Cell-free acylation of rat brain myelin proteolipid protein and DM-20.

Incubation of rat brain myelin with [3H]palmitic acid in the presence of ATP, CoA and MgCl2 or [14C]-palmitoyl-CoA in a cell-free system resulted in the selective labelling of 'PLP' [proteolipid protein; Folch & Lees (1951) J. Biol. Chem. 191, 807-817] and 'DM-20' [Agrawal, Burton, Fishman, Mitchell & Prensky (1972) J. Neurochem. 19, 2083-2089] which, after polyacrylamide-gel electrophoresis in SDS, were revealed by fluorography. These results provide evidence of the association of fatty acid-CoA ligase and acyltransferase in isolated myelin. Palmitic acid is covalently bound to PLP and DM-20, because 70 and 92% of the radioactivity was removed from proteolipid proteins after treatment with hydroxylamine and methanolic NaOH respectively. Incubation of myelin with [3H]palmitic acid in the absence of ATP, CoA, MgCl2, or all three, decreased incorporation of fatty acid into PLP to 3, 55, 18 and 2% respectively. The cell-free system exhibits specificity with respect to the chain length of the fatty acids, since myristic acid is incorporated into PLP at a lower rate when compared with palmitic and oleic acids. The acylation of PLP is an enzymic reaction, since (1) maximum incorporation of [3H]palmitic acid into PLP occurred at physiological temperatures and decreased with an increase in the temperature; (2) acylation of PLP with [3H]palmitic acid and [14C]palmitoyl-CoA was severely inhibited by SDS (0.05%); and (3) the incorporation of fatty acid and palmitoyl-CoA into PLP was substantially decreased by the process of freezing-thawing and freeze-drying of myelin. We have provided evidence that all of the enzymes required for acylation of PLP and DM-20 are present in isolated rat brain myelin. Acylation of PLP in a cell-free system with fatty acids and palmitoyl-CoA suggests that a presynthesized pool of non-acylated PLP and DM-20 is available for acylation.     

Fatty Acid Acylation of Rat Brain Myelin Proteolipid Protein In Vitro: Identification of the Lipid Donor

The immediate acyl chain donor for fatty acid esterification of proteolipid protein (PLP) was identified in an in vitro system. Rat brain total membranes, after removal of crude nuclear and mitochondrial fractions, were incubated with radioactive acyl donors, extracted with chloroform/methanol, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. In the presence of [3H]palmitic acid, CoA, ATP, and Mg2+, acylation of endogenous PLP occurred at a linear rate for at least 2 h. The radioactivity was associated with the protein via an ester linkage, mainly as palmitic acid. Omission of ATP, CoA, Mg2+, or all three reduced fatty acid incorporation into PLP to 44, 27, 8, and 4%, respectively, of the values in the complete system. Incubation of the membrane fraction with [3H]palmitoyl-CoA in the absence of CoA and ATP led to highly labeled PLP. These data demonstrate that activation of free fatty acid is required for acylation. Phospholipids and glycolipids were not able to acylate the PLP directly. Finally, when isolated myelin was incubated with [3H]palmitoyl-CoA in the absence of cofactors, only PLP was labeled, thus confirming the identity of palmitoyl-CoA as the direct acyl chain donor and suggesting that the acylating activity and the PLP pool available for acylation are both in the myelin.     

Rapid metabolism of fatty acids covalently bound to myelin proteolipid protein

Proteolipid protein (PLP), the major protein of central nervous system myelin, contains approximately 2 mol of covalently bound fatty acids. In this study, the in vivo turnover rate of the acyl chains bound to PLP was determined in 40-day-old rats after a single intracranial injection of [3H]palmitic acid. The apparent half-life of total fatty acids bound to PLP was approximately 7 days. After correction for acyl chain interconversion, the half-life of palmitate bound to PLP was only 3 days. This turnover rate is much more rapid than that of the protein moiety calculated under the same experimental conditions (t1/2 = 1 month). Additional evidence for the dynamic metabolism of acyl groups was provided by experiments in brain tissue slices which showed that acylation of PLP occurs in adult animals as well as during active myelination. Acylation of endogenous PLP in purified myelin and its subfractions was also studied during rat brain development using either [3H]palmitoyl-CoA or [3H]palmitic acid plus ATP and CoA. Labeling of endogenous PLP with [3H]palmitoyl-CoA was observed as early as 10 days postnatal and continued at the same rate throughout development. When [3H]palmitic acid was used as precursor in the presence of both ATP and CoA, esterification of myelin PLP occurred rapidly in adult animals, indicating that both nonacylated PLP and acyl-CoA ligase are present in myelin. Finally, pulse-chase experiments in a cell-free system showed that PLP-bound fatty acids turn over with a half-life shorter than 10 min. These observations are consistent with the concept that acylation of myelin PLP is a dynamic process involved mainly in myelin maintenance and function.     

In vitro acylation of myelin PLP and DM-20 in the quaking mouse brain

Both proteolipid proteins (PLP) and DM-20 were found to be present by the immunoblot technique in myelin isolated from quaking mouse brain; however, the relative concentration of these proteins in myelin from quaking brain was substantially reduced when compared to the control. Brain slices from littermate control and quaking mice were incubated with [³H]palmitic acid to determine the incorporation of fatty acid into myelin proteolipid proteins. Fluorography of gels containing myelin proteins from control and quaking mice brain revealed that both PLP and DM-20 were acylated. The incorporation of [³H]palmitic acid into quaking myelin PLP and DM-20 was reduced by 75% and 20% respectively of those in control brain. The significance of differential acylation of quaking myelin PLP and DM-20 is discussed with respect to availability of non-acylated pools of proteolipid proteins and the activities of acylating enzymes.     

Intramyelinic conversion of cerebrosides into acylgalactosylceramides

Acylgalactosylceramide (AGC) synthesis was measured in vivo, and in a cell free system. 24 hours post-injection of [³H]palmitic acid into rat brain, more than 60% of the AGC radioactivity was associated with an ester linkage. Isolated rat myelin was incubated in the presence of [¹⁴C]palmitic acid, 2mM ATP, 50 M CoA and 10 mM MgCl₂ and acylation of myelin cerebrosides occurred at a linear rate for at least 60 min. Incubation of isolated myelin under standard conditions with [³H] cerebrosides and [¹⁴C]palmitic acid produced double labeled AGC. Labeling of AGC was maximum at pH 7.5 and 37°C and appeared to be enzyme mediated inasmuch as it was reduced by myelin incubation with trypsin and drastically reduced by preheating the myelin for 5 min at 80°C. Omission of ATP, CoA, MgCl₂ or all three did not reduce fatty acid incorporation into AGC when compared to the values in the complete system. Addition of Triton X100 or Sodium Dodecyl Sulfate had little or no effect on the acylation of cerebrosides. Pulse chase experiments indicated that the reaction involved the net addition of fatty acid to the cerebrosides, rather than a rapid fatty acid exchange.     

Proteolipid protein and DM-20 are synthesized by Schwann cells,present in myelin membrane,but they are not fatty acylated

Proteolipid protein (PLP) and DM-20 were intensely labeled after immunoprecipitation of total cellular proteins and myelin proteins labeled with [³⁵S]methionine in nerve slices. These results provided evidence that PLP and DM-20 are incorporated into the myelin membrane following their synthesis in Schwann cells. In contrast, PLP and DM-20 were not fatty acylated after incubation of the nerve slices with [³H]palmitic acid, however, PO glycoprotein and 24kDa protein were heavily fatty acylated. The lack of fatty acylation of PLP and DM-20 in the peripheral nervous system suggests that fatty acyltransferase responsible for their acylation is absent or non-functional in the peripheral nervous system.     

Acylation of Rat Brain Myelin Proteolipid Protein with Different Fatty Acids

The acylation of rat brain proteolipid protein (PLP) with tritiated palmitic, oleic, and myristic acids was studied in vivo and in vitro and compared with the acylation of lipids. Twenty-four hours after intracranial injection of [3H]myristic acid, only 16% of the PLP-bound label appeared as myristic acid, with 66% as palmitic, 9% as stearic, and 6% as oleic acid, whereas greater than 63% of the label in total or myelin phospholipid was in the form of myristic acid. In contrast, after labelling with [3H]palmitic or oleic acids, 75% and 86%, respectively, of the radioactivity in PLP remained in the original form. When brain tissue slices were incubated for short periods of time, the incorporation of palmitic and oleic acids into PLP exceeded that of myristic acid by a factor of 8. In both systems and with all precursors studied, the label associated with PLP was shown to be in ester linkage. The results suggest a preferential acylation of PLP with palmitic and oleic acids as compared with myristic acid. This is consistent with the fatty acid composition of the isolated PLP.     

Myelin proteolipid protein is not esterified at the site of synthesis

Brain slices prepared from 20-day old rats were incubated with [³H]palmitic acid to study its incorporation into myelin proteins. After separation by SDS-PAGE, most of the label was found to be associated with the major proteolipid protein (PLP) and with the intermediate protein (I). The radioactivity measured in PLP at short incubation times was shown to be due to palmitic acid bound to the protein by ester linkages. Time-course incorporation of [³H]palmitic acid into PLP of fraction SN₄ (a myelin like membrane) and of purified myelin showed that the former was poorly labeled and no relationship of the type ‘precursor-product’ between these fractions could be detected. Incorporation of the fatty acid into PLP was not affected by inhibition of the synthesis or transport of myelin PLP with cycloheximide or colchicine, indicating that the pool of PLP that can be acylated must be larger than the extramyelin pool. Addition of unlabeled palmitic acid to the incubation medium, 30 min after the addition of [³H]palmitate, stopped the appearance of label in myelin PLP almost immediately, indicating that there is no significant extramyelin pool of PLP destined for transport into myelin. The results presented in this paper strongly suggest that esterification of PLP takes place in the myelin membrane or at a site very close to it.     

Simultaneously entry of palmitic acid into proteolipid protein in different myelin subfractions of rat brain

Rats of 20-days of age were injected intracranially with radioactive palmitic acid to study its incorporation into proteolipid protein (PLP) of myelin and myelin subfractions. At short times (120 min), the radioactivity present in PLP was shown to be due to palmitic acid bound to the protein by ester linkages. The specific radioactivity of palmitic acid labeled PLP was identical in all the myelin subfractions except the myelin-like fraction, in which it was lower, suggesting that the entry of the fatty acid into PLP of the different subfractions occurs simultaneously.Experiments using time staggered injections of ¹⁴C- and ³H-labeled palmitic acid also showed that entry of the fatty acid into PLP of the various subfractions was simultaneous. These results seem to indicate that the acylation of PLP occurs in the myelin membrane and that synthesis and transport of this protein are events unrelated to the acylation process.     

Accumulation of galactosylsphingosine (psychosine) does not interfere with phosphorylation and methylation of myelin basic protein in the twitcher mouse

In attempts to elucidate mechanisms of demyelination in the twitcher mouse (Twi), phosphorylation and methylation of myelin basic protein (MBP) were examined in the brainstem and spinal cord of this species. Phosphorylation of MBP in isolated myelin by an endogenous kinase and an exogenous [³²P]ATP was not impaired and protein kinase C activity in the brain cytosol was not reduced. When the methylation of an arginine residue of MBP was examined in slices of the brainstem and spinal cord, using [³H]methionine as a donor of the methyl groups, no difference was found between Twi and the controls. Radioactivity of the [³H] methionine residue of MBP of Twi was also similar to that of the controls. Thus, accumulation of psychosine in Twi does not interfere with the activity of endogenous kinase, methylation of MBP, and the synthesis and transport of MBP into myelin membrane.     

In vivo acylation of proteolipid protein and DM-20 in myelin and myelin subfractions of developing rat brain: Immunoblot identification of acylated PLP and DM-20

The acylation of proteolipid protein (PLP) was examined in myelin and myelin subfractions from rat brain during the active period of myelination. Proteolipid protein and DM-20 in myelin and myelin subfractions were readily acylated in developing rat brain 22 hours after intracerebral injection of [3H]palmitic acid. No differences in the relative specific activity of PLP in myelin from 9-, 15-, and 30-day-old rat brains was observed; however, the relative specific activity of PLP in the heavy myelin subfraction tended to be higher than that in the light myelin subfraction. The acylation of PLP was confirmed by fluorography of immuno-stained cellulose nitrate sheets, clearly establishing that the acylated protein is in fact the oligodendroglial cell- and myelin-specific protein, PLP. Since PLP is acylated in the 9-day-old animal, when little compact myelin is present, it is possible that the acylation of PLP is a prerequisite for the incorporation of this protein into the myelin membrane.     

Effect of ATP Depletion on the Palmitoylation of Myelin Proteolipid Protein in Young and Adult Rats

The present study was designed to determine whether the palmitoylation of the hydrophobic myelin proteolipid protein (PLP) is dependent on cellular energy. To this end, brain slices from 20- and 60-day-old rats were incubated with [3H]palmitate for 1 h in the presence or absence of various metabolic poisons. In adult rats, the inhibition of mitochondrial ATP production with KCN (5 mM), oligomycin (10 microM), or rotenone (10 microM) reduced the incorporation of [3H]palmitate into fatty acyl-CoA and glycerolipids by 50-60%, whereas the labeling of PLP was unaltered. Incubation in the presence of rotenone (10 microM) plus NaF (5 mM) abolished the synthesis of acyl-CoA and lipid palmitoylation, but the incorporation of [3H]palmitate into PLP was still not different from that in controls. In rapidly myelinating animals, the inhibition of both mitochondrial electron transport and glycolysis obliterated the palmitoylation of lipids but reduced that of PLP by only 40%. PLP acylation was reduced to a similar extent when slices were incubated for up to 3 h, indicating that exogenously added palmitate is incorporated into PLP by ATP-dependent and ATP-independent mechanisms. Determination of the number of PLP molecules modified by each of these reactions during development suggests that the ATP-dependent process is important during the formation and/or compaction of the myelin sheath, whereas the ATP-independent mechanism is likely to play a role in myelin maintenance, perhaps by participating in the periodic repair of thioester linkages between the fatty acids and the protein.     

Nitric Oxide Reduces the Palmitoylation of Rat Myelin Proteolipid Protein by an Indirect Mechanism

Brain slices from 20-day-old rats were incubated with [³H]palmitate for 2 hours in the absence or presence of the NO-donors S-nitroso-N-acetyl-penicillamine (SNAP), ethyl-2-[hydroxyimino]-5-nitro-3-hexeneamide (NOR-3), 4-phenyl-3-furoxan carbonitrile (PFC) and sodium nitroprusside (SNP). Each of these drugs reduced the incorporation of [³H]palmitate into myelin proteolipid protein (PLP) in a concentration-dependent manner, SNP being the most active. The effect of SNAP was prevented by the NO-scavenger PTIO (2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide). Furthermore, decayed-SNAP, sodium nitrite and N- nitrosopyrrolidine were inactive, suggesting that free NO and/or some of its direct oxidation products are the active molecular species. The amount of fatty acids bound to PLP and the rate of deacylation were unaffected by NO. Although NO diminished the number of thiols in brain and myelin proteins, with the formation of both nitrosothiols and disulfides, these changes did not parallel those in PLP acylation. In contrast, NO was effective at reducing the palmitoylation of brain and myelin lipids, and this effect along with that of PLP, was ascribed to a decrease in palmitoyl-CoA levels. The NO-induced reduction in acyl-CoA concentration was due to the decline in ATP levels, while the amount of [³H]palmitate incorporated into the tissue, the activity of palmitoyl-CoA ligase and palmitoyl-CoA hydrolase, and the concentration of CoASH were unaltered by the drugs. Experiments with endogenously-synthesized [¹⁸O]fatty acids confirmed that NO affects predominantly the ATP-dependent palmitoylation of PLP. In conclusion, the inhibitory action of NO on the fatty acylation of PLP is indirect and caused by energy depletion.     

The transfer of myristic and other fatty acids on lipid and viral protein acceptors in cultured cells infected with Semliki Forest and influenza virus.

[3H]Myristic and [3H]palmitic acid were compared as tracers for the fatty acylation of cellular lipids and viral glycoproteins in chicken embryo cells infected with fowl plague and Semliki Forest virus (SFV). Both of these substrates are incorporated into glycerolipids to a similar extent, whereas sphingolipids show much higher levels of palmitate than myristate after a 20 h labeling period. Both fatty acid species were found to be subject to metabolic conversions into longer chain fatty acids yielding 11.7% C16:0 from [3H]myristic and 11.8% C18:0 from [3H]palmitic acid. The reverse, a metabolic shortening of the exogenous acyl-chains yielding, for instance, significant levels of myristic acid from palmitic acid was not observed. Out of the various [3H]fatty acids present after in vivo labeling with [3H]myristic acid (C14:0) the elongated acyl-species arising from metabolic conversion (e.g., C16:0; C18:0) are preferred over myristic acid in the acylation of SFV E1 and E2 and of the influenza viral hemagglutinin (HA2). During acylation of exogenous E1 from SFV in vitro incorporation of palmitic acid from palmitoyl CoA exceeds that of myristic acid from myristoyl CoA by a factor of 37. This indicates that specificity for the incorporation of fatty acids into viral membrane proteins occurs at the level of the polypeptide acyltransferase(s).     

Fatty Acid Composition of Myelin Proteolipid Protein During Vertebrate Evolution

The hydrophobic myelin proteolipid protein (PLP) contains covalently bound long-chain fatty acids which are attached to intracellular cysteine residues via thioester linkages. To gain insight into the role of acylation in the structure and function of myelin PLP, the amount and pattern of acyl groups attached to the protein during vertebrate evolution was determined. PLP isolated from brain myelin of amphibians, reptiles, birds and several mammals was subjected to alkaline methanolysis and the released methyl esters were analyzed by gas-liquid chromatography. In all species studied, PLP contained approximately the same amount of covalently bound fatty acids (3% w/w), and palmitic, palmitoleic, oleic and stearic acids were always the major acyl groups. Although the relative proportions of these fatty acids changed during evolution, the changes did not necessarily follow the variations in the acyl chain composition of the myelin free fatty acid pool, suggesting fatty acid specificity. The phylogenetic conservation of acylation suggests that this post-translational modification is critical for PLP function.     

Effects of monensin on posttranslational processing of myelin proteins

Rat brain slices were incubated with [3H]palmitic acid and [14C]glycine to label the lipid and protein moieties, respectively, of myelin proteolipid protein (PLP). The effects of monensin on posttranslational processing of proteins were examined by measuring the appearance of [14C]glycine- and [3H]palmitate-labeled proteins in myelin and myelin-like fractions. At 0.01 and 0.10 microM, monensin did not appreciably affect total lipid or protein synthesis; higher concentrations caused increased inhibition. Monensin at 0.10 microM markedly decreased the appearance of [14C]glycine-labeled PLP in myelin, but had little effect on the 14C basic proteins or the incorporation of [3H]palmitic acid into total or myelin PLP. The same relative effect was apparent at higher monensin concentrations. In the myelin-like fraction, monensin at 0.10 microM also depressed entry of [14C]glycine into protein comigrating with PLP, and again had no effect on incorporation of [3H]palmitic acid. In addition, monensin increased the [3H]palmitate label associated with two high-molecular-weight proteins in the myelin-like fraction with no concomitant increase in [14C]glycine label.     

Recovery of Proteolipid Protein in Mice Heterozygous for the Jimpy Gene

We have measured levels and synthesis of proteolipid protein (PLP) and its transport into myelin in female mice heterozygous for the jimpy gene and in their normal female littermates. In both cord and cerebrum, jimpy carriers show deficits in PLP during development followed by compensation in adulthood. Recovery of PLP occurs earlier in cord than in brain. At 13 days levels of PLP in carriers compared to controls are reduced to 0.60 and 0.44, respectively, in cord and cerebrum. By 100 days, normal levels of PLP are attained in cord (1.13) whereas levels of PLP in cerebrum are only 0.78 of control. By 200 days full recovery occurs in cerebrum, with a ratio of 1.21, suggesting a possible over-compensation. The yield of myelin from cerebrum was reduced to 0.78 in carriers compared to controls at 17 days. In brain slices, incorporation of [3H]leucine into homogenate PLP from carriers is the same as in controls, whereas [3H]leucine incorporation into myelin PLP is reduced to 0.68 of control. These results indicate that synthesis of PLP in the carriers is normal at 17 days, but transport of PLP into myelin is reduced. Similarly, acylation of homogenate PLP is normal, whereas acylation of myelin PLP is reduced, as measured by incorporation of [3H]palmitic acid. Transport of PLP into myelin was compared to transport of MBP; transport of both proteins was equally decreased as indicated by the similar ratio of labeled PLP to MBP in myelin from carriers compared to noncarriers.(ABSTRACT TRUNCATED AT 250 WORDS)     

Autoacylation of myelin proteolipid protein with acyl coenzyme A

Rat brain myelin proteolipid protein (PLP) is known to contain long chain, covalently bound fatty acids. In the course of characterizing the mechanism of acylation, we found that the isolated PLP, in the absence of any membrane fraction, was esterified after incubation with [3H]palmitoyl coenzyme A (CoA). This observation demonstrated that the protein acts as both an acylating enzyme and an acceptor. Thus, acylation occurs by an autocatalytic process. The possibility of a separate acyltransferase that copurifies with PLP was essentially excluded by adding brain subcellular fractions to the reaction mixtures and by changing the isolation procedure. After deacylation, the protein was acylated at a 4-fold greater rate, suggesting that the original sites were reacylated. The palmitoyl-CoA concentration followed Michaelis kinetics, confirming that spontaneous acylation was not occurring. Pulse-chase experiments indicated that the reaction entails net addition of acyl groups. Although fatty acids are bound via an O-ester linkage, free SH groups are required in the reaction. Denaturation of the protein by sodium dodecyl sulfate or heat inhibits the reaction, whereas cerulenin has little or no effect. PO, the major protein in peripheral nerve myelin, is also an acylated protein, but it was not labeled upon incubation of either peripheral myelin or the isolated protein with [3H]palmitoyl-CoA, demonstrating that it is acylated by a different route. Several synthetic peptides derived from PLP sequences with sites known to be acylated in vivo as well as a series of deacylated PLP tryptic peptides were not labeled, indicating that integrity of the protein is required for acylation. Limited proteolysis and peptide mapping showed that the same sites are acylated in vitro or in vivo, suggesting that the autocatalytic acylation reaction is physiological.     

Acylation of myelin proteolipid protein in subcellular fractions of rat brainstem

The acylation of myelin proteolipid protein (PLP) and intermediate protein (IP) was investigated in an in vitro system of tissue slices prepared from actively myelinating rat brainstem. The incorporation of [³H]palmitate into the proteins in nine subcellular fractions including myelin and other cellular membranes which are actively involved in the synthesis and intracellular transport of the proteins was measured. More than 80% of [³H]palmitate-labeled proteins were recovered in myelin. The incorporation was highest in the heavy myelin and lowest in the light myelin subfraction. Appreciable acylation was also detected in the myelin-like fraction. On the other hand, the remaining fractions comprising a variety of endo- and ectomembranes, which harbored over 90% of newly synthesized PLP and IP as seen from [³H]leucine labeling showed practically no [³H]palmitate incorporation. The results indicate that the acylation of PLP and IP is a late event in their posttranslational processing and occurs only at their entry into the myelin sheath.     

Inositol acylation of a potential glycosyl phosphoinositol anchor precursor from yeast requires acyl coenzyme A.

Glycosyl phosphoinositol (GPI) anchors on proteins can be modified by palmitoylation of their inositol residue, which makes such anchors resistant to cleavage by phosphatidylinositol-specific phospholipase C (PI-PLC) (Roberts, W. L., Myher, J. J., Kuksis, A., Low, M. G., and Rosenberry, T.L. (1988) J. Biol. Chem. 263, 18766-18775). Mannosylated GPI lipids made in trypanosomal and mammalian cells can also be inositol-acylated, indicating that inositol acylation may be a normal step in GPI anchor synthesis. We find that Saccharomyces cerevisiae mutants blocked in dolichyl phosphate mannose synthesis accumulate a lipid that can be radiolabeled in vivo with [3H]myo-inositol, [3H]GlcN, and [3H]palmitic acid. This lipid is resistant to PI-PLC, yet sensitive to mild alkaline hydrolysis, and has been characterized as GlcN-phosphatidylinositol (PI), fatty acylated on its inositol residue. When yeast membranes are incubated with UDP-[14C] GlcNAc, 14C-labeled GlcNAc-PI and GlcN-PI are made. Addition of ATP and CoA, or of palmitoyl-CoA to incubations results in the synthesis of [14C]GlcN-(acyl-inositol)PI. This lipid is also made when membranes are incubated with [1-14C]palmitoyl-CoA and UDP-GlcNAc. We propose that acyl CoA is the donor in inositol acylation of GlcN-PI, and that GlcN-(acyl-inositol)PI is an obligatory intermediate in GPI synthesis.