Fatty acylation of the rat asialoglycoprotein receptor. The three subunits from active receptors contain covalently bound palmitate and stearate.

Rat hepatic asialoglycoprotein receptors (ASGP-Rs) are hetero-oligomers composed of three homologous glycoprotein subunits, designated rat hepatic lectins (RHL) 1, 2, and 3. ASGP-Rs mediate the endocytosis and degradation of circulating glycoconjugates containing terminal N-acetylgalactosamine or galactose, including desialylated plasma glycoproteins. We have shown in permeable rat hepatocytes that the ligand binding activity of one subpopulation of receptors (designated State 2 ASGP-Rs) can be decreased or increased, respectively, by ATP and palmitoyl-CoA (Weigel, P. H., and Oka, J. A.(1993) J. Biol. Chem. 268, 27186-27190). We proposed that a reversible and cyclic acylation/deacylation process may regulate ASGP-R activity during endocytosis, receptor-ligand dissociation, and receptor recycling. In the accompanying paper (Zeng, F-Y., and Weigel, P. H.(1995) J. Biol. Chem. 270, 21388-21395), we show that the ligand binding activity of affinity-purified State 2 ASGP-Rs is decreased by treatment with hydroxylamine under mild conditions consistent with these ASGP-Rs being fatty acylated in vivo. In this study, we used a chemical method to determine the presence of covalently-bound fatty acids in individual ASGP-R subunits. The affinity-purified ASGP-R preparations were separated by SDS-polyacrylamide gel electrophoresis under nonreducing conditions, and the gel slices containing individual RHL subunits were treated with alkali to release covalently bound fatty acids, which were subsequently analyzed by gas chromatography and confirmed by gas chromatography-mass spectrometry. Both stearic and palmitic acids were detected in all three receptor subunits. Pretreatment of ASGP-Rs with hydroxylamine before SDS-polyacrylamide gel electrophoresis reduced the content of both fatty acids by 66-80%, indicating that most of these fatty acids are attached to cysteine residues via thioester linkages. Furthermore, when freshly isolated hepatocytes were cultured in the presence of [3H]palmitate, all three RHL subunits in affinity-purified ASGP-Rs were metabolically labeled. We conclude that RHL1, RHL2, and RHL3 are modified by fatty acylation in intact cells.

Covalent binding of long chain saturated fatty acids occurs with a wide variety of membrane proteins post-translationally and significantly influences protein localization and/or function (1). Two of the most common modifications involve acylation with myristate and palmitate. Myristate is usually attached to an N-terminal glycine via an amide bond in a relatively stable linkage. In contrast, palmitate is attached either to cysteine residues via a thioester bond or to serine residues via an ester bond; both linkages, particularly the thioester, are hydrolyzed by alkali treatment (2). Palmitoylation of many proteins has been shown to be a dynamic process (3,4).
The hepatic asialoglycoprotein receptor (ASGP-R) 1 mediates the endocytosis of desialylated glycoproteins containing terminal galactose or N-acetylgalactosamine residues (5,6). The functional rat ASGP-R is a hetero-oligomer composed of three subunits, RHL1, RHL2, and RHL3, with molecular masses of 41,500, 49,000, and 54,000 Da, respectively (7). The amino acid sequences of all three subunits are closely related, and they are the products of two different genes (8). RHL1 is the major subunit of the ASGP-R, while RHL2 and RHL3, encoded by the same gene, are minor subunits and differ only in the type and amount of post-translational carbohydrate modification (9).
We and others have previously demonstrated that the ASGP ligands are endocytosed and intracellularly processed by two functionally different receptor populations via two distinct pathways (6,10). We have designated these two receptor populations as the State 1 ASGP-Rs and the State 2 ASGP-Rs. In intact cells, the State 2 ASGP-Rs undergo a transient inactivation/reactivation cycle during receptor recycling (11). This same cycle has been successfully reconstituted in permeable rat hepatocytes (12)(13)(14). In permeablized cells the State 2 ASGP-Rs are inactivated by the addition of ATP in a time-and temperature-dependent manner (12); these ATP-inactivated receptors can then be quantitatively reactivated by the addition of palmitoyl-CoA (13,14).
In the accompanying paper (15), we also demonstrate that the activity of one population of affinity-purified ASGP-Rs, the State 2 receptors, is selectively inactivated by treatment with hydroxylamine, a chemical frequently used to release thioester-linked fatty acids from proteins. These above results suggest that one or all ASGP-R subunits are modified by fatty acylation in vivo and that a reversible acylation/deacylation process may be involved in regulating the ligand-binding activity of ASGP-Rs as they function during receptor mediated endocytosis.
In this study, we have used gas chromatography-mass spectrometry to examine directly whether total (State 1 plus State 2) ASGP-Rs contain covalently-bound fatty acids. The results demonstrate that all three ASGP-R subunits contain covalently-linked palmitate and stearate. Treatment of each RHL subunit with hydroxylamine under mild conditions released both fatty acids in relatively large amounts, indicating that most fatty acids are attached via thioester linkages.
Preparation of Active ASGP-Rs-Isolated hepatocytes from male Sprague-Dawley rats (Harlan Breeding Laboratories, Houston, TX, and SasCo, Oklahoma City, OK) were prepared using a modified (17) collagenase perfusion procedure (18). The cells were suspended in medium 1/BSA, and incubated at 37°C for 1 h to increase and stabilize the total cell ASGP-R receptor activity (19). The active ASGP-Rs were purified from the preincubated hepatocytes by using asialo-orosomucoid-Sepharose as affinity ligand as described in the accompanying paper (15).
Electrophoresis-SDS-PAGE was performed as described by Laemmli (20). The samples were mixed with an equal volume of a 2-fold concentrated sample buffer containing 125 mM Tris-HCl, pH 6.8, 2% SDS (w/v), 20% (v/v) glycerol, and 0.01% (w/v) bromphenol blue, and incubated at room temperature for 20 min. Samples were not boiled in order to minimize the possibility of ester hydroylsis. The samples were then resolved by electrophoresis on 10% (w/v) acrylamide slab gels (10 ϫ 8 cm, Bio-Rad mini-gel apparatus) in the presence of 0.1% SDS. After electrophoresis, gels were stained with Coomassie Blue R-250 (1 mg/ml in 5% acetic acid, 25% methanol) for 5 min, and then destained with 5% acetic acid, 25% methanol until protein bands were clearly visible. To determine if hydroxylamine treatment releases covalently-bound fatty acids, the purified ASGP-R preparation was incubated with 1 M hydroxylamine or 1 M Tris (as a control) at pH 7.4 on ice for 2 h before the addition of sample buffer and SDS-PAGE.
Release of Covalently-bound Fatty Acids by Chemical Treatments-All glassware was new and thoroughly rinsed with chloroform/methanol (1:1, v/v) prior to use. HPLC-grade water was used throughout the procedure. Affinity-purified ASGP-R (400 g) was separated by SDS-PAGE, and after staining and destaining as described above, RHL1, RHL2, and RHL3 bands were cut from the gel, minced into small pieces, and placed into 15 ϫ 100-mm screw-cap vials. Controls with same size of blank gel or gel pieces containing 100 g of BSA were treated identically and placed in separate vials. The minced gel pieces were sequentially washed with 10 ml of water twice for 4 h each, 10 ml of 90% methanol, 10% water three times over 12 h, and 10 ml of 50% methanol in water twice for 8 h. The final wash was removed and the gel pieces were then treated with 2.5 ml of methanol/chloroform/water (2.5:1.25:1, v/v/v) twice each for 10 min. These extracts were pooled, transferred into a new vial, and water (1.0 ml) and chloroform (1.5 ml) were added. The contents were mixed vigorously and the biphasic solution was then allowed to separate. The chloroform phase was transferred to a new vial and dried under nitrogen gas.
To release ester-linked fatty acids from the RHL proteins, the minced gel pieces were dried under nitrogen after the final wash and 0.7 ml of 1.5 N NaOH was added to each vial. After shaking at 30°C for 3 h, the vials were cooled on ice and then neutralized with 0.3 ml of 6 N HCl. To each vial, 4.0 ml of chloroform/methanol (1:2, v/v) was added, and the vials were shaken at room temperature for 10 min. The chloroform/methanol extract was transferred to another new vial, each gel was washed with 1.5 ml of chloroform, and the wash was combined with each extraction. Water (1 ml) was added to each vial to obtain a two-phase solution, and the chloroform phase was transferred to a clean new vial and dried under nitrogen. The gels were also dried under nitrogen.
To hydrolyze amide-linked fatty acids, the dried, minced gel pieces were incubated with 1.0 ml of 6 N HCl at 110°C for 6 h in Teflon-lined screw-capped vials. After cooling to room temperature, 4.0 ml of meth-anol/chloroform (1:1) was added and the vials were shaken for 10 min. The extract was transferred to a new vial, the gel pieces were washed with 1.5 ml of chloroform, and this wash was added to the extract solution. Two phases were formed by addition of 1 ml of water and after vigorous mixing followed by phase separation the chloroform layer was transferred to a clean vial and dried under nitrogen.
Derivatization of Fatty Acids by BF 3 -Methanol-Fatty acids were derivatized to their corresponding methyl esters. BF 3 -methanol (0.5 ml) was added to each vial containing dried extracts, and the vial was flushed with nitrogen, capped, and incubated at 50°C for 30 min. The vials were cooled to room temperature and 0.5 ml of water and 2 ml of hexane were added. After mixing and a brief centrifugation, the upper hexane layer was transferred to a clean conical bottom screw-top vial. The aqueous phase was extracted again with 2.0 ml of hexane, which was removed, combined with the first upper phase, and dried under nitrogen.
Gas Chromatographic Analysis of Fatty Acid Methyl Esters-The dried residue was dissolved in 200 l of pesticide-grade hexane and the fatty acid methyl esters were analyzed using a Hewlett-Packard 5890 Gas Chromatograph with a DB 225 fused silica capillary column (J & W Scientific, Folsom, CA; 20 m long, 0.2 mm film thickness, under split mode) and flame ionization detector (21). The temperature of the injector port and detector was set at 200 and 300°C, respectively. Ultra high purity grade nitrogen, used as carrier gas at a flow rate 0.5 ml/min, was passed through molecular sieve (4 Å) and silica gel in order to remove the impurities of oxygen and moisture, respectively. The fatty acid methyl esters were analyzed using a thermal gradient from an initial column temperature of 150°C to a final temperature of 225°C with an increase of 10°C/min. The peak area of the fatty acid methyl esters was determined using the Hewlett-Packard 3396A integrator and the fatty acid methyl esters were quantitated by comparison to the response of known standards of fatty acid methyl esters.
Gas Chromatography-Mass Spectrometric Analysis of Fatty Acid Methyl Esters-To confirm the identity of the fatty acids, fatty acid methyl esters were also analyzed by GC-MS using a Varian 3400 gas chromatograph interfaced with a Fennigan Incos-50 mass spectrometer at the Analytical Chemistry Center, University of Texas Health Science Center, Houston, TX. The esters were separated on a DB 23 column (30 m long, 0.32-mm inner diameter) using a temperature gradient from 100 to 250°C with an increase of 10°C/min and then analyzed by electron impact (70 eV), as well as by chemical ionization (using methane as reagent gas) mass spectrometry.
Metabolic Labeling of Hepatocytes with [ 3 H]Palmitate-Freshly isolated rat hepatocytes were plated onto 60-mm dishes (ϳ2 ϫ 10 6 cells/ dish) and cultured at 37°C with 5% CO 2 in William's Medium E as described previously (17). The medium was supplemented with 10% fetal calf serum, 10 mM Hepes, pH 7.4, 2 mM L-glutamine, 50 milliunits/ml insulin, 50 g/ml gentamicin, 100 g/ml penicillin, 100 g/ml streptomycin, and 50 nM dexamethasone. For labeling experiments, cells were preincubated with medium without fetal calf serum at 37°C for 1 h. The labeling was then carried out in the medium containing 400 Ci of [ 3 H]palmitic acid and 10% fetal calf serum (dialyzed against phosphate-buffered saline). The active ASGP-Rs from radiolabeled cells were purified by affinity chromatogrphy using asialo-orosomucoid-Sepharose as described in Ref. 15, and separated by SDS-PAGE as described above.
Analysis of 3 H-Labeled Fatty Acids-The gel pieces containing 3 Hlabeled RHL1 were excised, and extensively washed with methanol/ water (1:1, v/v) and chloroform/methanol (1:1, v/v) sequentially until no radioactivity was detected in the wash. Gel peices were dried under nitrogen gas and incubated with 0.35 ml of 1.5 N NaOH at 30°C for 3 h. After cooling on ice, 0.15 ml of 6 N HCl was added, the released fatty acids were recovered by extracting twice with 2 ml of hexane. The hexane phases were pooled and dried under nitrogen gas. The samples were redissolved in 50 l of methanol. To identify radioactive fatty acids, 100 g each of unlabeled myristic acid, palmitic acid, and stearic acid were added as carrier and internal standards. Samples were analyzed by reverse-phase HPLC with a Vydac RP-300 C18 column (250 ϫ 4.6 mm, 5 m), eluted with acetonitrile/water (47:3, v/v) at a flow rate of 1 ml/min and monitored at 205 nm using a Shimadzo SPD-10A UV-Vis detector. The radioactivity in eluted fractions (1 ml) was determined using a Packard Tri-carb 2300TR liquid scintillation analyzer. The retention times of the radioactive peaks were compared with those of the standard fatty acids.
General-The protein content was measured by the method of Bradford (22) using BSA as standard. Western blotting was carried out as described elsewhere, using subunit-specific antibodies (23). For fluorography, the fixed gel was washed with water for 30 min, and then soaked in fluoro-HANCE solution at room temperature for 30 min, dried at 60 -70°C, and exposed to Kodak X-Omat film (Eastman Kodak Co.) at Ϫ70°C for 25 days.

RESULTS
Fatty acids can be covalently attached to proteins either to glycine residues via N-linked amide bonds, to cysteine residues via thioester bonds, or to serine residues via hydroxyester bonds (1,2). The thioester and hydroxyester linkages are labile to alkaline treatment. To determine if the ASGP-R is modified by fatty acylation, we used a direct chemical method to assess whether any of the ASGP-R subunits contain covalently bound fatty acids. Freshly purified ASGP-R, prepared from isolated rat hepatocytes by affinity chromatography, was examined immediately, since storage resulted in activity loss and potential deacylation as noted in the accompanying paper (15). To exclude the possibility that reducing agents such as ␤-mercaptoethanol or dithiothreitol may release some covalently-bound fatty acids, SDS-PAGE was carried out under nonreducing conditions and the samples were not boiled. Gel pieces containing the separated subunits RHL1, RHL2, and RHL3 were cut out, and subjected to extensive washing, extraction, and alkaline or acid hydrolysis as described under "Experimental Procedures." The released fatty acids were derivatized to their methyl esters and analyzed by GC-MS.
To assess the reliability of this method, we determined the optimal conditions for the washing steps, the alkali treatment to release ester-linked fatty acids, and the derivatization of fatty acids to their methyl esters. This was necessary in order to remove SDS and other components that could give contaminating GC peaks. In particular, we found that (i) extensive washing of the gel pieces with water and then methanol/water and (ii) performing the methanolysis with BF 3 -methanol at Ͻ60°C were both critical in order to reduce the background of observed GC peaks. A blank gel of the same size was used as negative control. Using our final conditions, this negative control gave small, but detectable peaks at positions corresponding to methyl palmitate and methyl stearate. The intensities of these peaks were very similar, if not identical, to those shown in the left panels of Fig. 1.
These background peaks were present even in gels on which no samples had been loaded and could not be eliminated by changing the experimental conditions. The intensities of these peaks were also significantly increased by increasing the gel size analyzed (data not shown). For this reason, we used the same size blank or subunit-containing gels (verified by weight) to minimize the effect of differing gel size on the results. To verify that SDS-PAGE and the washing procedure after SDS-PAGE could effectively remove noncovalently bound fatty acids, BSA which is known to contain many kinds of noncovalently bound fatty acids, was subjected to the same procedure. The result showed no detectable fatty acids in either the chloroform/methanol wash or the alkaline hydrolyzate in comparison with a blank gel (not shown), indicating that SDS-PAGE and the subsequent wash procedures (before alkali treatment) completely removed noncovalently bound fatty acids.
The analysis of alkaline hydrolysates of purified ASGP-R by GC-MS clearly demonstrated the presence of palmitate and stearate in each of the three RHL subunits (Figs. 1 and 2). One GC peak in these samples was identified as the methyl ester of palmitate by two criteria: First, its retention time (11.65 min) on gas chromatography was identical to that of standard methyl palmitate and this peak co-chromatographed with the standard methyl palmitate. Second, mass spectrometric analysis of this GC peak (retention time ϭ 11.65 min) gave a molecular ion (m/z ϭ 270), which is the expected mass of methyl palmitate. Furthermore, additional ion mass peaks at 74 (McLafferty ion), 87, and 143, are the expected major molecular fragment ions from the methyl esters of saturated fatty acids. The mass spectrum of the 11.65-min peak was identical to that of the standard methyl palmitate (Fig. 2A). The methyl ester of stearate was identified by the same criteria. All samples contained a GC peak with a retention time of 13.86 min, identical to that of standard methyl stearate, and GC-MS analysis revealed a molecular ion (m/z ϭ 298) and other major ions characteristic (21,24) of the methyl ester of stearate (Fig. 2B). In addition to methyl palmitate and methyl stearate, small amounts of the methyl esters of myristic, oleic, and linoleic acids were also detected by GC in the alkaline hydrolyzates of all three subunits (Fig. 1). Palmitate and stearate methyl esters were not detected in the chloroform/methanol wash of the gels containing ASGP-R subunits (Fig. 1, left panel). These esters were only found after processing the alkaline hydrolysates. After alkali treatment, the incubation of the gels containing receptor subunits with strong acid should release amide-linked fatty acids. The analysis of the acid hydrolysates of RHL1 and RHL2 by GC showed no detectable methyl esters of any fatty acid above background (data not shown), indicating the absence of amide-linked fatty acids in these two receptors subunits. RHL3, however, gave significant amounts of palmitate and stearate (Ն6 times background).
To determine if palmitate and stearate are attached to Cys residues of the receptor subunits via thioester linkages, the affinity-purified ASGP-Rs were treated prior to SDS-PAGE with hydroxylamine under mild conditions. Total ASGP-R was first treated with NH 2 OH, and then each RHL subunit was purified by SDS-PAGE and treated with alkali (Table I). Quantitation of the released fatty acids showed that hydroxylamine treatment released 73, 100, and 68% of the methyl palmitate that could be detected in RHL1, RHL2, and RHL3, respectively. Similarly, for methyl stearate 74, 100, and 79% of this fatty acid detectable in RHL1, RHL2 and RHL3, respectively, was released by NH 2 OH treatment. Tris treatment had no significant effect on any of the RHL subunits (Table I) thioester-linkages to Cys.
Taking the molar ratio of the subunits (RHL1:RHL2:RHL3 Ϸ 4:1:1) into account and assuming comparable yields, RHL2 and RHL3 contain about 4 -5 times more covalently bound fatty acids than RHL1 (Table I). Although the stoichiometries of fatty acids per RHL subunit have not been determined, these relative results were very similar in three independent experiments (Table I).
To confirm the above biochemical analyses, metabolic labeling experiments were performed to assess the incorporation of radiolabeled fatty acids into ASGP-R in intact hepatocytes. After cells were cultured in the presence of [ 3 H]palmitate for 16 h, ASGP-Rs were affinity-purified and analyzed by SDS-PAGE and fluorography. All three RHL subunits were labeled with [ 3 H]palmitate, although RHL1 had significantly more radioactivity than either RHL2 or RHL3 (Fig. 3). To verify that the 3 H radioactivity represents palmitate, RHL1 was isolated, treated with alkali, and the hydrolysate was analyzed by HPLC (Fig. 4). The majority (81%) of the released radoactivity comigrated with authentic palmitate confirming that this fatty acid was covalently incorporated into ASGP-Rs in vivo. The remaining radioactivity which migrated as small alkyl fragments (probably C2-C10) could be derived in part from O-acetylated sialic acids on RHL1. Similar metabolic labeling experiments using [ 3 H]stearate or [ 3 H]myristate did not show detectable incorporation of these fatty acids into any of the RHL subunits of affinity-purified ASGP-R (not shown). Ongoing experiments indicate that the human ASGP-R is also metabolically labeled when HepG2 or HuH7 cells are cultured in the presence of [ 3 H]palmitic acid. 2

DISCUSSION
A number of diverse proteins including viral proteins (25), the ␣ subunits of G proteins (26), the G protein-coupled receptors (27)(28)(29), and the transferrin receptor (30,31) have been shown to be modified by palmitate via thioester linkages. Palmitoylation has been suggested to play an important functional role in many proteins. For example, the palmitoylation of subunits of G proteins is involved both in membrane attachment as well as in modulating signaling capability (32,33). Palmitoylation of the neuronal growth cone protein GAP-43 reduces its ability to catalyze nucleotide exchange on a G protein (34). Many investigators have shown that palmitoylation/depalmitoylation of a variety of proteins is a dynamic process (3,4).
Although the present study shows for the first time that ASGP-Rs are fatty acylated, Stockert 3 found evidence over 10 years ago that [ 3 H]palmitate might be incorporated into the human ASGP-R. Our results here clearly demonstrate that all three subunits of the ASGP-R are modified by fatty acylation.
The GC-MS analysis shows the presence of both palmitate and TABLE I GC analyses of fatty acids released by alkali treatment of ASGP-Rs Freshly purified ASGP-Rs were subjected to SDS-PAGE, extraction, and alkaline hydrolysis as described under "Experimental Procedures." The released fatty acids were derivatized to their methyl esters, and analyzed by GC. The values for methyl palmitate and methyl stearate were obtained by comparing the area of corresponding peaks with those of known amounts of the standard fatty acid methyl esters (after subtracting the background for a blank gel of the same size). Three experiments were performed using independent ASGP-R preparations. To determine if pretreatment with hydroxylamine released fatty acids, equal amounts of ASGP-R from the same preparation (experiment 3) were incubated on ice for 2 h with buffer alone (10 mM Hepes, pH 7.4, 150 mM NaCl, 6.7 mM KCl, 0.025% Triton X-100), or buffer containing 1 M Tris or 1 M hydroxylamine. These treated samples were subsequently subjected to the same procedures as above to determine the remaining alkaline-releasable fatty acid content.  16 h, and active ASGP-Rs were purified as described under "Experimental Procedures." Nine-tenths of the purified ASGP-Rs (from one 60-mm dish containing about 2 ϫ 10 6 cells) were analyzed by SDS-PAGE and fluorography (lane 1), and one-tenth (lane 2) was analyzed by Western blotting and detected using a mixture of subunit-specific affinity-purified antibodies specific for RHL1 or RHL2/3.

FIG. 4. Metabolically labeled RHL1 contains authentic [ 3 H]palmitate.
Hepatocytes were cultured with [ 3 H]palmitate, and radiolabeled ASGP-Rs were purified and separated by SDS-PAGE. RHL1containing bands were excised, and subjected to extensive wash, alkali treatment, and extraction of released fatty acids. The hydrolysate was analyzed by reverse-phase HPLC as described under "Experimental Procedures." The retention times for co-chromatographed myristic acid (C14), palmitic acid (C16), and stearic acid (C18) are marked by arrows. stearate in each RHL subunit. The results from the GC-MS analyses were confirmed by the finding that all three subunits of the ASGP-R are metabolically labeled by [ 3 H]palmitic acid. Initial metabolic labeling studies with [ 3 H]stearate did not show incorporation into ASGP-Rs, although [ 3 H]palmitate was readily incorporated. In earlier studies, however, stearyl-CoA was as effective as palmitoyl-CoA in being able to reactivate State 2 ASGP-Rs that had been inactivated by ATP treatment of permeable cells (13). Furthermore, as shown in Table I and the accompanying paper (15), the sensitivity of fatty acid release by hydroxylamine treatment at neutral pH is indicative of a thioester linkage. Pretreatment with hydroxylamine released approximately 66 -100% of the stearate or palmitate in RHL1, RHL2, and RHL3 that could be released by alkaline hydrolysis. Therefore, Ն66% of the covalently bound fatty acids in alkali-labile linkages is probably present as thioesters. This percentage could actually be significantly greater if any fatty acyl migration occurs during receptor purification. Some thioesters could react with nucleophilic -NH 2 or -OH groups in proximal amino acids to produce amide-linked or esterified fatty acids. This may explain why virtually all of the metabolically incorporated [ 3 H]palmitate is released by mild NH 2 OH treatment (15) and, therefore, presumably present exclusively as thioesters.
The GC-MS and metabolic labeling results show apparent differences in the extent of RHL subunit fatty acylation, the fatty acids involved and the proportion of thioester linkages. It is premature to consider these differences significant, since the two methods are so different and each has inherent biases. For example, chemical analysis of purified total receptor should represent a steady-state incorporation of fatty acids mediated by every acylating and deacylation enzyme that recognizes ASGP-Rs from every part of the receptor trafficking pathway. Multiple cellular sites of modification (e.g. cell surface, endosome, or vesicles recycling back to the surface) and multiple fatty acylation sites within the ASGP-R likely make the overall modification pattern and process very complicated. Metabolic labeling studies, which are frequently not at a steady-state, can preferentially identify the fastest fatty acylation step for newly synthesized molecules or a site(s) of the most rapid fatty acyl turnover. More extensive studies will be required to characterize these processes in intact cells.
The cytoplasmic domains of each RHL subunit have a single Cys residue within 5 amino acids of the transmembrane domain, which is a likely site for fatty acylation. The same position in the transferrin receptor is palmitoylated (30,31,35). Studies are in progress to determine the site(s) of modification in the three RHL subunits. The three ASGP-R subunits are known to have different extents of glycosylation (9), and are also differentially modified by phosphorylation in their cytoplasmic domains (36,37). Our present results strongly suggest that the three subunits are also modified by fatty acylation to different extents. Thus, in the steady-state situation, all RHL subunits contain palmitate and stearate in similar ratios, but the amount of these two fatty acids in RHL1 is smaller than in RHL2 and RHL3. We report in the accompanying paper (15) that treatment of radiolabeled affinity-purified ASGP-Rs with hydroxylamine completely removes the [ 3 H]palmitic acid from RHL1, RHL2, and RHL3. The metabolic labeling experiments indicate that, in fact, only the State 2 ASGP-Rs (one of two receptor subpopulations that we have previously characterized) are palmitoylated and the vast majority of fatty acylation is via thioester bonds.
One population of ASGP-Rs (the State 2 ASGP-Rs) undergoes a transient inactivation-reactivation cycle during receptor recycling (6,(11)(12)(13)(14). Such an inactivation-reactivation cycle may provide a biological basis for the high efficiency of the ligand-receptor segregation step in this and in other recycling receptor systems (6). That fatty acyl-CoAs can regulate ASGP-R activity in permeable hepatocytes (13,14), suggests that a reversible fatty acylation-deacylation process could be the molecular basis of the State 2 receptor inactivation-reactivation cycle. The finding that one population of purified ASGP-Rs is selectively inactivated by hydroxylamine treatment, under mild conditions that releases essentially all of the covalently associated fatty acids, (15), supports this hypothesis. Thus, we conclude that the three ASGP-R subunits are modified by palmitate and stearate, and that acylation-deacylation of the ASGP-R could directly regulate its activity.
An important point about the fatty acylation results presented here is that we analyzed the total active ASGP-R pool, which contains both State 1 and State 2 receptors. It is likely that the State 1 ASGP-Rs are not fatty acylated; fatty acylation may occur exclusively in the State 2 ASGP-R population (15). Further studies will address whether fatty acylation in RHL1, RHL2, and/or RHL3 is the molecular basis for the two functionally distinct receptor populations and how fatty acylation/ deacylation regulates receptor activity.