Compared With Dietary Monounsaturated and Saturated Fat, Polyunsaturated Fat Protects African Green Monkeys From Coronary Artery Atherosclerosis
Abstract Atherogenic diets enriched in saturated, n-6 polyunsaturated, and monounsaturated fatty acids were fed to African green monkeys for 5 years to define effects on plasma lipoproteins and coronary artery atherosclerosis. The monkeys fed polyunsaturated and monounsaturated fat had similar plasma concentrations of LDL cholesterol, and these values were significantly lower than for LDL in the animals fed saturated fat. Plasma HDL cholesterol concentrations were comparable in animals fed saturated and monounsaturated fat and were significantly higher than in animals fed polyunsaturated fat. Thus, the monounsaturated fat group had the lowest LDL/HDL ratio. LDL particle size was largest in the saturated and monounsaturated fat groups, significantly larger than in the polyunsaturated fat group. LDL particle enrichment with cholesteryl oleate was the greatest in the animals fed monounsaturated fat, next greatest in the saturated fat–fed animals, and was least in the polyunsaturated fat–fed animals. Coronary artery atherosclerosis as measured by intimal area was less in the polyunsaturated fat compared with the saturated fat groups, was less in the animals fed polyunsaturated fat compared with the monounsaturated fat–fed animals, but did not differ between the monounsaturated and saturated fat groups. Cholesteryl ester, particularly cholesteryl oleate, accumulation in the coronary arteries was also similar between groups fed monounsaturated and saturated fat but was minimal in the animals fed polyunsaturated fat. In sum, the monkeys fed monounsaturated fat developed equivalent amounts of coronary artery atherosclerosis as those fed saturated fat, but monkeys fed polyunsaturated fat developed less. The beneficial effects of the lower LDL and higher HDL in the animals fed monounsaturated fat apparently were offset by the atherogenic shifts in LDL particle composition. Dietary polyunsaturated fat appears to result in the least amount of coronary artery atherosclerosis because it prevents cholesteryl oleate accumulation in LDL and the coronary arteries in these primates.
Presented in part at the Pennington Biomedical Research Symposium on Nutrition, Genetics, and Heart Disease, Baton Rouge, La, March 12-14, 1995.
- Received August 7, 1995.
- Accepted September 21, 1995.
The present studies address the question of whether substitution of monounsaturated fatty acids into the diet in place of saturated fatty acids will as effectively limit the development of CAA as the substitution of polyunsaturated fatty acids. For about three decades, the population in our country has undergone a gradual shift in dietary fatty acid composition with the substitution of vegetable oils rich in linoleic acid for animal fat.1 More recently, since the seminal studies of Mattson and Grundy,2 interest in monounsaturated fatty acids and their potential effectiveness as substitutes for saturated fatty acids has increased. While we are aware of many studies on the effectiveness of monounsaturated compared with polyunsaturated fatty acids on the plasma lipid and lipoprotein profiles associated with premature CHD,2 3 4 we are not aware of any studies in primates that directly compare monounsaturated fatty acids with polyunsaturated and saturated fatty acids for their effects on CAA per se. We are concerned about this because of the data showing that enrichment of plasma lipoproteins with cholesteryl oleate occurs in inverse proportion to cholesteryl linoleate in monkeys and the associated increase in CAA.5 6 7 Additionally, evidence shows lower proportions of cholesteryl linoleate in plasma lipoproteins in human patients suffering from CHD8 9 10 11 12 13 and an inverse relation between proportions of linoleate and oleate in serum lipids.8
To obtain data on CAA directly, we have used a primate model, the African green monkey, in which remarkable similarities to human beings in dietary fatty acid responsiveness of plasma lipoproteins14 15 16 17 and the association of high LDL-C and low HDL-C with exacerbated CAA7 16 17 18 19 have been documented. A disadvantage of this species is that atherosclerotic lesion development does not occur as rapidly as in other primate species, but a major advantage is that the atherosclerotic lesions are more similar to those seen in humans, both in distribution and in composition, than for other primate species.
Three groups of monkeys were selected that had equivalent mean±SD values for total cholesterol and HDL-C in response to a challenge with a diet containing saturated fat and cholesterol.20 After a monkey-chow washout period, the animals were then started on diets enriched with one of three individual fatty acids, ie, monounsaturated (18:1), polyunsaturated (18:2), or saturated (16:0) fatty acids. Two different periods of dietary fatty acid enrichment were actually performed, with the second period being the longest (≈3.5 years) and having the most striking degree of individual fatty acid enrichment, with over 70% of 18:1 or 18:2 and 40% 16:0. Data comparing the effects of the different diets on plasma lipid and lipoprotein endpoints are available.20 The present article details lipid and lipoprotein data collected throughout the experiment and presents the findings for CAA.
The animals studied included 44 feral adult male African green monkeys, Cercopithecus aethiops. The characteristics of the animal groups established for these studies have been documented in detail, and the compositions of the three different experimental diets are available.20 Briefly, the three experimental groups were selected to have similar plasma cholesterol concentrations in response to a common challenge diet that was enriched with saturated fat and cholesterol. After the challenge diet period and a 10-week washout period, each group was fed for about 5 years with experimental diets containing 35% of kilocalories as fat and 0.8 mg cholesterol/kcal; the diets differed only in fatty acid composition. Two diet periods with different degrees of fatty acid enrichments were used, with the most exaggerated being used for the longest period of the study (period 2), which included the final 42 months of the study. During period 2, the saturated fat diet (primary source of fat, palm oil) contained 40% of the fatty acids as palmitic acid. The monounsaturated fat diet (primary source of fat, oleic acid–enriched safflower oil) contained over 70% of the fatty acids as oleic acid. The polyunsaturated fat diet (primary source of fat, safflower oil) contained over 70% of the fatty acids as linoleic acid.
Body weights in each animal were measured monthly throughout the study. Average values (mean±SEM) rose slightly throughout the study (an average rate of 0.1 kg/y) and averaged 6.1±0.4, 6.1±0.4, and 6.5±0.6 kg for the saturated, monounsaturated, and polyunsaturated fat groups, respectively; no statistically significant differences among groups were found. Over the 5 years of the study, 2 animals from each diet group died. Necropsies were performed, and pneumonia (1), enteric disease (4), and kidney failure (1), each apparently unrelated to the diet exposure, were identified as causes of death. Animals were housed throughout the study in an AALAC (American Association for the Accreditation of Laboratory Animal Care)-accredited facility in adjacent individual cages in rooms where as many as 24 animals faced each other. Animals were fed weighed portions of food twice daily to assure that they received 100 kcal·kg−1·d−1 and raw carrots were given three times a week as a snack. In addition, a veternarian and/or an animal technician inspected all animals twice daily. All procedures were approved by the animal care and use committee of this institution.
Plasma was isolated from blood drawn into EDTA-containing vacuum tubes from animals that were sedated with ketamine hydrochloride (10 mg/kg) after an overnight fast.20 The frequency of blood sampling was bimonthly (occasionally more frequent) for routine measurements of total cholesterol,21 TGs,22 and HDL-C.23 Aliquots of plasma samples from seven selected bleedings across period 2, which had been stored at −80°C from the time of collection, were analyzed for apolipoproteins B, E, A-I, and A-II concentrations by enzyme-linked immunosorbent assays.24 25
On three different occasions during the last 36 months of the study, plasma was collected from each animal for a more detailed lipoprotein characterization. Isolation and separation of lipoproteins was done20 according to published methods,26 and LDL-C concentrations were measured directly. An agarose column packed with 4% agarose (BioGel A 15m, 200 to 400 mesh, BioRad, Inc) was used for lipoprotein separation when larger amounts of isolated purified lipoprotein were needed for chemical and physical characterizations; a Superose 6 column (Pharmacia Biotech, Inc) was used when cholesterol distributions were the primary endpoint. Measurement of LDL particle size, expressed as molecular weight in grams per micromole, was accomplished during the chromatographic separation of lipoproteins.27 Purity of lipoprotein classes isolated chromatographically is excellent26 ; albumin contamination is absent, and LDL and HDL do not contaminate each other.
Chemical endpoints were determined by using enzymatic assays after extraction and solubilization of the lipids in an aqueous buffer.28 For determination of individual CE compositions in lipid extracts, lipid classes were separated by using thin-layer chromatography with Empore TLC sheets (Analytichem International). The CE bands were scraped off the plate, and after saponification the fatty acids were methylated and separated by gas liquid chromatography.14
LDL Tm was measured by using a MicroCal ultrasensitive differential scanning calorimeter (MicroCal Inc) interfaced to a personal computer. LDL preparations from the agarose column were concentrated to ≥5 mg cholesterol/mL by using Centriflo ultrafiltration membranes (Amicon). Solutions were degassed under vacuum. The sample cell was filled with 1.3 mL of the LDL sample, and the reference cell was filled with an equal volume of column buffer (0.9% NaCl, 0.01% EDTA, and 0.01% NaN3, pH 7.4). The samples were preequilibrated at 5°C for 45 minutes before initiation of the heating scan from 5°C to 60°C at a rate of 90°C/h. During the preequilibration run the sample was maintained under a positive-pressure nitrogen atmosphere. Peak Tms were determined digitally from the computer output of each run. Tm accuracy of the instrument was monitored with dimyristoyl phosphatidylcholine vesicles.
For quantification of CAA, hearts were removed at necropsy and were perfusion-fixed in 10% neutral-buffered formalin at a pressure of 100 mm Hg for 1 hour. The LAD was then dissected free from the heart, cleaned of adventitia, and stored in formalin at 4°C until it was extracted in chloroform/methanol (2:1) for determination of lipid composition. The remainder of the heart was stored in formalin and used for histological evaluation of atherosclerosis. Five 3-mm serial sections were trimmed from each of the remaining two main coronary arteries (the right coronary and the left circumflex arteries), which were then dehydrated and embedded in paraffin. Slides were made with coronary arteries in cross section and were stained with Verhoeff-van Gieson’s stain. CAA was evaluated morphometrically as the intimal area (area in square millimeters between the IEL and the lumen) and the percent lumen stenosis (cross-sectional intimal area divided by the area within the IEL×100). For each animal values were expressed as the mean for the 10 sections of coronary artery evaluated per heart. In a separate study in African green monkeys, evaluation of mean intimal area in the right coronary artery, LAD, and left circumflex artery showed correlation coefficients >.85 when comparing intimal areas in any two arteries (L.L.R., J.K.S., unpublished data, 1995). The data suggest that the use of one of the coronary arteries for chemical analysis of atherosclerosis extent will not adversely affect the outcome of morphometric evaluations and will reflect a comparable extent of atherogenesis.
Several different statistical analyses were done. All group comparisons of outcome measures (or transformations) among diet groups were made by ANCOVA. This analysis was adjusted for prerandomization values of factors that were significantly associated with the endpoint of interest and were done to correct for chance imbalances among groups in prognostic factors and to reduce the variance of the estimate of group differences, thereby increasing the power of the comparisons. For atherosclerosis measurements, TPC response during the challenge diet period was a significant prognostic factor and was used as the covariate for the atherosclerosis endpoints. For the area inside the IEL of the coronary arteries, body weight was used as the covariate. Post hoc analyses were done with Fisher’s preferred least-significant differences test. In most cases, the probability of statistical significance is as indicated. Any difference greater than P=.1 was considered nonsignificant, but in most cases the probability is still indicated for reference. Pearson correlations and regression analyses were performed to determine significant prognostic factors.
The TPC data collected bimonthly for each group illustrate the constancy among the relative differences (Fig 1⇓). At all times the animals fed saturated fat had the highest TPC values, the values for animals fed monounsaturated fat were intermediate, and the animals fed polyunsaturated fat had the lowest TPC concentrations. When the animals were shifted during period 2 to diets with the greatest enrichments of 16:0, 18:1, or 18:2, the differences among the groups were accentuated. No change in the relative positioning was seen at any point of observation while the experimental diets were fed except when the cholesterol was mistakenly left out of one batch of the saturated fat diet early in period 2 (months 26 and 27). TPC was 422±9, 313±11, and 276±8 mg/dL for the saturated, monounsaturated, and polyunsaturated fat groups, respectively, during period 2, and 420±6, 345±9, and 293±8 mg/dL (mean±SEM) for the entire period of diet exposure. Repeated-measures ANCOVA showed that the difference between the saturated fat and monounsaturated fat groups was significant (P<.03) during period 2, as was the difference between the saturated and polyunsaturated fat groups (P<.002). The difference in TPC between the polyunsaturated and monounsaturated fat groups, although consistently observed, did not reach significance (P=.16 by ANCOVA).
Plasma triacylglycerol concentrations were also monitored bimonthly throughout the study. The values were very low for all groups, as is typical for monkeys fed high-fat diets, but the triacylglycerol concentrations for the polyunsaturated fat group were consistently the lowest, with the mean±SEM of the individual animal means being 15±3 mg/dL and the range in mean values on different dates being only 9 to 19 mg/dL. The means of the individual animal means for plasma triacylglycerol concentration in the saturated and monounsaturated fat groups were comparable, 26±4 and 24±7 mg/dL, with the range in mean values on different dates being 20 to 33 and 15 to 37 mg/dL for these diet groups, respectively. Repeated-measures ANCOVA showed that the triacylglycerol concentrations in the polyunsaturated fat group compared with those in the other two groups were significantly lower (P<.01) during period 2.
HDL-C data were gathered bimonthly throughout periods 1 and 2 (Fig 2⇓). The HDL-C concentration in the monounsaturated fat group consistently had the highest mean value, 87±2 mg/dL, and was significantly higher than that for the saturated fat group (75±1 mg/dL; P=.06) and that for the polyunsaturated fat group (48±1 mg/dL; P=.0001). The HDL-C in the polyunsaturated fat group was significantly lower than in the saturated fat group (P=.0008). The average of individual animal means for HDL-C in period 2 alone was 47±1 mg/dL for the polyunsaturated fat group, 90±2 mg/dL for the monounsaturated fat group, and 80±2 mg/dL for the saturated fat group.
ApoA-I and apoA-II concentrations were also measured in a total of seven plasma samples taken from each animal during period 2 (Table 1⇓). The means for apoA-I concentration were 200, 261, and 272 mg/dL for the polyunsaturated, saturated, and monounsaturated fat groups, respectively. The difference between apoA-I concentration in the polyunsaturated fat group and that in the saturated and monounsaturated groups was statistically significant as measured by repeated-measures ANCOVA. No significant difference in apoA-I concentration was found between the saturated and monounsaturated fat groups. For apoA-II, the values were low, averaging 18, 21, and 26 mg/dL for the polyunsaturated, saturated, and monounsaturated fat groups, respectively, with the only significant difference in apoA-II concentrations being between the polyunsaturated and monounsaturated fat groups.
Whole plasma apoB concentrations were also measured periodically on seven different occasions during period 2. The averages were 153, 125, and 114 mg/dL for the saturated, polyunsaturated, and monounsaturated fat groups, respectively; these values were significantly different by ANCOVA (Table 1⇑). Post hoc analyses showed that the apoB concentration for the monounsaturated fat group was significantly lower than for the saturated fat group, and apoB in the saturated fat group was significantly higher than in the polyunsaturated fat group. The values for apoB concentration in the monounsaturated and polyunsaturated fat groups did not differ.
ApoE concentrations were measured in the same plasma samples as for other apolipoproteins. The highest average apoE concentration (8.1 mg/dL) was found in the saturated fat group and was significantly higher than the apoE concentrations in the other two groups, which averaged <6 mg/dL (NS).
Plasma LDL was isolated and characterized in pure form. Dietary fatty acid effects on particle composition and size were numerous (Table 2⇓). The percentage of plasma cholesterol as LDL-C was lower in the monounsaturated fat group than in the other two groups, which had equivalent percentages. LDL-C concentrations in the polyunsaturated and monounsaturated fat groups were similar, with means of 166 and 184 mg/dL, respectively; the LDL-C concentration in the saturated fat group averaged 272 mg/dL. During the chromatographic isolation of LDL, the average LDL particle size was measured as LDL molecular weight. The LDL particles were large in the saturated fat group, were of comparable size in the monounsaturated fat group, and were smaller in the polyunsaturated fat group. The LDL-C–to-protein ratios, which generally reflect the particle size, were higher in the saturated and monounsaturated fat groups compared with those in the polyunsaturated fat group (Table 2⇓).
The LDL Tms were measured by differential scanning calorimetry (Table 2⇑). The saturated fat animals had the highest temperatures, the monounsaturated fat group had intermediate LDL Tm, and the polyunsaturated fat group had the lowest LDL Tm. Because the TG content is very low, Tm is determined in these LDL particles by the CE composition in the core lipid.
The LDL of the saturated fat animals contained about 650 CE molecules with saturated fatty acids, which was 1.7 to 1.8 times their number in the monounsaturated and polyunsaturated fat groups (Table 2⇑). The LDL particles from the monounsaturated fat group were greatly enriched in cholesteryl oleate molecules (≈2000 per LDL particle), which was a much higher number than for either of the other two groups. The LDL of the saturated fat group still had more than twice the number of cholesteryl oleate molecules than the LDL of the polyunsaturated fat animals. The number of cholesteryl linoleate molecules was highest in the LDL of the polyunsaturated fat group, being almost three times the number in the LDL of the monounsaturated fat animals and twice as many as the LDL of saturated fat animals.
CAA in each of the animals was quantified chemically and histologically. While none of the animals were maintained on a cholesterol-free diet, several studies have been completed in which low-cholesterol diet groups were maintained for as long as 5 years without CAA development, even in animals that were fully mature throughout the study.7 16 Therefore, we are confident that the atherosclerosis quantified for this study developed during the period of diet induction.
The data in Fig 3⇓ show the largest coronary artery lesions (based on intimal area) from an animal selected from each diet group. These lesions show the characteristics of atherosclerosis that developed in the animals of this study. The differences in size (measured as intimal area) for these three sections, 1.60, 0.76, and 0.31 mm2, closely paralleled the mean values for the CAIA for the saturated, monounsaturated, and polyunsaturated diet groups. The most advanced lesion was found in an animal from the saturated fat group. This is a very complicated lesion with a large amount of extracellular debris. Evidence of what appear to be several waves of lesion development are provided by the pattern of cellularity and necrosis. A large necrotic core is present in the lesion that includes calcification. The IEL is disrupted at a number of points, and medial damage is significant. The large size of the lesion has significantly narrowed the lumen of the artery.
The lesion from the animal fed monounsaturated fat is also an advanced lesion. Many aspects of lesion complication are apparent, including intimal accumulation of extracellular lipid, cholesterol clefts, macrophage infiltration, fibrous tissue accumulation, IEL disruption, and medial damage. The thickness of this lesion is not as great as for that of the saturated fat animal; this appears to be partly due to the ectasia (increase in coronary artery lumen diameter) that has occurred in this artery in response to the atherogenic process actively under way. Artery enlargement in the face of significant lesion development appears to have maintained lumen diameter at nearly a normal size. The lesion in the polyunsaturated fat–fed animal is also a complicated lesion with all the hallmarks of fibrous tissue accumulation, extracellular lipid accumulation, IEL disruption, and evidence of various cellular responses during lesion development. The extensiveness of this lesion is more limited than for those of the other two diet groups, covering only about 50% of the surface of the artery.
The amount of ectasia in each diet group was evaluated by examining the area (in millimeters squared) within the IEL. The average IEL area was smallest in the polyunsaturated fat group (0.97±0.1 mm2), larger in the monounsaturated fat group (1.34±0.12 mm2), and largest in the saturated fat group (1.51±0.16 mm2). ANCOVA, with body weight as the covariate to account for differences in animal size, indicated that the IEL area was significantly smaller in the polyunsaturated fat group than in either the saturated fat group (P=.005) or the monounsaturated fat group (P=.05). The average IEL areas in the monounsaturated fat and saturated fat groups were not significantly different (P=.35).
The data in Fig 4⇓ illustrate the average CAIA as a measure of the extent of atherosclerosis that developed in each of the animals. Because of the remarkable differences among individuals in atherogenic response (about 4 orders of magnitude), the intimal areas are plotted on a log scale. The heights of the bars, rank ordered from the worst to least according to the amount of intimal area, indicate that the extent of CAIA in the animals fed monounsaturated fat was similar to that in the animals fed saturated fat. In the polyunsaturated fat group, 7 of 12 (58%) animals had measurable CAIA, whereas 11 of 13 (85%) and 10 of 13 (77%) animals in the monounsaturated and saturated groups, respectively, had measurable CAIA. The difference in CAIA between the polyunsaturated fat– and saturated fat–fed animals was P=.07 by ANCOVA, while the level of significance for the difference in CAIA between the polyunsaturated fat– and monounsaturated fat–fed animals was P=.09. The comparison between the CAIA for monounsaturated and saturated fat animals showed no difference (P=.89). The patterns shown in Fig 4⇓ and the statistical analyses both support the conclusion that the extent of CAA was similar in the animals fed saturated and monounsaturated fats, and greater for both groups than that in the animals fed polyunsaturated fat.
A third morphometric measurement, percent lumen stenosis, was made. Percent stenosis, which is calculated from the intimal area divided by IEL area×100, estimates the degree of lumen narrowing as a result of lesion development. Stenosis is an indicator of the severity of atherosclerosis; the percent lumen stenosis of the most significant lesion from each of the study animals is graphed in Fig 5⇓. The most affected group was the saturated fat group, with 7 animals with >10% lumen narrowing, 4 with >20%, and with the most affected animal having about 80% lumen stenosis. In the monounsaturated fat group 4 animals exceeded 10% lumen stenosis and 3 were >25%. Three animals were >10% but only 1 was >20% in the polyunsaturated fat group; 5 animals in this group had <1% lumen narrowing. By ANCOVA, P=.08 for group difference; P=.05 for percent stenosis in the animals fed saturated versus polyunsaturated fat; P=.18 for percent stenosis in the animals fed saturated versus monounsaturated fat; and P=.6 for percent lumen stenosis in the animals fed monounsaturated versus polyunsaturated fat.
Data for CE accumulation in the LAD are shown for each of the animals in Fig 6⇓. CE concentration tended to be higher in the animals fed saturated fat (mean value, 10.8 mg/g) and monounsaturated fat (mean value, 9.6 mg/g) than in the animals fed polyunsaturated fat (mean value, 5.5 mg/g). The most affected animals in the monounsaturated fat group had a similar amount of CE as those in the saturated fat group. The pattern of coronary artery CE accumulation suggests that a diet effect was present, but the difference was not significant (P=.3 by ANCOVA). Although the amount of free cholesterol that accumulated in the coronary arteries is high in this primate species, no real trend for a difference between diet groups in free cholesterol concentration existed (P=.9 by ANCOVA).
Accumulation of individual CEs in the coronary arteries also was monitored (Table 3⇓). Two consistent diet-related differences were the greater amounts of CEs with saturated fatty acids (myristate, palmitate, and stearate) in the coronaries of the saturated fat–fed group and the greater amount of cholesteryl oleate accumulation in the coronaries of the monounsaturated fat–fed animals. More cholesteryl oleate accumulated than any other CE in the two diet groups with the most atherosclerosis. The amount of cholesteryl oleate that accumulated in the coronary arteries of the animals fed monounsaturated fat appeared to be as great or greater than that in the coronaries of the saturated fat animals. Four of the 5 animals with the higher accumulation of cholesteryl oleate (>10 mg/g) in the coronary arteries were in the monounsaturated fat group (Fig 7⇓). In the saturated fat group, 6 of 13 animals had cholesteryl oleate concentrations >5 mg/g. The amount of cholesteryl oleate in the coronary arteries of the polyunsaturated fat animals was modest by comparison, with the highest values being 3.2 and 3.3 mg/g.
Relations between the various lipoprotein and atherosclerosis measurements were examined with correlation analysis recognizing that group size was small. Various correlation matrices were constructed. With a few exceptions, depending on the endpoints being compared, the correlation coefficients were highest in the saturated fat group (Table 4⇓). The correlation between coronary artery CE concentration was positive and significant for comparisons of TPC, LDL-C, and LDL molecular weight. The correlation between coronary artery CE and HDL-C was negative, just below statistical significance. An interesting outcome is illustrated for the comparison of coronary artery CE with LDL Tm. The correlation was significant (r=.62, P<.01) in the monounsaturated fat group, lower (r=.4) in the saturated fat group, just missing significance, and no correlation was found in the polyunsaturated fat group (r=.1). The correlation was strongest in the monounsaturated fat group, the only group in which the LDL Tm spanned body temperature, indicating that the physical state of the core lipids of LDL may be an important endpoint in determining CAA extent.
The data in Fig 8⇓ illustrate the relation of LDL particle size with coronary artery cholesterol ester concentration as a measure of atherosclerosis extent. LDL size and CAA have been well correlated,5 7 16 and this was true in these studies. All groups in Fig 8⇓ appear to fit the same regression line, with more animals with higher LDL particle size being found in the saturated and monounsaturated fat groups. In no case did the regression lines for the individual groups appear to differ significantly. The regression line that fit the entire data set had a correlation coefficient of r=.7, and the correlation coefficients for individual groups were r=.78 for the polyunsaturated fat group, r=.7 for the monounsaturated fat group, and r=.64 for the saturated fat group. This comparison suggests that the extent of atherosclerosis is highly related to factors that affect LDL particle size, and cholesteryl oleate accumulation is a major contributor to this endpoint.
The goal of this study was to identify, in a relevant primate model, whether the recommendation to add monounsaturated fat to our diets to replace either saturated or polyunsaturated fat would positively benefit CAA, per se. We found that dietary enrichment of monounsaturated fatty acids had the effect of keeping LDL-C lower and HDL-C higher in the primate model, a very similar outcome to that described for humans by Mattson and Grundy.2 At the same time, many of the changes in LDL particle composition induced by monounsaturated fat were noted to be those associated with exacerbated atherosclerosis in the primate model. When CAA was quantified at the end of the 5 years of atherosclerosis induction, the atherosclerosis extent in the monounsaturated fat group was similar to that in the saturated fat group, while the polyunsaturated fat-fed animals had less (Figs 3⇑, 4⇑, 6⇑, and 7⇑). This outcome occurred despite the lower HDL-C in the polyunsaturated fat group (Fig 2⇑) and suggests that the changes in LDL particle composition induced by monounsaturated fat (larger LDL particles enriched in cholesteryl oleate with higher Tm) may be a more important determinant of atherogenicity than has been previously suspected.
To provide a thorough examination of the quality, quantity, and severity of atherosclerosis, we examined the coronary arteries for several morphometric endpoints, CE content, and types of CEs present. While variability was great and statistical significance in all these endpoints was not achieved, consistent trends were noted, including the accumulation in the coronaries of cholesteryl oleate as the major CE in the saturated and monounsaturated fat groups (Fig 7⇑ and Table 3⇑), ie, the groups with the most extensive atherosclerosis. Cholesteryl oleate is the primary product of tissue (arterial and hepatic) ACAT when oleyl coenzyme A is available as substrate,29 as it is for the saturated and monounsaturated fat groups. The acyl coenzyme A pools appear to have been enriched in linoleyl coenzyme A in the group fed polyunsaturated fat so that, in this case, the ACAT product was enriched in cholesteryl linoleate5 (Table 3⇑). LDL in the animals fed monounsaturated fat was greatly enriched in cholesteryl oleate, which shifted the LDL Tm above body temperature (Table 2⇑). If CE droplets within lesions have transitions above body temperature, the ability to hydrolyze and remove these droplets would be decreased.30 Both the LDL CEs that were deposited in the intima and those synthesized by arterial ACAT will be so affected in the monounsaturated and saturated fat groups but not the polyunsaturated fat group. We suspect that this physical-state difference may be a key contributor to the differences in atherosclerosis that we found. It may be noted that free cholesterol accumulation in the polyunsaturated fat group was equivalent to that in the other two groups (Fig 6⇑), perhaps suggesting that CE clearance from coronary arteries in this diet group is proceeding, and perhaps even more efficiently than in the other two groups based on the higher free cholesterol–to–cholesteryl ester ratio.
In the current climate, in which interest in LDL oxidation is high, the present findings are of particular interest. In monkeys fed the same fats as described here, in vitro oxidation of the LDL of animals fed mononsaturated fat is more difficult than oxidizing the LDL of monkeys fed polyunsaturated fat,31 even though the vitamin E contents of the LDL particles of the three diet groups are similar.32 Similar observations on the susceptibility to oxidation have been made with human LDL33 34 and are likely due to the dramatic shifts in fatty acid composition in LDL lipids induced by the type of dietary fat. The present findings suggest that avoiding dietary polyunsaturated fat by substituting monounsaturated fat as a means of protecting against LDL oxidation, as has been recommended,33 34 will be counterproductive for protection from CAA. Furthermore, enrichment of LDL with cholesteryl linoleate, as occurs when diets rich in linoleic acid are fed, has repeatedly been shown in humans to be associated with decreased CHD.8 9 10 11 12 13 Animals fed linoleic acid had a higher proportion of CEs as cholesteryl linoleate in LDL (Table 2⇑) and in the CEs in the coronary arteries (Table 3⇑). Yet, even with this enrichment in a more easily oxidized CE, the amount of atherosclerosis was less. Polyunsaturated fat–rich diets as compared with saturated fat–rich diets have been repeatedly shown to decrease CAA in primates6 7 16 18 35 ; no such effect of monounsaturated rich–diets in any primate has yet been demonstrated.
The origin of the changes in LDL composition induced by dietary monounsaturated fat has been suggested by isolated liver perfusion studies. CE secretion is increased when liver CE content is increased,36 37 liver CE secretion is proportional to hepatic ACAT activity,5 and liver-derived CEs reflect the fatty acid composition of the diet when comparing polyunsaturated and saturated fats.37 Likewise, CE enrichment of LDL is proportional to hepatic ACAT activity.5 Therefore, it is our hypothesis that the cholesteryl oleate–enriched LDL that accumulates as enlarged LDL particles in the plasma of monkeys fed monounsaturated fat are derived from cholesteryl oleate–enriched LDL precursor lipoproteins secreted by the livers of these animals.
The reason that plasma LDL-C concentrations were not as high in the monounsaturated fat group as in the saturated fat group may be explained by the hypothesis made by Woollett and colleagues38 based on work in hamsters. Compared with other diets, monounsaturated fatty acid–rich diets cause significantly higher liver CE concentrations in monkeys (L.L.R., J. Haines, R. Shah, J.K.S., unpublished data, 1995) and hamsters,38 likely as a result of enhanced ACAT that perhaps results from increased availability of the preferred substrate, oleyl coenzyme A. The apparent effect is to decrease the size of the putative hepatic regulatory pool of cholesterol, leaving hepatic LDL receptor levels higher and plasma LDL levels lower.38
In humans, LDL particle size is inversely related to plasma TG concentrations; monkeys fit at the extreme low end of the TG concentration spectrum, where the high lipoprotein lipase level found in nonhuman primates fed fat-rich diets appears to keep plasma TGs low.39 In monkeys, the CE-enriched LDL precursor particles secreted by the liver undergo less intravascular modification (TG-for-CE exchange, with subsequent TG removal via lipoprotein lipase40 ) than in humans, since the residence time of the TG-rich VLDL (and chylomicron) particles is less, even though CE transfer activity is high in monkeys41 and could easily facilitate this compositional change. Nevertheless, the extent of LDL-C enrichment (measured as LDL-C–to-protein in Table 2⇑) overlaps the range described as LDL-C–to-apoB for humans.42
A similar increase in secretion of hepatic CE in apoB-containing lipoproteins, presumably as ACAT-derived cholesteryl oleate, could occur in humans with increased risk for CHD. This likely would be expressed as a high cholesterol-to-apoB ratio in LDL, which has been identified in selected CHD patients.42 The presence in plasma of cholesteryl oleate–enriched LDL precursor lipoproteins as well as LDL, per se, appears more atherogenic from this study. This finding may be closely related to the reason that humans with higher percentages of cholesteryl linoleate (and lower proportions of cholesteryl oleate) in LDL have a lower incidence of CHD.8 9 10 11 12 13 Unfortunately, the measurement of CE composition has not been made in several of the more recent studies of human LDL,42 although LDL particles become enriched in cholesteryl oleate when diets rich in oleate are fed, indicating that humans and monkeys respond similarly.33 43
Finally, this is another study in which a group of monkeys with lower HDL-C concentrations (the polyunsaturated fat group) had less atherosclerosis than groups with significantly higher HDL.16 17 18 19 This is the first study in which we have shown that a diet group of monkeys with HDL-C concentrations higher than in other groups and LDL-C concentrations lower than in other groups was not protected by this lipoprotein profile. These findings suggest that the importance of HDL as a protective factor for CAA is secondary to the degree that LDL particles are atherogenic. The compositional changes in LDL particles induced by the monounsaturated fat–rich diet, even in the absence of elevated LDL-C concentrations, were sufficient to overcome higher, presumably more beneficial HDL levels.
Selected Abbreviations and Acronyms
|ACAT||=||acyl coenzyme A:acyltransferase|
|CAA||=||coronary artery atherosclerosis|
|CAIA||=||coronary artery intimal area|
|CHD||=||coronary heart disease|
|IEL||=||internal elastic lamina|
|LAD||=||left anterior descending coronary artery|
|MONO||=||monounsaturated fat diet|
|POLY||=||polyunsaturated fat diet|
|SAT||=||saturated fat diet|
|TPC||=||total plasma cholesterol|
This work was supported by National Institutes of Health grants HL 24736 and HL 49373 from the Heart, Lung, and Blood Institute. The authors wish to thank Dr Ed Hunter and the Institute for Edible Shortenings and Oils for the donation of the high–oleic acid safflower oil. The excellent technical assistance of Ramesh Shah, Jeff Haines, and Dr Martha Wilson is greatly appreciated, as is the help in manuscript preparation by Linda Odham. The assistance provided by Dr Bill Bullock in the pathological evaluations of the animals is also appreciated. The assistance of biostatistician Dr Timothy Morgan in experimental design and statistical analysis is also gratefully acknowledged.
Nutrition Monitoring in the United States. Hyattsville, Md: Public Health Service. US Dept of Health and Human Services; 1989:1255.
Mattson FH, Grundy SM. Comparison of effects of dietary saturated, monounsaturated and polyunsaturated fatty acids on plasma lipids and lipoproteins in man. J Lipid Res. 1985;26:194-202.
Mensink RP, Katan MB. Effect of dietary fatty acids on serum lipids and lipoproteins: a meta-analysis of 27 trials. Arterioscler Thromb. 1992;12:911-919.
Gustafsson I-B, Vessby B, Ohrvall M, Nydahl M. A diet rich in monounsaturated rapeseed oil reduces the lipoprotein cholesterol concentration and increases the relative content of n-3 fatty acids in serum of hyperlipidemic subjects. Am J Clin Nutr.. 1994;59:667-674.
Carr TP, Parks JS, Rudel LL. Hepatic ACAT activity in African green monkeys is highly correlated to plasma LDL cholesteryl ester enrichment and coronary artery atherosclerosis. Arterioscler Thromb. 1992;12:1274-1283.
Rudel LL, Carr TP. Modifications by dietary polyunsaturated fat of lipoproteins and atherosclerosis in primates. In: Serneri GGN, Gensin GF, Abbate R, Prisco D, eds. Atherosclerosis Inflammation and Thrombosis. Florence, Italy: Scientific Press; 1993:201-211.
Rudel LL, Johnson FL, Sawyer JK, Wilson MS, Parks JS. Dietary polyunsaturated fat modifies low density lipoproteins and reduces atherosclerosis of nonhuman primates with high and low diet responsiveness. Am J Clin Nutr.. 1995;62:463S-470S.
Kingsbury KJ, Brett C, Stovold R, Chapman A, Anderson J, Morgan DM. Abnormal fatty acid composition and human atherosclerosis. Postgrad Med J. 1974;50:425-440.
Schrade W, Boehle E, Frankfurt MD, Biegler R. Humoral changes in arteriosclerosis: investigations of lipids, fatty acids, ketone bodies, pyruvic acid, lactic acid, and glucose in the blood. Lancet. 1960;1:1409-1416.
Lewis B. Composition of plasma cholesterol ester in relation to coronary-artery disease and dietary fat. Lancet. 1958;1:71-73.
Wolfe MS, Parks JS, Morgan TM, Rudel LL. Childhood consumption of dietary polyunsaturated fat lowers risk for coronary artery atherosclerosis in African green monkeys. Arterioscler Thromb. 1993;13:863-875.
Babiak J, Lindgren F, Rudel LL. Effects of saturated and polyunsaturated dietary fat on the concentrations of HDL subpopulations in African green monkeys. Arteriosclerosis. 1988;8:22-32.
Rudel LL, Parks JS, Bond MG. Dietary polyunsaturated fat effects on atherosclerosis and plasma lipoproteins in African green monkeys. In: Scarpelli DG, Migaki G, eds. Current Topics in Nutrition and Disease in Comparative Pathobiology. New York, NY: Alan R Liss, Inc; 1986:501-523.
Rudel LL, Sawyer JK, Parks JS. Dietary fat, lipoprotein structure, and atherosclerosis in primates. Athero Rev. 1991;23:41-50.
Wolfe MS, Sawyer JK, Morgan TM, Bullock BC, Rudel LL. Dietary polyunsaturated fat decreases coronary artery atherosclerosis in a pediatric-aged population of African green monkeys. Arterioscler Thromb. 1994;14:587-597.
Parks JS, Kaduck-Sawyer J, Bullock B, Rudel LL. Effect of dietary fish oil on coronary artery and aortic atherosclerosis in African green monkeys. Arteriosclerosis. 1990;10:1102-1112.
Rudel LL, Haines JL, Sawyer JK. Effects on plasma lipoproteins of monounsaturated, saturated, and polyunsaturated fatty acids in the diet of African green monkeys. J Lipid Res. 1990;10:1873-1882.
Allain CC, Poon LS, Chan CSG, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem. 1974;20:470-475.
Fossati P, Prencipe L. Serum triglycerides determined colorimetrically with an enzyme that produces hydrogen peroxide. Clin Chem. 1982;28:2077-2080.
Lipid Research Clinics Program. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis. Washington, DC: Public Health Service; 1982.
Koritnik DL, Rudel LL. Measurement of apolipoprotein A-I concentration in nonhuman primate serum by enzyme-linked immunosorbent assay (ELISA). J Lipid Res. 1983;24:1639-1645.
Sorci-Thomas M, Wilson MD, Johnson FL, Williams DL, Rudel LL. Studies on the expression of genes encoding apolipoproteins B100 and B48 and the low density lipoprotein receptor in nonhuman primates: comparison of dietary fat and cholesterol. J Biol Chem. 1989;264:9039-9045.
Rudel LL, Pitts LL, Nelson CA. Characterization of plasma low density lipoproteins of nonhuman primates fed dietary cholesterol. J Lipid Res. 1977;18:211-222.
Billheimer JT, Gillies PJ. Intracellular cholesterol esterification. In: Esfahani M, Swaney JB, eds. Advances in Cholesterol Research. Caldwell, NJ: Telford Press; 1990:7-45.
Carr TP, Traber MG, Haines JL, Kayden HL, Parks JS, Rudel LL. Interrelationships of α-tocopherol with plasma lipoproteins in African green monkeys: effects of dietary fats. J Lipid Res. 1993;34:1863-1871.
Reaven P, Parthasarathy S, Grasse BJ, Miller E, Steinberg D, Witztum JL. Effects of oleate-rich and linoleate-rich diets on the susceptibility of low density lipoprotein to oxidative modification in mildly hypercholesterolemic subjects. J Clin Invest. 1993;91:668-676.
Reaven PD, Grasse BJ, Tribble DL. Effects of linoleate-enriched and oleate-enriched diets in combination with α-tocopherol on the susceptibilty of LDL and LDL subfractions to oxidative modification in humans. Arterioscler Thromb. 1994;14:557-566.
Sawyer JK, Johnson FL, Rudel LL. Atherosclerosis is less in cynomolgus monkeys fed polyunsaturated (n-6) fat in spite of lower HDL concentrations and larger LDL size. Circulation. 1989;80(suppl II):II-622. Abstract.
Johnson FL, St Clair RW, Rudel LL. Studies of the production of low density lipoproteins by perfused livers from nonhuman primates: effect of dietary cholesterol. J Clin Invest. 1983;72:221-236.
Johnson FL, St. Clair RW, Rudel LL. Effects of the degree of saturation of dietary fat on the hepatic production of lipoproteins in the African green monkey. J Lipid Res. 1985;26:403-417.
Woollett LA, Daumerie CM, Dietschy JM. Trans-9-octadecenoic acid is biologically neutral and does not regulate the low density lipoprotein receptor as the cis isomer does in the hamster. J Lipid Res. 1994;35:1661-1673.
Rudel LL, Star RJ. Species, diet, and gender differences in plasma postheparin lipolytic activities in nonhuman primates: relationships with plasma lipids and high density lipoproteins. Arteriosclerosis. 1990;10:350-357.
Quinet E, Tall A, Ramakrishnan R, Rudel L. Plasma lipid transfer protein as a determinant of the atherogenicity of monkey plasma lipoproteins. J Clin Invest. 1991;87:1559-1566.
Vega GL, Denke MA, Grundy SM. Metabolic basis of primary hypercholesterolemia. Circulation. 1991;84:118-128.
Reaven P, Parthasarathy S, Grasse BJ, Miller E, Almazan F, Mattson FH, Khoo JC, Steinberg D, Witztum JL. Feasibility of using an oleate-rich diet to reduce the susceptibility of low-density lipoprotein to oxidative modification in humans. Am J Clin Nutr. 1991;54:701-706.