© 1999 American Heart Association, Inc.
Atherosclerosis and Lipoproteins |
Dyslipidemia and Vascular Dysfunction in Diabetic Pigs Fed an Atherogenic Diet
From the Department of Food Science and Human Nutrition, College of Agriculture, Food and Natural Resources (J.L.D., J.D.S., G.A.W.), the Department of Veterinary Biomedical Sciences, School of Veterinary Medicine (M.H.L.), the Department of Biochemistry, College of Agriculture, Food and Natural Resources (G.A.W.), and the Dalton Cardiovascular Research Center and Department of Physiology, School of Medicine (J.L.P., M.H.L., M.S.), University of Missouri, Columbia.
Correspondence to Joseph L. Dixon, PhD, Dalton Cardiovascular Research Center, 122 Eckles Hall, University of Missouri, Columbia, MO 65211. E-mail dixonj{at}missouri.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—Diabetic patients typically have not only hyperglycemia but also dyslipidemia. Study of the pathogenic components of the diabetic milieu and mechanisms of accelerated atherosclerosis is hindered by inadequate animal models. A potentially suitable animal model for human diabetic dyslipidemia is the pig, because it carries a large fraction of total cholesterol in low-density lipoprotein (LDL), similar to humans. In this study, male Sinclair miniature pigs were made diabetic by destroying the insulin-producing cells of the pancreas with alloxan and then were fed a high fat and high cholesterol diet for comparison with pigs fed a nondiabetic high fat and high cholesterol diet and control pigs. Diabetic pigs exhibited hyperglycemia, but plasma urea nitrogen, creatinine, and transaminase levels were in the normal range, indicating no adverse effects on kidney and liver function. The lipoprotein profile in diabetic pigs was similar to that found in human diabetic patients and was characterized by hypertriglyceridemia (2.8-fold increase versus control and high fat–fed pigs) and a profound shift of cholesterol distribution into the LDL fraction (81%) versus the distribution in high fat–fed (64%) and control (57%) pigs. LDL particles were lipid-enriched and more heterogeneous in diabetic pigs. Apolipoprotein B was distributed among a much broader spectrum of LDL particles, and apolipoprotein E was partially redistributed from high-density lipoprotein to apolipoprotein B–containing lipoproteins in diabetic pigs. There was little change in apolipoprotein A-I distribution. Diabetic pigs showed several early signs of excess vascular disease. In diabetic pigs, 75% of the coronary artery segments showed contractile oscillations in response to prostaglandin F2
compared with 25% in high
fat–fed pigs and 10% in control pigs.
Endothelium-dependent relaxation of brachial arteries
was nearly abolished in diabetic pigs but unchanged in high fat–fed
versus control pigs. Carotid artery Sudan IV staining for fatty streaks
was significantly increased only in diabetic pigs. This porcine model
should provide insights into the etiology of human diabetic
dyslipidemia and facilitate study of peripheral
vascular and coronary artery disease in diabetic patients.
Key Words: Sinclair miniature swine • animal model • lipids • VLDL • LDL • HDL • cholesterol • triglycerides • endothelium • vascular smooth muscle • atherosclerosis • coronary arteries
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Several comprehensive reviews emphasize the severity of vascular disease in diabetes.1 2 3 4 Diabetes mellitus is the most widespread disease in industrialized nations, afflicting nearly 6% of the population in the United States.2 Large-scale clinical and epidemiological studies indicate that diabetes increases the risk of developing cardiovascular disease 2- to 6-fold1 3 and that 80% of all type 2 diabetics will die of an atherosclerotic event.4 Although the association between excess macrovascular disease and prolonged diabetes has been well documented, pathophysiological mechanisms underlying this relation are not clear.
There is debate about the relative roles of hyperglycemia and dyslipidemia in the excess coronary artery disease (CAD) associated with diabetes.5 Whereas the Diabetes Control and Complications Trial (DCCT) indicated that blood glucose is highly predictive of microvascular disease,6 the contribution of all the commonly measured risk factors can explain no more than 25% of the excess macrovascular CAD associated with diabetes.7 Recent reviews indicate the need for reexamination of traditional hypotheses,8 and increasing attention is being paid to altered plasma lipoprotein profile in the excess atherosclerosis associated with diabetes. Importantly, the plasma lipoprotein profile may be most critical,9 10 11 because at any total cholesterol level, diabetic individuals have 3- to 5-fold higher CAD mortality rates than do nondiabetic individuals.10
Type 2 diabetic patients typically do not have increased levels of LDL cholesterol.11 12 Instead, a major characteristic of the dyslipidemia of type 2 diabetes is hypertriglyceridemia,11 12 13 14 reflecting increased levels of VLDL triglyceride.11 14 The dyslipidemic profile also includes increased VLDL remnants, increased apoE in VLDL, increased small dense LDL,15 glycation of LDL,9 and decreased plasma HDL concentration.14 There is virtually uniform agreement that the highly abnormal lipid profile (diabetic dyslipidemia) in type 1 and type 2 diabetes should be highly atherogenic,1 8 9 14 16 17 but identification of the most critical components of the diabetic milieu that elicit excess vascular disease has been elusive.
It has long been appreciated that identification of pathogenic components of the diabetic milieu and our understanding of the mechanisms of excess vascular disease in diabetes have been limited by the lack of a suitable animal model.1 18 19 We have chosen the porcine model because it is widely accepted that pigs possess a cardiovascular system (particularly a coronary circulation) very similar to that of humans20 21 and that the chronic adaptations of the porcine coronary circulation in experimental CAD models are similar to those of human CAD patients.22 23 Importantly, the pig is a good model in which to study lipoprotein metabolism associated with hyperlipidemic diets,24 25 26 and alloxan- or streptozocin-treated pigs have been suggested by others to be a good animal model for metabolic studies of diabetes.27 28 Because previous studies in our laboratory showed that there were only minor changes in plasma lipids and vascular reactivity in diabetic pigs fed a conventional pig chow diet versus control pigs,29 in the present study we wished to compare diabetic pigs versus nondiabetic pigs when both were fed an experimental high fat, high cholesterol diet. We addressed 2 major criteria to determine the suitability of the porcine model: (1) the plasma lipid and apolipoprotein profile and (2) whether hyperglycemia and the altered lipid profile would result in excess vascular disease/dysfunction. We assessed altered vascular reactivity (dysfunction) because there is substantial evidence that increased vasoconstriction30 31 and impaired endothelium-dependent relaxation30 are early functional events that precede gross, structural atherosclerotic lesions in the later, established phases of atherosclerosis.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Experimental Design
The prevalence of insulin resistance in Western societies and the dietary habits that promote hyperlipidemia11 suggested to us that in our porcine model a high fat diet and the resulting hyperlipidemia may interact with impaired insulin-mediated actions to yield hyperglycemia and exacerbation of the lipoprotein profile.15 There are numerous molecular mechanisms for insulin resistance of the major insulin-responsive tissues, ie, skeletal muscle and adipose. In the present study, we impaired insulin action by simply decreasing plasma insulin levels by the destruction of pancreatic beta cells with alloxan.28 32 33 Alloxan-treated diabetic pigs (n=4) were fed a high fat, high cholesterol, atherogenic diet (Minipig chow supplemented with cholesterol, coconut oil, corn oil, and sodium cholate) as shown in Table 1
. The term "diabetic" is used in the
text for readability, but the figures have been labeled as "diabetic
high fat" to indicate more clearly that the alloxan-treated group was
fed the atherogenic diet. Because a previous study indicated that there
were minimal effects of diabetes alone on plasma lipids,29
an alloxan-treated group on normal pig chow was not included in the
present study. A second group of pigs was fed the atherogenic diet
(high fat, n=4), and a third group (n=5) was fed only Minipig chow
(Purina Mills, Inc). Pigs were fed twice daily and had free access to
water. Supplements for the atherogenic diet were obtained from Research
Diets.
|
Animals
All procedures involving animals were approved by the Animal
Care and Use Committee of the University of Missouri and complied fully
with those approved by the American Veterinary Medical Association
Panel on Euthanasia. Male Sinclair miniature swine between 9 and 12
months of age (sexually mature) were obtained from the Sinclair
Research Center (Columbia, Mo). Pigs were housed in a
temperature-controlled room (20°C to 22°C) with a 12-hour
light/dark cycle. Anesthesia was induced with the following
drugs given intramuscularly (in mg/kg): atropine 0.05, ketamine
20, and xylazine 2; the level of anesthesia was
subsequently maintained with isoflurane gas (up to 4%). A catheter was
placed in an ear vein for blood sampling and alloxan (or vehicle)
injection. Alloxan monohydrate (175 mg/kg, Aldrich Chemical Co,
Inc) was added to 0.9% NaCl, and the pH was adjusted to 7.0 with NaOH
to enable solubility and then sterilely filtered. The 40 to 60 mL of
alloxan solution was then administered intravenously over a
period of 
3 minutes.
The most critical time of care for the alloxan-treated pig was in the first 12 to 16 hours, when blood glucose levels fell to life-threatening levels. This initial hypoglycemia is most likely due to massive insulin release triggered by the cytotoxic effects of alloxan on the pancreatic beta cells.28 33 This serious decrease in blood glucose was avoided largely by providing food ad libitum and administering glucose intravenously. The transient hypoglycemic phase of the alloxan response was then followed by sustained hyperglycemia. For blood glucose measurements, a lancet was used to draw blood from an ear vein, and a drop was placed on a Accu-Check ADVANTAGE test strip, and glucose was measured with an Accu-Check monitor (Boehringer Mannheim Corp). Blood samples for glucose measurements were drawn 1 to 2 hours after meals twice per week for the 12 weeks of the study. The glycemic measures fructosamine and percent glycated plasma protein were conducted by the Diabetes Diagnostic Laboratory at the University of Missouri (for review, see References 34 and 3534 35 ). Plasma indicators for kidney function (urea nitrogen and creatinine) and liver function (alanine transaminase and aspartic transaminase) were assayed by the Veterinary Diagnostic Laboratory at the University of Missouri School of Veterinary Medicine.
Lipid Measures
Plasma was derived from blood samples taken via the anterior
vena cava from pigs fasted overnight before alloxan and/or diet
treatment and after 8 and 12 weeks of treatment. For total
cholesterol or triglyceride levels, plasma was
assayed directly by standard enzymatic kit (Sigma Chemical Co). For
lipoprotein cholesterol and triglyceride
levels, fresh plasma samples (1 mL) were chromatographed by
fast protein liquid chromatography (FPLC) on a Superose
6 column (HR 16, Pharmacia) and eluted with (in wt/vol) 0.9% NaCl,
0.01% Tris, 0.01% EDTA, and 0.02% sodium azide, pH 7.6. Fractions (2
mL) were collected and assayed for protein (A280)
and for cholesterol (standard enzymatic kit). For
lipoprotein analysis, the cholesterol and protein
profiles for every pig within a treatment group were averaged and
plotted versus fraction number (Figure 4). For VLDL, LDL, and
HDL lipid content, fractions from each pig corresponding to these
lipoproteins were collected and assayed for cholesterol and
triglyceride concentration by standard enzymatic assay.
These values are shown in Table 3.
|
|
Apolipoprotein Electrophoresis
Aliquots of plasma that was frozen and stored at -80°C from
each pig in a group were pooled and run from each group on the Superose
6 column. Fractions 11, 12, 16, 18, 20, 22, 24, 26, 28, 29, 30, and 32
from each group were boiled in electrophoresis buffer (4% SDS, 20%
glycerol, 10% 2-mercaptoethanol, and 0.125 mol/L Tris, pH 6.8) and
analyzed by SDS-PAGE on the same 3% to 15% gradient slab
gel.36 Transparency images were taken of the
Coomassie-stained gels by use of an Epson Expression 636 scanner. The
images were analyzed by using NIH Image (version 1.61). The
fraction with the greatest intensity for an apolipoprotein (apoB,
apoA-I, or apoE) was set arbitrarily at 1.0, and all other fractions
were expressed relative to this fraction. This analysis was for
the purpose of determining the relative distribution of an
apolipoprotein within a treatment group profile.
Isometric Tension Recordings
After 12 weeks of treatment, the pigs were again
anesthetized, and the chest was opened to achieve euthanasia.
The left circumflex coronary, brachial, and femoral arteries
were rapidly excised for contractile tension recordings as
described.37 38 The physiological
saline bathing the arteries contained (in mmol/L)
CaCl2 2.5, NaCl 131.5,
MgCl2 1.2, KCl 5, NaHCO3
25, NaH2PO4 1.2, and
glucose 10.1. Prostaglandin F2
(PGF2, 30 µmol/L [Lutalyse], Upjohn)
was used to preconstrict vessels for relaxation studies. Peak
relaxation was determined in response to cumulative concentrations of
the endothelium-dependent agonist bradykinin
(10-11 to 10-6
mol/L).
Sudan IV Staining
Quantification of fatty streaks was performed with Sudan IV
stain on the basis of modifications of 2 methods.39 40
Carotid arteries were harvested 5 mm above and below the
bifurcation of the main carotid, dissected free of excess connective
tissue and fat, rinsed with saline, and then fixed in 10% (vol/vol)
buffered formalin. Carotid arteries were opened longitudinally, then
exposed to 5 mg/mL Sudan IV in 70% (vol/vol) isopropanol for 15
minutes in a 37°C water bath, and rinsed 3 times with 70%
isopropanol. The arteries were then flattened under a glass plate and
photographed. The carotid artery stain ratio was derived by
analysis of digitally scanned photographs using ImagePro Plus
(Media Cybernetics, Inc). Red-stained areas (Figure 8, top) were
subjected to the binary process, and a semiautomated edge-tracking
function was used to delineate the stained area, which was expressed as
the ratio of stained area to total area.
|
Statistical Analyses
Sigmaplot and Sigmastat (Jandel Scientific) were used for
graphics and statistical analyses. Values are expressed as the
mean±SE. For blood glucose and plasma lipids, ANOVA was performed and
was followed by the Fisher multiple range test for post hoc
analysis. The percentage of arteries showing contractile
oscillations was analyzed with a
2 test; Sudan IV staining and
endothelium-dependent relaxation data were
analyzed with a Kruskal Wallis 1-way ANOVA on ranks test and
the Dunnett method for post hoc analysis. The significance
level chosen was P<0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Glycemia, Kidney Function, and Liver Function
Eight weeks after the administration of alloxan to miniature pigs, a nearly 6-fold increase in blood glucose was evident compared with the value in control pigs (Figure 1
). One of
the 4 diabetic pigs was given subcutaneous injections of 7 to 8 U
porcine isophane insulin for the first 2 weeks of the study. Because
these diabetic animals required virtually no exogenous insulin for
survival, some beta cell function must have remained. In response to an
intravenous glucose load (0.5 g/kg), plasma insulin
increased to 25 to 75 µU/mL from the fasting level of 4 µU/mL in
control pigs and increased to only 6 to 8 µU/mL from a fasting level
of 4 µU/mL in diabetic pigs. Postprandial blood glucose was also
increased significantly (27%) in the high fat–fed pigs. In a separate
group of diabetic pigs with similar blood glucose values, the glycemic
indices fructosamine and glycated plasma protein were increased 59%
and 30%, respectively, after only 1 week of diabetes (Figure 2). Although alloxan is a highly specific
pancreatic beta cell toxin, renal and liver toxicity is sometimes
noted.33 In our studies, if pigs develop renal toxicity,
it is apparent within 48 hours of alloxan treatment and is
characterized by enormous plasma urea nitrogen levels (>200 mg/dL) and
plasma creatinine (>10 mg/dL); because of the irreversible
damage, the pig is euthanized. All pigs given alloxan that completed
this study did not show nonspecific alloxan toxicity. As shown in Table 2, levels of plasma urea nitrogen and
creatinine in diabetic pigs were low and not different from
control levels, thus indicating normal kidney function. Furthermore,
plasma concentrations of the liver enzymes alanine transaminase and
aspartic transaminase in diabetic pigs were not different from control
concentrations, thus indicating normal liver function after alloxan
treatment.
|
|
|
Cholesterol and Triglyceride Concentrations
in Plasma and Lipoprotein Fractions
In pigs fed the control diet for 8 weeks (Table 3
), the fasting plasma total
cholesterol and triglyceride levels were
similar to control values reported in numerous studies conducted with
swine.21 24 25 26 40 41 Approximately 57% of the
cholesterol was in the LDL fraction, whereas 40% was in
the HDL fraction (Table 3). When pigs were fed an atherogenic
hyperlipidemic diet for 8 weeks, LDL, HDL, and total
plasma cholesterol levels increased. When alloxan-treated
pigs were concurrently fed the atherogenic diet, the total plasma
cholesterol level increased further; this increase was
primarily due to an increase in LDL cholesterol. VLDL
cholesterol was also increased in diabetic pigs compared
with control and high fat–fed pigs. The HDL cholesterol
level was slightly but not significantly lower in diabetic pigs
compared with pigs fed the high fat diet alone. The large increase in
cholesterol in the LDL fraction caused the
cholesterol distribution among lipoprotein particles in
diabetic pigs to differ greatly from that observed in control and high
fat–fed pigs. In diabetic pigs fed the high fat diet, 81% of the
cholesterol was in the LDL fraction and only 16% was in
the HDL fraction (Table 3). The percent cholesterol
found in the VLDL fraction remained small (2.5%) and was not
significantly different from that found in the other treatment
groups.
Total plasma triglyceride concentration did not change when
pigs were fed the hyperlipidemic diet but was
significantly greater in diabetic pigs (Table 3). Plasma lipids
were also measured at 12 weeks, at the time the pigs were killed for
study (Figure 3). Total
cholesterol values for high fat–fed and diabetic pigs were
no longer significantly different from each other at 12 weeks because
of greater variability in each group.
|
Lipoprotein Profiles by FPLC
The average cholesterol profile for each group is
shown in Figure 4. In control pigs (top,
left), plasma cholesterol was distributed primarily between
LDL and HDL. When pigs were fed the high fat diet alone, both LDL and
HDL cholesterol increased substantially (Figure 4, middle, left). For HDL, peak width increased more than peak height.
When pigs were rendered diabetic and concomitantly fed the high fat
diet, the cholesterol content of the LDL fraction increased
further. The peak for cholesterol in LDL in the diabetic
pigs was shifted 2 fractions to the left, indicating that the LDL
fraction was enriched in large LDL and cholesterol-rich
intermediate-density lipoprotein (IDL) or remnant particles (Figure 4, bottom, left). Cholesterol in individual HDL
fractions was either reduced or remained at the same level compared
with that seen in pigs fed the high fat diet alone. The
cholesterol content of the VLDL fraction either increased
greatly or appeared to merge with the LDL fraction, indicating the
presence of IDL. The average protein profile for each group (Figure 4, right) indicated that the total protein content of the LDL
fractions did not differ between control and high fat–fed pigs but
appeared to be slightly elevated in diabetic pigs. Diabetic VLDL
protein was increased greatly compared with VLDL in plasma from control
or high fat–fed pigs.
Apolipoprotein Profiles of Lipoprotein Fractions
The relative distributions of apolipoproteins among lipoprotein
fractions from Superose 6 chromatography were
analyzed by SDS-PAGE. A pooled plasma sample from each group
was run on the Superose 6 column, and aliquots from the indicated 12
fractions were electrophoresed on the same gel, stained, and scanned.
The relative distribution of an apolipoprotein within a profile was
plotted in Figure 5 after setting the
fraction with the maximal intensity to 1. Therefore, the data
represent relative distribution of an apolipoprotein in the
lipoprotein profile within a treatment group and cannot be used to
compare the levels of apolipoproteins among treatment groups. apoB
(Figure 5, left panels) in control plasma was distributed
between fractions 16 and 26, with a peak at fraction 20. In the high
fat–fed pigs, the apoB peak was shifted to the left. In the diabetic
pigs, the apoB peak was broader. No apoB was detectable in fractions
>26. apoA-I (Figure 5, middle panels) in control plasma was
distributed in fractions 28 to 32, with much smaller amounts in lower
fractions. Although there was a slight relative increase in apoA-I in
the lower fractions of plasma from high fat–fed and diabetic/high
fat–fed pigs, a majority of the apoA-I remained above fraction 26.
apoE (Figure 5, right panels) in control plasma was very low in
fractions <26. The peak for apoE was found in the HDL region at
fraction 28. In the high fat–fed pigs, a proportion of apoE was
redistributed from HDL to fractions 18 to 22; in the diabetic pigs, to
fractions 16 to 22. This region probably represents increased
IDL found in plasma from both these groups of pigs.
|
Taken together, the cholesterol and protein profiles indicate that the cholesterol content of LDL particles, but not the number, was increased in pigs fed the high fat diet alone. In pigs that were also diabetic, LDL cholesterol was further increased, whereas the percentage of cholesterol in the HDL fraction was decreased. The protein content of the LDL fraction was only slightly increased in the diabetic pigs, indicating that the number of LDL particles was not greatly increased in the diabetic state. The major change in diabetic pigs was the development of a more heterogeneous population of both triglyceride- and cholesterol-rich apoB-containing particles with an increased content of apoE.
Excess Vascular Disease
The first apparent difference in vascular function that we noted
was the oscillation in contractile tension of
coronary arteries from diabetic pigs after exposure to the
vasoconstrictor agonist PGF2 (Figure 6A). Coronary arteries from
control pigs show a monotonic increase in tension and stable plateau
for hours of recordings. Steady-state contractile tension
(taken as the average of the maximum oscillatory excursions where
applicable) elicited by PGF2 was significantly
increased in high fat–fed and diabetic groups compared with the
control group (Figure 6B). Isometric contractile responses were
studied in 8 segments of left circumflex coronary artery from
each pig. In the high fat–fed group, 25% of the segments (2 of 8
total) showed at least one oscillation in contractile
tension, whereas 75% of the arterial segments from
diabetic pigs (6 of 8 total) showed oscillations (Figure 6C). Nonparametric statistical analysis
showed significantly more segments oscillated in the diabetic group
compared with control and high fat–fed groups. The effect of diabetes
on contractile oscillations is particularly striking
because we found contractile oscillations in only 10% (1
of 10 total) control segments in the present study, and we have
never found similar oscillations in any of our other
studies on nondiseased coronary, brachial, or femoral arteries
involving hundreds of porcine arterial segments (eg, see
References 37, 38, and 4237 38 42 ). Furthermore, segments with oscillating
tension responses were found in every diabetic pig (Figure 6C).
Another quantitative assessment of the contractile
oscillations was the total number of
oscillations occurring in the segments over the 20-minute
duration of exposure of each arterial ring segment to
PGF2 (Figure 6D), which was
significantly greater only in arterial segments from
diabetic versus control pigs. In contrast, brachial arteries showed no
contractile oscillations in any of the groups. The altered
contractile responsiveness of coronary arteries was restricted
to the receptor-dependent agonist PGF2, in
view of the fact that the tension response to depolarization with
80 mmol/L K+ was not different between the
groups. Also, the brachial artery contractile response to
depolarization by 80 mmol/L K+ was not
different between the groups.
|
We further assessed vascular dysfunction by comparing the
endothelium-dependent relaxations of coronary
and brachial arteries. As shown in Figure 7A, left circumflex coronary
arteries preconstricted with PGF2 and then
exposed to cumulative concentrations of the
endothelium-dependent dilator bradykinin showed
relaxations that were not different between groups.
Endothelium-dependent relaxations of brachial artery
rings from diabetic pigs were significantly decreased compared with
those from high fat–fed and control pigs (Figure 7B). Maximal
relaxations to bradykinin in arteries from diabetic pigs were only
50% the maximal relaxation found in arteries from high fat–fed and
control pigs.
|
To assess fatty streaks as a more classical precursor of gross anatomic
atherosclerotic lesions,43 we used Sudan IV
staining.39 40 Figure 8
(top) shows the carotid artery bifurcation and adjacent areas. Staining
was virtually absent from control arteries of pigs fed a low fat diet,
whereas only the diabetic pigs showed a significant increase in
staining. Note also that the staining occurs near the bifurcation at
the top of the figure.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A previous 12-week study29 showed that diabetic pigs fed a conventional low fat, low cholesterol diet failed to exhibit altered lipoprotein levels or profiles or vascular dysfunction. Hypertriglyceridemia and altered LDL particle size are currently postulated to play a major role in the atherosclerosis observed in diabetes.11 44 45 For this reason, our goal was to study the effects of diabetes in animals fed a high fat, high cholesterol diet. The major findings of the present study were that the vascular disease/dysfunction and dyslipidemia that mimic human diabetic dyslipidemia were observed concurrently in diabetic pigs fed an atherogenic diet.
Dyslipidemia
Previous studies have indicated that the pig model would provide
insight into lipoprotein metabolism associated with the
diabetic state for several reasons. The pig has a lipoprotein
distribution similar to that found in humans and carries 50% to 60%
of total plasma cholesterol in LDL
particles.25 46 47 48 Numerous studies have shown that high
fat, high cholesterol diets cause large changes in the
plasma lipoprotein profile of the pig (eg, see References 24 and 2624 26 ).
For example, when diets high in saturated fat (15% to 20%) and
cholesterol (1% to 2%) were fed to pigs, plasma total
cholesterol increased from 50 to 100 mg/dL on control low
fat diets to 500 to 800 mg/dL on the high fat diet.25 47
Furthermore, within 3 to 8 months of feeding, these diets stimulate the
development of complex atherosclerotic lesions that are similar to
those seen in humans.21 26 49 50
The total blood cholesterol levels of pigs fed the high fat
diet alone in the present study were not as high as the levels
reported in some previous studies in which swine were fed a similar
high fat diet.25 26 51 A possibility for the lower levels
is that we used male pigs with gonads intact, which typically show
smaller increases in plasma cholesterol than do female
pigs,51 and castrated male pigs26 in response
to an atherogenic diet. Interestingly, sex differences have also been
noted in lipoprotein profiles of human diabetic
patients.15 Data from pigs fed a high fat diet show that
both the LDL and HDL cholesterol fractions increased in
this group. Such changes would be in line with observations made in
several models of
hyperlipoproteinemia.52 Under the
additional stress of diabetes, the LDL cholesterol peak was
further increased and became wider, indicating the presence of a more
heterogeneous population of IDL/LDL particles (Figure 4). Concurrently, HDL cholesterol was slightly, but
not significantly, decreased in diabetic pigs compared with the high
fat–fed pigs. Comparison of the Superose 6 profiles of protein and
cholesterol indicated that LDL cholesterol
increased to a much greater extent than did LDL protein in the diabetic
pigs. Therefore, the LDL fraction became
cholesterol-enriched, larger, and much more
heterogeneous in diabetic pigs compared with high fat–fed
pigs. Unlike plasma cholesterol, total plasma
triglycerides were not elevated when pigs were fed the high
fat diet alone (Table 3). This observation reflects the fact
that these animals have ample lipoprotein lipase activity and that the
hydrolysis of triglycerides in the VLDL fraction was
probably not saturated when pigs were fed the high fat diet. However,
plasma and VLDL triglycerides (data not shown) were
significantly increased in diabetic pigs. The increases in VLDL and
IDL/large LDL particles seen in diabetic pigs resembles in profile and
distribution those seen in hyperlipidemic
humans53 54 and may be the result of increases in the
production of or decreases in the catabolism of remnant-like
lipoprotein particles. The apolipoprotein profiles of the Superose 6
fractions indicated that apoB was shifted to larger particles in high
fat–fed pigs and shifted to both larger and smaller particles in
diabetic pigs (Figure 5). Although the formation of larger
particles (fractions 17 to 20, Figure 5) predominated in
diabetic pigs, the amount of smaller LDL particles (fractions 24 and
25, Figure 5) also increased in the diabetic state. More
complete fractionation of LDL and subsequent investigation into the
relative atherogenicity of each type of particle is required (and is
possible in this animal model) in order to ascribe atherogenic
blame.
ApoA-I remained largely in the HDL fraction in both high fat–fed and
diabetic pigs. Although a small left shoulder appeared in the
distribution of apoA-I in the Superose 6 fractions (Figure 5),
no new peaks or the appearance of HDLC25
became apparent on Superose 6 chromatography. apoE,
found almost exclusively in the HDL fraction in control pigs, was
partially redistributed to triglyceride- and
cholesterol-enriched particles in both high fat–fed and
diabetic pigs. Altogether, the present results suggest that the
plasma lipoprotein profile of pigs fed a high fat diet was further
modified to become more atherogenic in the diabetic state.
Excess Vascular Disease
Our findings clearly indicate that in pigs fed a high fat, high
cholesterol diet, diabetes accelerated vascular
disease/dysfunction1 in that (1) only diabetic pigs showed
increased contractile tension oscillations in
coronary arteries, (2) only diabetic pigs showed impaired
endothelium-dependent relaxation, and (3) only diabetic
pigs had significantly increased fatty streaks.
Altered vasomotor tone and increased Sudan staining of fatty streaks in
the diabetic pigs strongly argue that these indices of the early stages
of atherosclerosis will progress to mature lesions. The
significant increase in PGF2-induced
contractile oscillations in coronary arteries of
the high fat–fed pigs, together with a modest increase in fatty streak
formation (not statistically significant), is consistent with
the long-held prediction that coronary hyperreactivity precedes
formation of gross lesions in epicardial conduit
arteries.30 Indeed, Golenhofen et al31
suggested an association of spontaneous oscillatory activity with
proliferation of vascular smooth muscle cells in vasospasm and human
pathology. Thollon et al55 showed oscillations
of membrane potential in smooth muscle of porcine coronary
arteries with regenerating endothelium after balloon
injury. Furthermore, vascular smooth muscle cells isolated from
diabetic rats show an increased incidence of agonist-induced
intracellular Ca2+
oscillations.56 Finally, from a longitudinal
study of heart transplant recipients, Davis et al30
provided intravascular ultrasound evidence that decreased
endothelium-dependent vasodilation occurs in arteries
defined as structurally normal, ie, probably lacking fatty streaks.
Thus, the significant increase (P<0.05) in Sudan staining
for fatty streaks in carotid arteries (Figure 8) and the 3-fold
greater coronary contractile oscillations in our
diabetic pigs compared with the high fat–fed pigs are
consistent with the hypothesized progression of vascular
dysfunction to complex structural atherosclerotic lesions. Further
evidence for profoundly accelerated atherosclerosis in
the diabetic pigs was our finding that 9% of the carotid artery was
Sudan-stained after only 12 weeks (Figure 8) compared with 34
weeks required for comparable fatty streak development in the carotid
artery of nondiabetic miniature pigs fed an atherogenic
diet.57 Finally, the impaired
endothelium-dependent relaxation in brachial arteries
of only the diabetic pigs, not the high fat–fed pigs (Figure 7B), confirms a large body of evidence that
endothelial dysfunction occurs in diabetes (for review,
see Reference 5858 ). The lack of impaired
endothelium-dependent relaxation in the high fat–fed
pigs in the present study is different from the impairment found by
Cohen et al59 in hyperlipidemic pigs but
could be explained by several differences between the studies,
including the artery type, strain of pig, and
endothelium-dependent agonists. We were surprised that
bradykinin-induced relaxation was not impaired in coronary
arteries (Figure 7A). However,
endothelium-dependent relaxation has not been widely
studied in coronary arteries from the animal models (eg,
porcine and primate)58 that most closely mimic humans. Our
present study showing that diabetic vasculopathy is vascular bed
specific, taken together with the finding that the duration of diabetes
greatly influences endothelium-dependent
relaxation,60 points to the importance of studying
diabetic vasculopathy under highly controlled experimental
conditions.
Future work is needed with the porcine model in which vasoreactivity and morphological lesion formation are monitored in the diabetic high fat–fed pig over longer durations (>12 weeks to 20 or 30 weeks) during which CAD will have progressed further. These measures, combined with interventional studies to attenuate CAD, will provide more definitive evidence for altered vasoreactivity as a predictive tool for structural CAD. Indeed, mature calcified lesions were noted in diabetic high fat–fed pigs but not hyperlipidemic pigs after >20 weeks.61
It is too premature to ascribe the excess vascular dysfunction and
fatty streaks that were found in the present study (Figures 6 to 8) to altered VLDL, LDL, or triglycerides. We
have noted, however, that 12 weeks of diabetes, per se, did not alter
lipids or lipoproteins in pigs that were not fed an atherogenic diet
and that there was no vascular dysfunction.29 Thus, taken
together, these data argue for components of diabetic
dyslipidemia as important factors in accelerated
atherosclerosis, whereas blood glucose may either be an
"innocent bystander"5 or necessary, but not
sufficient, to induce atherosclerosis.
Need for a Suitable Animal Model: Advantages of the Porcine
Model
An American Diabetes Association task force made a recommendation
15 years ago to develop better animal models of accelerated
atherosclerosis associated with
diabetes.18 The lack of appropriate animal models is still
considered a limitation in the most recent reviews in the
field.8 19 Although many advances have been made in the
fields of insulin signaling and lipid metabolism by using
rodent and transgenic mouse models,62 it will be possible
to characterize CAD development in a porcine animal model that enables
serial measurements during the course of diabetes.
We destroyed pancreatic beta cells with alloxan to decrease plasma insulin levels. Although this experimentally straightforward maneuver resulted in a diabetic dyslipidemia that is profoundly similar to that in human diabetic patients,9 11 12 13 14 15 this is not identical to early stages of human type 2 diabetes characterized by insulin resistance and concomitant hyperinsulinemia. Instead, this model would best represent the later stages of type 2 diabetes, after the pancreatic beta cells have ceased to function. Importantly, type 1 diabetic patients with poor glycemic control also show dyslipidemia, including hypertriglyceridemia and LDL particle alterations.15 Thus, the porcine model very likely has relevance to the dyslipidemia in both type 1 and type 2 diabetes.
There are several important similarities between swine and humans
that reinforce the advantages of using swine: (1) Swine have
omnivorous habits; thus, they will consume a "human-type"
diet.63 (2) Metabolism of
foodstuffs,63 specifically lipoprotein
metabolism (References 24, 25, and 5124 25 51 and the present
report), is similar to that in humans. (3) Swine have a propensity
toward sedentary habits and obesity; unlike dogs, which will pace in
their cages if not adequately exercised, pigs are content to be
sedentary. Pigs are intensively used in obesity research with direct
relevance to humans.50 (4) Although the larger size of
even miniature swine (40 to 60 kg) has been considered by others to
be an absolute limitation to their use,32 the ability to
sample 30 to 50 mL of blood at weekly intervals is essential for the
lipid profiling that we have done. Furthermore, a large blood volume
enables the study of coagulation factors (eg,
platelets)64 that is not possible in smaller animals.
Very important, the size of the pig is similar to that of humans, thus
enabling trials of percutaneous catheter interventions
for revascularization with devices identical to
those used in humans.65 (5) The most important similarity
to humans is in the cardiovascular system,
specifically, coronary circulation and the susceptibility to
CAD. Similar to humans, pigs have few native coronary
collateral arteries22 23 ; the pharmacology of
coronary artery reactivity is similar to that in
humans66 ; and heart rate and, thus, metabolic
demand on the heart and cyclic changes in coronary blood flow
are also similar to those found in humans.22 27 Finally,
atherosclerotic lesions are morphologically similar to those in
humans.26 41 50 When fed low fat diets, swine develop
modest atherosclerosis, but on high fat diets, they
develop the full complement of atherosclerotic
lesions.21 26 The similar lipoprotein pattern further
suggests a similar etiology of atherosclerosis in swine
and humans.24 67 The above-mentioned major advantages plus
the tight experimental control of blood glucose and lipids in swine
will enable the most rigorous assessment of the independent effects of
these parameters on atherosclerosis.
Studies in humans are limited because both blood glucose and lipids are
changing and atherosclerosis is typically quantified by
extreme end points, such as mortality or myocardial infarction, rather
than more sensitive indices of the early progression of
atherosclerosis.15 68 We conclude that
this porcine model of diabetic dyslipidemia and excess
vascular disease/dysfunction will enable further studies to delineate
cellular and molecular mechanisms underlying dyslipidemia,
CAD, and peripheral atherosclerosis. This
animal model should be ideal for preclinical evaluation of
pharmacotherapy to prevent the excess CAD associated with diabetes.
|
Acknowledgments |
|---|
This study was supported by National Institutes of Health (NIH) grant HL-47586 and a University of Missouri Research Board grant to Dr Dixon; by an American Diabetes Association grant to Dr Weisman; by NIH grant HL-52490 to Dr Laughlin; and by the American Diabetes Association, NIH RR-13223, and Research Career Development Award (NIH HL-02872) to Dr Sturek. We thank Dr Guy Bouchard of the Sinclair Research Center (Columbia, Mo) for providing the Sinclair miniature pigs, Dr David E. Goldstein for measurement of fructosamine and glycated plasma protein, Karl Skala and Denise Stowers for their expert technical assistance, and Dr Hao Li for help with the Superose 6 chromatography. Brent Hill and Dr Douglas Bowles are gratefully acknowledged for critical reading of this manuscript.
Received May 11, 1998; accepted April 26, 1999.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Ruderman NB, Gupta S, Sussman I. Hyperglycemia, diabetes, and vascular disease: an overview. In: Ruderman N, Williamson J, Brownlee M, eds. Hyperglycemia, Diabetes, and Vascular Disease. New York, NY: Oxford University Press; 1992;3–20.
- Eastman RC, Cowie CC, Harris MI. Undiagnosed diabetes or impaired glucose tolerance and cardiovascular risk. Diabetes Care. 1997;20:127–128.[Medline] [Order article via Infotrieve]
- Kannel WB, McGee DL. Diabetes and cardiovascular disease: the Framingham study. JAMA. 1979;241:2035–2038.[Abstract]
- Garber AJ. The touch that kills. Clin Diabetes. 1995;13:57–58.
- Eschwège E, Balkau B, Fontbonne A. The epidemiology of coronary heart disease in glucose-intolerant and diabetic subjects. J Intern Med. 1994;736(suppl):5–11.
-
The Diabetes Control and Complications Trial
Research Group. The effect of intensive treatment of diabetes on the
development and progression of long-term complications in
insulin-dependent diabetes mellitus. N Engl J Med. 1993;329:977–986.
[Abstract/Free Full Text] - Pyörälä K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Metab Rev. 1987;3:463–524.[Medline] [Order article via Infotrieve]
- Semenkovich CF, Heinecke JW. The mystery of diabetes and atherosclerosis: time for a new plot. Diabetes. 1997;46:327–334.[Abstract]
-
Bierman EL. Atherogenesis in diabetes.
Arterioscler Thromb. 1992;12:647–656.
[Free Full Text] - Stamler J. Epidemiology, established major risk factors, and the primary prevention of coronary heart disease. In: Parmley WW, Chatterjee K, eds. Cardiology. Philadelphia, Pa: JB Lippincott; 1987:1–41.
- Steiner G. The dyslipoproteinemias of diabetes. Atherosclerosis. 1994;110(suppl):S27–S33.
- Reaven GM, Laws A. Insulin resistance, compensatory hyperinsulinaemia, and coronary heart disease. Diabetologia. 1994;37:948–952.[Medline] [Order article via Infotrieve]
- Sniderman A, Michel C, Racine N. Heart disease in patients with diabetes-mellitus. J Clin Epidemiol. 1992;45:1357–1370.[Medline] [Order article via Infotrieve]
- Howard BV, Howard WJ. Dyslipidemia in non-insulin-dependent diabetes mellitus. Endocr Rev. 1994;15:263–274.[Medline] [Order article via Infotrieve]
- Caixàs A, Ordónez-Llanos J, De Leiva A, Payes A, Homs R, Perez A. Optimization of glycemic control by insulin therapy decreases the proportion of small dense LDL particles in diabetic patients. Diabetes. 1997;46:1207–1213.[Abstract]
- Laakso M. Insulin resistance and coronary heart disease. Curr Opin Lipidol. 1996;7:217–226.[Medline] [Order article via Infotrieve]
- Pyorala K, Pedersen TR, Kjekshus J, Faergeman O, Olsson AG, Thorgeirsson G, The Scandinavian Simvastatin Survival Study (4S) Group. Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease: a subgroup analysis of the Scandinavian Simvastatin Survival Study (4S). Diabetes Care. 1997;20:614–620.[Abstract]
- Chobanian AV, Arquilla ER, Clarkson TB, Eder HA, Howard CF Jr, Regan TJ, Williamson JR. Cardiovascular complications. Diabetes. 1982;31:54–64.
- Barrett-Connor E. Does hyperglycemia really cause coronary heart disease? Diabetes Care. 1997;20:1620–1623.[Medline] [Order article via Infotrieve]
- Douglas WR. Of pigs and men and research. Space Life Sci. 1972;3:226–234.[Medline] [Order article via Infotrieve]
- Lee KT. Swine as animal models in cardiovascular research. In: Tumbleson ME, ed. Swine in Biomedical Research. New York, NY: Plenum Publishing Corp; 1986:1481–1496.
- White FC, Bloor CM. The pig as a model for myocardial ischemia. In: Tumbleson ME, ed. Swine in Biomedical Research. New York, NY: Plenum Publishing Corp; 1986:481–490.
- Schaper W, Schaper J. Collateral Circulation. New York, NY: Kluwer Academic Publishers; 1993.
- Jokinen MP, Clarkson TB, Prichard RW. Animal models in atherosclerosis research. Exp Mol Pathol. 1985;42:1–28.[Medline] [Order article via Infotrieve]
- Mahley RW, Weisgraber KH, Innerarity T, Brewer HB Jr, Assmann G. Swine lipoproteins and atherosclerosis: changes in the plasma lipoproteins and apoproteins induced by cholesterol feeding. Biochemistry. 1975;14:2817–2823.[Medline] [Order article via Infotrieve]
-
Bell FP, Gerrity RG. Evidence for an altered
lipid metabolic state in circulating blood monocytes under
conditions of hyperlipemia in swine and its implications in
arterial lipid metabolism. Arterioscler
Thromb. 1992;12:155–162.
[Abstract/Free Full Text] -
Hall JL, Stanley WC, Lopaschuk GD, Wisneski JA,
Pizzurro RD, Hamilton CD, McCormack JG. Impaired pyruvate oxidation but
normal glucose uptake in diabetic pig heart during
dobutamine-induced work. Am J Physiol. 1996;271:H2320–H2329.
[Abstract/Free Full Text] - Phillips RW, Panepinto LM. Swine as a model for human diabetes. In: Tumbleson ME, ed. Swine in Biomedical Research. New York, NY: Plenum Publishing Corp; 1986:549–560.
- Sturek M, Liu CY, Dick GM. Voltage-gated calcium channel current in coronary arterial smooth muscle cells from alloxan diabetic pigs. FASEB J. 1996;10:A660. Abstract.
-
Davis SF, Yeung AC, Meredith IT, Charbonneau F,
Ganz P, Selwyn AP, Anderson TJ. Early endothelial
dysfunction predicts the development of transplant coronary
artery disease at 1 year posttransplant. Circulation. 1996;93:457–462.
[Abstract/Free Full Text] - Golenhofen K, Mandrek K, Schaper W, Weinheimer G, Weigand W. Mechanical activity of isolated canine coronary arteries after coronary occlusion. Basic Res Cardiol. 1981;76:480–484.[Medline] [Order article via Infotrieve]
- Karasik A, Hattori M. Use of animal models in the study of diabetes. In: Kahn CR, Weir GC, eds. Joslin’s Diabetes Mellitus. Philadelphia, Pa: Lea & Febiger; 1994:317–350.
-
Rerup CC. Drugs producing diabetes through damage
of the insulin secreting cells. Pharmacol Rev. 1970;22:485–517.
[Free Full Text] - Goldstein DE, Little RR, Lorenz RA, Malone JI, Nathan D, Peterson CM. Tests of glycemia in diabetes. Diabetes Care. 1995;18:896–909.[Medline] [Order article via Infotrieve]
- Little RR, Goldstein DE. Long-term glucose monitoring with glycated proteins. Lab Med. 1992;23:533–538.
- Laemmli VK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680–685.[Medline] [Order article via Infotrieve]
-
Rapps JA, Jones AW, Sturek M, Magliola L, Parker
JL. Mechanisms of altered contractile responses to vasopressin and
endothelin in canine coronary collateral arteries.
Circulation. 1997;95:231–239.
[Abstract/Free Full Text] -
Bowles DK, Laughlin MH, Sturek M. Exercise
training alters the Ca2+ and contractile
responses of coronary arteries to endothelin. J Appl
Physiol. 1995;78:1079–1087.
[Abstract/Free Full Text] -
De la Llera-Moya M, Rothblat GH, Glick JM,
England JM. Etoposide treatment suppresses atherosclerotic plaque
development in cholesterol-fed rabbits. Arterioscler
Thromb. 1992;12:1363–1370.
[Abstract/Free Full Text] -
Cornhill JF, Barrett WA, Herderick EE, Mahley RW,
Fry DL. Topographic study of Sudanophilic lesions in
cholesterol-fed minipigs by image analysis.
Arteriosclerosis. 1985;5:415–426.
[Abstract/Free Full Text] -
Holvoet P, Theilmeier G, Shivalkar B, Flameng W,
Collen D. LDL hypercholesterolemia is
associated with accumulation of oxidized LDL, atherosclerotic plaque
growth, and compensatory vessel enlargement in coronary
arteries of miniature pigs. Arterioscler Thromb Vasc Biol. 1998;18:415–422.
[Abstract/Free Full Text] -
McAllister RM, Laughlin MH. Short-term exercise
training alters responses of porcine femoral and brachial arteries.
J Appl Physiol. 1997;82:1438–1444.
[Abstract/Free Full Text] -
Strong JP, Wissler RW, Gay S, Miller EJ, Daoud A,
Smith LC, Robertson AL, Mergner W, Chandler AB, McManus BM, Cornhill
JF, Hixson JE, McMahan A, McGill HC, Virmani R, Jagannathan SN. Natural
history and risk factors for early human atherogenesis. Clin
Chem. 1995;41:134–138.
[Abstract/Free Full Text] - Taskinen MR, Lahdenperä S, Syvänne M. New insights into lipid metabolism in non-insulin-dependent diabetes mellitus. Ann Med. 1996;28:335–340.[Medline] [Order article via Infotrieve]
- Laws A. Free fatty acids, insulin resistance and lipoprotein metabolism. Curr Opin Lipidol. 1996;7:172–177.[Medline] [Order article via Infotrieve]
- Thomas WA, Lee KT, Kim DN. Pathogenesis of atherosclerosis in the abdominal aorta and coronary arteries of swine in the first 90 days on a hyperlipidemic diet. In: Tumbleson ME, ed. Swine in Biomedical Research. New York, NY: Plenum Publishing Corp; 1986:1511–1525.
- Kim DN, Ho HT, Lawrence DA, Schmee J, Thomas WA. Modification of lipoprotein patterns and retardation of atherogenesis by a fish oil supplement to a hyperlipidemic diet for swine. Atherosclerosis. 1987;76:35–43.
- Chapman MJ, Goldstein S. Comparison of the serum low density lipoprotein and of its apoprotein in the pig, rhesus monkey and baboon with that in man. Atherosclerosis. 1976;25:267–291.[Medline] [Order article via Infotrieve]
- Reitman JS, Mahley RW, Fry DL. Yucatan miniature swine as a model for diet-induced atherosclerosis. Atherosclerosis. 1982;43:119–132.[Medline] [Order article via Infotrieve]
- Phillips RW, Panepinto LM, Spangler R, Westmoreland N. Yucatan miniature swine as a model for the study of human diabetes-mellitus. Diabetes. 1982;31:30–36.
- Reitman JS, Mahley RW. Changes induced in the lipoproteins of Yucatan miniature swine by cholesterol feeding. Biochim Biophys Acta. 1979;575:446–457.[Medline] [Order article via Infotrieve]
- Mahley RW. Atherogenic hyperlipoproteinemia: the cellular and molecular biology of plasma lipoproteins altered by dietary fat and cholesterol. Med Clin North Am. 1982;66:375–402.[Medline] [Order article via Infotrieve]
-
Kugiyama K, Doi H, Motoyama T, Soejima H, Misumi
K, Kawano H, Nakagawa O, Yoshimura M, Ogawa H, Matsumura T, Sugiyama S,
Nakano T, Nakajima K, Yasue H. Association of remnant lipoprotein
levels with impairment of endothelium-dependent
vasomotor function in human coronary arteries.
Circulation. 1998;97:2519–2526.
[Abstract/Free Full Text] - Marcoux C, Tremblay M, Frenderich A, Jacques H, Krimbou L, Nakajima K, Davignon J, Cohn JS. Plasma remnant-like particle lipid and apolipoprotein levels in normolipidemic and hyperlipidemic subjects. Atherosclerosis. 1998;139:161–171.[Medline] [Order article via Infotrieve]
-
Thollon C, Bidouard JP, Cambarrat C, Delescluse
I, Villeneuve N, Vanhoutte PM, Vilaine JP. Alteration of
endothelium-dependent
hyperpolarizations in porcine coronary
arteries with regenerated endothelium. Circ
Res. 1999;84:371–377.
[Abstract/Free Full Text] - Tam ES, Ferguson DG, Bielefeld DR, Lorenz JN, Cohen RM, Pun RY. Norepinephrine-mediated calcium signaling is altered in vascular smooth muscl