Unusual properties of retinyl palmitate hydrolase activity in rat liver1

These studies report the hydrolysis of retinyl palmitate with liver homogenates and homogenate fractions from retinol-depleted rats. The studies utilized an effective in vitro assay for retinyl palmitate hydrolase (RPH) activity, in which p g amounts of retinyl palmitate were employed as substrate, followed by the chromatographic separation and fluorescence assay of free and esterified retinol. RPH activity was maximal near pH 8 in Tris-maleate buffers, and required a bile salt for stimulation. Both cholate and taurocholate stimulated the reaction, whereas a number of other detergents tested were ineffective. The enzymatic activity showed an unusual subcellular distribution, with about 40% of total RPH activity recovered in the washed “nuclear” fraction (1,500g pellet) and about 30-35% in the 105,000 g supernatant. This unusual distribution was not observed for marker constituents for plasma membranes, nuclei, mitochondria, lysosomes, Golgi apparatus, or endoplasmic reticulum. Despite its enrichment in the “nuclear” fraction, RPH activity was not enriched in purified preparations of nuclei or plasma membranes. Thus, RPH activity was not localized in any single, characterized subcellular structure. Another striking feature of the hepatic RPH activity was its extreme variability from rat to rat as assayed in vitro. Both the unusual subcellular distribution and the marked variability in activity were not observed for a variety of other hepatic ester hydrolase activities examined. Of ten lipid and nonlipid esters tested as substrates, only the hydrolytic activities against cholesteryl oleate and phytyl oleate correlated with, and partly resembled, RPH activity in these respects. The results suggest that the observed RPH activity is relatively specific for the hydrolysis of retinyl palmitate, and may therefore be significantly involved in hepatic retinyl ester metabolism.-Harrison, E. H., J. E. Smith, and D. S. Goodman. Unusual properties of retinyl palmitate hydrolase activity in rat liver.]. Lipid Res. 1979. 20: 760-771. Supplementary key words binding protein * subcellular distribution * vitamin A deficiency Vitamin A * retinyl esters * retinolmin A is then stored in lipid droplets in the liver, in the form of long-chain retinyl esters, particularly as retinyl palmitate (1, 4). Vitamin A is mobilized from the liver in the form of the lipid alcohol retinol, bound to its specific transport protein, plasma retinol-binding protein (RBP) ( 5 , 6). Accordingly, prior to the mobilization of vitamin A from the liver, the stored retinyl esters must be hydrolyzed to form retinol. The retinol so produced forms a complex with a molecule of RBP, and the retinol-RBP complex is then secreted from the liver. RBP is a relatively small protein of approximately 20,000 daltons, with one binding site for one molecule of retinol. In plasma, RBP circulates mainly as a 1: 1 molar protein-protein complex with plasma prealbumin. Both the mobilization of vitamin A from the liver and the secretion of RBP are highly regulated processes, and are particularly regulated by the vitamin A status of the animal (7-9). Since RBP is mainly secreted from the liver as the holoprotein, it seems possible that the rates of hydrolysis of retinyl esters and of synthesis and secretion of RBP may be coordinated with each other in some way within the liver cell. From these considerations, it is clear that the enzymatic hydrolysis of retinyl esters in liver represents an important process in the overall metabolism of vitamin A in the body. Nevertheless, little information is available about this reaction. In earlier studies, in fact, a number of investigators were unable to demonstrate the in vitro hydrolysis of retinyl palmitate with liver homogenate preparations (IO, 11). In 1966, Mahadevan, Ayyoub, and Roels (12) reported that The liver is the major storage site for vitamin A in the body. Newly absorbed vitamin A reaches the liver mainly in the form of retinyl esters in association with chylomicron remnants (1, 2). Evidence exists that the hepatic uptake of vitamin A involves hydrolysis of the chylomicron retinyl esters, followed by intrahepatic reesterification of the free retinol (1,3). The vitaAbbreviations: RBP, retinol-binding protein; RPH, retinyl palmitate hydrolase. A partial report of this work was presented at the annual meeting of the Federation of American Societies for Experimental Biology, Chicago, Ill. in April, 1977 (43). The work described here forms part of a dissertation that was submitted to the Graduate School of Arts and Sciences of Columbia University in partial fulfillment of the requirements for the Ph.D. degree. To whom reprint requests and inquiries should be addressed. 760 Journal of Lipid Research Volume 20, 1979 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 1, 2013 w w w .j.org D ow nladed fom retinyl palmitate hydrolyzing activity was found in the “nuclear” and “mitochondrial-lysosome-rich” fractions of rat liver homogenates. The activity required the addition of a bile salt and was partially characterized with extracts of acetone powders of rat liver. The enzyme preparations hydrolyzed a variety of long-chain retinyl esters, with the greatest relative activity being seen with retinyl palmitate. No further information about this reaction has, however, been reported by these or other investigators. We now report the results of studies on the characteristics and subcellular localization of retinyl palmitate hydrolase activity in rat liver. These studies represent the first phase of a project that aims to explore in detail the enzymatic hydrolysis of long-chain retinyl esters by the liver. EXPERIMENTAL PROCEDURE Source of enzyme activity Homogenates and subcellular fractions from the livers of retinol-depleted rats were used throughout this study. Male, weanling rats of the Holtzmann strain were depleted of vitamin A by feeding them a vitamin A-deficient diet for 3-4 weeks; thereafter the rats were fed a diet supplemented with retinoic acid in order to maintain good growth and general health (7,s) . Retinol-depleted animals were used in order to avoid the problem posed by large and variable amounts of endogenous substrate (retinyl esters) present in the livers of normal animals. Preparation of subcellular fractions Homogenates of rat liver were prepared and fractionated by differential centrifugation according to the method of de Duve et al. (13) as recently modified by Amar-Costesec et al. (14). In this procedure as used, a crude pellet was prepared by low-speed centrifugation and washed (i.e., resuspended and recentrifuged) two times. The resulting pellet was resuspended to yield the “nuclear” or N fraction. The postnuclear supernatants were pooled to yield the cytoplasmic extract or E fraction. The cytoplasmic extract was then further fractionated to obtain the mitochondrial-lysosoma1 fraction (M + L) and the microsomal fraction (P). Each of these particulate fractions was washed two times before final resuspension. The final high-speed supernatant is referred to as the S fraction. Fractions enriched in plasma membranes were prepared from dilute, hypotonic liver homogenates by the method of Song et al. (15). Plasma membranerich fractions were also prepared directly from the N fraction by the method of Touster et al. (16). Purified preparations of cell nuclei were isolated directly from whole homogenates of rat liver by the method of Blobel and Potter (17). Assay for retinyl palmitate hydrolase (RPH) activity The standard RPH assay was conducted as follows. All procedures were conducted under subdued lighting. Buffer (50 mM Tris-maleate, pH 8), a-tocopherol (Sigma, 50 pg, added as an antioxidant, in 5 p1 of ethanol), sodium cholate (30 pmol), and retinyl palmitate (a gift of Hoffmann-LaRoche Inc., Nutley, NJ, 10 pg added in 100 PI of ethanol) were incubated with a source of enzyme in a final volume of 2 ml. Incubations were conducted under nitrogen at 37°C. Reactions were terminated by adding 2 ml of ethanol, and the resultant mixtures were extracted twice with 6.5 ml of hexane. The combined hexane extracts were concentrated and applied to a column of 0.5 g of alumina (Woelm, Eshwege, Germany) deactivated with 10% water. Retinyl palmitate was separated from the free retinol formed during the reaction by successive elution with 10 ml of 1% diethyl ether in hexane (to yield retinyl palmitate) and 10 ml of 20% ether in hexane (to yield retinol). Preliminary experiments confirmed that this procedure resulted in the complete and consistent separation of retinol from retinyl palmitate, with excellent recovery of both components. Each fraction was assayed fluorometrically for vitamin A by a modification (7) of the method of Thompson et al. (18). Enzyme activity was calculated as the percent hydrolysis of the recovered substrate. Under these standard assay conditions, 10% hydrolysis per hr corresponds to an activity of 2 nmol per hr. Total recovery of vitamin A substrate (esterified plus free retinol) was routinely 90% or more. Controls without enzyme were run with each experiment, and the values observed for activity (usually 2-4% hydrolysis) were subtracted from those of the test samples in the final calculations of enzyme activity. When various additions were made to the reaction mixtures, pilot experiments were conducted to assure that such additions did not interfere with the extraction or fluorometric assay of the vitamin A compounds. Assays for marker constituents The following constituents were assayed as markers for the indicated subcellular components: DNA (nuclei), cytochrome oxidase (mitochondria), acid phosphatase (lysosomes), 5’-nucleotidase (plasma membrane), galactosyltransferase (Golgi apparatus), glucose-6-phosphatase (endoplasmic reticulum), and RNA (rough endoplasmic reticulum). Glucose-6-phosphatase (1 9), 5’-nucleotidase ( 19), and cytochrome oxidase (20) were assayed according Harrison, Smith, and Goodman Hepatic retinyl palmitate hydrolase 761 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 1, 2013 w w w .j.org D ow nladed fom to the indicated published procedures. Acid phosphatase was assayed essentially as described by Glaumann and Dallner (21) except that Triton X-100 was not included in the reaction mixture. In order to release the full activity of this latent enzyme, all fractions were frozen and thawed three times prior to assay. For the determination of galactosyltransferase, the transfer of galactose from UDP-['4C]galactose to acid-hydrolyzed bovine submaxillary mucin was assayed in a manner similar to that described by Grimes (22). DNA and RNA were extracted from homogenates and fractions with perchloric acid (23) and estimated spectrophotometrically (24, 25). Assays for the hydrolysis of other esters The hydrolysis of a number of lipid and nonlipid monoesters was assessed in reaction mixtures nearly identical to that used in the assay of RPH. [ l-14C]Palmitic acid (56 pCi/pmol), [ l-'4C]oleic acid (59.7 pCi/pmol), and cholesteryl [ l-14C]oleate (2 1.5 pCilpmol) were obtained from Amersham-Searle (Arlington Heights, IL). Ethyl [ l-14C]palmitate and ethyl [ l-14C]oleate were prepared by'heating the appropriate l-14C-labeled fatty acid in acidic ethanol (2% H2S04 in absolute ethanol) at 70°C under nitrogen for 3 hr. Arachidonyl [ l-14C]palmitate and phytyl [ 114C]oleate were prepared by reacting the appropriate labeled fatty acid anhydride with the appropriate alcohol (Sigma), essentially as described by Lentz, Barenholz, and Thompson (26) for the synthesis of cholesteryl esters. The labeled fatty acid anhydrides were prepared by a microscale adaptation of the procedure of Selinger and Lapidot (27). The labeled product esters (arachidonyl palmitate and phytyl oleate) were purified by chromatography on small columns of alumina deactivated with 10% water. The purity of the several labeled esters was confirmed in each instance by chromatographic analysis. The enzymatic hydrolysis of cholesteryl oleate, ethyl palmitate, ethyl oleate, phytyl oleate, and arachidonyl palmitate was assessed by incubation of the labeled substrate ester with enzyme and extraction of the released free fatty acid into an alkaline aqueous phase (28). The incubations, carried out for 20-30 min at 37"C, contained 50 mM Tris-maleate, pH 8, 15 mM cholate, 0.002 pmol of substrate (added in 10 p1 of' ethanol), and an appropriately diluted homogenate fraction in a final volume of 0.2 ml. After incubation and extraction, a portion (1 ml) of the alkaline aqueous phase containing the released free fatty acid was added to 10 ml of ScintiVerse (Fisher Scientific Co.), and radioactivity was determined in a Packard liquid scintillation counter. Assays for the hydrolysis of p-nitrophenyl esters and 4-methylumbelliferyl esters were carried out at 37°C for 15 min in 2-ml reaction mixtures containing 50 mM Tris-maleate, pH 8, and 15 mM cholate. Substrates were added in 100 pl of ethanol (for acetate esters) or 100 p1 of acetone-ethanol 2:3 (v/v) (for palmitate esters). The substrate solutions of p-nitrophenyl esters were prepared at a concentration of 10 pmol/ml while those for the 4-methylumbelliferyl esters (Koch-Light Laboratories, Ltd., Colnbrook, Bucks, England) were used at 0.20 pmollml. Reactions with the p-nitrophenyl esters were terminated by the addition of 2 ml of ice-cold ethanol, and the hydrolysis of these esters was assayed by spectrophotometric determination of released p-nitrophenol at 400 nm (29). 4-Methylumbelliferyl esterase activity was assayed by fluorometric determination of released free 4-methylumbelliferone. For these assays, 0. l-ml aliquots of the reaction mixtures were each pipetted into 5 ml of ethanol, and the fluorescence of the resulting solutions was determined in an Aminco-Bowman spectrophotofluorometer at an excitation of 380 nm and an emission of 450 nm (30, 31). The hydrolysis of retinyl acetate was assayed exactly as described for retinyl palmitate except that the substrate was retinyl acetate (Eastman Organic Chemicals) in ethanol (5 pg retinol equivalentdml). Pilot experiments were conducted to develop suitable conditions for the conduct of each enzyme assay (i.e., for the assay of the enzymatic hydrolysis of each of the esters studied). All assays were carried out under conditions where product formation was proportional to the time of incubation and to the quantity of enzyme-containing protein in the reaction mixture. In determining the subcellular distributions of the activities, each fraction was assayed at three different protein concentrations.