The relationship of serum lipoproteins to atherosclerosis in the cholesterol-fed alloxanized rabbit

Serum triglycerides as a risk factor for cardiovascular diseases in type 2 diabetes mellitus: a systematic review and meta-analysis of prospective studies

Objective

The importance of triglycerides (TG) level as a risk factor for cardiovascular diseases (CVD) has been extensively investigated in the general population; however, their relationship in patients with type 2 diabetes mellitus (T2DM) is uncertain. We aimed to assess the association of TG with CVD in T2DM individuals.

Research design and methods

We searched bibliographic databases for studies published until June 2018, reporting on the relationship between TG and CVD in T2DM people. A random-effects model with inverse variance weighting was used to compute pooled estimates of the most fully adjusted risk ratios (RR) and corresponding 95% confidence intervals (CI) according to TG categories, unit TG, and logarithm (log) of TG for CVD.

Results

A total of 31 studies were included, involving 132,044 T2DM patients with 10,733 incident cardiovascular events. The pooled RR (95% CI) of CVD for an increase in baseline TG, log TG by 1-mmol/l and categorized in the highest vs. the lowest TG in T2DM were 1.06 (1.02, 1.09), 1.30 (1.18, 1.42) and 1.30 (1.16, 1.46), corresponding to a CVD risk increase of 6%, 30% and 30%, respectively. The pooled RR (95% CI) of CVD for per 1-mmol/L TG increment in eight studies and TG categories in three studies were 1.03 (0.98, 1.08) and 1.39 (0.92, 2.1) in T2DM patients adjusted for other lipids parameter, respectively.

Conclusions

In T2DM patients, an elevated triglyceride level cannot serve as an independent marker for an increased risk of cardiovascular events, but still, the higher serum TG levels tend to be associated with increased risks of CVD.

Electronic supplementary material

The online version of this article (10.1186/s12933-019-0851-z) contains supplementary material, which is available to authorized users.

Diabetes-induced accelerated atherosclerosis in swine

Patients with diabetes are at higher risk for atherosclerotic disease than nondiabetic individuals with other comparable risk factors. Studies examining mechanisms underlying diabetes-accelerated atherosclerosis have been limited by the lack of suitable humanoid animal models. In this study, diabetes was superimposed on a well-characterized swine model of atherosclerosis by injection of the beta-cell cytotoxin streptozotocin (STZ), resulting in a >80% reduction in beta-cells and an increase in plasma glucose to diabetic levels. Animals were maintained without exogenous insulin for up to 48 weeks. Plasma glucose and cholesterol levels and lesion extent and severity were quantified in swine with diabetes and hyperlipemia alone and in combination compared with controls. Diabetes had no effect on plasma cholesterol levels, but diabetic/hyperlipemic (D-HL) swine developed hypertriglyceridemia and showed a doubling in aortic sudanophilia over nondiabetic/hyperlipemic (N-HL) swine as early as 12 weeks (47.25 +/- 4.5 vs. 24.0 +/- 4.6%). At 20 weeks, coronary artery stenosis was significantly greater in D-HL than in N-HL animals (86 +/- 10 vs. 46 +/- 8%). Coronary lesions predominantly arose in the first 2-3 cm of the vessels and displayed humanoid morphology. Aortic lesions in D-HL swine had double the cholesterol content of those in N-HL swine, and incorporation of oleate into cholesteryl ester was significantly greater in grossly normal aortic areas of D-HL swine compared with N-HL and was attributed to similar elevated incorporation in monocytes. This large study demonstrates that a model of diabetes with humanoid characteristics, including hypertriglyceridemia and severe, accelerated atherosclerosis can be reproducibly induced and maintained in swine. This model should potentially be of great value in elucidating mechanisms underlying the accelerated atherosclerosis seen in human diabetic individuals.

Potential downregulation of HMG-CoA reductase after prolonged administration of P-407 in C57BL/6 mice

This study investigated the potential alteration in the amount of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase messenger RNA (mRNA) and lipoprotein lipase (LPL) mRNA in the livers of C57BL/6 mice after long-term (200 days) treatment with the nonionic surfactant called poloxamer 407 (P-407). Previously, P-407 has been used to produce a dose-controlled hyperlipidemic state in C57BL/6 mice with subsequent formation of atherosclerotic lesions. Five groups of mice were studied; controls (C); mice fed a standard chow diet enriched with only cholic acid (CH); mice fed the high-cholesterol, high-fat Paigen diet (HF); mice treated with 0.5 g/kg P-407 every third day (P); and mice administered 0.5 g/kg P-407 every third day while consuming a diet identical to that of mice in group CH (PC). Neither a significant (p < 0.05) weight loss nor alteration in liver enzymes (AST and ALT) were observed for any group throughout the study when compared with the control mice. Total plasma cholesterol (CHOL) was significantly elevated compared with controls for mice in groups HF, P, and PC, whereas total plasma triglycerides (TG) were significantly increased for mice in only groups P and PC. Long-term ingestion of a high-fat diet or a diet enriched in cholic acid resulted in a significant (p < 0.05) reduction in HDL-CHOL when compared with controls. Plasma samples assayed at 200 days for mice in groups HF and P showed a shift in the lipoprotein fraction distribution primarily to VLDL-CHOL as compared with mice in group C in which, as expected, most of the CHOL was contained in the HDL fraction. The biologic activity of HMG-CoA reductase assayed in hepatic microsomal homogenates was significantly reduced for mice in groups CH (p < 0.01), HF (p < 0.01), and PC (p < 0.05), but not for mice in group P, when compared with control. A statistical analysis of the data demonstrated significant (p < 0.05) reductions in the HMG-CoA reductase mRNA levels in hepatic tissue for all treatment groups relative to mRNA levels determined for mice in group C. In contrast, no treatment group demonstrated a significant difference in hepatic LPL mRNA levels when compared with mRNA levels determined for control animals. These data demonstrate that P-407 administration to C57BL/6 mice significantly decreased the amount of HMG-CoA reductase mRNA detected in liver.

Cholesterol and polyunsaturated acid enriched diet: effect on kinetics of the acrosome reaction in rabbit spermatozoa

In this study we examined the effect of cholesterol (Diet 2), cholesterol and fish oil (FO) polyunsaturated acid (Diet 3), and polyunsaturated acid (Diet 4) enriched diets upon the acrosome reaction (AR) of New Zealand White rabbit spermatozoa. Male rabbits fed with cholesterol alone or with FO increased their cholesterol and LDL-cholesterol serum levels after 15 days of diet. Ten semen samples were obtained after 2 months of diet. Our results suggest that hypercholesterolemia and hypertriglyceridemia in male rabbits could produce a decreased capacity of sperm AR after 4 h (0%, 0%, and 60% lower than the control), 6 h (0%, 68%, and 44%), or 8 h (58%, 52% and 32%) of incubation in capacitating medium. Another set of experiments were made with lysophosphatidylcholine (LPC), 80 micrograms/ml, and the same pattern of AR was seen. Nevertheless, the high cholesterol and total lipids (TL) levels in serum did not affect the cholesterol levels in seminal plasma (SP) but affect the SP total lipids. The diminished capacity of rabbit sperm to undergo the AR was not reverted by in vitro incubation with the Shinitsky medium for cholesterol depletion (MDC). These results indirectly suggest that the cholesterol/phospholipid ratio in hypercholesterolemic sperm is similar to that of controls and are in agreement with preliminary studies made in our laboratory that evidenced the same cholesterol/phospholipid ratio in rabbit sperm from hypercholesterolemic animals than from controls.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of pravastatin on serum cholesterol levels in hypercholesterolemic diabetic rabbits

Diabetes mellitus is associated with hyperlipidemia and increased risk of atherosclerosis. A diabetic animal model has been developed to study the effect of treatment with pravastatin, a potent HMG CoA reductase inhibitor, on plasma lipoprotein levels. Hypercholesterolemia was induced in alloxan diabetic and control rabbits by feeding a diet containing 25% casein and 10% hydrogenated coconut oil for 8 weeks. Feeding the casein-coconut oil diet to the diabetic group resulted in a 5-fold increase in serum cholesterol levels, which was not statistically different from the nondiabetic group fed this diet. However, in the diabetic group, there was more cholesterol in the VLDL fraction and less in LDL as compared to the nondiabetic group. Serum triacylglycerol levels in the diabetic rabbits were variable and ranged from 58-943 mg/dl. The diabetic and nondiabetic animals were then treated with pravastatin at a dose of 10 mg/kg per day for 21 days. In the nondiabetic group, pravastatin treatment significantly lowered serum and LDL cholesterol concentrations by 28.5% (52.3 mg/dl, P less than 0.05) and 36.2% (40.7 mg/dl, P less than 0.05) respectively, relative to the placebo group. Serum and VLDL triacylglycerol levels in the nondiabetic group were also significantly decreased following pravastatin treatment. In the diabetic group, serum and LDL cholesterol levels were decreased by 37.0% (69.1 mg/dl, P less than 0.05) and 52.7% (32.1 mg/dl, P less than 0.01), respectively, relative to the diabetics given the placebo. Pravastatin treatment did not adversely affect serum glucose levels. Thus, pravastatin treatment was effective in controlling the hypercholesterolemia present in these diabetic animals.

Hyperglycemia in normotriglyceridemic, hypercholesterolemic insulin-treated diabetic rabbits does not accelerate atherogenesis

Severely hyperlipidemic alloxan-diabetic cholesterol-fed rabbits were treated with different daily doses of insulin in order to study the effect of insulin on plasma lipids, lipoproteins and postheparin lipoprotein lipase activity. At plasma triglyceride levels of 15,000 mg/dl, untreated diabetic rabbits carried 73% (1950 mg/dl) of plasma total cholesterol in lipoproteins with a diameter larger than 75 nm (Sf greater than 400), 25% in smaller very low density lipoproteins (VLDL) and 1% in both low and high density lipoproteins (LDL, HDL). Insulin treatment greatly reduced plasma total cholesterol and triglyceride concentrations. The decrease of plasma total cholesterol concentration was paralleled by a decrease in the cholesterol of the largest lipoproteins (Sf greater than 400) and an increase in cholesterol of both smaller very low density lipoproteins and low density lipoproteins. At the same time, postheparin plasma lipoprotein lipase activity increased 2-8-fold. When plasma triglyceride levels were normalized by insulin treatment, the lipoprotein cholesterol distribution in diabetic cholesterol-fed rabbits was similar to that of normal cholesterol-fed rabbits. To study development of atherosclerosis, diabetic rabbits were cholesterol-fed and treated with insulin for eight weeks such that the triglyceride levels were normalized, but plasma glucose levels were still greatly elevated. Nondiabetic rabbits were cholesterol-fed simultaneously. Plasma cholesterol and triglyceride levels were similar in the two groups of rabbits, as well as cholesterol in Sf greater than 400 or smaller VLDL and cholesterol in HDL. However, LDL-cholesterol concentration in the insulin-treated diabetic rabbits was 1.5-2 times that in the nondiabetic rabbits. The two groups of rabbits developed similar degrees of atherosclerosis, as judged by aortic cholesterol content. Apparently, partially controlled diabetes in cholesterol-fed rabbits does not accelerate atherogenesis beyond that observed in nondiabetic cholesterol-fed rabbits.

Reduced atherogenesis in cholesterol-fed diabetic rabbits. Giant lipoproteins do not enter the arterial wall

In cholesterol-fed rabbits, alloxan-diabetes has an anti-atherogenic effect, which is associated with severe elevation of plasma triglyceride concentrations. To study this effect, we measured lipoprotein sizes and aortic permeability coefficients for cholesteryl ester and for albumin in hypertriglyceridemic diabetic cholesterol-fed rabbits and in normotriglyceridemic cholesterol-fed rabbits. With the same high cholesterol concentration in plasma, hypertriglyceridemic diabetic rabbits had 70% of plasma cholesterol in very large lipoproteins (diameter greater than 75 nm), whereas normotriglyceridemic rabbits had only about 10% of plasma cholesterol in these giant lipoproteins. The aortic permeability coefficients for cholesteryl ester in hypertriglyceridemic diabetic cholesterol-fed rabbits was only 10% to 50% of that in normotriglyceridemic cholesterol-fed rabbits. Aortic permeability coefficients for albumin did not differ significantly between the hypertriglyceridemic and normotriglyceridemic rabbits. The results suggest that the large size of a major fraction of plasma lipoproteins in the hypertriglyceridemic diabetic cholesterol-fed rabbits is responsible for the relatively low aortic permeability coefficient for cholesteryl ester from plasma and hence for reduced atherogenesis in these animals.

Relationship of an abnormal plasma lipoprotein to protection from atherosclerosis in the cholesterol-fed diabetic rabbit

Alloxan-diabetic rabbits develop a pronounced hypercholesterolemia and hypertriglyceridemia in response to cholesterol feeding. Despite higher levels of plasma cholesterol, these animals have much less atherosclerosis than cholesterol-fed nondiabetics. To determine whether this effect is due to properties of the lipoproteins, we compared chemical, physical, and metabolic characteristics of a very low density lipoprotein (VLDL) fraction (d less than 1.019 g/ml) from the diabetic and nondiabetic cholesterol-fed rabbits. The molar ratio of triglyceride to cholesteryl ester in the particles from diabetic animals ranged from 2:1 to 6:1, and this ratio remained constant in subfractions from individual rabbits. Triglyceride from nondiabetic control animals was a minor component. Differential scanning calorimetry showed a distinct order-disorder phase transition for cholesteryl ester at approximately 42 degrees C in the fractions from control animals, whereas in fractions from most of the diabetics no such transition was observed, indicating that both triglyceride and cholesteryl ester are present in the core of the same particle. The relative amount of apoprotein E in particles from diabetic animals was much less than that of cholesterol-fed controls. The ability of the lipoproteins from both groups to stimulate cholesteryl ester formation in mouse peritoneal macrophages also was tested. Lipoproteins from cholesterol-fed controls stimulated cholesteryl ester formation in a dose-dependent manner, but particles from the diabetic group had little or no effect. The results suggest that the presence of unusual VLDL particles in diabetic cholesterol-fed rabbits is responsible, at least in part, for the reduced incidence of atherosclerosis in this animal model.

The relationship of abnormal circulating insulin levels to atherosclerosis

The evidence linking insulin with atherosclerosis can be divided into two parts. First, there is evidence that a proportion of subjects who have atherosclerosis or who are at risk of developing atherosclerosis hav elevated circulating insulin levels. The high insulin levels may be associated with another metabolic abnormality such as obesity, hypertriglyceridaemia, uraemia or consumption of oral contraceptives, may be inappropriate to the blood sugar levels as in mild diabetes, or may be of exogenous origin as in insulin-treated diabetics. The tissues of these subjects are exposed to high concentrations of insulin, and it seems reasonable to suggest that elevated insulin levels may have pathological effects on these tissues. Secondly, there is increasing evidence that the arterial wall is an insulin sensitive tissue. Exposure of arterial tissue to insulin results in proliferation of smooth muscle cells, inhibition of lipolysis, and synthesis of cholesterol, phospholipid and triglyceride. Chronic exposure to high concentrations of insulin results in the development of lipid filled lesions similar to those of early atherosclerosis. Thus, insulin has the ability to promote changes in the artery which, in the long term, may progress to atherosclerosis. The two lines of evidence together suggest that high levels of circulating insulin may have a role in the development of atherosclerosis.

The lipid metabolism of the arterial wall and its abnormalities in diabetes

Arteries are not simply conduits for the transport of blood, but consist of metabolically active tissue which has the capacity to synthesize all the components of the atherosclerotic lesion. The smooth muscle cell appears to be the most important metabolically active cell in the arterial wall. There is little information on arterial metabolism in human diabetes. Experimental diabetes depressed all aspects of arterial lipid metabolism and this effect is reversed by insulin. Insulin promotes changes in arterial metabolism which are similar to those an atherosclerosis. Thus the relationship of human diabetes to the metabolism of the arterial wall is complex and little understood.

Distribution of cholesterol and triglycerides among lipoprotein fractions in fat-fed rabbits at different levels of serum cholesterol

The serum lipoproteins of rabbits given semisynthetic cholesterol-free diets containing coconut oil or butter or a conventional rabbit chow supplemented with cholesterol, were studied by preparative ultracentrifugation and electrophoresis. (1) All three diets elevated the total cholesterol level but only the coconut oil diet markedly increased the triglyceride (TG) content in addition. All ultracentrifugation fractions showed elevated cholesterol/TG ratios, and this was especially evident for the cholesterol diet. In the hyperlipidemic rabbits cholesterol was therefore mainly transported in lipoproteins with a changed lipid composition. (2) The lipid levels of the "HDL" fraction were more or less unaffected by the lipid concentration in whole serum. In the total serum cholesterol ranges 150-500 (coconut oil diet) and 100-300 mg/100 ml (cholesterol diet), most cholesterol was transported as "LDL" cholesterol. This latter fraction reached maximum cholesterol concentrations of about 350 (coconut oil diet) and 400 mg/100 ml (cholesterol diet) at total cholesterol levels of approximately 600 and 1200 mg/100 ml serum, respectively. At still higher levels of total cholesterol, the whole increment was concentrated to the "VLDL" fraction. (3) With semisynthetic diets in the whole cholesterol range 250-400 mg/100 ml it was possible, with respect to cholesterol, to induce fairly similar concentrations and distributions to those seen in man, with about 60% transported as "LDL", 30% as "VLDL" and 10% as "HDL" cholesterol with the coconut oil diet and 65%, 20% and 15%, respectively, with the butter diet. Such experimental conditions seem to be suitable for testing the hypocholesterolemic action of drugs intended for human hyperlipidemia Type II. (4) Compared with earlier investigations on rabbits, the present distribution study suggests that the degree of aortic lipid infiltration in cholesterol-fed rabbits is better related to the levels of "LDL" than to "VLDL" cholesterol.

Actions of vitamins A and E and some nicotinic acid derivatives on plasma lipids and on lipid infiltration of aorta in cholesterol-fed rabbits

Marked hypercholesterolemia and moderate lipid infiltration of the aorta were induced by feeding rabbits a diet containing 1% cholesterol + 3% corn oil for 70 days. In the liver the concentration and pool size of cholesterol increased and those of triglycerides (TG) decreased. On dietary addition of vitamin A and vitamin E (44 000 I.U. and 125 mg respectively, once daily for 5 days a week) the following changes were noted in comparison with the fat-fed rabbits not receiving extra addition of vitamins. There was a slight decrease of the levels of plasma cholesterol and an increase of those of plasma TG. The liver cholesterol concentration increased but, according to the concomitant reduction of the liver weight, there was no significant change in lever cholesterol or TG pools. In the aorta the vitamins markedly reduced the lipid infiltrated area as well as the cholesterol content. Both niceritrol** and S-2040 [pyridine-2,5-dicarboxylic acid di(beta-pyridylcarbinol ester)] in a dietary concentration of 0.5% decreased plasma cholesterol by about 20%. This reduction, as well as that induced by the vitamins, was confined to the VLDL-fractions only. S-2040 slightly reduced the cholesterol accumulation in the aorta. In rabbits given both the vitamins and niceritrol or S-2042 there was an additive reduction of plasma cholesterol. Here the nicotinic acid derivatives were partly able to counteract the increases of plasma TG induced by the vitamins. In the aorta the combination vitamins + S-2042 but not that of vitamins + niceritrol tended to give a better protection than the vitamins alone. On a normal diet vitamins A + E significantly increased the liver cholesterol concentration and pool and decreased the liver TG pool, but did not affect the other parameters. Possible mechanisms for the prophylactic action of the vitamins against lipid infiltration of the aorta of cholesterol-fed rabbits are discussed.

The effect of alloxan diabetes on experimental cholesterol atherosclerosis in the rabbit. IV. The effect of insulin therapy on the inhibition of atherosclerosis in the alloxan-diabetic rabbit

Experiments were undertaken to ascertain whether the previously demonstrated inhibition of the development of experimental aortic atherosclerosis in alloxan-diabetic rabbits fed cholesterol was due to the injection of alloxan per se or to the existence of the diabetic state produced by alloxan. It was established that, by treating the diabetic state with insulin, the diabetic state could be ameliorated and the inhibitory effect obviated. It was therefore concluded that the inhibitory phenomenon was not due to the injection of alloxan per se but that it was associated with one or more factors that characterize the alloxan diabetic state in the rabbit and that are reversible by insulin therapy. In the course of the experiment it was demonstrated that the inhibitory effect was apparent in cholesterol-fed diabetic rabbits whether or not their diet was supplemented with vegetable oil. The previously reported metabolic abnormalities of the diabetic animals were confirmed. It was established that suitable treatment of the cholesterol-fed diabetic animals with insulin would bring all the metabolic aberrations, including those of the serum lipids, into reasonably close correspondence with those observed in non-diabetic rabbits fed cholesterol.

The mechanism of alloxan protection in experimental atherosclerosis

Experiments were performed to compare the effects of cholesterol feeding in (a) control rabbits, (b) alloxan-diabetic rabbits, and (c) rabbits injected with alloxan while the pancreas was temporarily occluded from the circulation. The alloxan-diabetic rabbits consumed significantly higher quantities of cholesterol and food and had serum cholesterol and lipoprotein (S(f) 5-9 and S(f) 16-30) concentrations significantly increased over the control levels. They failed to show a commensurate increase in the degree of atherosclerosis. Rabbits in which the diabetogenic action of alloxan was prevented by temporary occlusion of the pancreas from the circulation during its administration developed grades of hypercholesterolemia, hyperlipoproteinemia, and atherosclerosis not significantly different from the controls. The results are interpreted as indicating that the effects of alloxan on tissues other than the pancreas do not protect against experimental atherosclerosis produced by cholesterol feeding.