Atherosclerosis and Lipoproteins |
From the CNR Institute of Clinical Physiology and Department of Internal Medicine, University of Pisa, Pisa, Italy, and Istituto di Semeiotica e Nefrologia Medica (U.G., A.F.-P.), University of Verona, Verona, Italy.
Correspondence to Ele Ferrannini, MD, CNR Institute of Clinical Physiology, Via Savi 8, 56126 Pisa, Italy. E-mail ferranni{at}ifc.pi.cnr.it
| Abstract |
|---|
|
|
|---|
Key Words: insulin LDL cholesterol oxidation free radicals
| Introduction |
|---|
|
|
|---|
In vitro studies have shown that insulin induces
H2O2 formation in human
peritoneal adipocytes, macrophages, and brain stem
cells.11 12 13 In suspended human fat cells,
physiological amounts of insulin stimulate
H2O2 release via activation
of a membrane-bound dehydrogenase, an effect that is still seen in
isolated adipocyte plasma membranes and is not prevented by blockade of
the insulin receptor tyrosine kinase.11 The latter
findings have led to the idea that hydrogen peroxide generation may be
an alternative insulin-signaling pathway in adipocytes. In diabetic
rats, the production of free radicals is stimulated by the
intraperitoneal administration of
insulin.14 Recent clinical studies have demonstrated that
the urinary excretion of 8-iso-prostaglandin
F2
, a bioactive product of
arachidonic acid peroxidation, is increased in patients
with diabetes and that this abnormality is improved by vitamin E
supplementation.15 In studies carried out in our
laboratory in healthy volunteers, acute insulin administration induced
a decrement in total plasma vitamin E concentrations.16
Because vitamin E is the main tissue scavenger of free
radicals,17 one interpretation of that finding was that
acute insulin administration would enhance oxidative stress. However,
because other parameters of free-radical activation were
not concomitantly measured, an alternative explanation was that the
reduction of circulating vitamin E levels may be due to accelerated
transport from the intravascular to the extravascular space, which is
analogous to the insulin effect on LDL cholesterol
transport.18
The present study was therefore undertaken to establish whether acute euglycemic hyperinsulinemia within the physiological range has a pro-oxidant activity in vivo and whether such an effect may be related to insulin sensitivity.
| Methods |
|---|
|
|
|---|
|
The study protocol was approved by the Institutional Ethics Committee, and the purpose, nature, and risks involved in the study were explained to all subjects before obtaining their written consent to participate.
Protocol
All subjects were studied at 9 AM after an overnight
(10- to 12-hour) fast. A 20-gauge polyethylene catheter was inserted
into an antecubital vein for infusion of test substances. A wrist vein
was cannulated retrogradely with another catheter, and the hand was
placed in a warming box (60°C) for arterialized blood
sampling. After 30 minutes of stabilization, saline was infused for 120
minutes (time-control period). Next, insulin was infused at the rate of
7 pmol · min-1 ·
kg-1 (1 mU ·
min-1 · kg-1)
while maintaining plasma glucose constant at the basal level by the
euglycemic insulin clamp technique.19 Briefly,
4 minutes after starting exogenous insulin, an infusion of glucose (as
a 20% D-glucose solution) was begun at an initial rate of
10 µmol · min-1 ·
kg-1; thereafter, the glucose infusion rate was
adjusted every 5 minutes on the basis of on-line measurements of plasma
glucose concentrations by using a computerized
algorithm.19 Before the start of the study and at timed
intervals during the clamp, arterialized blood samples were
obtained for the measurement of plasma glucose and insulin levels. At
times -120, 0, and 120 minutes, arterialized blood samples
were drawn for the measurement of the following variables: lipid
profile, vitamin E content in LDL, cell-mediated malondialdehyde (MDA)
generation in LDL, and the susceptibility of LDL
cholesterol to in vitro copper sulfateinduced
oxidation.
Procedures
For lipoprotein separation, whole blood was collected into
Vacutainer tubes (Becton Dickinson) containing EDTA (1 mg/mL) and
immediately centrifuged at 2000 rpm for 20 minutes at 4°C.
Plasma was stored at 4°C and processed for lipoprotein separation
within 1 day. Lipoprotein was isolated by sequential
ultracentrifugation in NaBr solution20
containing EDTA (1 mg/mL) and stored at 4°C. To minimize the
possibility of LDL oxidation during isolation, all solutions used in
this process were deoxygenated by argon bubbling. LDL was
stored under nitrogen at 4°C in a sterile dark environment and
processed within 1 day.
Susceptibility of LDL to oxidation by copper was evaluated by measuring the length of the lag phase on the basis of the development of fluorescence during copper-catalyzed LDL oxidative modification, as previously described.21 High-performance liquid chromatography was used to measure vitamin E content in LDL cholesterol, as previously described.22 Cell-mediated LDL oxidative modification was also determined. Human umbilical vein endothelial cells were isolated as described22 and used at passages 2 to 4. LDL oxidation was prepared by adding 1.5 mL of serum-free F-12 medium containing 200 µg/mL protein to each well of human umbilical vein endothelial cells and incubating at 37°C for 24 hours. The extent of LDL oxidation was determined by measuring the concentration of thiobarbituric acidreactive substances, as described.23 Briefly, aliquots of the incubation mixture containing 200 µg LDL were removed and added to tubes containing 0.05 mL of 2% butylated hydroxytoluene, 2 mL of 0.67% thiobarbituric acid, and 10% trichloroacetic acid (2:1). The tubes were heated at 100°C for 10 minutes and then cooled and centrifuged at 2500 rpm for 10 minutes. The absorbance of the supernatant fraction was read at 532 nm, and the quantification was achieved by comparison with a standard curve of MDA equivalents generated by acid-catalyzed hydrolysis of 1,1,3,3-tetraethoxypropane.23 All measurements were made in duplicate.
Cholesterol and triglyceride levels in plasma were determined by Technicon Autoanalyzer II methodology. Plasma glucose was measured by the glucose oxidase method with a glucose analyzer (Beckman Instruments), and insulin was measured by radioimmunoassay (Linco Research).
Data Analysis
Because previous studies have demonstrated that hepatic glucose
production is fully suppressed during the second hour of an
insulin clamp with the insulin dose used in the present
studies,19 whole-body glucose utilization (insulin
sensitivity) was calculated from the infusion rate of exogenous glucose
after correction for changes in glucose levels in a distribution volume
of 250 mL · kg-1. Glucose disposal rates
were averaged over the second hour of the euglycemic clamp
study, during which nearsteady-state conditions prevailed, and
expressed per kilogram of lean body mass.
Data are given as the mean±SEM; LDL oxidation parameters and vitamin E concentrations are given as median and interquartile range. Mean group values were compared with the use of the Wilcoxon signed rank test (2 groups) or Friedman test (3 groups). Simple and multiple regression analyses were carried out by standard methods.
| Results |
|---|
|
|
|---|
|
Lag phase, MDA generation, and LDL vitamin E content were all interrelated at baseline (lag phase versus vitamin E, r=0.73, P<0.0001; lag phase versus MDA, r=0.85, P<0.0001; and vitamin E versus MDA, r=0.69, P=0.0004). Furthermore, the lag phase was significantly shorter in subjects with higher mean blood pressure levels (r=0.42, P<0.05).
The lag phase of in vitro LDL oxidation was unchanged after 2 hours of
saline infusion, whereas it decreased significantly after insulin
infusion (Table 3
). This effect of
insulin was seen in 20 of 23 subjects (Figure
). Likewise, MDA
generation and LDL vitamin E content were not altered by saline
infusion, whereas MDA generation was significantly increased and LDL
vitamin E content was significantly decreased at the end of the clamp
(Table 3
). The observed changes in oxidative
parameters were consensual. Thus, the reduction in lag
phase after insulin was directly related to the decrement of LDL
vitamin E content (r=0.55, P<0.01) and tended to
be related to the rise in MDA generation (r=0.36,
P=0.09). A greater insulin-induced reduction in lag phase
was associated with higher fasting triglyceride levels both
in whole plasma and in the VLDL fraction (r=0.46,
P<0.03 for both). When the study group was subdivided
according to the median serum triglyceride concentration,
the insulin-induced decrease in lag phase was directly related
(P=0.01) to baseline serum triglycerides in the
11 subjects with higher triglyceride levels
(2.04±0.28 mmol/L) but not in those with lower
triglyceride levels (0.84±0.08 mmol/L,
P=NS). In contrast, none of the insulin-induced changes in
oxidative parameters was related to insulin sensitivity or
any other anthropometric or metabolic variable.
|
|
| Discussion |
|---|
|
|
|---|
700 pmol/L) that we used in these
clamp studies. It must be noted that insulin promotes LDL cholesterol uptake by upregulating LDL receptors.18 If large, more buoyant LDL particles were preferred to small, dense LDL molecules in such a process, insulinization would be associated with an enrichment of plasma with small, dense LDL particles, intrinsically more susceptible to oxidation.25 26 In the present study, insulin did cause a small decrease in serum LDL concentrations; in addition, the possibility that insulin may cause an in vivo selection of lipoprotein has been recently documented in studies measuring LDL size after an oral glucose load.27
Higher serum triglyceride levels are generally associated with a predominance of small, dense LDL.26 In the present study, the lack of correlation between fasting triglyceride levels and LDL lag phase could be due to the relatively small sample size or could reflect the fact that very few subjects were hypertriglyceridemic (only 3 subjects >2.3 mmol/L) or could be a combination of these 2 factors. On the other hand, the presence of higher triglycerides, in whole serum or the VLDL fraction, was associated with an enhanced effect of insulin, suggesting that higher triglycerides do signal an increased susceptibility of LDL to oxidative stimuli. Insulin also induced a small decrement (4% on average) in circulating triglyceride levels. Thus, one could surmise either that the insulin-induced shortening of the lag phase would have been even larger had the triglycerides remained unchanged or that the kinetics of the insulin effect was faster than any secondary effect mediated by changes in triglycerides. In general, the nature of the relation between triglycerides and LDL oxidizability is imperfectly understood. On the one hand, when plasma triglycerides are reduced with intensive exercise and dieting or by fibrate therapy, in vitro LDL cholesterol oxidizability is reduced.28 On the other hand, recent studies have shown that if triglyceride levels are acutely increased (by an infusion of Intralipid [triglyceride emulsion]), the susceptibility of LDL cholesterol to oxidation is decreased rather than increased.29 From these results, it has been suggested that enriching the LDL molecule with triglycerides reduces its oxidizability, whereas the subsequent hydrolysis of triglycerides within LDL can lead to atherogenic changes. Collectively, these data suggest that one should distinguish between the acute and chronic effects of lipid exchange between lipoprotein particles on their susceptibility to oxidation. Clearly, additional work is needed to gain insight into these mechanisms.
The insulin effect observed in the present studies was small in
size (averaging 7%). However, the experiment was designed to test
whether insulin alone (ie, without hyperglycemia) at
physiological concentrations had any acute effects
on LDL oxidizability. On the other hand, chronic
hyperinsulinemia, such as prevails in
insulin-resistant individuals, may translate into a larger
insulin pro-oxidant effect. In this respect, it is relevant that
treatment of type 2 diabetic patients with troglitazone, a
thiazolidinedione compound with an
-tocopherol moiety,
leads to an improvement in insulin sensitivity (and plasma insulin
levels) and a reduction in LDL susceptibility to
oxidation.23 The extent to which the ex vivo measurements
of LDL cholesterol oxidation reflect the in vivo oxidative
damage of lipoproteins has been generally uncertain. Although this
uncertainty has not prevented a wide application of the methodology to
a number of clinical conditions,26 30 31 the presence of
oxidized LDL in vivo has recently been demonstrated in human serum by
NMR spectroscopy.32 Importantly, the results of this assay
agreed well with the simultaneously measured in vitro LDL
oxidizability, indicating that the latter does reflect a biological
phenomenon.
A series of studies using human aortic intimal cells was reviewed by Sobenin et al,30 who demonstrated that sera from type 2 diabetic patients show increased atherogenic properties, as measured by the ability of LDL cholesterol to accumulate in these cells. The modifications to the LDL molecule included nonenzymatic glycosylation and loss of sialic acid. Some of these atherogenic modifications of the LDL cholesterol particle can be induced by an excessive activity of free radicals.33 In vivo, hyperglycemia has been invariably regarded as the principal culprit of the increased free radical production observed in type 2 diabetic patients.34 35 In line with this notion, in type 2 diabetic patients the total plasma antioxidant capacity, erythrocyte lipid peroxidation activity, the susceptibility of LDL cholesterol to oxidation,36 37 38 and the excretion of oxidized products of arachidonic acid15 are all increased, in proportion to the severity of hyperglycemia.39 40 Even in nondiabetic subjects, the relative hyperglycemia induced by an oral glucose tolerance test stimulates the production of free radicals.41 Furthermore, increased reactive oxygen species activity has been documented in hyperinsulinemic nonhyperglycemic conditions, such as obesity,42 essential hypertension,5 smoking,3 and dyslipidemia.43 Also, a long latency between the histological evidence of atherosclerosis and the onset of hyperglycemia in type 2 diabetes is the rule. Atherogenesis is likely to set in when the only abnormality (in future type 2 diabetic or hypertensive patients) is insulin resistance/hyperinsulinemia and its metabolic correlates.44 Activation of peroxidation (lipid-conjugated dienes and MDA) is already present in the initial lesion of atherosclerosis, the fatty streak.45
In conclusion, insulin at physiological doses is associated with increased LDL peroxidation independent of the presence of hyperglycemia. This result supports the possibility that insulin resistance may be implicated in the genesis of the early atherosclerotic lesions (and/or the progression of atherosclerosis) through the attendant hyperinsulinemia.
Received April 22, 1999; accepted May 31, 1999.
| References |
|---|
|
|
|---|
and
platelet activation in diabetes mellitus: effects of improved
metabolic control and vitamin E supplementation.
Circulation. 1999;99:224229.This article has been cited by other articles:
![]() |
T. Weinbrenner, H. Schroder, V. Escurriol, M. Fito, R. Elosua, J. Vila, J. Marrugat, and M.-I. Covas Circulating oxidized LDL is associated with increased waist circumference independent of body mass index in men and women Am. J. Clinical Nutrition, January 1, 2006; 83(1): 30 - 35. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Fortuno, G. San Jose, M. U. Moreno, O. Beloqui, J. Diez, and G. Zalba Phagocytic NADPH Oxidase Overactivity Underlies Oxidative Stress in Metabolic Syndrome Diabetes, January 1, 2006; 55(1): 209 - 215. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1999 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |