Ldl Receptor

Prostaglandins, Leukotrienes and Essential Fatty Acids (1997) 57(4 & 5), 411-418 @ Harcourt Brace & Co Ltd 1997

S a t u r a t e d f a t t y acids and LDL r e c e p t o r modulation in humans and monkeys K. C. Hayes, P. Khosla, T. Hajri, A. Pronczuk Foster Biomedical Research Laboratory, Brandeis University, Waltham MA 02254, USA

Summary It has been known for 40 years that dietary saturated fat (SAT FAT) increases plasma cholesterol, including LDL-C and HDL-C. In humans, where LDL-C is typically > 90 mg/dl this SAT FAT effect largely reflects changes in LDL-C pool size. The original human studies suggested that LDL-C expansion during SAT FAT consumption reflected reduced LDL clearance (LDL receptor activity) in hyperlipemics and increased LDL production rates in normolipemics (LDL-C < 100 mg/dl). This dual explanation is supported by data from several animal models where specific saturated fatty acids (SFAs) have been the focus. However, the situation is complicated by the fact that polyunsaturated fatty acids (PUFAs) oppose SFAs, i.e. PUFAs decrease LDL-C and increase LDL receptor (LDLr) activity, so the effect of SAT FAT intake may represent the combined influence of increased SFAs and decreased PUFAs. In fact, careful scrutiny of primate data suggests a negligible effect of saturated fat on LDL clearance (and receptor activity) in the absence of dietary cholesterol when PUFA intake is adequate (5-10%en) and the lipoprotein profile is relatively normal (LDL-C < 90 mg/dl), i.e. normolipemic situations at the time of dietary intervention. In such cases increases in LDL-C due to SFAs (particularly 12:0+14:0) appear to reflect LDL overproduction associated with a shift in cholesterol from tissues to the plasma cholesteryl ester (CE) pool (both LDL-C and HDL-C) without altering whole-body cholesterol balance. The reason for this shift, which is accompanied by an increase in the plasma oleic/linoleic CE ratio, is unknown but may reflect a decreased rate of CE hydrolysis by the liver. When individuals or animals are rendered hyperlipemic by other factors (e.g. chronic caloric and dietary cholesterol excesses in humans or by cholesterol feeding in animals) specific SFAs (particularly 16:0) can contribute to decreased LDL; activity initiated by a primary factor, such as dietary cholesterol. However, LDLr down-regulation by dietary cholesterol greatly exceeds any contribution from SFAs. INTRODUCTION

It is generally appreciated that dietary saturated fatty acids (SFAs) raise total plasma cholesterol (C, both LDL-C and HDL-C). A current school of thought suggests that the major effect of SFAs is to depress LDL receptors (LDLr) and decrease clearance of LDL-C from plasma. 1-3 However, this conclusion may be premature when one considers the general response across species and the particulars of data generated in human subjects. In fact, the fundamental role of SFAs is probably not on LDLr activity but is much broader in scope. For example, SFAs increase plasma cholesterol in most species with the increase being distributed according to the LDL or HDL profile of the species under investigation. In humans the primary increase is in LDL-C. In macaque monkeys it represents a balanced increase in LDL-C and HDL-C, Correspondence to: K. C. Hayes, Tel. 001 781 736 2051; Fax. 001 781 736 2054

while gerbils and hamsters develop a major increase in HDL-C. 4 In other words, SFAs increase plasma cholesterol in the dominant lipoprotein pool of the species in question, if at all. In humans fed SFAs the expanding LDL-C pool could reflect several factors: 1) increased production of LDL; 2) decreased LDL clearance causing a 'backup' in plasma, or; 3) simply having LDL particles produced and cleared at normal rates, but each particle packed with more cholesterol per particle. When these particles circulate, each would carry more cholesterol to expand the plasma LDL-C pool. A combination of all three mechanisms is possible, but most studies have focused on production and clearance by tracing LDL apoB with 1125radiolabelling. To better appreciate the published findings, it is instructive to consider which fatty acids affect the LDL-C pool most demonstrably within the context of two basic factors that influence the LDL-C response, in general. 411

412

Hayes et al

KEY TO METABOLIC FACTORS IMPACTING LIPOPROTEIN SETPOINT

LDL-C (mg/dl) 24o

4 VLDL LP~"~TG /~VER

~,.j-~v v

~

/ CHOLESJER_OLLDL_I ) BLOCKS LD~

I N 3b 3a~ HDL IDL / ~ /CETP

1. LDL, activity is key 2. VLDL output 3a. LDL formation 3b. HDL formation 4. LPL removes TG $. CETP adds to LDL 6. CE formation 7. Bile acid synthesis

14:0

M I N I M A L 18:2 T H R E S H O L D BELOW W H I C H LDL, A C T I V I T y IS EXCEEDINGLY VULNERABLE A N D L D L - C RISES I N RESPONSE T O C E R T A I N S A T U R A T E S

2oo MAXIMUM HEPATIC LDLr A C T I V I T Y INDUCIBLE BY n 6 FUFA (UppER "[HRESHOLD)

18o 36:0

leo 14o

18:0

loo

L Al~ taty ,~Arry A O ~ ~ MO~-OR- LVSS EQUAL A ~ T m MAX. T ~ L D OFl~en

8o eo

I

I

~

I

I

J

I

I

~

I

I

I

4o

1

2

3

4

5

6

7

8

9

10

11

12

20 o

L L

Fig. 1

Metabolicfactors (indicated as 1-7) contributing to the

lipoprotein setpoint, or profile, of an individual, are depicted. In humans, where LDL-C predominates, the LDLr activity (1) is critical as it ultimately affects LDL clearance and pool size. But LDL also indirectly depends on the rate of VLDL 'substrate' production by the liver (2) and its conversion to IDL (3a) and HDL (3b) when acted upon by lipoprotein lipase (LPL) in adipose and muscle (4). If the LDLr is up-regulated (as in rats), then IDL clears directly to the liver with minimal LDL formation. An accumulation of IDL-C or LDL-C is augmented by active transfer of cholesteryl esters from HDL by CETP (5). Once delivered to the liver, the free cholesterol can be 'removed' by re-esterification via ACAT (6) (probably only in species that accumulate hepatic CEs) or converted to bile acids (7) for excretion (extremely efficient in rat livers). Hepatic cholesterol accumulation 'blocks' the LDLr causing the plasma pool of LDL to expand. The balance among all these variables establishes the lipoprotein setpoint. Dietary fat and cholesterol influence many of these parameters.

The primary factor is the inherent lipoprotein profile (setpoint) of the host (Fig. 1), and the second relates to those dietary factors (e.g. 18:2 and dietary cholesterol) that modify this setpoint or alter its 'threshold'. This combination of genetics and diet modulate the impact of SFAs on cholesterol metabolism. 4,s Without first reconciling these two variables, experimental results are typically confusing and often uninterpretable. The lipoprotein setpoint has been discussed elsewhere, 5 but briefly it reflects the various metabolic parameters dictating the plasma lipoprotein profile of an individual. An elevated setpoint would encompass hyperlipidemia with elevated LDL-C, triglycerides (TGs), and low HDL-C, whereas someone with a low setpoint would have a total cholesterol (TC) less than 180 mg/dl, low LDL-C and TGs with normal to elevated HDL-C. The setpoint is determined, in part, by dietary factors, a lower setpoint being favored by dietary soluble fiber, vegetable protein, and a low-cholesterol intake. Although both SFAs and polyunsaturated fatty acids (PUFAs) can affect the setpoint, we have previously underscored the specific relationship and importance of dietary intake of linoleic acid (18:2 n-6), but abbreviated as 18:2) and the 18:2 threshold (Fig. 2).4.e The intake of 18:2 seems especially critical because it affects how specific SFAs influence cholesterol metabolism. When the available 18:2 is below its required 'threshold' for an individual or population, cholesterol-

Fig. 2 The scheme depicts the putative dynamics between the relative importance of linoleic acid (18:2, as percent dietary energy) and modulation of LDL-C, which in turn is based on antagonism between 18:2 and primarily 14:0. Only 18:2 is perceived as exerting a positive influence on (i.e. increases) LDL receptor (LDLr) activity in humans. By shifting the typical 18:2 intake from low (3%en) to moderate (7%en) the hepatic clearance of LDL-C is increased (lowering LDL-C and plasma cholesterol) to counter the cholesterolelevating influence of 12:0+14:0, which is probably via LDL overproduction (see text). Consumption of 16:0 exerts a negative effect (increases LDL-C) only if the 18:2 intake is too low (below threshold, as in cholesterol feeding), probably because VLDL production increases causing increased LDL formation (see Fig. 1). Other factors affecting cholesterol metabolism presumably affect the 18:2 threshold requirement, accounting for individual and population differences in the threshold.

emia readily develops as SFAs are increased. However, once the 18:2 threshold is exceeded, it becomes increasingly difficult to elicit a SFA response, and adding more 18:2 to the diet fails to decrease TC appreciably. These points can be demonstrated in the hamster where the fatty acid (FA) influence on LDL receptors has been investigated. 1,2 Based largely on this hamster model fed a substantial amount of cholesterol, current dogma suggests that SFAs (12:0, 14:0, 16:0) raise cholesterol primarily by decreasing LDL-receptor (LDL~)activity, thereby expanding the LDL-C pool. 3 Our working hypothesis would argue that dietary cholesterol depresses the LDLr (raising the lipoprotein setpoint) and thereby exacerbates (or indirectly elicits) the SFA effect. On the other hand, dietary 18:2 can lower the lipoprotein setpoint by enhancing LDLr activity.2'5 A common misconception that emanates from discussions of the Spady-Dietschy hamster model is that SFAs alone are responsible for the decreased LDL-receptor activity. However, only in the hamster fed chow with 0.12% added cholesterol is this SFA effect demonstrable, as pointed out by the authors themselves/ In the absence of dietary cholesterol in chow-fed hamsters (and several other species, as well) it is difficult to demonstrate a difference between SFAs and polyunsaturated fatty acids (PUFAs) or monounsaturated fatty acids {MUFAs) on LDL-C or LDLr activity, suggesting that SFAs per se do not depress LDL~ activity. Nonetheless in certain species, such as the gerbil and cebus monkey (and probably in certain humans), a SFA-induced increase in LDL-C pool size does occur (in response to specific fatty acids) without the interaction of cholesterol feeding. Even

Prostaglandins, Leukotrienes and Essential Fatty Acids (1997) 57(4 & 5), 411-418

© Harcourt Brace & Co Ltd 1997

Saturated fatty acids and LDL receptor modulation in humans and monkeys

413

Table I LDL apoB kinetics for humans fed different dietary fats Study design (ref)

LDL-C mg/dl

LDL apoB pool size mg/kg

LDL apoB FCR pools/day

LDL apoB PR mg/kg/day

Turner et aV (< 150 mg cholesterol/day, P/S 0.2 vs 8, 18:2@ 4 vs 32%en) normolipemics (n = 7, x-over) butter 122" 33 0.35 11.5* safflower 98 ~ 28 0.39 10.5 hyperlipemics (n = 8, x-over) butter 271 * safflower 219

72* 51

0.23 0.27*

15.4* 12.7

Shepherd et al 8 (400 mg cholesterol/day, P/S 0.25 vs 4, 18:2@ 5 vs 24%en) (n = 8, x-over) milk fat 134 36 0.32 11.5 corn oil 104" 31" 0.35* 10.8 Cortese et al 9 (340 mg cholesterol/day, P/S 0.30 vs 25%en as fat) (n = 8, x-over) High-SAT FAT 210" 37* 0.26 9.6* Low-SAT FAT 175 26 0.30* 7.7 (365 mg cholesterol/day, P/S 0.12 vs 3.8, 18:2@ 2.5 vs 25%en) (n = 3, x-over) butter 164" 38* 0.37 15.3 sunflower 129 31 0.35 11.8 *significant difference attributed to dietary fat; tsignificant difference between normolipemics and hyperlipemics fed safflower oil.

in these species (as pointed out below), it is not always clear from the study design whether SFAs alone are to blame, or whether a simultaneous decrease in PUFAs is a factor. When evaluated in isolation, both human and nonhuman primate studies are inconclusive on this subject. Even collectively they fail to provide strong support for direct modulation of LDLr by SFAs and, in fact, appear to disfavor any such conclusion when details of the various designs are scrutinized. HUMAN STUDIES

Turner et al 7 were among the first to illustrate the concept that the lipoprotein setpoint affects apoB kinetics when dietary fat saturation is manipulated in humans (Table 1). Two dietary fat extremes were fed to individuals with different lipoprotein setpoints, i.e. normolipemics (n= 7) or hyperlipemics (n = 8). Fat contributed 40% en, with one diet based on butterfat, the other safflower oil. Because the ratios between polyunsaturated and saturated fatty acids (P/S ratios) represented extremes (at 0.2 vs 8 ), the lower TC during the safflower oil diet could represent the effect of high PUFAs or removal of SFAs, or both. It is not possible to make that distinction from the design. Dietary cholesterol was < 150 mg/day, minimizing any complicating effect it might have had. When consuming the safflower oil diet, both groups of subjects decreased their LDL-C by 20%, and LDL apoB by 15% (normolipemics) or 28°/o (hyperlipemics), but only the hyperlipemic group revealed a 15% decrease (P< 0.05) in LDL © Harcourt Brace & Co Ltd 1997

apoB fractional catabolic rate (FCR) when fed butter fat, reflective of a LDLr clearance problem. However, in both groups during the butter fat period the production rate of LDL apoB (PR) increased significantly by 10% (normo) and 20% (hyper). This suggests that excess synthesis, not clearance, was the predominant problem associated with dietary saturated fat (SAT FAT), with impaired clearance possibly being secondary to an initially elevated lipoprotein setpoint (hyper group only). A similar dietary fat challenge was conducted by Shepherd et al,8 and the design also included a crossover between mflkfat and corn oil (Table 1). During corn oil intake, LDL-C declined by 22% and LDL apoB pool size by 14% coupled with a 9% increase in FCR without affecting the production rate. However, 400 mg cholesterol was consumed/day, which would support the notion that altered (depressed) LDG by cholesterol allowed for upregulation and increased FCR by PUFAs, as delineated by the hamster model. 1-3 This study emphasizes the importance of monitoring dietary cholesterol when interpreting results, even in humans. A third human study in hypercholesterolemic subjects adds to our understanding. 9 Two groups of subjects were evaluated, again for LDL apoB kinetics (Table 1). The first group (n=8) was used to compare high and low fat intakes. A high-SFA diet (46% en, P/S 0.3 based on butter and meat fat) depressed the FCR and increased the LDL apoB pool size and production rate relative to the same fat composition (P/S 0.3) at low-fat intake (20% en). This result identified the possibility that absolute intake of a SFA or 18:2, or even total fat intake, represent potential

Prostaglandins, Leukotrienes and Essential FattyAcids (1997) 57(4 & 5), 411-418

414

Hayes et al

Table 2 LDL apoB kinetics for monkeys fed different dietary fats

Study design (ref)

Khosla and Hayes 1° rhesus no cholesterol 12:0 + 14:0 16:0 + 18:1 Khosla et al 1~ rhesus 0.07% cholesterol Am Fat Blend 0.03% cholesterol Step 112:0 + 14:0 Step 116:0-rich Khosla and Hayes ~3cebus no cholesterol 16:0-rich 18:1 -rich 18:2-rich Khosla and Hayes TM cebus no cholesterol 16:O-rich 18:l-rich 0.3% cholesterol 16:0-rich 18:l-rich Khosla and Hayes ~ecebus 0.08% cholesterol Am Fat Blend 0.04% cholesterol Step 1-16:0 Step I-trans 18:1 Nicolosi et a117cebus no cholesterol coconut oil corn oil 0.1% cholesterol coconut oil corn oil

LDL-C mg/dl

LDL apoB FCR pools/day

LDL apoB PR mg/kg/day

(n = 4) 73 57

14 0.86 7* 0.79 (n = 6, paired design) 50

113 82*

44 0.99 36* 1.12 (n = 10, x-over)

44 40

50 53 48

20 1.29 20 1.38 21 1.22 (n = 12, x-over)

25 22 26

53 47

18 17

1.72 t 1.71t

30 29

40 t 34 t

1.06* 1.27

41t 42 t

136 t 117 t

0.80*

12.0 5.5*

143

39

(n = 11, x-over) 76 *t

59

1.12 *t

33

64 61 (n = 5)

52 53

1.34 1.35

31 34

126" 38

35* 20

1.10 1.88*

39 38

155" 58 t

41" 22

0.71t 1.45*

29 32

Stucchi et al ~8cynomolgus 0.1% cholesterol coconut oil 468* butter + oils 324* corn oil 47 Brousseau et al2° cynomolgus 0.1% cholesterol 12:0 + 14.0-rich 219" 18:1 -rich 191 18:2-rich 160 Hunt et a121cynomolgus 0.01% cholesterol SATs POLYS 0.06% cholesterol SATS POLYs 0.50% cholesterol SATs POLYs Sorci-Thomas et a122African no cholesterol lard safflower 0.4% cholesterol lard safflower

LDL apoB pool size mg/kg

(n = 6-13) 90 0.42 70 0.44 16* 1.20* (n = 10, x-over) 112" 81 72

36 30 18*

0.53 0.51 0.62*

55* 41 42

(n = 3-5) 95 66*

na na

0.75 0.75

19 15*

194 142*

na na

0.49 0.47

25 14*

248 221" Green

na na

0.17 0.15 [LDLr mRNA] (n = 4-5) 88 2.4 80 1.9

26 22*

68 59 163 t 120 *t

137 t 101 ,t

1.0" 1.2

na na na na

*significant difference attributed to dietary fat; tsignificant difference due to dietary cholesterol.

Prostaglandins, Leuketrienes and Essential Fatty Acids (1997) 57(4 & 5), 411-418

© Harcourt Brace & Co Ltd 1997

Saturated fatty acids and LDL receptor modulation in humans and monkeys

complicating variables between published study designs. In addition, data were presented for a second group of subjects (n = 3) where the LDL-C and apoB pool size were increased by an extreme SAT FAT diet based on butter (P/S 0.12, 2.5 %en as 18:2) vs an extreme PUFA FAT based on sunflower oil (P/S 3.8, 25%en as 18:2). Too few subjects were studied, but no effect was noted on FCR, and only a trend for increased LDL apoB production rate was observed with the butter diet. However, cholesterol intake at 365 mg/ day during the butter diet represented a potential bias. Thus, the few h u m a n studies that evaluated the impact of dietary fat on LDL turnover (LDLr activity) are complicated by the extremes in SFAs vs PUFAs and the limited type of SFAs fed (all butter-based). In essence, a large divergence in 18:2 %en intake between the SFA and PUFA diets, as well as the varied cholesterol load resulted in no clear trend for SFA/PUFA effects, let alone any specific saturated fatty acid effects, on LDL apoB FCR or LDL apoB production rates. NONHUMAN

PRIMATES

To address certain shortcomings in the h u m a n studies we conducted five studies in monkeys (two in rhesus and three in cebus) where these variables were carefully controlled. The first experiment was in rhesus monkeys ~° where the 18:2 0/oen was held relatively constant (4O/0en)between two cholesterol-free diets in which the P/S ratios were relatively low (0.17 and 0.35) and the major exchange was between specific SFAs (12:0 + 14:0) for (16:0) + 18:1 (= 18:1 n-9, oleic acid). This was achieved by replacing a coconut oil-soybean oil mixture with palm oil-soybean oil providing 3 l%en from total fat. The results (Table 2) revealed a trend for increased LDL-C during 12:0 + 14:0 consumption, coupled with a doubling of both the LDL apoB pool and production rates without affecting LDLr activity, i.e. FCR was not altered. These results suggested that the major effect of specific SFAs (i.e. 12:0 + 14:0) was to increase the production rate of LDL apoB rather than affect clearance when the 18:2 %en was adequate and constant between diets and no diet cholesterol was present to down-regulate the LDL receptors. A second rhesus study ~1 confirmed the first. In this experiment all monkeys initially received an American fat blend (38%en as fat, 16%en as SFAs with cholesterol at 180 mg/1000 kcal) and then were split into two groups to receive one of two Step I diets (30%en, 10%en SFAs, with 75 mg/1000 kcal of cholesterol). In one of these reducedfat diets the SFAs were predominately 16:0, in the other mostly 12:0 + 14:0. The LDL kinetics (Table 2) indicated that the main diet effect was attributed to reduced cholesterol consumption, which caused the FCR to increase equally during both Step I diets. No differential effect of © Harcourt Brace & Co Ltd 1997

415

16:0 vs 12:0 + 14:0 was seen on FCR, but 12:0 + 14:0 did result in a greater LDL-C and LDL apoB pool size because the production rate of LDL apoB increased. Because rhesus are less sensitive to saturated fat than cebus monkeys, 12 the latter species was challenged with 40%en fat in cholesterol-free diets that contained saturates, monounsaturates or polyunsaturates from palm oil, high- 18:1 safflower oil, or high- 18:2 safflower, respectively. 13 The 18:2 %en was 4, 6, and 29, respectively. Somewhat unexpectedly LDL kinetics revealed absolutely no differences in LDL-C, apoB pool size, or production rate between these three dietary fatty acid profiles, again in the absence of cholesterol and with adequate 18:2 intake in all cases (Table 2). It was concluded that when the LDLr activity and lipoprotein setpoint are normal and 18:2 is at (or above) its required threshold at the time of intervention (i.e. lipoprotein metabolism is unstressed), no effect of a 16:0-rich SAT FAT (unlike 12:0 + 14:0-rich fats) can be detected on LDL metabolism, even in a fatty acid sensitive species. As a follow-up study, cebus were again challenged with either a 16:0-rich or 18:l-rich fat in the presence or absence of 0.3% cholesterol using a crossover design. TM A significant depression in LDLr activity occurred that was attributable to cholesterol consumption per se (Table 2). Furthermore, only when cholesterol was added to these fats did LDL-C and apoB pool size increase significantly in the 16:0-rich diet group relative to the 18:l-rich diet group; and although the apoB production rate increased with cholesterol feeding, the increase was similar for both fatty acid profiles. The increases in LDL-C and apoB pool size during the 16:0-rich fat were explained by the greater depression in FCR (LDLr activity) associated with cholesterol intake. Thus, only in the presence of dietary cholesterol (which depresses LDLr activity and increases the lipoprotein setpoint and 18:2 threshold) could one distinguish between 16:0 and 18:1. This is analogous to the cholesterol-fed hamster model where 18:1 induced greater LDL-C clearance than 16:0 or 14:0Y ,~5 The impact of dietary cholesterol on LDLr activity is further reflected in a final cebus experiment 1~ that also compared Step I diets (16:0-rich vs trans 18:l-rich containing 30%en from fat, 0.03% cholesterol) with an American fat blend (38% en as fat, 0.08% cholesterol, high in 12:0 + 14:0). Again the importance of dietary cholesterol (and possibly lower fat intake) was apparent on LDL kinetics because both Step I diets reduced LDL-C and the apoB pool size, while increasing the FCR: However, the production rate of LDL apoB was unaffected by the modestly lower cholesterol intake of the two Step I diets, even though 12:0 + 14:0 was also substantially reduced from the control diet period. Collectively these data confirm that differentiating between 16:0 and 18:1 depends on initial dietary choles-

Prostaglandins, Leukotrienes and Essential FattyAcids (1997) 57(4 & 5), 411-418

416

Hayes et al

terol down-regulation of LDLr. In the absence of dietary cholesterol, our LDL kinetic data indicate that certain SFAs, namely 12:0 + 14:0, can increase LDL apoB, LDL-C, and LDL apoB production without altering LDL apoB FCR (LDLr activity). Thus, certain SFAs appear to alter LDL metabolism independent of any depression in clearance, probably by increasing apoB production. Furthermore, these details can only be elaborated when cholesterol and 18:2 %en are carefully controlled and individual SFAs are accounted for during the dietary fat exchanges. OTHER M O N K E Y STUDIES

The background information detailed above facilitates interpretation of other reports concerning LDL~ activity in n o n h u m a n primates. For example, the initial report on apoB kinetics in cebus m o n k e y s was conducted after 5-10 years of feeding either coconut oil (without added 18:2) or corn oil with or without 0.1% dietary cholesterol (four diet groups). The P/S ratios were extraordinarily polar at 0.02 vs 4.7 with 18:2 intake estimates of 0.6% vs 17%en (Table 2). A P/S ratio of 0.02 is extremely low and approaches essential fatty acid deficiency, whereas a P/S of 4.7 is extremely high and much above the 18:2 threshold. Even in the absence of dietary cholesterol, the authors found that coconut oil Without added 18:2 greatly expanded LDL-C and LDL apoB pools. This was linked to reduced LDL FCR (both LDL~ activity and nonreceptor uptake were depressed) without any change in LDL apoB production. The altered LDL metabolism described during coconut oil intake may have reflected a lack of 18:2 coupled with a high intake of 12:0 + 14:0 or, during corn off intake, a distortion caused by excess 18:2. However, the latter seems unlikely because high 18:2 intake in normolipemic humans 7-9 and cebus monkeys ~3 had no effect on FCR. When 0.1% cholesterol was added, neither fat group revealed a change in LDL production rate, even though the added cholesterol depressed FCR and expanded LDL-C slightly. A drawback with the design is that without better control over 18:2 and no control over individual variation among monkeys (no crossover design), the true effect of specific SFAs or PUFAs cannot be realized. In a second study by the same authors 18 cynomolgus monkeys were used to again compare coconut oil (12:0 + 14:0-rich) and corn oil with a butter-based diet rich in 16:0 + 18:0 (Table 2). The P/S ratios were 0.02, 4.9 and 0.54 while the %en from 18:2 was 0.7, 23, and 7, respectively. M1 three diets contained 0.1% cholesterol. The marked rise in LDL-C demonstrates the extreme sensitivity of cynos to a dietary cholesterol load, at least in the presence of saturated fat. As commented on elsewhere 19the degree of dietary cholesterol overload depressed LDL~so severely (raised the lipoprotein setpoint) that it essentially pre-

empted the possibility of distinguishing between the relative ability of specific SFAs to modulate LDLr activity. Only the high 18:2 intake (23%en) associated with corn oil was able to overcome the LDI~ depression (increasing FCR and essentially normalizing LDL-C and LDL apoB kinetics). In a better designed study also using cynomolgus monkeys, these authors 2° fed 10 monkeys three different fats (containing saturates, monounsaturates and polyunsaturates) in a crossover design. Exquisite care was taken with fat composition, blending 5-6 fats per diet in order to exactly exchange 10% en as 12:0 + 14:0 for 18:1 or 18:2 within purified diets containing 30%en as fat, again with 0.1% cholesterol (230mg/1000kcal). As a result of the careful fat formulation, the 18:2 %en was within a realistic dietary range at 5.7 %en in the 12:0 + 14:0-rich diet, 4.7 %en for the 18:l-rich diet, and 17.7 %en for the 18:2-rich diet. The P/S ratios were 0.4, 1.1, 3.4, respectively. The much lower cholesterolemia (compared to Stucchi et al~a) reflected by the lower LDL-C, clearly demonstrates the power of 18:2 to lower LDL-C in the cholesterol-fed cynomolgus. For example, the difference of 0.7 vs 5.7 %en as 18:2 between the roughly comparable 12:0 + 14:0-rich diets in the two experiments, decreased LDL-C more than 50% in the second instance. But more important to our discussion is the clear demonstration that the rise in LDL-C due to 12:0 + 14:0 (even in the presence of moderately severe cholesterol loading) was attributed to increased LDL apoB production, not depressed FCR (i.e. not an LDL~ clearance problem) when compared to the 18:l-rich (neutral) diet. Furthermore, only a high 18:2, intake (18 %en) was able to counter the depressed FCR that had resulted from the high dietary cholesterol down-regulation of LDL receptors. Thus, while the production rate of apoB was not different between 18:1 and 18:2, and only increased with 12:0 + 14:0, the clearance rate (LDLr activity) was not different between 12:0 + 14:0 and 18:1, only improving with the 18:2-rich diet. The importance of 18:2 is further demonstrated in another study involving cynomolgus monkeys. 21 The design represented a reasonable P/S comparison (0.5 vs 0.9) based on two different blends of the same four fats (butter, beef tallow, safflower oil, corn oil), which provided adequate control over 18:2 at 8% vs 11%en (both presumably above the 18:2 threshold at low cholesterol intakes). Graded intakes of cholesterol were applied, but the six diet groups (2 fats, 3 levels of cholesterol) contained only 3 to 5 monkeys per group without any crossover. Nonetheless, LDL-C and the apoB production rate were both increased by the SAT FAT diet without any effect on FCR (LDLr activity) relative to POLYs during a neglible intake of cholesterol (0.01%). At higher dietary cholesterol intakes (0.06, 0.50%) LDL-C steadily increased with both fats, with the differential effect between saturates and polyunsaturates being sustained (Table 2). ApoB

Prostaglandins, Leukotrienes and Essential Fatty Acids (1997)57(4 & 5), 411-418

©Harcourt Brace & Co Ltd 1997

Saturated fatty acids and LDL receptor modulation in humans and monkeys

production also increased with cholesterol feeding, but the effect was moderated by PUFAs. Similar to all such studies, FCR was depressed dramatically by cholesterol feeding, but a differential effect between SFAs and PUFAs was not observed. As with Brousseau et al,20 the implication is that SFAs have no effect per se on LDL apoB clearance when 18:2 is adequate; and 18:2 adequacy depends, in part, on the cholesterol load. Apparently at 0.06% cholesterol and above, even 1 l%en from 18:2 was insufficient to overcome the cholesterol suppression of LDLr activity in cynos. On the other hand, certain SFAs can increase apoB production, and this stimulation may be exacerbated when dietary cholesterol is added. Thus, the primary explanation for an increase in LDL-C due to SFAs in the absence of dietary cholesterol, and possibly even the main effect of SFAs in the presence of dietary cholesterol, would seem to be overproduction of LDL rather than impaired LDLr clearance, with the overproduction deriving largely from 12:0 + 14:0 intake. The importance of the 18:2 threshold on the SFA effect is further supported by data from African Green monkeys = fed lard or safflower oil-rich diets (40% en from fat) with or without 0.4% cholesterol (four diet groups). Although the P/S ratio comparison (0.3 vs 2.2) was rather exaggerated, the SAT FAT diet actually supplied a reasonable 18:2 intake (6% en) compared to 24%en from 18:2 for the safflower-rich fat blend. Without added cholesterol, no difference was noted in LDL-C, plasma apoB, or an index of LDLr activity, i.e. hepatic LDL~ mRNA abundance (Table 2). As expected, adding the rather large cholesterol load depressed LDLr (mRNA), but still no difference was detected between fats, i.e. no depression in LDL~was attributable specifically to SFAs. Again, one must conclude that in a normocholesterolemic situation with unencumbered lipoprotein metabolism (lipoprotein setpoint normal) and even under a cholesterol load that expands the LDL-C pool, a 16:0 + 18:0-rich fat source (lard) does not exaggerate the depressed LDLr activity as long as 18:2 intake is adequate (at or above threshold). In summary, saturated fat does not appear to be a factor in reducing clearance of LDL when lipoprotein metabolism is normal. Normal lipoprotein metabolism in humans depends, in part, on adequate 18:2 intake and minimal (< 300 mg/day) intake of dietary cholesterol. On the other hand, certain normally consumed saturated fats (rich in 12:0 + 14:0) in the context of a low-cholesterol diet are able to increase the circulating LDL-C pool in certain species or individuals, the primary cause being an overproduction of LDL With increasing dietary cholesterol, down-regulation of LDLr occurs, raising the lipoprotein setpoint. Under these circumstances 18:2 is the only fatty acid to consistently lower LDL-C by increasing the fractional catabolic rate (LDL~ activity). However, ff LDLr activity is already normal or if the dietary cholesterol load © Harcourt Brace & Co Ltd 1997

417

is too large, an 18:2-induced decrease in LDL-C does not occur at reasonable 18:2 intakes (up to 12%en). Under most dietary circumstances SFAs probably do not affect LDLr activity other than by displacing 18:2, i.e. depressing 18:2 intake below its critical threshold requirement. ADDENDUM

The results of White and coworkers (this meeting) are not consistent with our findings, but it should be noted that two of their model assumptions need careful consideration. The first pertains to triglyceride molecular structure (TG-MS) and the second to the hamster's response to fatty acids as being representative of humans. Feeding mono-acyl TGs in order to focus on a single fatty acid assumes that TG structure and the relative load of a given fatty acid presented in this conformation have no effect on fatty acid utilization or lipoprotein (LP) metabolism. Numerous reports indicate otherwise, i.e. at least some structured TGs have unique effects when supplied in large quantity, particularly those involving saturated fatty acids. Certain structural arrangements of TGs (e.g. fatty acid pairings) may be more important than others, but we are unable at this point to designate the relative importance of each. Thus, until experiments are conducted to delineate specific attributes of given TG-MS, it is only prudent that modeling experiments that explicitly or implicitly utilize animal models to influence our thinking about human LP metabolism as affected by dietary fatty acids, should present TG forms encountered in human diets. Having said that, the applicability of these data to the h u m a n situation is questionable. The results presented are of interest for potential clues concerning the impact of mono-acyl glycerides, but in the end one is left with the discordant gap between these results and actual data derived from humans consuming natural fats containing greater or lesser amounts of the specific fatty acids in question. In real diet situations these fatty acids are normally found in combination with other fatty acids on the same TG molecule. Keep in mind that with rare exception humans (or any species) seldom consume any fatty acid in its mono-acyl TG form. Thus, the peculiar influence of a large intake of trimyristin in these hamsters may represent the simultaneous selective exclusion of 18:2, 18:3 (aqinolenic acid) or even 18:1 and 16:0, which are very likely needed concomitantly for the synthesis of a specific phospholipid or regulatory diacylglycerol. If one uses structured TGs to pursue the latter hypothesis, the present experiments have merit, but as examples of experiments implicitly designed to elucidate the role of fatty acids in the normal physiology of human (or mammalian) LP metabolism, they come up short. In essence, the application of elegant molecular biology in tissues that are without the confines of normal physi-

Prostaglandins, Leukotrienes and Essential Fatty Acids (1997) 57(4 & 5), 411-418

418

Hayes et al

o l o g y g e n e r a t e s results t h a t are m o s t difficult to i n t e r p r e t at best, a n d p r o b a b l y n o t a practical e x a m p l e of h u m a n LP m e t a b o l i s m as affected b y diet. It is t r u e t h a t feeding real fats (containing m u l t i p l e fatty acid mixtures) does n o t p e r m i t c o n c l u s i o n s specific to i n d i v i d u a l fatty acids, b u t utilizing m u l t i p l e p e r m u t a t i o n s a n d feeding r e g i m e n s w i t h carefully s e l e c t e d fat b l e n d s does allow for t h e practical isolation or e m p h a s i s of indiv i d u a l d i e t a r y fatty acids (and fats) in a f a s h i o n similar to t h a t e n c o u n t e r e d o n a daily basis. The latter e x p e r i m e n t a l a p p r o a c h is s o m e w h a t tedious, b u t at least t h e h o s t r e s p o n s e is w i t h i n t h e b o u n d s of n o r m a l p h y s i o l o g y a n d has t h e p o t e n t i a l of real-life applicability to h u m a n s , a s s u m i n g t h a t t h e l i m i t a t i o n of t h e a n i m a l m o d e l u n d e r i n v e s t i g a t i o n is a p p r e c i a t e d a n d a c k n o w l e d g e d in t h e interpretation. As i n d i c a t e d earlier, t h e p r e s e n t h a m s t e r e x p e r i m e n t s cause one to w o n d e r w h e t h e r t h e effect of t r i p a l m i t i n r e p r e s e n t s 16:0 p e r se or a n o v e r a b u n d a n c e of 16:0 in t h e relative a b s e n c e of 18:2. Because this line of q u e s t i o n i n g leads to a n infinite n u m b e r of h y p o t h e s e s a n d experim e n t s in proof, it seems m u c h m o r e logical experimentally to c o n t r o l a m a x i m u m of t h e s e possibilities from t h e o u t s e t b y i n c l u d i n g at least t h o s e fatty acids k n o w n to b e r e q u i r e d for n o r m a l p h y s i o l o g i c a l events, i n c l u d i n g certain essential fatty acids a n d (surprisingly) e v e n a certain level of saturates. Note t h a t t h e s e diets c o n t a i n no essential fatty acids, o n l y 18:1 a n d s a t u r a t e d fatty acids. Because t h e m e t a b o l i s m of fatty acids (and t h e i r i m p a c t on LPs) is d e m o n s t r a b l y affected b y o t h e r factors affecting t h e l i p o p r o t e i n profile (e.g. diet cholesterol) it b e c o m e s e x c e e d i n g l y i m p o r t a n t n o t o n l y to c o n t r o l t h e level of cholesterol c o n s u m e d , b u t to define t h e b a s a l i m p a c t of cholesterol i n g e s t e d b y t h e m o d e l w i t h i n t h e c o n t e x t of a fatty acid load. For example, it h a s b e e n s h o w n t h a t all s a t u r a t e d fatty acids (except 18:0) a p p e a r e q u a l l y cholest e r o l e m i c in t h e D i e t s c h y m o d e l once sufficient cholesterol is fed to d o w n - r e g u l a t e LDL receptors. In h a m s t e r s fed purified diets, this occurs at a b o u t 0.04% cholesterol, so b o t h t h e 0.12% a n d 0.25% c h o l e s t e r o l l o a d s h e r e i n w o u l d b e e x p e c t e d to o v e r s h a d o w , e v e n mask, a n y influence t h a t i n d i v i d u a l fatty acids m i g h t have. REFERENCES 1. Spady D. K., Dietschy J. M. Dietary saturated triacylglycerols suppress hepatic low density lipoprotein receptor activity in the hamster. Proc Natl Acad Sci USA 1985; 82: 4526-4530. 2. Woolett L. A., Spady D. K., DietschyJ. M. Saturated and unsaturated fatty acids independently regulate low density lipoprotein receptor activity and production rate. JLipid Res 1992; 33: 77-88. 3. Dietschy D. K. Experimental mechanism: regulation of plasma LDL cholesterol. Am J Clin Nutr 1995; 62: 679S-688S. 4. Hayes K. C., Pronczuk A., Khosla P. A rationale for plasma cholesterol modulation by dietary fatty acids: modeling the human response in animals. J Nutr Biochem 1995; 6:188-194.

5. Hayes K. C. Saturated fats and blood lipids: new slant on an old story. CanJ Cardio11995; 11: 39G-46G. 6. Hayes K. C., Khosla P. Dietary fatty acid thresholds and cholesterolemia. FASEB J 1992; 6: 2600. 7. Turner J. D., Le N-A., Brown W. V. Effect of changing dietary fat saturation on low density lipoprotein metabolism in man. Am J Physio11981; 241: E57-E63. 8. Shepherd J., Christopher J. P., Grundy S. M., Yeshurun D., Gotto A. M., Taunton O. D. Effects of saturated and polyunsaturated fat diets on the chemical composition and metabolism of low density lipoproteins in man. J Lipid Res 1980; 21:91-99. 9. Cortese C., Levy Y., Janus E. D., et al. Modes of action of lipidlowering diets in man: studies of apolipoprotein B kinetics in relation to fat consumption and dietary fatty acid composition. Eur J Clin Invest 1983; 79-85.

10. Khosla P., Hayes K. C. Dietary fat saturation in rhesus monkeys affects LDL concentrations by modulating the independent production of LDL apolipoprotein B. Biochim Biophys Acta 1991; 1083: 46-56.

11. Khosla P., Hajri T., Pronczuk A., Hayes K. C. Decreasing dietary lauric and myristic acid improves plasma lipids more favorably than decreasing dietary palmffic acid in rhesus monkeys fed AHA Step 1 diets. J N u t r 1977; 127: 5255-5309. 12. Pronczuk A., Patton G. M., Stephen Z. F., Hayes K. C. Species variation ill the atherogenic profile of monkeys: relationship between dietary fats, lipoproteins, and platelet aggregation. Lipids 1991; 26: 213. 13. Khosla P., Hayes K. C. Comparison between dietary palmitate (16:0), oleate (18:1) and linoleate (18:2) on plasma lipoprotein metabolism in cebus and rhesus monkeys fed cholesterol-free diets. A m J Clin Nutr 1992; 3~: 51. 14. IGlosla P., Hayes K. C. Dietary palmitic acid raises plasma LDL cholesterol relative to oleic acid only at a high intake of cholesterol. Biochim Biophys Acta 1993; 1210:13-22. 15. Woolett L. A., Spady D. K., Dietschy J. M. Mechanisms by which saturated triacyglycerols elevate the plasma low density lipoprotein-cholesterol concentration in hamsters. J Clin Invest 1989; 84:119-128. 16. Khosla P., Haiti T., Pronczuk A., Hayes K. C. Replacing dietary palmitic acid with elaidic acid (t-18:1A9) depresses HDL and increases CETP activity in cebus monkeys. J N u t r 1997; 127: 5345-5365. 17. Nicolosi R. J., Stucchi A. F., Kowala M. C., Hennessy L. K., Hegsted D. M., Schaefer E. J. Effect of dietary fat saturation and cholesterol on LDL composition and metabolism. Arteriosclerosis 1990; 10:119-128. 18. Stucchi A. F., Terpstra A. H. M., Nicolosi R. J. LDL receptor activity is down-regulated similarly by a cholesterol-containing diet high in palmitic acid or high in lauric and myristic acids in cynomolgus monkeys. J N u t r 1995; 125: 2055-2063. 19. Hayes K. C., Khosla J. P. Dietary saturated fatty acids and LDL receptor activity. J Nutr 1996 (letter); 126:1000-1001. 20. Brouseau M. E., Stucchi A. F., Vespa D. B., Schaefer J., Nicolosi R.J. A diet enriched in monounsaturated fats decreases low density lipoprotein concentrations in cynomolgus monkeys by a different mechanism than does a diet enriched in polyunsaturated fats. J N u t r 1993; 123: 2049-2058. 21. Hunt C. E., Funk G. M., Vidmar T. J. Dietary polyunsaturated to saturated fatty acids ratio alters hepatic LDL transport in cynomolgus macaques fed low cholesterol diets. J N u t r 1992; 122: 1960-1970. 22. Sorci-Thomas M., Wilson, M. D.,Johnson F. L., Williams D. L., Rudel L. L. Studies on the expression of genes encoding apolipoproteins B100 and B48 and the low density lipoprotein receptor in nonhuman primates. JBiol Chem 1988; 264: 9039-9045.

Prostaglandins, Leukotrienes and Essential Fatty Acids (1997) 57(4 & 5), 411-418

© Harcourt Brace & Co Ltd 1997