Biosynthesis of l-Alkyl-2-acetyl-sn-glycero-3-phosphocholine (Platelet-activating Factor) from 1-Alkyl-2-acyl-sn-glycero-3-phosphocholine by Rat Alveolar Macrophages PHOSPHOLIPASE AS AND ACETYLTRANSFERASE ACTIVITIES DURING PHAGOCYTOSIS AND IONOPHORE STIMULATION*

l-Alkyl-2-acyl-sn-glycero-3-phosphocholine (alkyl- acyl-GPC) comprises 11% of the total phospholipids of rat alveolar macrophages. This endogenous pool of alkylacyl-GPC was prelabeled by incubating the macrophages with [1,2-3H]alkyllyso-GPC (54 Ci/mmol), which enters the cells and is acylated. The effect of various stimuli on the synthesis and release into the media of labeled alkylacetyl-GPC (platelet-activating factor) from the cells was used to establish the role of inactive alkylacyl-GPC as a precursor of the biologically active derivative. A phagocytic agent (zymosan, 100 pg/ml) and an ionophore (A23187,2 PM) stimulated the release of both alkylacetyl-GPC and alkyllyso-GPC into the media at the expense of cellular alkylacyl-GPC. Phospholipase Az activity (at pH 4.5 and in 1 mM EDTA) was also increased in the media. The stimulatory effect of zymosan and the ionophore on alkylacetyl-GPC re- lease was prevented by mepacrine (0.1 mM), an agent that inhibits the release of fatty acids from phospholip- ids. These data indicate that phospholipase

3 To whom correspondence should be addressed. in phospholipid metabolism in the lung (for reviews, see Refs. 1 and 2). One potentially important aspect of their role may be the biosynthesis and/or catabolism of inflammatory mediators derived from phospholipids. As part of our study of pulmonary phospholipid metabolism, we have investigated the biosynthesis and cellular release of alkylacetyl-GPC,' i.e. platelet-activating factor.
Alkylacetyl-GPC is a biologically active phospholipid with potent platelet-activating and antihypertensive activities (for reviews, see Refs. [3][4][5]. It has also been implicated as a mediator in immune responses including inflammation and anaphylaxis (6). A number of investigators have demonstrated that alkylacetyl-GPC is released by a variety of tissues and cell types, e.g. basophils (7), platelets (8), neutrophils (91, and macrophages (10-12). Specific enzymatic reactions involved in the biosynthesis of alkylacetyl-GPC have been documented in some cell types. These reactions include the formation of the ether bond by alkyldihydroxyacetone-P synthase (13, 14), transfer of phosphocholine to 1-alkyl-2-acetyl-sn-glycerol by a cholinephosphotransferase (14, 15), and acetylation of alkyllyso-GPC by an acetyltransferase (15)(16)(17)(18)(19). It is not known which of these reactions is quantitatively the most important in the biosynthetic pathway(s) leading to the synthesis of alkylacetyl-GPC in macrophages or any other cell. However, it is clear that acetyltransferase is markedly stimulated by agents known to evoke platelet-activating factor responses (18, 19).
In theory, the stimulation of the release of alkylacetyl-GPC could be due to a combined enzymatic process: 1) activation of phospholipase A2, which causes an increase in the concentration of the precursor alkyllyso-GPC; and 2) stimulation of acetyltransferase activity, which catalyzes the acetylation of the lyso precursor. Our approach to investigate this hypothesis for the biosynthesis of alkylacetyl-GPC has been to prelabel the endogenous pool of alkylacyl-GPC by incubating macrophages with 1-[1',2'-'HH]alkyllyso-GPC of high specific activity, and then to determine the effect of various stimuli or inhibitors on the synthesis and release of alkylacetyl-GPC into the media. We demonstrate here that alkylacyl-GPC is an effective precursor of the acetylated bioactive phospholipid and that stimulation of rat alveolar macrophages by zymosan and the ionophore A23187 increases the release of alkylacetyl-GPC and alkyllyso-GPC (derived from alkylacyl-GPC) from the alveolar macrophages. The release is correlated with an increase in intracellular acetyltransferase and an increase in ' The abbreviations used are: alkylacetyl-GPC, I-alkyl-2-acetyl-snglycero-3-phosphocholine; TPA, 12-0-tetradecanoylphorbol-13-acetate; BSA, bovine serum albumin. were incubated for 15 min at 37 "C in a final volume of 1 ml with 30 nmol of hexadecyllyso-GPC, 100 nmol of ["Hlacetyl-CoA (4000 dpm/nmol), and 0.1 mmol of Tris-HCl (pH 6.9). The incubation mixture was extracted and the radioactive products were analyzed by thin layer chromatography.
Thin layer chromatography of the products showed that the major ,'H-and '%-labeled lipid co-chromatographed and had a mobility identical with that of a standard of alkylacetyl-GPC.
Student's t test was used to calculate probabilities of significance.

Phospholipid
Composition-The phospholipid composition of rat alveolar macrophages is given in Table  I. The aikylacyl-GPC fraction represents a substantial proportion (>lO%) of the total phospholipids of these cells and was readily labeled by incubating macrophages with [ 1,2-"HIalkyllyso-GPC ( Fig. 1). After a 4-h incubation, greater than 80% of the radioactivity recovered in the cells was in the alkylacyl-GPC fraction, less than 5% was in the alkyllyso-GPC fraction, and only traces of radioactivity (tO.l%) were in the alkylace- tyl-GPC fraction. The remaining 11-15% of radioactivity recovered in other fractions was not further characterized. These results demonstrate that alveolar macrophages have an efficient mechanism for the uptake and acylation of the alkyl lysophospholipid for storage as an inactive precursor (alkylacyl-GPC) of the bioactive phospholipid (alkylacetyl-GPC).
Cellular Metabolism of Alkylacyl-GPC-Exposure of macrophages labeled with ['H]alkylacyl-GPC to a phagocytic agent (zymosan) or a calcium ionophore (A23187) led to an increase in the appearance of labeled metabolites in the media and a commensurate loss of radioactivity from the cells (Fig.  2 A ) . At the beginning, the rate of loss of radioactivity from the cells was highest in cells exposed to the ionophore. However, after an initial lag of approximately l h, the rate of loss of label from the cells exposed to zymosan also increased Minutes substantially, so that after a 1.5-h incubation, cells exposed to either stimuli had 8-15% less total radioactivity than did the control cells (Fig. 2 A ) . This loss of cellular radioactivity was reflected primarily by a -20% reduction in the radioactivity in the alkylacyl-GPC fraction of phospholipids of cells exposed to the stimuli for 1.5 h (Fig. 2B). In contrast, the rate of loss of radioactivity from the intracellular alkyllyso-GPC fraction was approximately one-fourth that of the alkylacyl-GPC fraction and was not appreciably altered by exposures to zymosan or the ionophore. Thus, the radioactivity released from the stimulated macrophages into the media was derived from the cellular pool of alkylacyl-GPC. Chromatographic analysis of the radioactive phospholipids released into the media demonstrated that ["H]alkylacetyl-GPC accumulated in the media of cells exposed to ionophore and at a slower rate and to a lesser extent than in the media of cells exposed to zymosan (4.0 and 2.5%, respectively, of the radioactivity in the media at 60 min, Fig. 3A). Ionophore and zymosan also stimulated the release of alkyllyso-GPC, the predominant labeled phospholipid in the media (67 and 9076, respectively, of the total at 60 min, Fig. 3B). Phorbol myristate acetate, in contrast to the ionophore and zymosan, had no effect on the quantity of ["Hlalkylacetyl-GPC found in the media (0.5% of the total, Fig. 3A), although it did stimulate the accumulation of alkyllyso-GPC (93% of the total "H, Fig.  3B).
Significant quantities of alkylacyl-GPC (29% of the total at 60 min) were also detected in the media of cells exposed to the ionophore (Fig. 3C). The reason for the appearance of this labeled alkylacyl phospholipid in the media is not clear. It apparently is not due to cell lysis since exposure of macrophages to A23187 did not affect the release of lactate dehydrogenase or acetyltransferase into the media (data not shown).
Phospholipase Az Actiuity-The decrease in label from the intracellular alkylacyl-GPC pool from cells exposed to zymosan or ionophore and the accumulation of labeled alkyllyso-GPC in the media would appear to be the result of a stimulation of phospholipase A2 by these stimuli. Addition of a phospholipase A2 inhibitor, mepacrine, to the incubation media of cells exposed to zymosan or the ionophore prevented the loss of radioactivity from the cellular pool of alkylacyl-GPC (Table 11). Furthermore, the inhibitor also caused a sharp reduction (60%) in the amount of alkylacetyl-GPC accumulated in the media of cells exposed to the ionophore and to a lesser extent (50%) in cells exposed to zymosan (Table  11). Bromophenacyl bromide (0.1 mM), another phospholipase inhibitor, also had a similar effect on the conversion of alkylacyl-GPC to alkylacetyl-GPC (data not shown).
The agents that stimulated the release of alkyllyso-GPC from cells were also tested for their effect on phospholipase AS activity. The A:! activity when assayed at pH 4.5 with 1 mM EDTA in the incubations was approximately 1.4-to 1.9fold higher in the media from the cells exposed to zymosan, ionophore, or TPA than in the media from cells not exposed to the stimuli (Table 111). However, the total activity of the cells and media combined was not significantly increased by the stimuli. We interpret these results to indicate that such agents increase the amount of phospholipase released from the cells into the media rather than directly stimulating phospholipase AB activities. The activit.y of a phospholipase A2 (27) measured a t neutral pH in the presence of calcium in either the media or cells was not substantially altered by the stimuli (Table 111).
In order to gain insight on whether the phospholipase A:, activity affected by zymosan, TPA, or the ionophore was of lysosomal origin, we also examined the effect of these stimuli on the release to the media of acid phosphatase, an enzyme of lysosomal origin (28). The total amount of phosphatase activity recovered was not significantly altered by the zymosan or ionophore treatment. Between 82 and 94% of the total phosphatase activity was intracellular (Table IV). However, after exposure of the cells to zymosan or the ionophore, a significant   TABLE I1 Effect of zymosan, ionophore A23187, a n d mepacrine on intracellular alkylacyl-GPC and the release of alkylacetyl-GPC from alveolar macrophages Cells prelabeled as described in Fig. 2 were incubated for 2 h with BSA/RPMI media or BSA/RPMI media containing either zymosan (200 pg/ml)  increase occurred in the relatively small amount of phosphatase activity released into the media (Table IV). Thus, these agents appear to stimulate the release of lysosomal enzymes from the macrophages. TPA caused a moderate stimulation (3540%) of phosphatase activity in both the cells and media, which suggests that TPA can activate acid phosphatase as well as affect its release from macrophages.
Influence of Acetate on Mepacrine Inhibition of the Zymosan-and Ionophore-stimulated Release of Alkylacetyl-GPC-The effect of acetate on the accumulation of alkylacetyl-GPC in the media of cells exposed to stimuli and mepacrine (Table V)    '' See Table I1 for the inhibitory effects of mepacrine on stimulus.   biosynthesis of the bioactive lipid. Quantitative assessment of the effect by acetate on the accumulation of alkylacetyl-GPC was hampered by a large variation in the magnitude of the responses in different cell populations. However, certain consistent features of these responses to acetate were apparent. Addition of acetate alone was ineffective in stimulating alkylacetyl-GPC accumulation and had only slight effect on the responses in the ionophore-or zymosan-treated cells (Table V). However, the presence of acetate in the media prevented the inhibitory effects of mepacrine on the release of alkylacetyl-GPC from macrophages after stimulation by zymosan and A23187 ( Table V). The extent that alkylacetyl-GPC accumulated in the media of cells exposed to TPA (with and without mepacrine) was not increased by acetate ( Table  V) .
Acetyltransferase Actiuity-Homogenates prepared from cells that had been incubated for 1 h in media containing zymosan or the caIcium ionophore contained a higher level of acetyltransferase activity than homogenates from cells not exposed to the stimuli (Table VI); TPA was relatively ineffective in affecting acetyltransferase activity. In addition to the stimulatory effect of ionophore A23187 on intact cells, the ionophore also directly stimulated acetyltransferase activity in cell homogenates when the assay was done in the absence of added calcium (Table VII). The activity in the homogenate was inhibited by EDTA, whereas the activity was not stimulated above control values by the addition of calcium (Table  VII) .

DISCUSSION
Alveolar macrophages form and release alkylacetyl-GPC and alkyllyso-GPC when exposed to phagocytic or ionophoretic stimuli (Ref. 29 and present study). We interpret that the accumulation of the labeled alkyl lysophospholipid found in the media at the expense of labeled cellular alkylacyl-GPC is due to an increase in the activity of a phospholipase A2. This enzyme appears to be closely coupled with the acetyltransferase activity that is responsible for the formation of alkylacetyl-GPC released from the cells stimulated with zymosan or the ionophore (Fig. 4). The effects of metabolic inhibitors further strengthen the proposed precursor role of the inactive form of the ether phospholipid, alkylacyl-GPC, and the role of phospholipase A2 in the formation of the important lysophospholipid that serves as the substrate for the acetyltransferase in the synthesis of alkylacetyl-GPC. Mepacrine and bromophenacyl bromide prevented both the decrease in radioactivity in cellular alkylacyl-GPC and the accumulation of radiolabeled alkylacetyl-GPC in the media associated with stimulation of macrophages. It is recognized that mepacrine and bromophenacyl bromide may have other actions that could affect the cellular metabolism of phospholipids. These include inhibition of phospholipase C (30) and the capacity of mepacrine to form derivatives with phosphatidylethanolamine (31). However, the ability of these inhibitors to prevent the decrease in the labeled pool of cellular alkylacyl-GPC associated with the accumulation of alkyllyso-GPC, the expected product of a deacylation reaction, is consistent with the ability of mepacrine and bromophenacyl bromide to inhibit fatty acid release via phospholipase A2 in intact cells (30,32,33).
In macrophages, at least two types of phospholipase A2 activities have been described that utilize diacyl-GPC as a substrate: one has a pH optimum of 4.5, does not require added calcium, and is lysosomal; the other has a pH optimum of 8.5, requires added calcium, and is microsomal (27,34,35). In the present investigation when assayed under optimal conditions with 1,2-["H]alkylacyl-GPC as a substrate, both types of phospholipase AZ activities were also detected, although at a substantially lower level than reported for the deacylation of diacyl-GPC labeled with arachidonate at the sn-2 position (27). It is possible that the phospholipase A2 that utilizes alkyl phospholipids as substrates is not the same enzyme that hydrolyzes diacyl-GPC. However, the phospholipase AS activity associated with the alkyllyso-GPC release from macrophages does appear to be of lysosomal origin since agents that stimulated the accumulation of ["Hlalkyllyso-GPC and phospholipase Az in the media also caused a corresponding increase in acid phosphatase activity. The A2 activity released into the media had characteristics similar to those reported for lysosomal phospholipase A2 (27,34,37). Furthermore, the stimulated release of a phospholipase A:! and of the lysosomal enzyme acid phosphatase into the media from cells exposed to zymosan reflects the abiIity of inflammatory agents to stimulate the release of lysosomal enzymes from macrophages (36). Nevertheless, we cannot discount the possibility that other nonlysosomal phospholipases not measurable under our assay conditions are also affected by the stimuli.
Clearly, if alkylacetyl-GPC is derived from alkylacyl-GPC as our results indicate, acetylation of the lysophospholipid intermediate is a necessary step and may represent an important regulatory point in the synthesis of the biologically active phospholipid. Wykle et al. (16) have reported the presence of a specific acetyltransferase in a variety of tissues that is capable of catalyzing the acetylation of alkyllyso-GPC. This activity has also been recently demonstrated in murine peritoneal macrophages (18), and in polymorphonuclear leukocytes (15,19) and eosinophils from humans (19). Moreover, acetyltransferase activity has been shown to be modulated by calcium ionophore A23187 (19) and zymosan (15,18). The mechanism for the in vivo regulation of acetyltransferase activities in rat alveolar macrophages, as well as other cells, has yet to be established. The stimulation of acetyltransferase by a calcium ionophore in intact cells and the inhibition of its activity in homogenates by a calcium chelating agent as observed in the present study suggests a regulatory function of calcium. On the other hand, our observation that the ionophore and not calciumper se stimulated acetyltransferase in homogenates suggests that the ionophore can assert a direct effect on the enzyme rather than just on calcium availability.
The regulatory significance of the acetylation reaction in alveolar macrophages is illustrated by the effect of acetate upon the stimulation of the formation and cellular release of alkylacetyl-GPC by ionophore A23187 under conditions in which deacylation is partially inhibited by mepacrine. The fact that, when acetate is added to the incubation media, the inhibitory effect of mepacrine on the formation of alkylacetyl-GPC is prevented suggests that the rate-limiting step in t,he formation of the bioactive phospholipid from alkylacyl-GPC is the acetylation of alkyllyso-GPC rather than the deacylation of alkylacyl-GPC. Indeed, an increased formation of alkyllyso-GPC without a parallel increase in acetyltransferase activity may not be sufficient to stimulate alkylacetyl-GPC formation. This is apparent with TPA, a membrane perturbant that increases the level of the lysophospholipid precursor, but unlike the zymosan or ionophore does not appreciably stimulate acetyltransferase and, therefore, does not induce an accumulation of alkylacetyl-GPC in the media even in the presence of added acetate.
We conclude that in rat alveolar macrophages both phospholipase Ar and acetyltransferase behave as synergistic enzymes in the synthesis of alkylacetyl-GPC from alkylacyl-GPC, with alkyllyso-GPC as an intermediate (Fig. 4). The acetyltransferase activity would appear to be an important regulatory enzyme in the biosynthesis of platelet-activahg factor.