staining (24 hours),and B shows the total amount of PAI-1mRNA (the sum of that in the 3.2- and

- 1.3 kb 2.2-kb bands).

B70

60

X- 50

<c 40z

cc 30E

20

10

0

24 48 24 48

Control Proinsulin

10% to 20% of immunuoreactive insulin-like molecules(IRI).33M34 In nonobese patients with NIDDM, the frac-tional contributions increase to 14% for proinsulin and upto 45% for des(31,32)proinsulin under both fasting condi-tions [0.018 nmol/L for proinsulin and 0.056 pmol/L fordes(31,32)proinsulin] and after an oral glucose load. Inobese subjects with NIDDM, the two precursors of insulinaccount for as much as 66% of total IRI.14"15,35 Thus, muchof the conventionally measured IRI reflects contributionsof precursors rather than insulin per se in patients withNIDDM who are not being treated with insulin. Theconcentrations of agonists used in the present study wereselected to be consistent with concentrations of precursorsseen in patients with NIDDM. In vivo, hepatocytes may beexposed to even higher concentrations (as high as 7nmol/L) of the precursors of insulin because portal venousblood drains directly from the pancreas into the liver.36

:S-1 8S

24 48 24 48 24 48 h

des(31,32) des(64,65)Proinsulin Proinsulin C-Peptide (10 nM)

"Hyperinsulinemia" (actually increased IRI), knownto occur in subjects with NIDDM or other conditionscharacterized by insulin resistance including obesity andhypertension, appears to be a risk factor for coronaryartery disease.37 In patients with NIDDM, the increasedconcentrations of IRI in plasma are paralleled by in-creased PAI-1.23 Such patients exhibit dysfunction ofpancreatic p-cells characterized by elaboration of in-creased quantities of precursors of insulin that cancomprise more than 50% of IRI in blood.38,39 Proinsulinis cleared more slowly from the circulation than insulin,thereby intensifying hyperproinsulinemia.'7Some investigators have obtained information sug-

gesting that the "hyperinsulinemia" in NIDDM actuallyreflects hyperproinsulinemia and hyperdes(31,32)proinsulinemia as judged from results of assays withhighly specific immunoassays for each moiety.'4'5,3940

A

c

0

c

._co

C

70

CL

0r-

0am

0U)

CJC0)

Co

0)CD10r

a:(D

(0(D

0

0r-

(_0

Q

CL

0

PAI-1mRNA

GAPmRNA

rRNA

,j ...111111 ...

.....................

................



ownloaded from



0 1510 0 1510

des(31,32) des(64,65)Proinsulin Proinsuln

(,ugimi)

(10 nM)

FIG 6. Bar graph of plasminogen activator inhibitor type-1 (PAl-i)protein measured in the conditioned media of highly differentiatedhuman hepatoma cells (Hep G2) exposed to 10 nmol/L concen-

trations of insulin, proinsulin, des(31,32)proinsulin, or des(64,65)proinsulin for 24 hours in the presence of increasing concentra-tions of a monoclonal antibody to the insulin receptor.

Thus, "hyperinsulinemic" patients may, in fact, be in-sulin deficient.40 In view of these considerations, it is ofparticular interest that plasma PAI-1 declines (sic) inpatients with type II diabetes given exogenous insulin,perhaps because concentrations in plasma of proinsulinand other precursors of insulin decline as stimulation ofthe pancreatic/-cells by hyperglycemia declines. In fact,concentrations of PAI-1 in plasma in type II diabeticpatients correlate with concentrations of proinsulin anddes(31,32)proinsulin rather than with concentrations ofinsulin per se.18The markedly (up to 18-fold) increased frequency of

cardiovascular events observed when proinsulin was

administered to patients for at least 1 year17 is consis-tent with a deleterious effect of hyperproinsulinemia on

fibrinolysis mediated by increased PAI-1. Under suchconditions, the concentration of proinsulin in plasma inthe proinsulin-treated patients was as high as 8 nmol/L,a concentration similar to that used in the present study.The results of the present study are consistent with

the hypothesis that both the increased risk of cardiovas-cular disease associated with insulin resistance and"hyperinsulinemia" (defined in terms of increased IRI)and the association of increased PAI-1 with increasedIRI may be attributable, in part, to effects of precursors

of insulin on the fibrinolytic system and specifically on

PAI-1 synthesis. They are consistent with the dichotomybetween plasma PAI-1 in patients with NIDDM whoare given exogenous insulin compared with plasmaPAI-1 in those who are not18 and with the parallelchanges in plasma PAI-1 and plasma proinsulin seen inpatients with NIDDM not given exogenous insulin.They are consistent also with the possibility that pre-

cursors of insulin may contribute to vascular disease byimpairing fibrinolysis, consequently increasing exposure

of vessel walls to intermittent platelet activation andrelease of platelet- and clot-associated mitogens. Theysuggest that reduction of atherogenesis in patients withNIDDM and in those with other conditions character-ized by insulin resistance may be possible by suppres-

sion of elaboration of precursors of insulin or by phar-macologic antagonism of the direct effects of precursors

of insulin on PAI-1 synthesis.

AcknowledgmentsDr Thomas Nordt dedicates this research to his father,

Dieter Nordt (1937-1992).

This work was supported in part by National Heart, Lung,and Blood Institute grant HL-17646, SCOR in Coronary andVascular Diseases. Dr Thomas K. Nordt is supported by theDeutsche Forschungsgemeinschaft in Germany (No. 214/1-1).The authors thank Dr Kevin J. Klassen for collaboration inobtaining the aortic segments; Dr Hiroko Noda-Heiny forassistance in preparing the smooth muscle cells; Dr HirofumiSawa for assistance in preparing the PAI-1 cDNA probe used;Dr Bruce H. Frank, Lilly Research Laboratory, Eli LillyCompany, for human proinsulin, des(31,32)proinsulin, anddes(64,65)proinsulin; the Tissue Division of Mid-America Eyeand Tissue Bank for the aortic segments from organ donors;Drs Satoshi Fujii and Ronald Gingerich for helpful discus-sions; Denise Nachowiak and Jeffrey Labuda for superbtechnical assistance; and Lori Dales and Kelly Hall for secre-tarial support.

References1. Jacoby RM, Nesto RW. Acute myocardial infarction in the diabetic

patient: pathophysiology, clinical course and prognosis. JAm CollCardiol. 1992;20:736-744.

2. Juhan-Vague I, Alessi MC, Vague P. Increased plasma plas-minogen activator inhibitor 1 levels: a possible link between insulinresistance and atherothrombosis. Diabetologia. 1991;34:457-462.

3. Schneider DJ, Nordt TK, Sobel BE. Attenuated fibrinolysis andaccelerated atherogenesis in type II diabetic patients. Diabetes.1993;42:1-7.

4. Sobel BE, Hirsh J. Principles and practice of coronary throm-bolysis and conjunctive treatment. Am J Cardiol. 1991;68:382-388.Editorial.

5. Wiman B, Hamsten A. The fibrinolytic enzyme system and its rolein the etiology of thromboembolic disease. Semin Thromb Hemost.1990;16:207-216.

6. Aznar J, EstelIes A, Tormo G, Sapena P, Tormo V, Blanch S,Espafia F. Plasminogen activator inhibitor activity and other fibri-nolytic variables in patients with coronary artery disease. Br HeartJ. 1988;59:535-541.

7. Olofsson BO, Dahlen G, Nilsson TK. Evidence for increased levelsof plasminogen activator inhibitor and tissue plasminogen activatorin plasma of patients with angiographically verified coronary arterydisease. Eur Heart J. 1989;10:77-82.

8. Bick RL, Baker WF. Tissue plasminogen activator inhibitor type 1(PAI-1) abnormalities in patients with precocious and prematurecoronary artery disease. Blood. 1992;80:507a. Abstract.

9. Graf S, Beckmann R, Christ G, Geppert A, Kaindl F, Binder BR,Huber K. Possible causes of a defective fibrinolytic system in youngpatients with coronary artery disease. J Am Coll Cardiol. 1993;21:211A. Abstract.

10. Verheugt FWA, ten Cate JW, Sturk A, Imandt L, Verhorst PMJ,van Eenige MJ, Verwey W, Ross JP. Tissue plasminogen activatoractivity and inhibition in acute myocardial infarction and angio-graphically normal coronary arteries. Am J Cardiol. 1987;59:1075-1079.

11. Hamsten A, Wiman B, de Faire U, Blomback M. Increased plasmalevels of a rapid inhibitor of tissue plasminogen activator in youngsurvivors of myocardial infarction. N Engl J Med. 1985;313:1557-1563.

12. Hamsten A, de Faire U, Walldius G, Dahlen G, Szamosi A,Landou C, Blomback M, Wiman B. Plasminogen activatorinhibitor in plasma: risk factor for recurrent myocardial infarction.Lancet. 1987;2:3-9.

13. Huber K, Jorg M, Probst P, Schuster E, Lang I, Kaindl F, BinderBR. A decrease in plasminogen activator-1 activity after successfulpercutaneous transluminal coronary angioplasty is associated witha significantly reduced risk for coronary restenosis. ThrombHaemost. 1992;67:209-213.

14. Temple RC, Clark PMS, Nagi DK, Schneider AE, Yudkin JS,Hales CN. Radioimmunoassay may overestimate insulin in non-insulin-dependent diabetics. Clin Endocrinol. 1990;32:689-693.

15. Nagi DK, Hendra TJ, Ryle AJ, Cooper TM, Temple RC, ClarkPMS, Schneider AE, Hales CN, Yudkin JS. The relationships ofconcentrations of insulin, intact proinsulin and 32-33 split pro-insulin with cardiovascular risk factors in type 2 (non-insulin-dependent) diabetic subjects. Diabetologia. 1990;33:532-537.

16. Spradlin CT, Galloway JA, Anderson JH. Apparent increase incoronary heart disease with human proinsulin (HPI): random-ization failure or HPI effect? Diabetologia. 1990;33(suppl):A60.Abstract.

60 F

Lg 400

<a 20 r

J, L

-r J1,

0 ]L--- -I-I-T--F--L----1I0 1 510 0 1 510 0 1 510

Control lnsuln Proinsulin

IT,II,



ownloaded from



17. Galloway JA, Hooper SA, Spradlin CT, Howey DC, Frank BH,Bowsher RR, Anderson JH. Biosynthetic human proinsulin: reviewof chemistry, in vitro and in vivo receptor binding, animal andhuman pharmacology studies, and clinical trial experience.Diabetes Care. 1992;15:666-692.

18. Jain SK, Nagi DK, Slavin BM, Lumb PJ, Yudkin JS. Insulintherapy in type 2 diabetic subjects suppresses plasminogenactivator inhibitor (PAI-1) activity and proinsulin-like moleculesindependently of glycemic control. Diabetic Med. 1993;10:27-32.

19. Emeis JJ, Kooistra T. Interleukin 1 and lipopolysaccharide inducean inhibitor of tissue-type plasminogen activator in vivo and incultured endothelial cells. J Exp Med. 1986;163:1260-1266.

20. Chmielewska J, Wiman B. Determination of tissue plasminogenactivator and its 'fast' inhibitor in plasma. Clin Chem. 1986;32:482-485.

21. Kessler SW. Rapid isolation of antigens from cells with a staphy-lococcal protein A-antibody absorbant: parameters of the inter-action of antibody-antigen complexes with protein A. J Immunol.1975;115:1617-1624.

22. Feinberg AP, Vogelstein B. A technique for radiolabeling DNArestriction endonuclease fragments to high specific activity. AnalBiochem. 1983;132:6-13.

23. Thomas PS. Hybridization of denatured RNA and small DNAfragments transferred to nitrocellulose. Proc NatlAcad Sci U S A.1980;77:5201-5205.

24. Kooistra T, Sprengers ED, van Hinsbergh VWM. Rapid inacti-vation of the plasminogen-activator inhibitor upon secretion fromcultured human endothelial cells. Biochem J. 1986;239:497-503.

25. Loskutoff DJ. Regulation of PAI-1 gene expression. Fibrinolysis.1991;5:197-206.

26. Akiyama T, Ishida J, Nakagawa S, Ogawara H, Watanabe S, ItohN, Shibuya M, Fukami Y. Genistein, a specific inhibitor oftyrosine-specific protein kinases. JBiol Chem. 1987;262:5592-5595.

27. Podskalny JM, Takeda S, Silverman RE, Tran D, Carpentier JL,Orci L, Gordon P. Insulin receptors and bioresponses in a humanliver cell line (Hep G-2). Eur J Biochem. 1985;150:401-407.

28. Alessi MC, Juhan-Vague I, Kooistra T, Declerck PJ, Collen D.Insulin stimulates the synthesis of plasminogen activator inhibitor1 by the human hepatocellular cell line Hep G2. Thromb Haemost.1988;60:491-494.

29. Kooistra T, Bosma PJ, Tons HAM, van den Berg AP, Meyer P,Princen HMG. Plasminogen activator inhibitor 1: biosynthesis and

mRNA level are increased by insulin in cultured human hepa-tocytes. Thromb Haemost. 1989;62:723-728.

30. Schneider DJ, Sobel BE. Augmentation of synthesis of plas-minogen activator inhibitor type 1 by insulin and insulin-likegrowth factor type I: implications for vascular disease in hyperin-sulinemic states. Proc Natl Acad Sci U S A. 1991;88:9959-9963.

31. Given BD, Cohen RM, Shoelson SE, Frank BH, Rubenstein AH,Tager HS. Biochemical and clinical implications of proinsulin con-version intermediates. J Clin Invest. 1985;76:1398-1405.

32. Rhodes CJ, Lincoln B, Shoelson SE. Preferential cleavage of des-31,32-proinsulin over intact proinsulin by the insulin secretorygranule type II endopeptidase. JBiol Chem. 1992;267:22719-22727.

33. Sobey WJ, Beer SF, Carrington CA, Clark PMS, Frank BH, GrayIP, Luzio SD, Owens DR, Schneider AE, Siddle K, Temple RC,Hales CN. Sensitive and specific two-site immunoradiometricassays for human insulin, proinsulin, 65-66 split and 32-33 splitproinsulins. Biochem J. 1989;260:535-541.

34. Bowsher RR, Wolny JD, Frank BH. A rapid and sensitive radio-immunoassay for the measurement of proinsulin in human serum.Diabetes. 1992;41:1084-1090.

35. Heding LG. Determination of total serum insulin (IRI) in insulin-treated diabetic patients. Diabetologia. 1972;8:260-266.

36. Misbin RI, Merimee TJ, Lowenstein JM. Insulin removal byisolated perfused rat liver. Am J Physiol. 1976;230:171-177.

37. Bierman EL. Atherogenesis in diabetes. Arterioscler Thromb. 1992;12:647-656.

38. Porte D. P-Cells in type II diabetes mellitus (Banting Lecture1990). Diabetes. 1991;40:166-180.

39. Ward WK, LaCava EC, Paquette TL, Beard JC, Wallum BJ, PorteD. Disproportionate elevation of immunoreactive proinsulin intype 2 (non-insulin-dependent) diabetes mellitus and in experi-mental insulin resistance. Diabetologia. 1987;30:698-702.

40. Temple RC, Carrington CA, Luzio SD, Owens DR, Schneider AE,Sobey WJ, Hales CN. Insulin deficiency in non-insulin-dependentdiabetes. Lancet. 1989;1:293-295.

41. Nordt TK, Klassen KJ, Schneider DJ, Sobel BE. Augmentation ofsynthesis of plasminogen activator inhibitor type-1 in arterial endo-thelial cells by glucose and its implications for local fibrinolysis.Arterioscler Thromb. In press.

42. Klassen KJ, Nordt TK, Schneider DJ, Sobel BE. Constitutivebiosynthesis of plasminogen activator inhibitor type-1 (PAI-1) bycultured human aortic endothelial cells independent of insulin.Coronary Artery Dis. 1993;4:713-719. by guest on M

arch 21, 2018http://circ.ahajournals.org/

Dow

nloaded from


T K Nordt, D J Schneider and B E Sobelinsulin. A potential risk factor for vascular disease.

Augmentation of the synthesis of plasminogen activator inhibitor type-1 by precursors of

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1994 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.89.1.321

1994;89:321-330Circulation.

http://circ.ahajournals.org/content/89/1/321World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://circ.ahajournals.org/content/89/1/321

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


Documents

Activator Inhibitor Type-1 by Precursors of Insulin