Brief Critical Reviews
July 7 992: 195-206
How Does Fish Oil Lower Plasma Triglycerides? Studies using dietary fish oil, safflower oil, and palm oil suggest that the decrease in plasma triacylglycerol due to dietary fish oils is somehow related to a decrease in the capacity of the liver to hydrolyze phosphatidate, which then affects microsomal synthesis of triacylglycerol from diacylglycerol.
Marine oils, rich in n - 3 fatty acids, appear to be more effective in lowering the concentration of plasma triacylglycerols than vegetable oils, generally rich in n - 6 fatty acids. The mechanisms responsible for the hypotriglyceridemic effects of fish oils are currently under intense investigation. Among the mechanisms suggested as possibly responsible for the effect of fish oils on plasma triacylglycerol levels are inhibition of the synthesis and secretion of triacylglycerol-containing lipoproteins, inhibition of fatty acid synthesis, enhancement of fatty acid oxidation, decreased activity of enzymes responsible for esterification of fatty acids, and changes in the ratios of the fatty acid esters formed. A recent study assessed the relative contribution of some of these mechanisms to the total hypotriglyceridemia produced by fish oils by comparing the effects of three different dietary oils (marine, safflower, and palm) on fatty acid oxidation and glycerolipid synthesis in rat liver.' The investigators simultaneously studied beta- and peroxisomal oxidation and key enzymatic activities in triacylglycerol (diacylglycerol acyltransferase and phosphatid a t e p h o s p h o h y d r o l a s e ) a n d phospholipid (cytidylyltransferase) biosynthesis. Young male rats were maintained for either ten days or one month on diets that differed only in the type of fat (fish, safflower, or palm oil). Food intake was allowed until 12 hours before the rats were killed. No significant differences were observed among groups in food intake, body weight gain, liver weight, or liver protein content. The plasma triacylglycerol levels of the fish oil group were 50% lower than those of the palm oil group, whereas This review was prepared by John G. Coniglio, Ph.D., at the Department of Biochemistry, Vanderbilt University, Nashville, TN 37232.
Nutrition Reviews, Vol. 50, No. 7
those of the safflower oil group were between the other two. The hepatic triacylglycerol was not affected by dietary treatment. Palmitic and oleic acids were the predominant fatty acids in plasma and liver of palm oil-fed rats, whereas linoleic acid was predominant in the safflower oil-fed group. Docosahexaenoic acid was present in small amounts in plasma and liver of safflower oil- and palm oil-fed rats, but both docosahexaenoic and eicosapentaenoic acids were abundant in fish oil-fed animals. In this group, the ratio of eicosapentaenoic to docosahexaenoic acid in plasma but not in liver reflected the dietary ratio. Beta- and peroxisomal oxidation were assessed in freshly prepared whole-liver homogenates incubated with l-'4C-palmitoyl CoA in the presence and absence of potassium cyanide. In this assay, mitochondrial beta-oxidation activity was observed as cyanide-sensitive activity (using conditions optimized for mitochondrial oxidation), and peroxisoma1 oxidation was due to cyanide-insensitive activity (using conditions optimized for peroxisomal activity). In both ten- and 30-day studies, no significant differences were observed among the three groups in mitochondrial beta-oxidation. In the tenday study, peroxisomal fatty acid oxidation in the fish-oil group was about 45% higher than in the other two groups, but the proportion of total oxidation that was peroxisomal was unchanged. However, peroxisomal oxidation activity in the fish-oil group studied at 30 days was 25-50% lower than that observed after ten days, and the rates were not different among the three experimental groups. In previous studies2 it has been suggested that diversion of polyunsaturated fatty acids (particularly eicosapentaenoic and docosahexaenoic) from pathways of esterification was a major cause of the fish oil-induced lowering of triacylglycerol levels. A major alternative pathway through fatty acid oxidation had been demonstrated in perfused liver^.^ In the latter studies, hepatic release of VLDL (very low density lipoproteins)-triacylglycerols was depressed by fish but not by safflower oil, and ketogenesis was increased in livers of fish oil-fed rats, compared to those of controls and safflower oil-fed rats, which were similar. Liver peroxisomes oxidized polyunsaturated fatty acids such as arachidonic and eicosapentaenoic acids, but mitochondria
195
appeared to be the major site of oxidation of these c o m p o ~ n d sAlthough .~ Halminski et al.' observed a higher rate of peroxisomal oxidation in fish oil-fed rats vs. those fed safflower or palm oil, the proportion of total oxidation by peroxisomes was similar in the three groups. From these data, the investigators inferred that fish oil had no specific effect on peroxisomes but, rather, increased both mitochondrial and peroxisomal oxidation. These results should be cautiously interpreted, since the fatty acid substrate used had a shorter chain length than fatty acids normally considered as natural substrates for peroxisomal oxidation and that the rates among the three groups of rats were not different at the end of the 30-day feeding period. Diacylglycerol acyltransferase was measured in whole homogenates and all subcellular fractions using ''C-myristoyl CoA and diolein. Whole homogenates stored at -70°C for 12 months suffered no substantial loss of activity, according to the authors. The activity of this enzyme was not affected by the type of fat in the diet, and in the rats fed 30 days the activity was twofold greater than that observed after ten days of feeding. The lack of an effect of fish oil on enzymatic activity indicated that this enzyme was not involved in the effect of the oil on synthesis of triacylglycerols. However, the aut h o r cautioned ~~ that inhibition of the activity of this enzyme by eicosapentaenoic acid had been reported in studies using cultured rat hepatocytes. Such inhibition could result in decreased synthesis of triacylglycerols due to reduced activity of this enzyme. Many differences in experimental conditions between the two studies preclude specific comparisons. Phosphatidate hydrolysis, also measured in whole homogenates and all subcellular fractions, was determined by release of inorganic phosphate from dimyristoyl phosphatidylcholine. The fish oilfed rats had the lowest activity of the three groups (19% lower than the safflower oil-fed group, which in turn had an activity 22% lower than the palm oil-fed rats). Pooled data from all three groups significantly and positively correlated with plasma concentrations of triacylglycerols. (See Figure 1.) This was interpreted to indicate that the rate of hepatic phosphatidate hydrolysis was a good predictor of plasma triacylglycerol concentration. A decrease of 10-35% in the capacity to hydrolyze phosphatidate occurred in all groups in the 30-day feeding study. However, the activity of the rats fed fish oil continued to be lower than that of the other two groups, which at this time had similar activities. Because the rate of phosphatidate hydrolysis is considered a key intermediate process in the synthesis of glycerolipids, this enzymatic step might be involved with the plasma triacylglycerol-lowering 196
z.ol 0
E E
1.5
Y
In
-m 0-
!
0.0 300
1
400
500
600
700
1
I
800
900
Phosphatidate Hydrolysis nrnol/(rnin * g liver)
Figure 1. Correlation of plasma triglyceride concentration with liver phosphatidate hydrolysis in rats fed different dietary lipids for ten days. Points correspond to individual animals (n = 9/dietary group): closed squares, open squares, and closed triangles represent fish oil-, safflower oil-, and palm oil-fed animals, respectively; r = 0.68, p < 0.001. Reprinted with permission.'
effect of marine oils. Decreased phosphatidate hydrolase activity had also been reported in livers of rats perfused with albumin-bound docosahexaenoic and eicosapentaenoic acids6 and in liver microsomes of rats fed a commercial diet supplemented with fish oiL7 Safflower oil, as well as fish oil, affected the activity of phosphatidate hydrolase in the studies of Halminski et al. ; I therefore, the authors suggested that the effect might depend on the degree of unsaturation of the fat. Phosphoryl choline cytidylyltransferase activity was measured by the conversion of phosphoryl[methyl ''C]-choline to radiolabeled cytidine diphosphocholine by liver homogenates and subcellular fractions. The activity was significantly but only slightly lower in the rats fed fish oil for ten days than that in the other two groups (which had similar values). Feeding the diets for 30 days resulted in lower activities in all groups, but the values for the fish oil-fed rats were still the lowest of the three groups. The results observed in these studies with fish oil agree with previous findings of decreased synthesis of phosphatidyl choline in hepatocytes preincubated with eicosapentaenoic acid.* Because safflower oil and palm oil had similar effects on the activity of this enzyme, the action of fish oil was apparently due specifically to the n - 3 acids and not merely to unsaturation. Although the effect was smaller than that on phosphatidate hydrolase activity, it may be important in regulating VLDL synthesis and secretion, in which process phosphatidyl choline is an essential component. Nutrition Reviews, Vol. 50,No. 7
Activities of enzymes that are important in the processes of VLDL-triacylglycerol synthesis and secretion could possibly be affected by translocation from cytosol to the site of activity in membranes. However, in these studies, no evidence supporting this possibility was obtained. As a result of their studies, Halminski et al.' suggested that the decrease in plasma triacylglycer01s due to dietary fish oils is somehow related to a decrease in the capacity of the liver to hydrolyze phosphatidate, which would then affect microsomal triacylglycerol synthesis from diacylglycerol. However, the precise mechanisms by which fish oils mediate these effects are still unknown, and it appears that both the degree of unsaturation and the precise position of the double bonds may play a role in these effects. Other proposed mechanisms cannot be excluded. The continued exploration of the effect of individual fatty acids on key enzymes involved in the complex process of synthesis and secretion of all components of the lipoproteintriacylglycerol particle remains worthwhile. Halminski MA, Marsh JB, Harrison EH. Differential effects of fish oil, safflower oil and palm oil on fatty acid oxidation and glycerolipid synthesis in rat liver. J Nutr 1991;121:1554-61 Wong S, Reardon M, Nestel P. Reduced triglyceride formation from long-chain polyenoic fatty acids in rat hepatocytes. Metabolism 1985;34:900-5 Wong SH, Nestel PJ, Trimble RP, Storer GB, lllman
RJ, Topping DL. Adaptive effects of dietary fish and safflower oil on lipid and lipoprotein metabolism in perfused rat liver. Biochim Biophys Acta 1984;792: 1039
4. Christensen E, Hagve T, Christophersen BO. Mitochondrial and peroxisomal oxidation of arachidonic and eicosapentaenoic acid studied in isolated liver cells. Biochim Biophys Acta 1986;879: 313 2 1 5. Rustan AC, Nossen JO, Christiansen EN, Drevon CA. Eicosapentaenoic acid reduces hepatic synthesis and secretion of triacylglycerol by decreasing the activity of acylcoenzyme A:l,2-diacylglycerol acyltransferase. J Lipid Res 1988;29:1417-26 6. Wong S, Marsh JB. Inhibition of apolipoprotein synthesis and phosphatidate phosphohydrolase activity by eicosapentaenoic and docosahexaenoic acids in the perfused rat liver. Metabolism 1988;37: 1177-81 7. Marsh JB, Topping DL, Nestel PJ. Comparative effects of dietary fish oil and carbohydrate on plasma lipids and hepatic activities of phosphatidate phosphohydrolase, diacylglycerol acyltransferase and neutral lipase activities in the rat. Biochim Biophys Acta 1987;922:23%43 8. Strum-Odin R, Adkins-Finke 8 , Blake WL, Phinney SD, Clarke SD. Modification of fatty acid composition of membrane phospholipids in hepatocyte monolayer with n-3, n-6 and n-9 fatty acids and its relationship to triacylglycerol production. Biochim Biophys Acta 1987;921:378-91
Cellular Retinol-Binding Protein Functions in the Regulation of Retinoid Metabolism A microsomal retinyl ester hydrolase from rat tissues is activated by apo-cellular retinol-binding protein, which thus regulates the availability of retinol either for binding to serum retinolbinding protein or for metabolic oxidation. Microsomal retinal synthesis was found to occur in rat tissues with holo-cellular retinol-binding protein as substrate. Retinal could be further oxidized to retinoic acid by a cytosolic fraction.
A principle that emerged some years ago and states that retinoids do not exist in tissues in the free state but are always protein-bound has now been extended: retinoids reacting as intermediates in metabolic pathways do not diffuse through the aqueous This review was prepared by George Wolf, D.Phil., at the Department of Nutritional Sciences, University of California, Berkeley, CA 94720. Nutrition Reviews, Vol. 50, No. 7
medium but, rather, are transported and react while bound to specific proteins. These low-molecularweight proteins belong to a family of lipid-binding proteins and include fatty acid- and steroid-binding proteins that function in transport and metabolism of their ligands.' For example, the reduction of retinal (formed by cleavage of p-carotene) occurs with retinal bound to cellular retinol-binding protein (type 11) (CRBPII);2 the esterification of retinol in liver takes place with retinol bound to CRBP;3 in the eye, 11-cis-retinal is reduced to 1 1-cis-retinol while bound to cellular retinal-binding p r ~ t e i nRe.~ cently, Boerman and Napoli' discovered a third function for CRBP: the regulation of a metabolic reaction, namely, the hydrolysis of retinyl esters to retinol. Retinol is released into the circulation in combination with serum retinol-binding protein (RBP) and transthyretin. This release is tightly regulated, depending on the availability of free (unesterified) ret197
July 7 992: 195-206
How Does Fish Oil Lower Plasma Triglycerides? Studies using dietary fish oil, safflower oil, and palm oil suggest that the decrease in plasma triacylglycerol due to dietary fish oils is somehow related to a decrease in the capacity of the liver to hydrolyze phosphatidate, which then affects microsomal synthesis of triacylglycerol from diacylglycerol.
Marine oils, rich in n - 3 fatty acids, appear to be more effective in lowering the concentration of plasma triacylglycerols than vegetable oils, generally rich in n - 6 fatty acids. The mechanisms responsible for the hypotriglyceridemic effects of fish oils are currently under intense investigation. Among the mechanisms suggested as possibly responsible for the effect of fish oils on plasma triacylglycerol levels are inhibition of the synthesis and secretion of triacylglycerol-containing lipoproteins, inhibition of fatty acid synthesis, enhancement of fatty acid oxidation, decreased activity of enzymes responsible for esterification of fatty acids, and changes in the ratios of the fatty acid esters formed. A recent study assessed the relative contribution of some of these mechanisms to the total hypotriglyceridemia produced by fish oils by comparing the effects of three different dietary oils (marine, safflower, and palm) on fatty acid oxidation and glycerolipid synthesis in rat liver.' The investigators simultaneously studied beta- and peroxisomal oxidation and key enzymatic activities in triacylglycerol (diacylglycerol acyltransferase and phosphatid a t e p h o s p h o h y d r o l a s e ) a n d phospholipid (cytidylyltransferase) biosynthesis. Young male rats were maintained for either ten days or one month on diets that differed only in the type of fat (fish, safflower, or palm oil). Food intake was allowed until 12 hours before the rats were killed. No significant differences were observed among groups in food intake, body weight gain, liver weight, or liver protein content. The plasma triacylglycerol levels of the fish oil group were 50% lower than those of the palm oil group, whereas This review was prepared by John G. Coniglio, Ph.D., at the Department of Biochemistry, Vanderbilt University, Nashville, TN 37232.
Nutrition Reviews, Vol. 50, No. 7
those of the safflower oil group were between the other two. The hepatic triacylglycerol was not affected by dietary treatment. Palmitic and oleic acids were the predominant fatty acids in plasma and liver of palm oil-fed rats, whereas linoleic acid was predominant in the safflower oil-fed group. Docosahexaenoic acid was present in small amounts in plasma and liver of safflower oil- and palm oil-fed rats, but both docosahexaenoic and eicosapentaenoic acids were abundant in fish oil-fed animals. In this group, the ratio of eicosapentaenoic to docosahexaenoic acid in plasma but not in liver reflected the dietary ratio. Beta- and peroxisomal oxidation were assessed in freshly prepared whole-liver homogenates incubated with l-'4C-palmitoyl CoA in the presence and absence of potassium cyanide. In this assay, mitochondrial beta-oxidation activity was observed as cyanide-sensitive activity (using conditions optimized for mitochondrial oxidation), and peroxisoma1 oxidation was due to cyanide-insensitive activity (using conditions optimized for peroxisomal activity). In both ten- and 30-day studies, no significant differences were observed among the three groups in mitochondrial beta-oxidation. In the tenday study, peroxisomal fatty acid oxidation in the fish-oil group was about 45% higher than in the other two groups, but the proportion of total oxidation that was peroxisomal was unchanged. However, peroxisomal oxidation activity in the fish-oil group studied at 30 days was 25-50% lower than that observed after ten days, and the rates were not different among the three experimental groups. In previous studies2 it has been suggested that diversion of polyunsaturated fatty acids (particularly eicosapentaenoic and docosahexaenoic) from pathways of esterification was a major cause of the fish oil-induced lowering of triacylglycerol levels. A major alternative pathway through fatty acid oxidation had been demonstrated in perfused liver^.^ In the latter studies, hepatic release of VLDL (very low density lipoproteins)-triacylglycerols was depressed by fish but not by safflower oil, and ketogenesis was increased in livers of fish oil-fed rats, compared to those of controls and safflower oil-fed rats, which were similar. Liver peroxisomes oxidized polyunsaturated fatty acids such as arachidonic and eicosapentaenoic acids, but mitochondria
195
appeared to be the major site of oxidation of these c o m p o ~ n d sAlthough .~ Halminski et al.' observed a higher rate of peroxisomal oxidation in fish oil-fed rats vs. those fed safflower or palm oil, the proportion of total oxidation by peroxisomes was similar in the three groups. From these data, the investigators inferred that fish oil had no specific effect on peroxisomes but, rather, increased both mitochondrial and peroxisomal oxidation. These results should be cautiously interpreted, since the fatty acid substrate used had a shorter chain length than fatty acids normally considered as natural substrates for peroxisomal oxidation and that the rates among the three groups of rats were not different at the end of the 30-day feeding period. Diacylglycerol acyltransferase was measured in whole homogenates and all subcellular fractions using ''C-myristoyl CoA and diolein. Whole homogenates stored at -70°C for 12 months suffered no substantial loss of activity, according to the authors. The activity of this enzyme was not affected by the type of fat in the diet, and in the rats fed 30 days the activity was twofold greater than that observed after ten days of feeding. The lack of an effect of fish oil on enzymatic activity indicated that this enzyme was not involved in the effect of the oil on synthesis of triacylglycerols. However, the aut h o r cautioned ~~ that inhibition of the activity of this enzyme by eicosapentaenoic acid had been reported in studies using cultured rat hepatocytes. Such inhibition could result in decreased synthesis of triacylglycerols due to reduced activity of this enzyme. Many differences in experimental conditions between the two studies preclude specific comparisons. Phosphatidate hydrolysis, also measured in whole homogenates and all subcellular fractions, was determined by release of inorganic phosphate from dimyristoyl phosphatidylcholine. The fish oilfed rats had the lowest activity of the three groups (19% lower than the safflower oil-fed group, which in turn had an activity 22% lower than the palm oil-fed rats). Pooled data from all three groups significantly and positively correlated with plasma concentrations of triacylglycerols. (See Figure 1.) This was interpreted to indicate that the rate of hepatic phosphatidate hydrolysis was a good predictor of plasma triacylglycerol concentration. A decrease of 10-35% in the capacity to hydrolyze phosphatidate occurred in all groups in the 30-day feeding study. However, the activity of the rats fed fish oil continued to be lower than that of the other two groups, which at this time had similar activities. Because the rate of phosphatidate hydrolysis is considered a key intermediate process in the synthesis of glycerolipids, this enzymatic step might be involved with the plasma triacylglycerol-lowering 196
z.ol 0
E E
1.5
Y
In
-m 0-
!
0.0 300
1
400
500
600
700
1
I
800
900
Phosphatidate Hydrolysis nrnol/(rnin * g liver)
Figure 1. Correlation of plasma triglyceride concentration with liver phosphatidate hydrolysis in rats fed different dietary lipids for ten days. Points correspond to individual animals (n = 9/dietary group): closed squares, open squares, and closed triangles represent fish oil-, safflower oil-, and palm oil-fed animals, respectively; r = 0.68, p < 0.001. Reprinted with permission.'
effect of marine oils. Decreased phosphatidate hydrolase activity had also been reported in livers of rats perfused with albumin-bound docosahexaenoic and eicosapentaenoic acids6 and in liver microsomes of rats fed a commercial diet supplemented with fish oiL7 Safflower oil, as well as fish oil, affected the activity of phosphatidate hydrolase in the studies of Halminski et al. ; I therefore, the authors suggested that the effect might depend on the degree of unsaturation of the fat. Phosphoryl choline cytidylyltransferase activity was measured by the conversion of phosphoryl[methyl ''C]-choline to radiolabeled cytidine diphosphocholine by liver homogenates and subcellular fractions. The activity was significantly but only slightly lower in the rats fed fish oil for ten days than that in the other two groups (which had similar values). Feeding the diets for 30 days resulted in lower activities in all groups, but the values for the fish oil-fed rats were still the lowest of the three groups. The results observed in these studies with fish oil agree with previous findings of decreased synthesis of phosphatidyl choline in hepatocytes preincubated with eicosapentaenoic acid.* Because safflower oil and palm oil had similar effects on the activity of this enzyme, the action of fish oil was apparently due specifically to the n - 3 acids and not merely to unsaturation. Although the effect was smaller than that on phosphatidate hydrolase activity, it may be important in regulating VLDL synthesis and secretion, in which process phosphatidyl choline is an essential component. Nutrition Reviews, Vol. 50,No. 7
Activities of enzymes that are important in the processes of VLDL-triacylglycerol synthesis and secretion could possibly be affected by translocation from cytosol to the site of activity in membranes. However, in these studies, no evidence supporting this possibility was obtained. As a result of their studies, Halminski et al.' suggested that the decrease in plasma triacylglycer01s due to dietary fish oils is somehow related to a decrease in the capacity of the liver to hydrolyze phosphatidate, which would then affect microsomal triacylglycerol synthesis from diacylglycerol. However, the precise mechanisms by which fish oils mediate these effects are still unknown, and it appears that both the degree of unsaturation and the precise position of the double bonds may play a role in these effects. Other proposed mechanisms cannot be excluded. The continued exploration of the effect of individual fatty acids on key enzymes involved in the complex process of synthesis and secretion of all components of the lipoproteintriacylglycerol particle remains worthwhile. Halminski MA, Marsh JB, Harrison EH. Differential effects of fish oil, safflower oil and palm oil on fatty acid oxidation and glycerolipid synthesis in rat liver. J Nutr 1991;121:1554-61 Wong S, Reardon M, Nestel P. Reduced triglyceride formation from long-chain polyenoic fatty acids in rat hepatocytes. Metabolism 1985;34:900-5 Wong SH, Nestel PJ, Trimble RP, Storer GB, lllman
RJ, Topping DL. Adaptive effects of dietary fish and safflower oil on lipid and lipoprotein metabolism in perfused rat liver. Biochim Biophys Acta 1984;792: 1039
4. Christensen E, Hagve T, Christophersen BO. Mitochondrial and peroxisomal oxidation of arachidonic and eicosapentaenoic acid studied in isolated liver cells. Biochim Biophys Acta 1986;879: 313 2 1 5. Rustan AC, Nossen JO, Christiansen EN, Drevon CA. Eicosapentaenoic acid reduces hepatic synthesis and secretion of triacylglycerol by decreasing the activity of acylcoenzyme A:l,2-diacylglycerol acyltransferase. J Lipid Res 1988;29:1417-26 6. Wong S, Marsh JB. Inhibition of apolipoprotein synthesis and phosphatidate phosphohydrolase activity by eicosapentaenoic and docosahexaenoic acids in the perfused rat liver. Metabolism 1988;37: 1177-81 7. Marsh JB, Topping DL, Nestel PJ. Comparative effects of dietary fish oil and carbohydrate on plasma lipids and hepatic activities of phosphatidate phosphohydrolase, diacylglycerol acyltransferase and neutral lipase activities in the rat. Biochim Biophys Acta 1987;922:23%43 8. Strum-Odin R, Adkins-Finke 8 , Blake WL, Phinney SD, Clarke SD. Modification of fatty acid composition of membrane phospholipids in hepatocyte monolayer with n-3, n-6 and n-9 fatty acids and its relationship to triacylglycerol production. Biochim Biophys Acta 1987;921:378-91
Cellular Retinol-Binding Protein Functions in the Regulation of Retinoid Metabolism A microsomal retinyl ester hydrolase from rat tissues is activated by apo-cellular retinol-binding protein, which thus regulates the availability of retinol either for binding to serum retinolbinding protein or for metabolic oxidation. Microsomal retinal synthesis was found to occur in rat tissues with holo-cellular retinol-binding protein as substrate. Retinal could be further oxidized to retinoic acid by a cytosolic fraction.
A principle that emerged some years ago and states that retinoids do not exist in tissues in the free state but are always protein-bound has now been extended: retinoids reacting as intermediates in metabolic pathways do not diffuse through the aqueous This review was prepared by George Wolf, D.Phil., at the Department of Nutritional Sciences, University of California, Berkeley, CA 94720. Nutrition Reviews, Vol. 50, No. 7
medium but, rather, are transported and react while bound to specific proteins. These low-molecularweight proteins belong to a family of lipid-binding proteins and include fatty acid- and steroid-binding proteins that function in transport and metabolism of their ligands.' For example, the reduction of retinal (formed by cleavage of p-carotene) occurs with retinal bound to cellular retinol-binding protein (type 11) (CRBPII);2 the esterification of retinol in liver takes place with retinol bound to CRBP;3 in the eye, 11-cis-retinal is reduced to 1 1-cis-retinol while bound to cellular retinal-binding p r ~ t e i nRe.~ cently, Boerman and Napoli' discovered a third function for CRBP: the regulation of a metabolic reaction, namely, the hydrolysis of retinyl esters to retinol. Retinol is released into the circulation in combination with serum retinol-binding protein (RBP) and transthyretin. This release is tightly regulated, depending on the availability of free (unesterified) ret197
