1013
Reduction in Triacylglycerol Levels by Fish Oil Correlates with Free Fatty Acid Levels in ad libitum Fed Rats David A. Otto*, Janet K. Baltzell I and Joseph T. Wooten Department of Research, The Baptist Medical Centers, Birmingham, Alabama 35211 Rats were fed (for 2 or 6 wk) purified diets containing lard (LD) or menhaden oil (MO) at two levels of dietary fat, L~, at 11.5 and 20.8% of energy in the low fat (LF) and the medium fat (MF) diets, respectively. Following the diet period, rats were sacrificed after either an overnight fast or after ,mlnterrupted a d libitum feeding. The studies were designed to investigate the dependence of our previously reported effects of MO, Le. the reduction of plasma free f a t t y acid (FFA) levels and accumulation of hepatic triacylglycerols, on the dietary fat concentration and the nutritional state of the animal at the time of sacrifice. Reductions in plasma triacylglycerol and cholesterol levels in MO-fed relative to LD-fed rats were observed under all conditions. FFA levels were consistently reduced by M(~ feeding at both dietary fat concentrations, but only when blood was sampled from a d libitum fed rats. Under these conditions there was a significant positive relationship between plasma FFA and triacylglycerol concentrations. R~ duction in plasma FFA levels m a y be an additional mechanism associated with the triacylglycerol-lowering effect of fish oil (FO). The L F and M F MO diets caused a rise in plasma glucose levels with no significant change in insulin concentration, indicating that the reduction of FFA by MO was not related to changes in insulin concentration or insulin sensitivity. The MO diets had no effect on skeletal muscle or epididymal adipose tissue lipopr~ tein lipase activity, demonstrating that catabolism of triacylglycerol-rich lipoproteins contributes little, if any, to the MO-dependent reductions of plasma triacylglycerol and FFA. The previously reported accumulation of hepatic triacylglycerols after high fat (HF; 30% of energy) MOfeeding was not observed with the L F or M F MO diets, suggesting t h a t the apparent direct inhibition of triacylglycerol secretion by FO imposes a rate-llmitation only when feeding H F diets. Lipids 27, 1013-1017 (1992). In earlier studies (1,2) we observed significant but inconsistent reductions in plasma free fatty acid (FFA) levels in fish oil (FO)-fed, overnight fasted rats. An effect on plasma FFA by FO was also observed by Singer et aL (3), who reported that dietary r polyunsaturated fatty acid (n-3 fatt y acids) supplementation in hyperlipidemic patients resulted in a marked decrease in FFA during a standard glucose tolerance test that was associated with a decline in serum triacylglycerol levels. The authors (3) suggested that the decrease in FFA indicated reduced lipolysis and that it might contribute to the triacylglycerol-lowering effect of n-3 fatty acids. We and others also demonstrated a decrease in *To whom correspondence should be addressed at Department of Research, The Baptist Medical Centers, 701 Princeton Avenue, Birmingham, AL 35211. 1Present address: NRICGP, CSRS/USDA, 901 D Street S.W., Suite 323, Aerospace Building, Washington, D.C. 20250-2200. Abbreviations: CO, corn oil; FFA, free fatty acids; FO, fish off; HF, high fat; LD, lard; LF, low fat; LPL, lipoprotein lipase; MF, medium fat; MO, menhaden oil.
adipose tissue mass in FO-fed rats (2,4,5) that was accompanied by a decrease in adipocyte volume without a decrease in fat cell number (4,5}. Additionally, we found that this decrease in adipose tissue mass was positively correlated with the FO-mediated changes in plasma FFA (2), suggesting a causal relationship. Since a reduction of plasma FFA may reflect an important mechanism involved in FOmediated effects, we investigated why this effect of FO has not been routinely observed in the numerous studies conducted with dietary FO. Specifically, we considered whether observing a reduction in plasma FFA levels was dependent upon the concentration of d i e t a r y n-3 fatty acids or the nutritional state (overnight fasted vs. ad libitum fed rats) of the animals at the time of sacrifice We also investigated the relationship between plasma FFA and the lowering of plasma triacylglycerol levels by FO. A preliminary account of this work has been presented (6).
MATERIALS AND METHODS Experimental design and diets. Male Sprague-Dawley rats (225-250 g; VAF/Plus, Charles River Laboratories, Inc., Raleigh, NC) were housed in suspended transparent polycarbonate cages with stainless steel wire bottoms in an animal facility maintained at 22 + 1~ 70% humidity and with a 12-h light/dark cycle. Rats were acclimated to the surroundings and fed a standard AIN-76A purified diet (7) (Table 1, low fat diet with corn oil as the fat source; Research Diets, Inc., New Brunswick, NJ) for I wk. Animals had free access to deionized water. Following the acclimation period, rats were matched for weight and assigned to respective groups (6 rats/group), based on the diet, the length of feeding, and the nutritional state of the animal at sacrifice. Rats were assigned to one of four diet groups. These included low fat (LF; 11.5% of energy) and medium fat (MF; 20.8% of energy) diets, prepared with either lard (LD) or menhaden oil (MO; vacuum stripped MO provided by the National Institutes of Health, Bethesda, MD) as the major fat source (Tables I and 2). All diets contained a minimum of 4% of energy from corn oil (CO) to prevent essential f a t t y acid deficiency. Thus, LD and MO provided 7.5 and 16.8% of energy in the L F and M F diets, respectively. The purified diets were based on the AIN-76A diet (7) and contained identical ingredients except for the source of fat. In the M F diets, the additional fat energy was balanced by an isocaloric reduction in carbohydrate (sucroseJcorn starch, 2:1, w/w). Therefore. the nutrient (protein, vitamins and minerals)/energy (g/kJ) ratios were the same in all diets. As in our previous studies (1,2), the diets were also balanced for cholesterol (8.96 mg/1000 kJ) and/3-sitosterol {76.8 rag/ 1000 kJ) relative to the energy content of the diets, and for the antioxidant content (a- and y-tocopherols, and tertbutylhydroquinone; 1.5, 1.2 and 0.2 mg/g fat) relative to the amount of fat in the diets. The importance of including antioxidante in the diets is emphasized by a recent report by Haglund et al. (8) indicating t h a t an FO diet rich in vitamin E (1.5 IU/g fat) had a greater triacylglycerollowering effect t h a n one with only 0.3 IU/g. Our diets LIPIDS, Vol. 27, no. 12 (1992)
1014
D.A. OTTO ET AL. TABLE 2
TABLE 1
Fatty Acid Composition of Lard and Menhaden Oil
Composition of Purified Diets a
Low fat Energy g/kg (%) 50.0 11.5
Medium fat Energy g/kg (%) 95.0 20.8
Ingredients Fat b Sucrose/corn starch (2:1, w/w) 650.0 66.6 589.0 57.3 Casein (alcohol extracted) 200.0 20.5 211.0 20.5 Cellulose 50.0 0.0 52.6 0.0 DL-Methionine 3.0 0.0 3.2 0.0 Salt mixc 35.0 0.4d 36.8 0.4d Vitamin mixe 10.0 1.0d 10.5 1.0d Choline bitartrate 2.0 0.0 2.1 0.0 aDiets were based on the AIN-76A diet. Medium fat diets were prepared by isocaloric substitution of fat for carbohydrate; thus the nutrient (protein, vitamins and minerals)/kJ ratios were unchanged from the low fat diet. The energy contents of the respective diets were 16.4 and 17.3 kJ/g for the low and medium fat diets. bAll diets contained a minimum of 4% of energy from corn oil to prevent essential fatty acid deficiency. The remaining fat energy was supplied by either lard or menhaden oil. Thus, lard and menhaden oil provided 7.5 and 16.8% of energy in the low and medium fat diets, respectively. Menhaden oil was provided by the Fish Oil Test Materials Program of the National Institutes of Health. CSalt mix (in g/kg of mixture} calcium phosphate, dibasic, 500; magnesium oxide, 24; potassium citrate, monohydrate, 220; potassium sulfate, 52; sodium chloride, 74; chromium potassium sulfate, 0.55; cupric carbonate, 0.3; potassium iodate, 0.01; ferric citrate, 6.0; manganous carbonate, 3.5; sodium selenite, 0.01; zinc carbonate, 1.6; sucrose, 118.03. dAs sucrose. evitamin mix (in g/kg of mixture): Vitamin A palmitate (500,000 IU/g), 0.8; Vitamin D3 (400,000 IU/g), 0.25; Vitamin E acetate (500 IU/g), 10.0; menadione sodium bisulfate, 0.08; biotin (1%), 2.0; cyanocobalamin(0.1%), 1.0; folic acid, 0.2; niacin, 3.0; calciumpantothenate, 1.6; pyridoxine HC1, 0.7; riboflavin, 0.6; thiamin HC1, 0.6; sucrose, 979.17.
contained approximately 1.4 IU of vitamin E/g fat along with tert-butylhydroquinone The MO diets were prepared in our laboratory by addition of MO to a basal diet as previously described (1,9}. All other diets were customprepared by a commercial vendor (Research Diets, Inc.). Diet cups were placed in the cages such that the diets were not exposed to direct light. All diets were replaced every third or fourth day (9). Rats were fed one of the four diets for 2 wk (12 rats/diet group) or 6 wk (6 rats/diet group}. At the end of the feeding period, half of the animals from each group were sacrificed in the morning after uninterrupted ad libitum feeding. The other half were sacrificed after an overnight fast. The data from 6 wk fed rats are presented in the text only when relevant to the discussion. Plasma and tissue collection. A t the time of sacrifice, rats were anesthetized with pentobarbital (5 mg/100 g body wt). Blood was drawn from the abdominal aorta into a syringe containing 0.25 mL of 0.2 M ethylene diaminetetraacetic acid, p H 7.4, and plasma was collected by centrifugation at 500 X g (4~ for 20 rain. Immediately after drawing blood, livers were rapidly removed, weighed and freeze-clamped with aluminum tongs precooled in liquid nitrogen (10). The thin layer of frozen liver tissue was later ground into a fine powder under liquid LIPIDS, Vol. 27, no. 12 (1992)
Fatty acid 14:0 14:1 16:0 16:1n-9 16:1n-7 18:0 18:1n-9 18:2n-6 18:3n-3 20:1 20:5n-3 22:1n-9 + n-ll 22:5n-3 22:6n-3 Saturated Monounsaturated Polyunsaturated n-6 n-3 P/S ratio
Lard 1.4 0.1 24.5 3.3 15.0 42.8 10.5 1.0 0.6 40.9 46.8
% of Total fatty acids Menhaden oil 7.6 0.1 16.6 0.2 10.3 3.1 8.1 1.2 1.2 2.3 14.8 0.5 2.5 8.7 27.3 21.5
10.5 1.0 0.28
1.2 27.2 1.04
nitrogen. Soleus muscles and epididymal fat pads were collected, weighed and immediately frozen by dropping into liquid nitrogen. All samples were stored at - 8 0 ~ until assayed. Analytical procedures. P l a s m a t r i a c y l g l y c e r o l , cholesterol and glucose were measured using Baker kits with the Encore Chemistry S y s t e m centrifugal analyzer (Baker I n s t r u m e n t s Corp., Allentown, PA). The assay for FFA (Amano International E n z y m e kit reagents, Troy, VA) was adapted for use on the Encore chemistry System (1). Plasma insulin was determined by radioimmunoassay using a commercial kit (Ciba C o m i n g Diagnostics Corp., Medfield, MA). Liver triacylglycerol was extracted and assayed as previously reported (1). Soleus muscle and epididymal adipose tissue lipoprotein lipase (LPL) activities were measured as described earlier (2). Statistical analysis. Statistical differences between LD and MO diet groups were determined by the Student's ttest. Correlation between measured parameters was determined by least squares regression analysis. All statistical analyses were performed utilizing the SYSTAT statistical software package (SYSTAT, Inc., Evanston, IL).
RESULTS AND DISCUSSION
Plasma triacylglyceroL cholesterol and free fatty acids. The reduction in plasma triacylglycerol and cholesterol levels in MO-fed relative to LD-fed rats was observed consistently under all conditions of dietary fat concentration and nutritional state (Table 3). In this study we considered whether the normal rise in plasma FFA t h a t occurs with fasting might sometimes mask effects of FO on FFA, and thus, explain the inconsistencies of our previous observations with overnight fasted rats (1). After 2 wk of feeding, a significant decrease in plasma FFA was observed in the MO- vs. LD-ad libitum fed rats at both fat concentrations. In contrast, after an overnight fast, FFA were reduced
1015
F I S H OIL, F F A A N D P L A S M A T R I A C Y L G L Y C E R O L S
TABLE 3
Plasma Metabolites from a d libitum Fed and Overnight Fasted R a t s a
Nutritional state
Diets Fat concentration
ad libitum F e d
Low fat Medium fat
Overnight fasted
Low fat Medium fat
Fat source
Triacylglycerols (mM)
Free f a t t y acids (~M)
Cholesterol (mM)
Glucose (mM)
Insulin (~U/mL)
LD MO LD MO
3.35 1.25 4.25 1.74
+-- 0.11 +-- 0.20 b _+ 0.29 +_ 0.07 b
2.40 1.86 2.45 1.47
+_ 0.18 +-- 0.13 b __ 0.47 __ 0.16 b
316 207 348 168
_+ + +_ +_
35 21 b 34 16 b
8.9 10.4 9.0 9.9
-+ +_ +-
0.4 0.3 b 0.4 0.2 b
70 65 93 75
++ +-+
8 6 9 8
LD MO LD MO
1.90 0.69 1.44 0.78
+_ _ _ +_
1.78 1.60 1.99 1.29
+ 0.13 +- 0.16 --+ 0.18 + 0.08 b
418 410 499 343
__ __ + +_
41 36 52 18 b
8.1 8.7 5.9 8.2
+ + ++_
0.2 0.6 0.3 0.7 b
66 47 56 44
++_ + _+
6 8 4 3
0.21 0.06 b 0.12 0.06 b
a R a t s were fed t h e r e s p e c t i v e d i e t s for 2 w k before sacrifice. V a l u e s are t h e m e a n -- S E (n -- 6). A b b r e v i a t i o n s : M O , m e n h a d e n oil; LD, lard. b s i g n i f i c a n t difference b e t w e e n MO- a n d L D - f e d a n i m a l s , P < 0.05.
only in the MF MO group. After 6 wk this effect was again significant in ad l i b i t u m fed rats (with LF diets, 429 _+ 54 vs. 186 +_ 23 ~M FFA; and with MF diets, 338 + 92 vs. 176 +_ 21 in LD-fed and MO-fed rats, respectively) but not seen at all in the overnight fasted rats (with LF diets, 406 +_ 13 vs. 410 +_ 38, and with MF diets, 463 +_ 45 vs. 413 +__53 in LD-fed and MO-fed rats, respectively). In our earlier study (1) we made similar observations with high fat (HF) diets (30% of energy; identical to diets described in Table 1 with an isocaloric substitution of fat for carbohydrate). After uninterrupted ad l i b i t u m feeding of HF CO, LD and MO diets for 2 wk, FFA were significantly reduced in the MO-fed rats relative to the CO- and LD-fed rats (435 __ 34, 448 __ 68 and 201 _+ 43 ~M in CO-, LD- and MO-fed rats, respectively; P < 0.05; data from ad libitum fed rats were not previously reported). However, like the LF- and MF-fed rats after an overnight fast, there were no differences between the three H F diet groups (351 + 28, 392 +_ 102 and 377 +_ 83 ~M in CO-, LD- and MOfed rats, respectively). Thus, the data clearly demonstrate that blood must be sampled in the fed state in order to consistently observe the effects of dietary FO on plasma FFA. There was a highly significant correlation between plasma triacylglycerol and FFA levels in ad l i b i t u m fed rats (Fig. 1; data from Table 3). We also observed this with ad libitum fed rats receiving HF diets [r -- 0.74, P < 0.006, n = 12; from the experiment of (1), but data not previously shown]. These data strongly suggest that along with other mechanisms, the lowering of plasma FFA levels by FOfeeding may contribute significantly to the decrease in plasma triacylglycerols by decreasing the supply of substrate for hepatic triacylglycerol synthesis. This is consistent with the recent report of Singer et al. (3) and focuses attention to possible effects of dietary FO on adipose tissue lipolysis, as has been observed by Parrish et al. (5). P l a s m a insulin and glucose. We previously reported that feeding HF MO diets followed by an overnight fast resuited in a reduction in plasma insulin levels with an accompanying reduction or no change in plasma glucose levels (1,2). This is consistent with an improvement in peripheral insulin sensitivity. In the present experiment with the LF and MF MO diets, insulin levels were only
~ " 6.0
" •5.0 4.0
5 ej
~, 3.0 2.0 .~ff'/'~ -- _
I--
1.0 CD
f
C') L F - M E ~ H A D ~ ~J
CqL
A LF-L~D
9 MF-I.ARD
O.O
l
i
I
,
I
|
!
i
1 O0 200 300 400 PLASMA FREE FATTY ACIDS (p,m o l / L )
500
F I G . 1. Correlation between p l a s m a free fatty acids a n d p l a s m a triacylglycerols from a d libitum f e d rats [data f r o m Table 3; for all data, r = 0.71, P < 0.001, n = 23; for low f a t (LF) diets, r = 0.68, P = 0.022, n = 11; and for m e d i u m f a t (MF) diets, r = 0.92, P < 0.001, n = 12].
slightly lowered or unchanged relative to those in LD-fed rats (Table 3; approaching significance only in overnight fasted rats adapted to the LF or MF MO diets, P --- 0.074 and 0.059, respectively). Under these conditions, plasma glucose was generally elevated (Table 3). The rise in glucose may have been in response to the slight reductions in plasma insulin levels, but it is also possible that insulin sensitivity (peripheral or hepatic) actually worsened with LF and MF MO-feeding. The data clearly show that the lower plasma FFA levels seen with MO-feeding at all concentrations cannot be explained by either an increase in insulin concentration or improved insulin sensitivity at the site of the adipose tissue. L P L and adipose tissue mass. Unlike our earlier study (2) in which we observed a significant elevation of soleus muscle LPL in H F MO-fed rats, feeding LF and MF MO diets had no effect on soleus muscle or epididymal adipose tissue LPL activities from rats sacrificed in either the fed or overnight fasted state (data not shown). Since these diets did cause significant reductions in plasma triacylglycerol and FFA levels (Table 3), it is unlikely that regulation of these enzyme activities contributes significantly LIPIDS, Vol. 27, no. 12 (1992)
1016
D.A. OTTO E T AL. TABLE 4 Liver Weight and Triacylglycerol from a d libitum Fed and Overnight Fasted Rats a
Diets Fat concentration Low fat
Liver weight Liver Fat /Body weight triacylglycerols Nutritional state source (g) (• 1 0 0 ) (~maol'liver -1) ad libiturn Fed LD 16.8 4- 0.7 4.3 +_ 0.1 429 +-- 61 MO 16.4 4- 0.5 4.2 4- 0.1 259 4- 37b Medium fat LD 17.9 4- 0.3 4.4 4- 0.1 673 +- 11~ MO 18.4 4- 0.4 4.6 4- 0.1 b 411 4- 51 Overnight fasted Low fat LD 13.0 4- 0.7 3.4 4- 0.1 250 4- 35 MO 13.1 4- 0.8 3.4 4- 0.2 254 4- 37 Medium fat LD 11.3 4- 0.6 2.9 4- 0.1 342 4- 67 MO 13.3 4- 0.4b 3.4 4- 0.1 b 270 4- 33 aRats were fed the respective diets for 2 wk before sacrifice. Values are the mean 4- SE (n = 6). MO, menhaden oil; LD, lard. bSignificant difference between MO- and LD-fed animals, P < 0.05. to either of these effects of FO, at least at the FO concentrations used in this study. Although not statistically significant, there was a tendency for the L F and M F MOfed rats (both ad l i b i t u m fed and overnight fasted) to have reduced epididymal adipose tissue wt/body wt ratios (not shown} as observed previously by us (2; with H F diets) and others (4,5). In our earlier study we demonstrated that this reduction in adipose tissue mass was positively correlated with the FO-dependent decrease in FFA levels. It is, thus, reasonable to propose t h a t the reduction in FFA is linked to the decrease in adipose tissue mass. Parrish e t al. (5) reported t h a t the decrease in adipose tissue mass was accompanied by a decrease in lipoprotein binding and basal lipolysis in epididymal adipocytes, and an increase in hormone-stimulated lipolysis in both epididymal and perirenal adipocytes. These effects together with the hypotriglyceridemic effect of FO could explain the decrease in adipose tissue mass and the related decrease in plasma FFA. L i v e r w e i g h t and triacylglycerol levels. In our earlier s t u d y (1), we observed t h a t feeding H F MO diets caused significant hepatomegaly t h a t was due entirely to hypertrophy. The current s t u d y demonstrates t h a t this effect of FO, which is reflected in an increase in the liver wt/body wt ratio was also observed in the MF, but not the L F diet groups regardless of the nutritional state of the rats at sacrifice (Table 4). We (1) and others (11) have suggested that the hepatomegaly is related to peroxisomal proliferation induced by FO-feeding. We also reported earlier (1) t h a t there was an elevation in liver triacylglycerols caused by H F MO-feeding concomitant with the lowering of plasma triacylglycerol levels, suggesting a primary inhibition of hepatic triacylglycerol secretion by MO, independent of triacylglycerol synthesis. Our suggestion is supported by the recent in vitro experiment by Lang and Davis (12) indicating t h a t eicosapentaenoic acid and docosahexaenoic acid decreased the secretion of both triacylglycerols and apo B by cultured rat hepatocytes independent of their synthesis, but only at the highest concentration tested (1 mM f a t t y acid]0.13 mM albumin). Consistent with this, we did not see an accumulation of liver triacylglycerols after feeding rats L F or M F diets (Table 4). In fact, there was a significant reduction in liver triacylglycerol levels in MOfed rats (Table 4; ad l i b i t u m fed only) under these dietary conditions, resulting in a highly significant relationship LIPIDS, Vol. 27, no. 12 (1992)
between the plasma and liver triacylglycerol concentrations (data from Tables 3 and 4; r = 0.71, P < 0.001, n = 24). Thus, our data suggest that very low density lipoprotein secretion p e r se may be rate-limiting, but only with H F diets when the supply of exogenous FFA may exceed the capacity of the liver to dispose of them. With the MF and L F diets, the rate of hepatic triacylglycerol synthesis appears to limit secretion. Conclusion. The present s t u d y offers an explanation for the apparent variable response of plasma FFA to feeding dietary FO. Our data indicate that this is due to the nutritional state of the animal at the time of blood sampling. The FO-induced reduction of plasma FFA was consistently observed in ad l i b i t u m fed animals at all fat levels (LF, M F and HF), resulting in a strong positive relationship between plasma FFA and triacylglycerols. This strongly suggests t h a t lowered plasma FFA levels by FO-feeding may play a significant role in the triacylglycerol-lowering effects by limiting the availability of substrate for hepatic triacylglycerol synthesis. We have shown t h a t the reduction in FFA levels by MO-feeding was not related to changes in the insulin concentration or insulin sensitivity. The previously observed effect of the H F MO diet on skeletal muscle L P L was not seen with the L F or M F diets, suggesting t h a t catabolism of triacylglycerol-rich lipoproteins contributes little, if any, to the FO-dependent reductions of plasma triacylglycerols or FFA. The current s t u d y indicates t h a t the apparent direct inhibition of triacylglycerol secretion by FO (1) imposes a rate-limitation only when feeding H F diets. At the lower fat concentrations, hepatic triacylglycerol synthesis appears to limit secretion.
ACKNOWLEDGMENTS This research was supported by the National Institutes of Health Grant Number 1 R01 HL39246 from the National Heart, Lung and Blood Institute. Vacuum stripped MO was provided by the Fish Oil Test Materials Program, Division of Nutrition Research Coordination, Building 31, Room 4B63, National Institutes of Health, Bethesda, MD 20892. REFERENCES 1. Otto, D.A., Tsai, CE., Baltzell, J.K., and Wooten, J.T. (1991) Biochim. Biophys. Acta 1082, 37-48. 2. Baltzell, J.K., Wooten, J.T., and Ott~ D.A. (1991) Lipids 26, 289-294.
1017
FISH OIL, FFA AND PLASMA TRIACYLGLYCEROLS 3. Singer, P., Wirth, M., and Berger, I. (1990) Atherosclerosis 83, 167-175. 4. Parrish, C.C., Pathy, D.A., and Angel, A. (1990) Metabolism 39, 217-219. 5. Parrish, C.C., Path),, D.A. Parkes, J.G., and Angel, A. (1991) J. Cell. PhysioL 148, 493-502. 6. Otto, D.A., Baltzell, J.K., and Wooten, J.T. (1991) FASEB J. 5, A1303. 7. American Institute of Nutrition (1977)J. Nutr. 107, 1340-1348. 8. Haglund, O., Luostarinen, R., Wallin, R., Wibell, L., and Saldeen, T. (1991)J. Nutr. 121, 165-169.
9. Tsai, C.E., Wooten, J.T., and Otto, D.A. (1989) Nutr. Res. 9, 673-678. 10. Veech, R.L., Guynn, R.W, and Veloso, D. (1972)Biochem. J. 127, 387-397. 11. Yamazaki, R.K., Shen, T., and Schade, G.B. 0987) Biochim. Biophys. Acta 920, 62-67. 12. Lang, C.A., and Davis, R.A. (1990)J. LipidRes. 31, 2079-2086. [Received February 10, 1992, and in revised form September 14, 1992; Revision accepted October 15, 1992]
LIPIDS, Vol. 27, no. 12 (1992)
Reduction in Triacylglycerol Levels by Fish Oil Correlates with Free Fatty Acid Levels in ad libitum Fed Rats David A. Otto*, Janet K. Baltzell I and Joseph T. Wooten Department of Research, The Baptist Medical Centers, Birmingham, Alabama 35211 Rats were fed (for 2 or 6 wk) purified diets containing lard (LD) or menhaden oil (MO) at two levels of dietary fat, L~, at 11.5 and 20.8% of energy in the low fat (LF) and the medium fat (MF) diets, respectively. Following the diet period, rats were sacrificed after either an overnight fast or after ,mlnterrupted a d libitum feeding. The studies were designed to investigate the dependence of our previously reported effects of MO, Le. the reduction of plasma free f a t t y acid (FFA) levels and accumulation of hepatic triacylglycerols, on the dietary fat concentration and the nutritional state of the animal at the time of sacrifice. Reductions in plasma triacylglycerol and cholesterol levels in MO-fed relative to LD-fed rats were observed under all conditions. FFA levels were consistently reduced by M(~ feeding at both dietary fat concentrations, but only when blood was sampled from a d libitum fed rats. Under these conditions there was a significant positive relationship between plasma FFA and triacylglycerol concentrations. R~ duction in plasma FFA levels m a y be an additional mechanism associated with the triacylglycerol-lowering effect of fish oil (FO). The L F and M F MO diets caused a rise in plasma glucose levels with no significant change in insulin concentration, indicating that the reduction of FFA by MO was not related to changes in insulin concentration or insulin sensitivity. The MO diets had no effect on skeletal muscle or epididymal adipose tissue lipopr~ tein lipase activity, demonstrating that catabolism of triacylglycerol-rich lipoproteins contributes little, if any, to the MO-dependent reductions of plasma triacylglycerol and FFA. The previously reported accumulation of hepatic triacylglycerols after high fat (HF; 30% of energy) MOfeeding was not observed with the L F or M F MO diets, suggesting t h a t the apparent direct inhibition of triacylglycerol secretion by FO imposes a rate-llmitation only when feeding H F diets. Lipids 27, 1013-1017 (1992). In earlier studies (1,2) we observed significant but inconsistent reductions in plasma free fatty acid (FFA) levels in fish oil (FO)-fed, overnight fasted rats. An effect on plasma FFA by FO was also observed by Singer et aL (3), who reported that dietary r polyunsaturated fatty acid (n-3 fatt y acids) supplementation in hyperlipidemic patients resulted in a marked decrease in FFA during a standard glucose tolerance test that was associated with a decline in serum triacylglycerol levels. The authors (3) suggested that the decrease in FFA indicated reduced lipolysis and that it might contribute to the triacylglycerol-lowering effect of n-3 fatty acids. We and others also demonstrated a decrease in *To whom correspondence should be addressed at Department of Research, The Baptist Medical Centers, 701 Princeton Avenue, Birmingham, AL 35211. 1Present address: NRICGP, CSRS/USDA, 901 D Street S.W., Suite 323, Aerospace Building, Washington, D.C. 20250-2200. Abbreviations: CO, corn oil; FFA, free fatty acids; FO, fish off; HF, high fat; LD, lard; LF, low fat; LPL, lipoprotein lipase; MF, medium fat; MO, menhaden oil.
adipose tissue mass in FO-fed rats (2,4,5) that was accompanied by a decrease in adipocyte volume without a decrease in fat cell number (4,5}. Additionally, we found that this decrease in adipose tissue mass was positively correlated with the FO-mediated changes in plasma FFA (2), suggesting a causal relationship. Since a reduction of plasma FFA may reflect an important mechanism involved in FOmediated effects, we investigated why this effect of FO has not been routinely observed in the numerous studies conducted with dietary FO. Specifically, we considered whether observing a reduction in plasma FFA levels was dependent upon the concentration of d i e t a r y n-3 fatty acids or the nutritional state (overnight fasted vs. ad libitum fed rats) of the animals at the time of sacrifice We also investigated the relationship between plasma FFA and the lowering of plasma triacylglycerol levels by FO. A preliminary account of this work has been presented (6).
MATERIALS AND METHODS Experimental design and diets. Male Sprague-Dawley rats (225-250 g; VAF/Plus, Charles River Laboratories, Inc., Raleigh, NC) were housed in suspended transparent polycarbonate cages with stainless steel wire bottoms in an animal facility maintained at 22 + 1~ 70% humidity and with a 12-h light/dark cycle. Rats were acclimated to the surroundings and fed a standard AIN-76A purified diet (7) (Table 1, low fat diet with corn oil as the fat source; Research Diets, Inc., New Brunswick, NJ) for I wk. Animals had free access to deionized water. Following the acclimation period, rats were matched for weight and assigned to respective groups (6 rats/group), based on the diet, the length of feeding, and the nutritional state of the animal at sacrifice. Rats were assigned to one of four diet groups. These included low fat (LF; 11.5% of energy) and medium fat (MF; 20.8% of energy) diets, prepared with either lard (LD) or menhaden oil (MO; vacuum stripped MO provided by the National Institutes of Health, Bethesda, MD) as the major fat source (Tables I and 2). All diets contained a minimum of 4% of energy from corn oil (CO) to prevent essential f a t t y acid deficiency. Thus, LD and MO provided 7.5 and 16.8% of energy in the L F and M F diets, respectively. The purified diets were based on the AIN-76A diet (7) and contained identical ingredients except for the source of fat. In the M F diets, the additional fat energy was balanced by an isocaloric reduction in carbohydrate (sucroseJcorn starch, 2:1, w/w). Therefore. the nutrient (protein, vitamins and minerals)/energy (g/kJ) ratios were the same in all diets. As in our previous studies (1,2), the diets were also balanced for cholesterol (8.96 mg/1000 kJ) and/3-sitosterol {76.8 rag/ 1000 kJ) relative to the energy content of the diets, and for the antioxidant content (a- and y-tocopherols, and tertbutylhydroquinone; 1.5, 1.2 and 0.2 mg/g fat) relative to the amount of fat in the diets. The importance of including antioxidante in the diets is emphasized by a recent report by Haglund et al. (8) indicating t h a t an FO diet rich in vitamin E (1.5 IU/g fat) had a greater triacylglycerollowering effect t h a n one with only 0.3 IU/g. Our diets LIPIDS, Vol. 27, no. 12 (1992)
1014
D.A. OTTO ET AL. TABLE 2
TABLE 1
Fatty Acid Composition of Lard and Menhaden Oil
Composition of Purified Diets a
Low fat Energy g/kg (%) 50.0 11.5
Medium fat Energy g/kg (%) 95.0 20.8
Ingredients Fat b Sucrose/corn starch (2:1, w/w) 650.0 66.6 589.0 57.3 Casein (alcohol extracted) 200.0 20.5 211.0 20.5 Cellulose 50.0 0.0 52.6 0.0 DL-Methionine 3.0 0.0 3.2 0.0 Salt mixc 35.0 0.4d 36.8 0.4d Vitamin mixe 10.0 1.0d 10.5 1.0d Choline bitartrate 2.0 0.0 2.1 0.0 aDiets were based on the AIN-76A diet. Medium fat diets were prepared by isocaloric substitution of fat for carbohydrate; thus the nutrient (protein, vitamins and minerals)/kJ ratios were unchanged from the low fat diet. The energy contents of the respective diets were 16.4 and 17.3 kJ/g for the low and medium fat diets. bAll diets contained a minimum of 4% of energy from corn oil to prevent essential fatty acid deficiency. The remaining fat energy was supplied by either lard or menhaden oil. Thus, lard and menhaden oil provided 7.5 and 16.8% of energy in the low and medium fat diets, respectively. Menhaden oil was provided by the Fish Oil Test Materials Program of the National Institutes of Health. CSalt mix (in g/kg of mixture} calcium phosphate, dibasic, 500; magnesium oxide, 24; potassium citrate, monohydrate, 220; potassium sulfate, 52; sodium chloride, 74; chromium potassium sulfate, 0.55; cupric carbonate, 0.3; potassium iodate, 0.01; ferric citrate, 6.0; manganous carbonate, 3.5; sodium selenite, 0.01; zinc carbonate, 1.6; sucrose, 118.03. dAs sucrose. evitamin mix (in g/kg of mixture): Vitamin A palmitate (500,000 IU/g), 0.8; Vitamin D3 (400,000 IU/g), 0.25; Vitamin E acetate (500 IU/g), 10.0; menadione sodium bisulfate, 0.08; biotin (1%), 2.0; cyanocobalamin(0.1%), 1.0; folic acid, 0.2; niacin, 3.0; calciumpantothenate, 1.6; pyridoxine HC1, 0.7; riboflavin, 0.6; thiamin HC1, 0.6; sucrose, 979.17.
contained approximately 1.4 IU of vitamin E/g fat along with tert-butylhydroquinone The MO diets were prepared in our laboratory by addition of MO to a basal diet as previously described (1,9}. All other diets were customprepared by a commercial vendor (Research Diets, Inc.). Diet cups were placed in the cages such that the diets were not exposed to direct light. All diets were replaced every third or fourth day (9). Rats were fed one of the four diets for 2 wk (12 rats/diet group) or 6 wk (6 rats/diet group}. At the end of the feeding period, half of the animals from each group were sacrificed in the morning after uninterrupted ad libitum feeding. The other half were sacrificed after an overnight fast. The data from 6 wk fed rats are presented in the text only when relevant to the discussion. Plasma and tissue collection. A t the time of sacrifice, rats were anesthetized with pentobarbital (5 mg/100 g body wt). Blood was drawn from the abdominal aorta into a syringe containing 0.25 mL of 0.2 M ethylene diaminetetraacetic acid, p H 7.4, and plasma was collected by centrifugation at 500 X g (4~ for 20 rain. Immediately after drawing blood, livers were rapidly removed, weighed and freeze-clamped with aluminum tongs precooled in liquid nitrogen (10). The thin layer of frozen liver tissue was later ground into a fine powder under liquid LIPIDS, Vol. 27, no. 12 (1992)
Fatty acid 14:0 14:1 16:0 16:1n-9 16:1n-7 18:0 18:1n-9 18:2n-6 18:3n-3 20:1 20:5n-3 22:1n-9 + n-ll 22:5n-3 22:6n-3 Saturated Monounsaturated Polyunsaturated n-6 n-3 P/S ratio
Lard 1.4 0.1 24.5 3.3 15.0 42.8 10.5 1.0 0.6 40.9 46.8
% of Total fatty acids Menhaden oil 7.6 0.1 16.6 0.2 10.3 3.1 8.1 1.2 1.2 2.3 14.8 0.5 2.5 8.7 27.3 21.5
10.5 1.0 0.28
1.2 27.2 1.04
nitrogen. Soleus muscles and epididymal fat pads were collected, weighed and immediately frozen by dropping into liquid nitrogen. All samples were stored at - 8 0 ~ until assayed. Analytical procedures. P l a s m a t r i a c y l g l y c e r o l , cholesterol and glucose were measured using Baker kits with the Encore Chemistry S y s t e m centrifugal analyzer (Baker I n s t r u m e n t s Corp., Allentown, PA). The assay for FFA (Amano International E n z y m e kit reagents, Troy, VA) was adapted for use on the Encore chemistry System (1). Plasma insulin was determined by radioimmunoassay using a commercial kit (Ciba C o m i n g Diagnostics Corp., Medfield, MA). Liver triacylglycerol was extracted and assayed as previously reported (1). Soleus muscle and epididymal adipose tissue lipoprotein lipase (LPL) activities were measured as described earlier (2). Statistical analysis. Statistical differences between LD and MO diet groups were determined by the Student's ttest. Correlation between measured parameters was determined by least squares regression analysis. All statistical analyses were performed utilizing the SYSTAT statistical software package (SYSTAT, Inc., Evanston, IL).
RESULTS AND DISCUSSION
Plasma triacylglyceroL cholesterol and free fatty acids. The reduction in plasma triacylglycerol and cholesterol levels in MO-fed relative to LD-fed rats was observed consistently under all conditions of dietary fat concentration and nutritional state (Table 3). In this study we considered whether the normal rise in plasma FFA t h a t occurs with fasting might sometimes mask effects of FO on FFA, and thus, explain the inconsistencies of our previous observations with overnight fasted rats (1). After 2 wk of feeding, a significant decrease in plasma FFA was observed in the MO- vs. LD-ad libitum fed rats at both fat concentrations. In contrast, after an overnight fast, FFA were reduced
1015
F I S H OIL, F F A A N D P L A S M A T R I A C Y L G L Y C E R O L S
TABLE 3
Plasma Metabolites from a d libitum Fed and Overnight Fasted R a t s a
Nutritional state
Diets Fat concentration
ad libitum F e d
Low fat Medium fat
Overnight fasted
Low fat Medium fat
Fat source
Triacylglycerols (mM)
Free f a t t y acids (~M)
Cholesterol (mM)
Glucose (mM)
Insulin (~U/mL)
LD MO LD MO
3.35 1.25 4.25 1.74
+-- 0.11 +-- 0.20 b _+ 0.29 +_ 0.07 b
2.40 1.86 2.45 1.47
+_ 0.18 +-- 0.13 b __ 0.47 __ 0.16 b
316 207 348 168
_+ + +_ +_
35 21 b 34 16 b
8.9 10.4 9.0 9.9
-+ +_ +-
0.4 0.3 b 0.4 0.2 b
70 65 93 75
++ +-+
8 6 9 8
LD MO LD MO
1.90 0.69 1.44 0.78
+_ _ _ +_
1.78 1.60 1.99 1.29
+ 0.13 +- 0.16 --+ 0.18 + 0.08 b
418 410 499 343
__ __ + +_
41 36 52 18 b
8.1 8.7 5.9 8.2
+ + ++_
0.2 0.6 0.3 0.7 b
66 47 56 44
++_ + _+
6 8 4 3
0.21 0.06 b 0.12 0.06 b
a R a t s were fed t h e r e s p e c t i v e d i e t s for 2 w k before sacrifice. V a l u e s are t h e m e a n -- S E (n -- 6). A b b r e v i a t i o n s : M O , m e n h a d e n oil; LD, lard. b s i g n i f i c a n t difference b e t w e e n MO- a n d L D - f e d a n i m a l s , P < 0.05.
only in the MF MO group. After 6 wk this effect was again significant in ad l i b i t u m fed rats (with LF diets, 429 _+ 54 vs. 186 +_ 23 ~M FFA; and with MF diets, 338 + 92 vs. 176 +_ 21 in LD-fed and MO-fed rats, respectively) but not seen at all in the overnight fasted rats (with LF diets, 406 +_ 13 vs. 410 +_ 38, and with MF diets, 463 +_ 45 vs. 413 +__53 in LD-fed and MO-fed rats, respectively). In our earlier study (1) we made similar observations with high fat (HF) diets (30% of energy; identical to diets described in Table 1 with an isocaloric substitution of fat for carbohydrate). After uninterrupted ad l i b i t u m feeding of HF CO, LD and MO diets for 2 wk, FFA were significantly reduced in the MO-fed rats relative to the CO- and LD-fed rats (435 __ 34, 448 __ 68 and 201 _+ 43 ~M in CO-, LD- and MO-fed rats, respectively; P < 0.05; data from ad libitum fed rats were not previously reported). However, like the LF- and MF-fed rats after an overnight fast, there were no differences between the three H F diet groups (351 + 28, 392 +_ 102 and 377 +_ 83 ~M in CO-, LD- and MOfed rats, respectively). Thus, the data clearly demonstrate that blood must be sampled in the fed state in order to consistently observe the effects of dietary FO on plasma FFA. There was a highly significant correlation between plasma triacylglycerol and FFA levels in ad l i b i t u m fed rats (Fig. 1; data from Table 3). We also observed this with ad libitum fed rats receiving HF diets [r -- 0.74, P < 0.006, n = 12; from the experiment of (1), but data not previously shown]. These data strongly suggest that along with other mechanisms, the lowering of plasma FFA levels by FOfeeding may contribute significantly to the decrease in plasma triacylglycerols by decreasing the supply of substrate for hepatic triacylglycerol synthesis. This is consistent with the recent report of Singer et al. (3) and focuses attention to possible effects of dietary FO on adipose tissue lipolysis, as has been observed by Parrish et al. (5). P l a s m a insulin and glucose. We previously reported that feeding HF MO diets followed by an overnight fast resuited in a reduction in plasma insulin levels with an accompanying reduction or no change in plasma glucose levels (1,2). This is consistent with an improvement in peripheral insulin sensitivity. In the present experiment with the LF and MF MO diets, insulin levels were only
~ " 6.0
" •5.0 4.0
5 ej
~, 3.0 2.0 .~ff'/'~ -- _
I--
1.0 CD
f
C') L F - M E ~ H A D ~ ~J
CqL
A LF-L~D
9 MF-I.ARD
O.O
l
i
I
,
I
|
!
i
1 O0 200 300 400 PLASMA FREE FATTY ACIDS (p,m o l / L )
500
F I G . 1. Correlation between p l a s m a free fatty acids a n d p l a s m a triacylglycerols from a d libitum f e d rats [data f r o m Table 3; for all data, r = 0.71, P < 0.001, n = 23; for low f a t (LF) diets, r = 0.68, P = 0.022, n = 11; and for m e d i u m f a t (MF) diets, r = 0.92, P < 0.001, n = 12].
slightly lowered or unchanged relative to those in LD-fed rats (Table 3; approaching significance only in overnight fasted rats adapted to the LF or MF MO diets, P --- 0.074 and 0.059, respectively). Under these conditions, plasma glucose was generally elevated (Table 3). The rise in glucose may have been in response to the slight reductions in plasma insulin levels, but it is also possible that insulin sensitivity (peripheral or hepatic) actually worsened with LF and MF MO-feeding. The data clearly show that the lower plasma FFA levels seen with MO-feeding at all concentrations cannot be explained by either an increase in insulin concentration or improved insulin sensitivity at the site of the adipose tissue. L P L and adipose tissue mass. Unlike our earlier study (2) in which we observed a significant elevation of soleus muscle LPL in H F MO-fed rats, feeding LF and MF MO diets had no effect on soleus muscle or epididymal adipose tissue LPL activities from rats sacrificed in either the fed or overnight fasted state (data not shown). Since these diets did cause significant reductions in plasma triacylglycerol and FFA levels (Table 3), it is unlikely that regulation of these enzyme activities contributes significantly LIPIDS, Vol. 27, no. 12 (1992)
1016
D.A. OTTO E T AL. TABLE 4 Liver Weight and Triacylglycerol from a d libitum Fed and Overnight Fasted Rats a
Diets Fat concentration Low fat
Liver weight Liver Fat /Body weight triacylglycerols Nutritional state source (g) (• 1 0 0 ) (~maol'liver -1) ad libiturn Fed LD 16.8 4- 0.7 4.3 +_ 0.1 429 +-- 61 MO 16.4 4- 0.5 4.2 4- 0.1 259 4- 37b Medium fat LD 17.9 4- 0.3 4.4 4- 0.1 673 +- 11~ MO 18.4 4- 0.4 4.6 4- 0.1 b 411 4- 51 Overnight fasted Low fat LD 13.0 4- 0.7 3.4 4- 0.1 250 4- 35 MO 13.1 4- 0.8 3.4 4- 0.2 254 4- 37 Medium fat LD 11.3 4- 0.6 2.9 4- 0.1 342 4- 67 MO 13.3 4- 0.4b 3.4 4- 0.1 b 270 4- 33 aRats were fed the respective diets for 2 wk before sacrifice. Values are the mean 4- SE (n = 6). MO, menhaden oil; LD, lard. bSignificant difference between MO- and LD-fed animals, P < 0.05. to either of these effects of FO, at least at the FO concentrations used in this study. Although not statistically significant, there was a tendency for the L F and M F MOfed rats (both ad l i b i t u m fed and overnight fasted) to have reduced epididymal adipose tissue wt/body wt ratios (not shown} as observed previously by us (2; with H F diets) and others (4,5). In our earlier study we demonstrated that this reduction in adipose tissue mass was positively correlated with the FO-dependent decrease in FFA levels. It is, thus, reasonable to propose t h a t the reduction in FFA is linked to the decrease in adipose tissue mass. Parrish e t al. (5) reported t h a t the decrease in adipose tissue mass was accompanied by a decrease in lipoprotein binding and basal lipolysis in epididymal adipocytes, and an increase in hormone-stimulated lipolysis in both epididymal and perirenal adipocytes. These effects together with the hypotriglyceridemic effect of FO could explain the decrease in adipose tissue mass and the related decrease in plasma FFA. L i v e r w e i g h t and triacylglycerol levels. In our earlier s t u d y (1), we observed t h a t feeding H F MO diets caused significant hepatomegaly t h a t was due entirely to hypertrophy. The current s t u d y demonstrates t h a t this effect of FO, which is reflected in an increase in the liver wt/body wt ratio was also observed in the MF, but not the L F diet groups regardless of the nutritional state of the rats at sacrifice (Table 4). We (1) and others (11) have suggested that the hepatomegaly is related to peroxisomal proliferation induced by FO-feeding. We also reported earlier (1) t h a t there was an elevation in liver triacylglycerols caused by H F MO-feeding concomitant with the lowering of plasma triacylglycerol levels, suggesting a primary inhibition of hepatic triacylglycerol secretion by MO, independent of triacylglycerol synthesis. Our suggestion is supported by the recent in vitro experiment by Lang and Davis (12) indicating t h a t eicosapentaenoic acid and docosahexaenoic acid decreased the secretion of both triacylglycerols and apo B by cultured rat hepatocytes independent of their synthesis, but only at the highest concentration tested (1 mM f a t t y acid]0.13 mM albumin). Consistent with this, we did not see an accumulation of liver triacylglycerols after feeding rats L F or M F diets (Table 4). In fact, there was a significant reduction in liver triacylglycerol levels in MOfed rats (Table 4; ad l i b i t u m fed only) under these dietary conditions, resulting in a highly significant relationship LIPIDS, Vol. 27, no. 12 (1992)
between the plasma and liver triacylglycerol concentrations (data from Tables 3 and 4; r = 0.71, P < 0.001, n = 24). Thus, our data suggest that very low density lipoprotein secretion p e r se may be rate-limiting, but only with H F diets when the supply of exogenous FFA may exceed the capacity of the liver to dispose of them. With the MF and L F diets, the rate of hepatic triacylglycerol synthesis appears to limit secretion. Conclusion. The present s t u d y offers an explanation for the apparent variable response of plasma FFA to feeding dietary FO. Our data indicate that this is due to the nutritional state of the animal at the time of blood sampling. The FO-induced reduction of plasma FFA was consistently observed in ad l i b i t u m fed animals at all fat levels (LF, M F and HF), resulting in a strong positive relationship between plasma FFA and triacylglycerols. This strongly suggests t h a t lowered plasma FFA levels by FO-feeding may play a significant role in the triacylglycerol-lowering effects by limiting the availability of substrate for hepatic triacylglycerol synthesis. We have shown t h a t the reduction in FFA levels by MO-feeding was not related to changes in the insulin concentration or insulin sensitivity. The previously observed effect of the H F MO diet on skeletal muscle L P L was not seen with the L F or M F diets, suggesting t h a t catabolism of triacylglycerol-rich lipoproteins contributes little, if any, to the FO-dependent reductions of plasma triacylglycerols or FFA. The current s t u d y indicates t h a t the apparent direct inhibition of triacylglycerol secretion by FO (1) imposes a rate-limitation only when feeding H F diets. At the lower fat concentrations, hepatic triacylglycerol synthesis appears to limit secretion.
ACKNOWLEDGMENTS This research was supported by the National Institutes of Health Grant Number 1 R01 HL39246 from the National Heart, Lung and Blood Institute. Vacuum stripped MO was provided by the Fish Oil Test Materials Program, Division of Nutrition Research Coordination, Building 31, Room 4B63, National Institutes of Health, Bethesda, MD 20892. REFERENCES 1. Otto, D.A., Tsai, CE., Baltzell, J.K., and Wooten, J.T. (1991) Biochim. Biophys. Acta 1082, 37-48. 2. Baltzell, J.K., Wooten, J.T., and Ott~ D.A. (1991) Lipids 26, 289-294.
1017
FISH OIL, FFA AND PLASMA TRIACYLGLYCEROLS 3. Singer, P., Wirth, M., and Berger, I. (1990) Atherosclerosis 83, 167-175. 4. Parrish, C.C., Pathy, D.A., and Angel, A. (1990) Metabolism 39, 217-219. 5. Parrish, C.C., Path),, D.A. Parkes, J.G., and Angel, A. (1991) J. Cell. PhysioL 148, 493-502. 6. Otto, D.A., Baltzell, J.K., and Wooten, J.T. (1991) FASEB J. 5, A1303. 7. American Institute of Nutrition (1977)J. Nutr. 107, 1340-1348. 8. Haglund, O., Luostarinen, R., Wallin, R., Wibell, L., and Saldeen, T. (1991)J. Nutr. 121, 165-169.
9. Tsai, C.E., Wooten, J.T., and Otto, D.A. (1989) Nutr. Res. 9, 673-678. 10. Veech, R.L., Guynn, R.W, and Veloso, D. (1972)Biochem. J. 127, 387-397. 11. Yamazaki, R.K., Shen, T., and Schade, G.B. 0987) Biochim. Biophys. Acta 920, 62-67. 12. Lang, C.A., and Davis, R.A. (1990)J. LipidRes. 31, 2079-2086. [Received February 10, 1992, and in revised form September 14, 1992; Revision accepted October 15, 1992]
LIPIDS, Vol. 27, no. 12 (1992)
