INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

A number of integral membrane proteins are modified after their synthesis by the covalent attachment of long-chain fatty acids (mostly palmitic acid) to one or more cytoplasmically oriented cysteine residues (for review, see Refs. 1-6). In the majority of the cases, the chemically bound acyl chains turn over much faster than the protein backbone, implying that palmitoylation is a regulatory modification. In fact, fatty acylation of this and other types of proteins has been shown to be modulated by physiological (7-9) or pharmacological stimuli (10-14). To date, the metabolic features of palmitoylation have only been studied by labeling cultured cells with [³H]palmitic acid, and the half-life of the palmitate has been estimated from the disappearance of the protein-bound radioactivity after isotopic dilution with the unlabeled fatty acid. Unfortunately, labeling experiments using [³H]palmitic acid are difficult to interpret, particularly when considering the possibility that exogenous and endogenous palmitate may not have equal access to the fatty acid donor pools. Furthermore, since the specific radioactivity of the donor pool of palmitate used for protein palmitoylation cannot be estimated, it is not possible to determine the number of protein molecules participating in such rapid deacylation-reacylation cycles. Consequently, the radioactivity that becomes associated with a polypeptide during the course of an experiment could either represent the periodic repair of thioester linkages on a few protein molecules or the physiologically relevant exchange of the fatty acids on many molecules.

In the central nervous system, proteolipid protein (PLP)¹ accounts for more than 50% of the total myelin protein (15). Although the specific function of this tetraspan membrane protein is unclear, its physiological importance is demonstrated by the requirement of normal PLP synthesis during myelination (16). PLP contains between 2 and 3 mol of fatty acids, mainly palmitic, oleic, and stearic acid (17), which are bound to several intracellular cysteine residues via labile thioester linkages (18-20). Studies employing [³H]palmitate have shown that the attachment of fatty acid to the polypeptide backbone takes place close to or within the myelin membrane (21, 22). The half-life of the palmitate bound to PLP measured in vivo was found to be approximately 3 days, a value significantly smaller than that of the protein backbone (t1/2 > 30 days) (23). Moreover, pulse-chase experiments in cell-free systems have shown that PLP-bound palmitate turns over within a few minutes (23). The occurrence of dynamic palmitoylation of PLP in myelin is also supported by the presence of substantial levels of PLP acylesterase activity in this subcellular fraction (24). However, the occurrence of PLP in a metabolically stable membrane such as myelin makes it difficult to envision the function that such a dynamic modification may have, and therefore it raises questions as to how many molecules do indeed participate in deacylation-reacylation cycles.

To specifically address the question regarding the number of fatty acids being incorporated into PLP during a period of time, we used the elegant technique of H₂¹⁸O exchange initially developed by Schmid and co-workers (25-27) to determine phospholipid acyl chain turnover in macrophages. In this study, we incubated rat brain white matter slices in medium containing H₂¹⁸O, a molecule that readily equilibrates into the cells and participates in all normal hydrolytic reactions. As fatty acyl-esters in phospholipids are hydrolyzed in the presence of H₂¹⁸O, the isotopic oxygen becomes either the hydroxyl or the carbonyl oxygen of the resultant free fatty acid (FFA) (Fig. 1). The carbonyl ¹⁸O-labeled fatty acids are then activated to fatty acyl-CoA, and reesterified creating phospholipids and acylproteins with the stable isotope incorporated at the carbonyl oxygen of the newly formed oxyester and thioester linkage, respectively. The incorporation of these ¹⁸O-labeled fatty acids into PLP and lipids is quantitatively determined via GLC/MS of the methyl esters released by alkaline methanolysis. Using this technique, we found that a significant proportion of the palmitate (7.6%) and stearate (2.5%) in PLP from 20-day-old animals are incorporated during the course of 6 h. These values cannot be attributed to changes in the stoichiometry of acylation or to the acylation of newly synthesized protein, but to the replacement of unlabeled fatty acids for ¹⁸O-labeled fatty acids. The value of this approach is also evident when considering that, whereas significant amounts of radioactivity are incorporated into PLP in the adult (60-day-old) animal when [³H]palmitic acid is used as tracer (23), only a minute proportion of PLP molecules are palmitoylated when using endogenously generated ¹⁸O-labeled fatty acids. To the best of our knowledge, this is the first time that the metabolism of acyl chains in palmitoylated proteins has been studied with a stable isotope technique.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Materials-- Sprague-Dawley rats of either 20 or 60 days of age were purchased from Harlan Sprague-Dawley (Madison, WI). H₂¹⁸O (95-98% enrichment) was obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA) and was used at a 50% (v/v) concentration in the incubation medium. [³H]₂O (10 mCi/ml) and L-[1-¹⁴C]methionine (59.2 mCi/mmol) were from American Radiolabeled Chemicals, Inc. (St. Louis, MO). Cycloheximide (lot 21H0093), phospholipase A₂ (from Naja naja venom), and the penicillin-streptomycin mixture were acquired from Sigma. 1-Methyl-3-nitro-1-nitrosoguanidine used to generate diazomethane was from Aldrich. All other chemicals were of the highest purity available.

Incubation of Tissue Slices-- Twenty- or 60-day-old rats were killed by decapitation. The brain was immediately removed from the animal, and both the cerebral cortex and the cerebellum were dissected out and discarded. The remaining tissue, mostly white matter, was sliced into sections (400 µm × 400 µm). Tissue sections (600 mg) were then transferred to a flask containing 2 × Krebs-Ringer bicarbonate buffer, pH 7.4, supplemented with 20 mM glucose and diluted to 1 × buffer with either H₂¹⁸O (experimental) or H₂¹⁶O (control). In some experiments, 100 µM cycloheximide was added prior to the incubation to inhibit the synthesis of myelin PLP. The slices were incubated for up to 6 h in a 37 °C shaker bath under a constant atmosphere of 95% O₂ and 5% CO₂. One hundred units of penicillin and 100 µg of streptomycin in 0.9% NaCl were added to the medium to inhibit bacterial growth during the longer incubation periods. After incubation, samples were removed from the bath and placed on ice. The supernatant was aspirated, and the tissue slices were homogenized in 6 ml of cold 0.32 M sucrose using a 15-ml Wheaton homogenizer.

Myelin Preparation-- Myelin was isolated from the homogenate by the method of Norton and Poduslo (28). Briefly, the homogenate was layered onto 6 ml of 0.85 M sucrose and spun for 60 min at 100,000 × g at 4 °C in a Beckman L7 Ultracentrifuge. After centrifugation, the 0.32 M/0.85 M interphase was collected and osmotically shocked by the addition of 10 volumes of cold water. Purified myelin membranes were collected by centrifugation at 17,000 × g for 20 min at 4 °C in a Beckman model J21B centrifuge. The pellet was suspended in 1 ml of cold water. A 200-µl aliquot of each sample was removed for lipid analysis.

Isolation of Myelin Proteolipid Protein-- Before the isolation of PLP, KCl was added to the myelin suspension to a final concentration of 0.1 M. Proteolipids were extracted from myelin with 19 volumes of chloroform:methanol (2:1, v/v) (29). The total lipid extracts were washed once with water and once with methanol:water (1:1, v/v). After the addition of 0.5 volume of chloroform, the extracts were concentrated under vacuum and the protein was precipitated with cold acetone. Finally, proteins were collected by low speed centrifugation and the pellets were dissolved in 0.5 ml of chloroform:methanol:0.1 M HCl (1:1:0.05, by volume). Samples were analyzed on a Sephadex LH-60 column (1 cm × 50 cm; Pharmacia, Uppsala, Sweden) equilibrated and eluted with the same solvent mixture (30). PLP, free of non-covalently bound lipids, eluted at the void volume. PLP-containing fractions were combined, and the amount of protein was calculated using an absorption coefficient E^1% at 280 nm of 13.6.

Gas Liquid Chromatography-- Purified PLP plus a known amount of the internal standard nonadecanoate methyl ester were dried under N₂, and the covalently bound fatty acids were transesterified via alkaline methanolysis. For this purpose, the protein was left standing overnight at room temperature in chloroform:0.21 M NaOH in methanol (2:1, v/v). The solution was then neutralized with 0.2 volume of 0.35 M acetic acid and washed twice with methanol:water (1:1, v/v). The resultant fatty acid methyl esters (FAMEs) were dried under N₂ and dissolved in 10 µl of hexane. Aliquots of 1-2 µl were analyzed by gas-liquid chromatography using a Hewlett Packard 5890 Series II gas chromatograph (Kennett Square, PA) equipped with a fused silica Megabore DB-225 column (15 m × 0.53 mm; J&W, Folsom, CA), a flame ionization detector, and an integrator. Peaks were identified by the use of standard FAMEs. The area under each peak was considered proportional to the mass of each methyl ester within the sample, and quantities were calculated based on the amount of internal standard added. Appropriate column blanks were analyzed in parallel and subtracted out from the sample runs.

Gas Liquid Chromatography/Mass Spectrometry-- To determine the proportion of each FAME carrying ¹⁸O, samples were separated on a 30-m capillary DB-225 column with a Hewlett Packard 5890 Series II Plus gas liquid chromatograph coupled to a Hewlett Packard 5972A mass selective detector. ¹⁸O atom percent excess in each fatty acid methyl ester was calculated using the McLafferty rearrangement ion (m/z = 74 and m/z = 76 for ¹⁶O-containing and ¹⁸O-containing methyl esters, respectively) (31). Control samples incubated in buffer prepared with H₂¹⁶O were used to correct for the natural abundance of ¹⁸O (about 1%). The data were evaluated utilizing the following equation (24).

(Eq. 1)

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the H218O technique for assessing the metabolism of PLP acyl chains. See the Introduction for a detailed description. Note that, during the formation of fatty acyl-CoA, the hydroxyl oxygen is lost, and consequently the 18O specific activity of the acyl chain donor is only one-half of that of the FFA. Jagged lines represent the aliphatic, saturated or unsaturated, hydrocarbon chains of fatty acids. R represents the glyceryl moiety of membrane glycerolipids. CX denotes cycloheximide.

Lipid Analysis-- Phospholipids were extracted from myelin with chloroform:methanol (1:1, v/v). Insoluble material was removed by low speed centrifugation, and 0.5 volume of chloroform was added to the total lipid extracts. The extracts were washed once with water and once with methanol:water (1:1, v/v), and were then dried under N₂. Phospholipids were separated by thin-layer chromatography on Silica Gel G plates using chloroform:methanol:water:acetic acid (60:50:4:1, by volume) as the developing solvent. Lipid classes were visualized with iodine vapor, scraped, and eluted from the silica with chloroform:methanol:water (30:50:20, by volume). PC, PS, and PE were derivatized by alkaline methanolysis (32) and analyzed via GLC/MS as described above. FFAs were isolated from an aliquot of the same total lipid extract by thin-layer chromatography on Silica Gel G plates developed with hexane:diethyl ether:acetic acid (60:40:2, by volume). Spots were visualized with 2,7-dichlorofluerescein, scraped, and eluted. Finally, fatty acids were methylated by reaction with diazomethane in ether (33) and analyzed via GLC/MS. It is noteworthy that the presence of acid in the various running solvents does not result in the loss of ¹⁸O, since similar results were obtained when lipids were separated using neutral solvents.

Synthesis of Myelin Proteolipid Protein-- Brain white matter slices from 20-day-old rats were incubated in Krebs-Ringer bicarbonate buffer as described above with 4 µCi of [¹⁴C]methionine in the absence or presence of 100 µM cycloheximide. After incubation, myelin PLP was isolated and the radioactivity incorporated into the protein was determined by liquid scintillation counting. Aliquots of the total homogenate were used to determine the specific radioactivity of free methionine by amino acid analysis using the PICO-TAG method (Waters, Milford, MA) and liquid scintillation counting. Specific radioactivity of [¹⁴C]methionine in the slices ranged between 2.9 and 4.3 nCi/nmol. To calculate the proportion of newly synthesized PLP, the specific radioactivity of PLP (expressed as nanocuries of ¹⁴C-radioactivity incorporated/nmol of PLP) was divided by the number of moles of methionine/mol of PLP (i.e. 4) and by the corresponding specific radioactivity of methionine (nanocuries of ¹⁴C radioactivity/nmol of methionine). Finally, these values were multiplied by 100 to express the data as percent of newly synthesized PLP (Fig. 4).

Statistical Analysis-- Statistical significance was determined using the Student's unpaired t test and a one-way analysis of variance calculated with MINITAB data analysis software (release 1.1).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Incorporation of ¹⁸O-Labeled Fatty Acids into PLP-- Myelin proteolipid protein represents a family of structurally related gene products, the most abundant of which are the major PLP and the DM-20 protein (15). Both species are S-acylated with a similar profile of long-chain fatty acids (20, 34), and the metabolic behavior of these acyl chains is comparable (35). Furthermore, in cell-free systems, both proteins are non-enzymatically acylated (35) and equally susceptible to enzymatic deacylation (24). Hence, in this study, the metabolism of the chemically bound fatty acids was studied in the mixture of these species. Prior to assessing the incorporation of endogenously generated fatty acids into PLP, it was necessary to determine whether or not the protein's acyl chain composition changed during incubation. As shown in Table I, the amount and proportion of fatty acids in PLP isolated from white matter slices that had been incubated for 1 and 6 h are similar. The protein contains approximately 3% (w/w) covalently bound fatty acids; palmitic, oleic, and stearic acid account for >79% of the total acyl chains. PLP from 60-day-old rats also has an equivalent amount of fatty acids but a higher proportion of palmitic acid than that isolated from younger animals. No differences were observed between incubated and non-incubated tissue (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Incorporation of 18O into palmitic and stearic acid derived from PLP and myelin FFAs. Brain white matter slices from 20-day-old rats were incubated in Krebs-Ringer bicarbonate buffer containing 50% H218O for different periods of time. After incubation, myelin PLP and FFAs were isolated and the amount of 18O incorporated into palmitic and stearic acids was determined by GLC/MS. % 18O enrichment values were calculated as described under "Experimental Procedures." Data represent the mean ± S.E. of three to five experiments. Closed and open symbols correspond to values for palmitic and stearic acid, respectively.

Incorporation of ¹⁸O into Myelin FFAs-- ¹⁸O incorporation into both palmitic and stearic acids from myelin increased with time, resulting in 67.2 and 104.6% ¹⁸O enrichment at 6 h, respectively (Fig. 2B). However, although the proportion of ¹⁸O-labeled palmitic acid reached the maximum at 3 h, the incorporation of the stable isotope into stearic acid continued to increase during the course of the incubation. Nevertheless, the high % ¹⁸O enrichment values indicate that most of the myelin FFAs participate in deacylation-reacylation cycles. It is important to note that the time-dependent uptake of ¹⁸O into free fatty acids is due to the steady acyl chain turnover of membrane lipids and not to a slow entry of isotopically labeled water into cells. Experiments in which tissue slices were incubated [³H]₂O revealed that the radiolabel equilibrated with the tissue within the first 5 min (data not shown).

Incorporation of ¹⁸O-Labeled Fatty Acids into Myelin Phospholipids-- Palmitic and stearic acids derived from myelin PC and PE exhibited a gradual increase in the amount of ¹⁸O (Fig. 3, A and B). After the 6 h of incubation, and upon correction for the changes in the specific activity of the myelin FFAs, palmitic and stearic acid in PC achieved 8.7% and 4.7% ¹⁸O enrichment, respectively (Fig. 3D). Over the same period, palmitic and stearic acids in PE reached an ¹⁸O enrichment of 10.2 and 6.6%, respectively (Fig. 3E). Both the uncorrected and corrected % ¹⁸O enrichment values for PS-derived palmitic and stearic acids were maximal at 1 h and remained unchanged during the course of incubation (Fig. 3, C and F). This constant ¹⁸O incorporation is difficult to interpret, since in myelin this lipid is synthesized mainly by base-exchange reactions with PC and PE (39). ¹⁸O incorporation into the fatty acid moiety of sphingomyelin was not analyzed because the acid hydrolysis that would be required to release the amide-linked fatty acid would result in the quantitative loss of ¹⁸O label (26).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Incorporation of 18O into palmitic and stearic acid derived from myelin phospholipids. Experiments were carried out as described in Fig. 2. Individual myelin phospholipids were isolated by thin-layer chromatography, and the amount of 18O incorporated into palmitic and stearic acids was determined by GLC/MS. % 18O enrichment values were calculated as described under "Experimental Procedures." Data represent tha mean ± S.E. of three to five experiments. Closed and open symbols correspond to values for palmitic and stearic acid, respectively. Asterisks denote the values that are significantly different (p < 0.05) from that at 1 h.

¹⁸O Incorporation in the Presence of Cycloheximide-- PLP can acquire ¹⁸O-labeled fatty acids via acyl chain turnover and/or via the acylation of newly synthesized molecules (Fig. 1). To distinguish between these possibilities, white matter slices were incubated with H₂¹⁸O for 4 h in the absence and presence of cycloheximide, a potent inhibitor of the translocation reaction in protein synthesis. As shown in Table II, 100 µM cycloheximide had no effect on the incorporation of ¹⁸O-labeled fatty acids into PLP. In contrast, the inhibitor reduced the incorporation of [¹⁴C]methionine into PLP by >80% (Fig. 4), indicating that acylation occurs mostly on the preexisting protein molecules. This conclusion is supported by the finding that, under the same incubation conditions, the rate of PLP synthesis (Fig. 4) was approximately 5 times and 20 times lower than the uncorrected (Fig. 2A) and corrected rate of incorporation of [¹⁸O]palmitic acid into PLP (Fig. 2C), respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Synthesis of myelin PLP in tissue slices. Brain white matter slices from 20-day-old rats were incubated in 2 ml of Krebs-Ringer bicarbonate buffer containing 4 µCi of [14C]methionine for different periods of time. After incubation, myelin PLP was isolated and the amount of 14C radioactivity incorporated into the protein was determined by liquid scintillation counting. The percent of newly synthesized PLP molecules was calculated as described under "Experimental Procedures." Data represent the mean ± S.E. of three experiments. Closed and open symbols correspond to values obtained in the absence and presence of 100 µM cycloheximide, respectively. Asterisks denote the values that are significantly different (p < 0.05) from that at 1 h.

Pulse-Chase Experiments-- The above results suggest that ¹⁸O-labeled fatty acids are likely to be incorporated via the deacylation-reacylation of a proportion of preexisting protein molecules. To examine the possibility that the half-life of PLP-bound fatty acids is within the time-frame of our experiments, pulse-chase experiments were carried out. For this purpose, white matter slices were incubated in buffer containing 50% H₂¹⁸O for 3 h, after which the medium was removed and replaced with buffer containing 100% H₂¹⁶O. Slices were then incubated for an additional 3 h, to allow for the disappearance of the isotopically labeled fatty acids. As shown in Table III, during the 3-h chase, the proportion of ¹⁸O-containing free palmitic and stearic acid in myelin decreased by approximately 67% and 45%, respectively. The half-life of palmitate calculated from the decay values was 1.9 h, a value very close to that obtained from the ¹⁸O-incorporation curve (t1/2 = 1.7 h). In contrast to FFAs, the proportion of ¹⁸O-labeled palmitate in PLP did not diminish during the chase period. Similarly, no decay of the ¹⁸O label was observed in the fatty acids derived from myelin phospholipids (Table III).

Developmental Differences in ¹⁸O Enrichment-- Incubation of white matter slices from adult (60-day-old) rats in physiological buffer containing 50% H₂¹⁸O for 4 h resulted in 0.20 ± 0.11% ¹⁸O enrichment into PLP-derived palmitic acid (Table II). When the incubation was extended to 6 h, the % ¹⁸O enrichment was 0.18 ± 0.07, a value not different from that at 4 h. This could be a result of either a constant proportion of molecules acquiring ¹⁸O in the older animals, or the increased error obtained with such low incorporations. Nonetheless, the % ¹⁸O enrichment values were at least 10 times lower than those obtained in the younger animals. The limited incorporation of the [¹⁸O]palmitic acid into PLP in the older animals is not the result of a lower specific activity of the myelin free palmitic acid pool, since there were no differences in isotope incorporation into this FFA between the two ages (Table II). Incorporation of [¹⁸O]palmitic acid into PC and PE was also dramatically decreased, consistent with the low metabolic activity of the mature brain (Table II).

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

In this study, we have determined the minimal amount of fatty acids that are incorporated into myelin PLP in a particular period of time. These experiments are novel in that they assess the palmitoylation of a protein utilizing the endogenous fatty acids, thus eliminating the problems associated with the entry of the [³H]palmitic acid into cells, its equilibration with specific metabolic pools, and its interconversion into other radiolabeled metabolites. We have found that an amount close to 3% of palmitate and 1% of the stearate bound to PLP is incorporated in 6 h and 4 h, respectively. These are minimal rates of incorporation since they do not include adjustments for the specific activity of the donor fatty acid pool. When these values are corrected by changes in the specific activity of the myelin FFAs, the incorporation of palmitic and stearic acid into PLP increased to approximately 7.6% and 2.6%, respectively. Since PLP has six potential acylation sites (20), the amount of fatty acids incorporated into the protein cannot be equated to the number of protein molecules that are being acylated. Consequently, it was not possible to distinguish whether only 7.6% of PLP molecules exchange all of their palmitate or whether a larger proportion of the molecules becomes palmitoylated to a lesser extent.

Both this study and those using [³H]palmitic acid (21, 22) have shown that acylation of myelin PLP is not affected by cycloheximide. The persistence of normal levels of palmitate incorporation long after the arrest of the protein synthesis can be attributed to: (a) palmitoylation occurring at a site temporally distant from the rough endoplasmic reticulum of the oligodendroglial cell, where PLP synthesis takes place; (b) an increase in the stoichiometry of acylation; or (c) fatty acid turnover. The first of these possibilities is rather unlikely since the rate of PLP acylation with ¹⁸O-labeled fatty acids exceeds by at least 5-fold the rate at which the newly synthesized protein molecules are accumulated into myelin. The second alternative can also be ruled out because the amount of fatty acids covalently bound to PLP remains unchanged during the incubation and throughout development. Thus, the data can solely be explained by exchange of unlabeled fatty acids for ¹⁸O-fatty acids. Interestingly, the amount of [¹⁸O]palmitic acid incorporated into PLP did not decrease upon replacement of H₂¹⁸O by H₂¹⁶O, indicating that the half-life of the protein-bound palmitate is greater than that of free palmitic acid (t1/2 ~ 2 h). Because of this relatively low turnover rate, incubation for 6 h was insufficient to achieve isotopic equilibrium between the PLP-bound palmitic acid and the myelin free [¹⁸O]palmitic acid. However, the rate of uptake of ¹⁸O into PLP-derived palmitic acid showed a gradual decrease during the course of the incubation. Assuming that this decline was caused by the approach to the equilibrium, it is possible to calculate both the maximal number of fatty acids that can be incorporated and the half-life of the protein-bound palmitic acid. When the corrected ¹⁸O incorporation values shown in Fig. 2C were fitted to a first-rate equation, the curve reached a limit at 12-14% ¹⁸O enrichment with a half-life of 4-5 h.

Profound differences in the incorporation of ¹⁸O-labeled fatty acids into PLP were observed between young and adult animals. The % ¹⁸O enrichment in palmitic and stearic acid derived from PLP was greatly diminished in the older, slowly myelinating animals. This reduction is too large to be only explained by the 3-fold increase in the concentration of PLP, and therefore protein-derived ¹⁶O-fatty acids, that occurs between 20 and 60 days of age (40). Consequently, the age-associated changes are, to a large extent, due to a reduction in the number of PLP molecules undergoing palmitoylation with age. The limited incorporation of [¹⁸O]palmitic acid into PLP in the older animal was an unexpected finding since there are no noticeable developmental differences in acylation of PLP with [³H]palmitic acid (23). One possibility is that the presence of exogenously added palmitic acid may have influenced the normal fatty acid metabolism. However, we found that the addition of [¹⁸O]palmitic acid to an incubation medium containing 50% H₂¹⁸O does not change the % ¹⁸O enrichment of PLP-derived palmitic acid (data not shown). This result also suggests that the radioactivity normally incorporated into PLP in the adult, slowly myelinating animal represents almost negligible amounts of palmitate. In light of this new finding and contrary to our original view (23), it is fair to hypothesize that PLP palmitoylation plays some role in myelin formation and/or compaction rather than in the maintenance of this membrane.

Little is known regarding the subcellular site of protein palmitoylation, and it appears to depend upon the protein in question. Initial studies have shown that acylation of viral and cellular membrane glycoproteins occurs in membranes from the endoplasmic reticulum/Golgi complex (41-44). However, for proteins that participate in rapid deacylation-reacylation cycles, the attachment of the fatty acid is likely to take place at the plasma membrane, where the presence of protein acyltransferase activity has been recently demonstrated (45-50). In the case of PLP, studies involving biosynthetic labeling with [³H]palmitic acid and cellular fractionation have suggested that acylation occurs at a locus close to or within the myelin membrane (22, 23). Both the dynamic nature of PLP acylation and the finding that palmitoylation is reduced in adult animals suggest that the reaction is likely to occur in specialized regions of the myelin sheath, such Schmidt-Lantermann incisures; inner, outer, and paranodal loops; and the network of cytoplasmic channels, where PLP may be accessible to the acylating/deacylating machinery as well as fatty acyl-CoA. Experiments combining ¹⁸O labeling and fractionation of the myelin membranes are being undertaken to localize more precisely the subcompartment(s) where acylation takes place.

In this study, we also showed that a significant proportion of the acyl chains in myelin phospholipids acquire ¹⁸O during the course of the incubation. No attempts were made to determine whether the uptake of ¹⁸O into phospholipid-derived fatty acids takes place by the de novo synthesis through the acylation of glycerophosphate and/or via deacylation-reacylation reactions. However, the high levels of isotope incorporation relative to the proportion of membrane lipids synthesized during incubation suggests that the exchange of fatty acids on preexisting phospholipid molecules constitutes a major mechanism. In general, ¹⁸O enrichment profiles of PLP-fatty acids resemble those of the major myelin phospholipids. In both cases, the incorporation of ¹⁸O was time-dependent, unaffected by cycloheximide, and greatly reduced in the adult animal. At present, the reason(s) for the similarities in the metabolic behavior of PLP and phospholipid acyl chains are unknown. The possibility, however, that phospholipids could be contaminating the PLP preparations can be safely excluded based on the following observations. (a) Addition of tritiated phospholipids, glycolipids, or palmitic acid to unlabeled PLP prior to chromatography on Sephadex LH-60 does not result in the appearance of radioactivity in the protein peak (30); (b) chemical analysis of isolated PLP yields less than 0.035% (w/w) lipid phosphorous (30); (c) incubation of isolated PLP with phospholipase A2 does not remove the protein-bound fatty acid (24); and (d) the fatty acid composition of PLP does not resemble that of any myelin lipids (34).

As mentioned above, a benefit of the H₂¹⁸O method developed by Schmid and co-workers (25-27) is that it eliminates the problems associated with the slow entry of [³H]palmitic acid into cells and its equilibration with specific metabolic pools. This becomes evident when considering recent findings revealing that some fatty acid stores in the cell are resistant to labeling upon incubation with radiolabeled fatty acids (51). However, in our opinion, the major advantage of this stable isotope technique is that it allows one to estimate the minimal number of molecules being modified during the course of an experiment. Knowledge of acylation rates and the proportion of protein molecules that undergo palmitoylation is critical when making biological assumptions regarding the function of the modification. Consequently, it would be of considerable interest to extend this approach to study the palmitoylation of other proteins. Although the use of H₂¹⁸O to conduct a systematic study of protein palmitoylation is somewhat limited by the necessity for substantial amounts of purified protein (nanomoles), with the advent of effective expression systems in eukaryotic cells, heavy isotope labeling should become increasingly feasible.