|
|
||||||||
Metabolic Research Group, VA Medical Center and University of Kentucky, Lexington, Kentucky 40511
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
The critical role for oxidized LDL in atherosclerosis is reviewed elsewhere (7-10). Formation of fatty streaks within arterial walls is initiated by oxidation of LDL occurring predominantly within the subendothelial space (SES). Lipoprotein lipids and apoproteins are oxidized within the SES as a result of exposure to the free radicals produced by arterial wall cells (11). Oxidized LDL are taken up by macrophages that eventuate as foam cells, precursors of fatty streaks (11, 12).
Twenty years ago Zilversmit (13) suggested that atherosclerosis is a postprandial phenomenon with lipoprotein particles in the postprandial state contributing to atherosclerosis. Because of consumption of regular meals throughout the day, most western people are predominantly in a postprandial state. Most studies of lipoprotein metabolism, however, have focused on the fasting state because it is thought to be more reproducible and better-defined baseline for metabolic studies (14). Studies of LDL oxidation have also concentrated on the fasting state although postprandial lipemia may exhibit greater atherogenic potential via increases in the postprandial plasma triglyceride levels (15). Levels of triglycerides in the LDL fraction are increased in the postprandial state (14, 16) which, in turn, increases the susceptibility of LDL to oxidative modification (17).
Data on LDL oxidation in the postprandial state are still limited, and kinetics of LDL oxidation in the diabetic condition are not well delineated. The present study was undertaken to determine whether the postprandial LDL from diabetic subjects exhibits alterations that might predispose them to accelerated atherosclerosis as compared to controls. The kinetics of LDL oxidation in fasting and postprandial states of diabetic and control subjects were compared.
These comparisons suggested that oxidation of LDL from the postprandial state of diabetic subjects was accelerated compared to their fasting state. Furthermore, LDL from diabetic subjects in the postprandial state appeared to be oxidized more than postprandial LDL from control subjects. The results suggest that the deviant pattern of LDL oxidation displayed in the postprandial state of diabetic subjects may be a critical contributor to the cardiovascular risks in these individuals.
| Materials and Methods |
|---|
|
|
|---|
Test Meal.
All subjects were asked to eat a light supper on the day before the study and abstain from alcohol consumption. Following a 12-hr overnight fast, subjects came in as outpatients and had their first phlebotomy of 50 ml of blood. They they were fed a test meal containing 56 g of cornflakes, 250 ml of milk, 250 ml of cream, and 10 g of granulated sugar. Subjects were asked to eat this meal as quickly as possible. The meal provided 960.3 kilocalories and 84.7 g of fat. Six hr after the meal, subjects returned for their second phlebotomy of 50 ml of blood. During the 6 hr after the meal, subjects were instructed to consume only water, coffee, or tea without sugar or cream. The two blood samples were used for the analysis of glucose, glycohemoglobin, and serum lipids, and for lipoprotein isolation. Postprandial studies were scheduled so that one control subject and one or two diabetic subjects had tests on the same day. This ensured that all diabetes samples were processed concurrently with control samples.
Serum Measurements.
The Lipid Research Laboratory performed all serum measurements at the VA Medical Center, Lexington, KY. The Abbott Quick-Start glucose kit (Abbott Laboratories, North Chicago, IL) was used to analyze serum glucose concentrations. Glycosylated hemoglobin (HbA1c) concentrations were determined by the Abbott IMxR glycated hemoglobin test (Abbott Laboratories, normal values, 4.2%6.4%). The Abbott Spectrum analyzer was used for the analysis of all the serum lipid concentrations (Abbott Laboratories). LDL-cholesterol concentration was determined by the Friedewald formula (18). LDL-cholesterol concentrations were not calculated if serum triglyceride values were >4.5 mM.
Isolation of LDL Fraction by Ultracentrifugation.
A 50-ml sample of whole blood was collected in tubes containing 0.1% EDTA; plasma was obtained by centrifugation at 2500 rpm for 15 min. Lipoprotein fractions were isolated by a modification of previously described techniques (19). Plasma density was raised to 1.09 g/ml by the addition of KBr. The adjusted plasma was placed in Beckman ultracentrifuge tubes along with saline EDTA (0.1% EDTA) and ultracentrifuged at 50,000 rpm at 4°C for 11.25 hr using a Vti 60 rotor in a Beckman model L8-80 ultracentrifuge. The VLDL-LDL fraction recovered by tube slicing was adjusted to a density of 1.3 g/ml, layered under 27 ml of saline EDTA (0.1% EDTA), and ultracentrifuged at 50,000 rpm for 2.5 hr at 4°C. The LDL fraction was identified visually and obtained by slicing the tubes at the top third portion. Following slicing, each recovered LDL fraction was dialyzed extensively in 4 liters of phosphate buffer saline (PBS) pH 7.4 to remove the EDTA and salts. The dialyzed LDL fractions were stored under N2 at 4°C until further assays were performed.
Measurement of Oxidation Parameters.
Oxidation studies were performed on isolated and dialyzed LDL fractions from all subjects. Methods used for the estimation of the different oxidation parameters are as follows.
In vitro copper oxidation at 37°C.
Susceptibility of LDL to in vitro oxidation was assessed using copper as a pro-oxidant. Protein content of isolated LDL samples was estimated by the Lowry method (20). Susceptibility of LDL fraction to in vitro oxidation was assessed by incubating 50 µg of LDL protein with 5 µM Cu in a final volume of 1 ml of PBS. LDL oxidation was continuously monitored at 37°C for 2 hr, and conjugated dienes (CD), formed as a result of the oxidation, were measured every 5 min at 234 nm (21) in a temperature-controlled Spectronic Genesys spectrophotometer. After 2 hr, oxidation was stopped by addition of 25 µl of 40 mM EDTA and 40 µl of 1 mM BHT. The lag phase (minutes) was determined from the oxidation curves obtained, as an intercept of the slope of the oxidation curve and the baseline (21).
Measurement of lipid peroxides and thiobarbituric acid reactive substances formation during LDL oxidation.
Formation of lipid peroxides during LDL oxidation was measured spectrophotometrically at 365 nm based on the method described by El-Saadani et al. (22). Lipid peroxides formed during LDL oxidation were quantified using the molar absorptivity of iodine measured at 365 nm, which is 2.46 x 104 M-1 cm-1. Thiobarbituric acid reactive substances (TBARS) were estimated by the method of Kosugi and Kikugawa (23).
Estimation of the percentage reduction in the free amino groups.
The percentage reduction in amino groups as a consequence of derivatization of the lysine residues during oxidation was determined by modification of the method described by Steinbrecher (24). The amount of amino groups remaining following oxidation was estimated based on reactivity of the free amines with trinitrobenzene sulfonic (TNBS) acid. Oxidized LDL, 25 µg was incubated with 1 ml of 4% NaHCO3 and 50 µl of 0.1% TNBS in a shaking water bath at 37°C. After 1 hr, the reaction was stopped with 100 µl of 1 N HCl and 100 µl 10% sodium dodecyl sulfate, and absorbance was measured at 340 nm. Native LDL sample, 25 µg, was treated in the same way and from this the percentage reduction in the free amines was determined. Valine was used as a standard.
Electrophoresis.
Relative electrophoretic mobility (REM) of all samples was determined by agarose gel electrophoresis as described by Noble (25). A 1% agarose gel was placed in an electrophoresis cell immersed in 0.05 M barbital buffer (pH 8.6). Samples were mixed with saturated sucrose and then loaded onto the gel. The gel was run at 90 volts for 45 min, following which it was fixed in 5% acetic acid in 75% ethanol for 30 min. After another 30 min of drying, the gels were stained with Sudan Black until well-stained distinct bands were obtained. Migration of all samples was measured as distance traveled from origin to center of the lipoprotein band. The REM of the oxidized samples was determined by dividing the mobility of modified LDL by mobility of native unoxidized LDL.
A sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed to determine purity of isolated apo B-100 LDL fraction. A 4%20% polyacrylamide gradient gel with 0.1% SDS was cast along with a 3% stacking gel. Samples were boiled with SDS containing ß-mercaptoethanol and loaded into wells at a concentration of 5 µg/well. Electrophoresis was carried out at 200 volts (constant voltage) for 45 min. Samples were stained with Coomasie blue stain (0.8%) for 30 min and then destained for 1 hr. Along with LDL fractions, a molecular marker was used as a standard.
Statistical Analysis.
All results have been expressed as mean ± SEM. Single factor analysis of variance (ANOVA) test was used to analyze differences statistically between diabetic subjects and controls. The paired t test was used for analyses within the group, in the fasting and the postprandial states. Additionally, several parameters were tested for correlations using regression analysis. Significance of comparative tests was determined by P values. The r values were used to determine correlations between parameters. A two-tailed P value
0.05 was considered significant. Statistical analyses were performed using Microsoft Excel (Version 5.0, Microsoft, Redmond, WA).
| Results |
|---|
|
|
|---|
-blockers (one subject), and anti-inflammatory agents (one subject).
Diabetic subjects were significantly older (P < 0.001) than controls and also had a significantly higher BMI (P = 0.017). They had very poor glycemic control; serum glucose (P < 0.001) and glycohemoglobin concentrations (P < 0.001) were significantly higher than values for control subjects (Table I)
.
|
The most significant difference between diabetic and control subjects was in the serum triglyceride concentrations. Whereas all diabetic subjects had serum triglyceride concentrations <3.4 mM during screening, at the time of the investigation, five subjects had serum triglyceride concentrations >3.4 mM (range: 3.55.8 mM). Because of the limited number of diabetic subjects, LDL oxidation data from the hypertriglyceridemic subjects were included in the analysis. Serum triglyceride concentrations were significantly higher in diabetic than control subjects in both fasting (P = 0.004) and postprandial samples (P < 0.001). Postprandially, the increments in serum triglyceride concentrations were significant for the control (P = 0.007) and diabetic subjects (P = 0.002). Postprandial triglyceride concentrations were 65% higher than fasting concentrations in control subjects and 55% higher than in diabetic subjects.
Oxidation Studies.
Figure 1
illustrates the LDL oxidation kinetics for the four groups of subjects. Rates of CD formation were similar for fasting and postprandial LDL samples from control subjects although CD values were higher at 120 min for fasting samples. The CD concentration in the fasting samples from diabetic subjects was similar to control fasting samples. There were no differences in the concentration of CD in the postprandial samples from diabetic subjects when compared to postprandial samples from control subjects (Table II)
. The lag times cannot be estimated from Figure 1
because averaging values changes the slope of the curve; lag times provided in Table II
represent the average of individual lag times.
|
|
In control subjects, there were no significant differences in lag phase between fasting and postprandial samples. In diabetic subjects, the lag phase decreased significantly from 40 min in fasting samples to 33 min in the postprandial samples (P = 0.003). Comparison between control and diabetic subjects in the fasting state showed that diabetic subjects had a shorter lag phase compared to controls although this difference was not significant. However, in the postprandial state, lag phase was significantly shorter in diabetic subjects than in controls (P = 0.005).
In control subjects, there were no significant differences in lipid peroxide values between fasting and postprandial states. In diabetic subjects, also, there were no significant differences in lipid peroxide values between fasting and postprandial states. Comparison between control and diabetic subjects in fasting and postprandial states showed that diabetic subjects had similar levels of lipid peroxides compared to control subjects (Table II)
.
In control subjects, there were no significant differences in TBARS values between fasting and postprandial states. However, in diabetic subjects, TBARS values were significantly higher in postprandial samples compared to fasting samples (P = 0.006). Comparing control and diabetic subjects, TBARS values were slightly but not significantly higher in diabetic subjects than in controls with fasting state (P = 0.5) and were significantly higher in the postprandial state (P = 0.006). Despite variations in the TBARS observed, the percentage reduction in the free amino groups was very similar in diabetic and control subjects in the fasting and postprandial states.
Isolated LDL fractions were subjected to SDS-PAGE to ascertain the purity of the samples. The unoxidized LDL sample from control and diabetic subjects, fasting and postprandial, represented a single band indicating that the isolated fractions contained only LDL particles (data not shown).
In control subjects, the REM of oxidized fasting samples was similar to the REM of oxidized postprandial samples. In diabetic subjects, as well, the REM of the oxidized fasting and postprandial samples did not differ significantly. Comparison between control and diabetic subjects showed that REM of oxidized samples form diabetic subjects in the fasting and postprandial states were greater than REM of oxidized samples from controls although the differences were not significant (Table II)
.
Figure 2
illustrates results of regression analysis between serum triglyceride concentrations and lag phase in control and diabetic subjects. In diabetic subjects, serum triglyceride concentrations showed a negative correlation with lag phase in fasting (r = -0.53, P = 0.06) and postprandial states (r = -0.80, P = 0.002). In controls, this correlation was absent in both the fasting (r = 0.01, P = 0.97) and postprandial states (r = 0.027, P = 0.93). In diabetic subjects, there was a negative correlation between serum cholesterol concentrations and lag phase in both the fasting (r = -0.41, P = 0.16, not significant) and postprandial states (r = -0.68, P = 0.01). In control subjects, however, we did not observe a significant correlation between these two parameters in the fasting or postprandial state. We also observed that fasting serum triglyceride concentrations correlated significantly with postprandial values in diabetic subjects (r = 0.89, P < 0.001). Age was negatively correlated to lag phase in fasting diabetic subjects (P = 0.034) but not in postprandial diabetic subjects or control subjects (fasting or postprandial).
|
| Discussion |
|---|
|
|
|---|
We did not quantitate the lipid and apoprotein concentrations of LDL from these subjects. We hypothesize that higher serum triglyceride levels in diabetic subjects may have contributed to alterations in LDL oxidation. With hypertriglyceridemia, LDL from diabetic individuals is triglyceride-enriched (28) and has profound changes in core and surface composition (29). Such triglyceride-rich LDL particles, in turn, have a proneness to oxidative modification (17). Increased triglyceride content can increase the fluidity of LDL particles (30) and, therefore, accessibility of oxidants to neutral core lipids within the LDL particle. Although we did not assess LDL compositional changes from diabetic or control subjects, we demonstrated a negative correlation between serum triglycerides and lag phase in diabetic subjects in fasting and postprandial states, suggesting that the elevated serum triglyceride levels in diabetic subjects may affect oxidation of LDL. This relationship however was not seen in control subjects.
Our results, consistent with those of Regnstrom et al. (17), also revealed a significant inverse correlation between serum cholesterol concentrations and lag phase in the postprandial state of diabetic subjects, suggesting that a rise in the serum cholesterol concentrations may affect the lag phase of LDL oxidation. Oxidation of LDL as related to cholesterol content is indicated to be indirect through a common link to polyunsaturated fatty acids (PUFA) (31). The effect of serum cholesterol levels may be mediated via the increased content of PUFA in the cholesterol ester fractions in the LDL cholesterol, which has also been demonstrated to be increased in the LDL fractions of the diabetic subjects (28).
Increased serum triglyceride levels and decreased HDL cholesterol are frequently accompanied by the presence of small dense LDL (32, 33). Serum triglyceride concentrations have, in fact, been reported to be major determinants of the properties of LDL in NIDDM (28, 32, 34). The elevated levels of serum triglycerides in our diabetic subjects could have induced formation of small dense LDL (32, 34, 35). Lewis et al. (35) documented that fasting hypertriglyceridemia in NIDDM is an important predictor of postprandial lipids and lipoprotein abnormalities with potential atherogenic effects. We speculate that one such abnormality in the LDL fractions induced by the hypertriglyceridemia in our diabetic subjects may have been a decrease in the particle size, causing acceleration in the oxidative process (36).
Our study provides a basis for further investigations that might furnish important answers about the existence of the phenomenon of accelerated oxidation in diabetic subjects. If these findings are confirmed and extended to other sex and age groups in the diabetic population, LDL oxidation would need to be taken into account as a critical risk factor for coronary heart disease complications in hypertriglyceridemic diabetic subjects.
| Footnotes |
|---|
1 To whom requests for reprints should be addressed at Medical Service, 111C, 2250 Leestown Road, Lexington, KY 40511. E-mail: jwandersmd{at}aol.com ![]()
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. Peter and A. Rees Review: Postprandial glycaemia and cardiovascular risk The British Journal of Diabetes & Vascular Disease, January 1, 2008; 8(1): 8 - 14. [Abstract] [PDF] |
||||
![]() |
P. N Bavenholm and S. Efendic Postprandial hyperglycaemia and vascular damage - the benefits of acarbose Diabetes and Vascular Disease Research, September 1, 2006; 3(2): 72 - 79. [Abstract] [PDF] |
||||
![]() |
A. Ceriello Postprandial Hyperglycemia and Diabetes Complications: Is It Time to Treat? Diabetes, January 1, 2005; 54(1): 1 - 7. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Beckman, A. B. Goldfine, M. B. Gordon, L. A. Garrett, J. F. Keaney Jr., and M. A. Creager Oral antioxidant therapy improves endothelial function in Type 1 but not Type 2 diabetes mellitus Am J Physiol Heart Circ Physiol, December 1, 2003; 285(6): H2392 - H2398. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. J. Sampson, N. Gopaul, I. R. Davies, D. A. Hughes, and M. J. Carrier Plasma F2 Isoprostanes: Direct evidence of increased free radical damage during acute hyperglycemia in type 2 diabetes Diabetes Care, March 1, 2002; 25(3): 537 - 541. [Abstract] [Full Text] [PDF] |
||||
![]() |
A Liguori, P Abete, J.M Hayden, F Cacciatore, F Rengo, G Ambrosio, D Bonaduce, M Condorelli, P.D Reaven, and C Napoli Effect of glycaemic control and age on low-density lipoprotein susceptibility to oxidation in diabetes mellitus type 1 Eur. Heart J., November 2, 2001; 22(22): 2075 - 2084. [Abstract] [PDF] |
||||
![]() |
R. C. Ruhe and R. B. McDonald Use of Antioxidant Nutrients in the Prevention and Treatment of Type 2 Diabetes J. Am. Coll. Nutr., October 1, 2001; 20(90005): 363S - 369. [Abstract] [Full Text] |
||||
![]() |
J. W. Anderson, E. C. Konz, and D. J. A. Jenkins Health Advantages and Disadvantages of Weight-Reducing Diets: A Computer Analysis and Critical Review J. Am. Coll. Nutr., October 1, 2000; 19(5): 578 - 590. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |