Purification and Characterization of 1- O -Acylceramide Synthase, a Novel Phospholipase A 2 with Transacylase Activity*

A novel pathway for ceramide metabolism, 1- O -acylce-ramide formation, was previously reported (Abe, A., Shayman, J. A., and Radin, N. S. (1996) J. Biol. Chem. 271, 14383–14389). In this pathway a fatty acid in the sn -2 position of phosphatidylethanolamine or phosphatidylcho-line is transferred to the 1-hydroxyl position of ceramide. An enzyme that catalyzes the esterification of N -acetyl-sphingosine was purified from the postmitochondrial supernatant of calf brain through consecutive steps, including ammonium sulfate fractionation, DEAE-Sephacel, phenyl-Sepharose, S-Sepharose, Sephadex G-75, concanavalin A-agarose, and heparin-Sepharose chromatography. The molecular mass of the enzyme was determined to be 40 kDa by gel filtration on Sephadex G-75. The enzyme bound to concanavalin A-agarose column was eluted with the buffer containing 500 m M a -methyl- D -man-nopyranoside. Further purification by heparin-Sepha-rose chromatography resulted in separation of two peaks of enzyme activity. Coincidence between the transacylase activity and a stained protein of a molecular mass of 40 kDa was observed, as determined by SDS-polyacrylamide gel electrophoresis and recovery after separation over an acidic native gel. The second peak of activity from the heparin-Sepharose chromatography represented a purification of

Recently ceramide has been identified as a candidate lipid mediator regulating cell proliferation, growth, differentiation, development, and apoptosis (1,2). In many studies, proof of concept for a biological role of ceramide has relied on the use of N-acetylsphingosine (NAS), 1 a truncated and membrane-permeable homologue of endogenous ceramides. Reliance on NAS is problematic since this compound is subject to metabolism when exogenously applied to cultured cells. We observed that MDCK cells and NIH 3T3 cells actively metabolized and converted NAS to other sphingolipids and resulted in a considerable increase in levels of sphingosine, natural (long chain) ceramide, and glucosylceramide (3)(4)(5).
Previously reported work on the metabolism of NAS in MDCK cell homogenates revealed the formation of a unique, alkali-labile lipid, 1-O-acylceramide (5). An enzyme activity that catalyzed esterification of a hydroxyl group at C1 position of NAS was described. The enzyme displayed an acidic pH optimum and was mainly recovered in a soluble fraction after centrifugation at 100,000 ϫ g. In the presence of NAS and using liposomes consisting of phosphatidylcholine, [2-14 C]arachidonylphosphatidylethanolamine and sulfatide, the enzyme catalyzed both the transacylation of NAS and the release of arachidonic acid from phosphatidylethanolamine. In the absence of NAS, the release of arachidonic acid was still observed. The enzyme predominantly catalyzes the deacylation or transacylation of both phosphatidylethanolamine and phosphatidylcholine. Additionally, Ca 2ϩ was not required for the enzyme activity. This enzyme appeared to be a novel Ca 2ϩindependent PLA 2 . In the present paper, we report the purification and characterization of this novel acid phospholipase A 2 with transacylase activity from calf brain.

Polyacrylamide Gel Electrophoresis
Polyacrylamide gel electrophoresis in the absence of SDS was performed by the modified method of Reisfeld et al. (7) on 10% acrylamide gels at pH 4.3. Polyacrylamide gel electrophoresis in the presence of SDS was performed by the method of Laemmli and Favre (8) on 12% acrylamide gel. The protein bands were stained with Coomassie Brilliant Blue R-250 or using a silver staining kit (Bio-Rad). The protein content was determined by the bicinchroninic acid protein assay (Pierce) with bovine serum albumin as a standard. The solution with a low concentration of protein was concentrated by the deoxychoratetrichloroacetic acid precipitation method (9) and then assayed.
When the transacylase activity was extracted from a polyacrylamide gel, two gels were loaded with the partially purified enzyme (Step 7, see below) in 0.02% methyl green plus 10% glycerol. Electrophoresis was performed in a cold room. One gel was sliced into 2-mm segments. Each gel was minced by a glass homogenizer with 650 l of 50 mM sodium citrate (pH 4.5) and vortexed overnight at 4°C. The samples were centrifuged for 10 min at 2,000 ϫ g to remove the gel particles. The transacylase activity was measured for each sample using 400 l of supernatant. The other gel was equilibrated for 2 h against 2.3% SDS, 5% 2-mercaptoethanol, 10% glycerol, 62.5 mM Tris-HCl (pH 6.8) and then electrophoresed to a slab polyacrylamide gel (12% separating gel) in the presence of SDS.

Enzyme Assay
The assay conditions are described in the figure and table legends. In general, liposomes consisting of DOPC (70 mol%), PE (30 mol%), and dicetylphosphate (10 mol%) were used as acyl group donors. Constituent lipids in chloroform were mixed and dried under a stream of nitrogen. Fifty mM sodium citrate (pH 4.5) was added to the dried lipids at a volume 1 ml/1.28 mol of lipid phosphorus. The lipids were dispersed into the buffer for 8 min in an ice water bath using a probe sonicator. The donor liposomes (64 nmol of phospholipid) were incubated with 5 nmol of [ 3 H]NAS (10,000 cpm), 5 g of bovine serum albumin and the transacylase activity at 37°C in a total volume of 500 l of 38-40 mM sodium citrate (pH 4.5). The reaction was terminated by adding 3 ml of chloroform/methanol (2/1). The activity was determined as described in the previous paper (5). In the assay using nonradioactive NAS, the lipids were extracted and chromatographied as in the use of radioactive NAS. The activity was determined by image scan analysis of O-acyl-NAS band obtained by charring.

Purification of Transacylase from Calf Brain
Postmitochondrial Supernatant (Step 1)-Fresh calf brains were stored at Ϫ80°C as thin blocks after removal of pia mater. Frozen brains (600 g) were partly allowed to thaw in the cold room overnight and then cut in small pieces, after which all manipulations were performed at 4°C. A 30% homogenate in 10 mM Tris-HCl (pH 7.4 at room temperature), 1 mM EDTA, 0.5 mM PMSF (buffer A) was prepared by a Polytron for 1 min at 20,000 rpm. The mixture was centrifuged at 14,000 ϫ g for 30 min. The pellet was reextracted as described above.
Ammonium Sulfate Precipitation (Step 2)-Ammonium sulfate was added to the postmitochondrial supernatant to 20% saturation (10.6 g/100 ml). After stirring for 1 h, the precipitate was removed by centrifugation at 14,000 ϫ g for 30 min. The supernatant was adjusted to 70% saturation with ammonium sulfate (31.2 g/100 ml) to precipitate the enzyme. After standing overnight, the clear supernatant above the precipitate was siphoned off. The precipitate was collected by centrifugation at 10,000 ϫ g for 30 min, dissolved in 60 ml of buffer A, and dialyzed for 2 days against 5 times 4 liters of 25 mM Tris-HCl (pH 7.4, at room temperature), 1 mM EDTA (buffer B). The protein was adjusted to less than 20 mg/ml with buffer B and centrifuged at 100,000 ϫ g for 1 h.

DEAE-Sephacel Chromatography (
Step 3)-The supernatant was applied to a column (5 ϫ 22 cm) of DEAE-Sephacel preequilibrated with buffer B at a flow rate of 100 ml/h. Unbound protein was eluted with 2 liters of buffer B. The enzyme activity was recovered in the unbound fraction.
Phenyl-Sepharose Chromatography (Step 4)-The unbound fraction was adjusted to 3 M NaCl and then applied to a column (2.5 ϫ 20 cm) of phenyl-Sepharose preequilibrated with buffer B containing 3 M NaCl (buffer C) at a flow rate of 80 ml/h. Unbound material was eluted with 1000 ml of buffer C. Additionally the column was rinsed by 450 ml of buffer B containing 1.5 M NaCl (buffer CЈ) and then by 400 ml of buffer B containing 0.15 M NaCl (buffer D). A small portion of the enzyme activity was released from the column during rinsing by buffer D. Most enzyme activity was eluted by 400 ml of buffer D containing 50% (w/v) ethylene glycol (buffer E) and kept at Ϫ80°C.
The same procedure, steps 1-4, was separately carried out 4 times. Then each fraction kept at Ϫ80°C after the phenyl-Sepharose chromatography step was combined and dialyzed against 10 mM sodium phosphate (pH 6.0), 1 mM EDTA (buffer F).
S-Sepharose Chromatography (Step 5)-The dialysate was applied to a column (1.4 ϫ 20 cm) of S-Sepharose preequilibrated with buffer F at a flow rate of 40 ml/h. Unbound protein was eluted by buffer F. The bound protein was eluted by a 480-ml linear gradient of NaCl from 0 -0.3 M NaCl in buffer F. The enzyme activity was found in fractions between 60 and 130 mM NaCl as two peaks. Each peak was separately collected and named "S-Sepharose fast" and "S-Sepharose slow" by following the order of elution.
Sephadex G-75 Chromatography (Step 6)-Both fractions, S-Sepharose fast and S-Sepharose slow, were concentrated by Centriprep-10 (Amicon) and then applied separately to a column (2.6 ϫ 90 cm) of Sephadex G-75 preequilibrated with 0.15 M NaCl, 25 mM Tris-HCl (pH 7.4), 1 mM EDTA (buffer D) at a flow rate of 40 ml/h. Protein was eluted with buffer D. For S-Sepharose slow, a peak of the enzyme activity appeared at Ve/Vo ϭ 1.41. For S-Sepharose fast, two small peaks of the enzyme activity were observed at Ve/Vo ϭ 1.12 and 1.41. The active fractions obtained from S-Sepharose slow were collected and used for further purification.
Concanavalin A-Agarose Chromatography (Step 7)-The fraction obtained in Step 6 was adjusted to 1 mM CaCl 2 and 1 mM MnCl 2 and applied to a column (0.85 ϫ 1.8 cm) of Con A-agarose preequilibrated in 500 mM NaCl, 25 mM Tris-HCl (pH 7.4), 1 mM CaCl 2 , and 1 mM MnCl 2 (buffer G) at a flow rate 1.5 ml/h. Unbound protein was eluted by 10 ml of buffer G. The column was rinsed by buffer G including 10 mM ␣-methyl-D-mannopyranoside. During the rinse, a small portion of the enzyme activity was released from the column. Most enzyme activity was eluted by 10 ml of buffer G including 500 mM ␣-methyl-D-mannopyranoside (buffer H) and possessed a very high specific activity in the transacylase assay. The active fraction eluted with buffer H was collected, adjusted to 77.3 M of Triton X-100, and then dialyzed against 10 mM Tris-HCl (pH 7.4), 1 mM EDTA, 77.3 M (w/v) Triton X-100 (buffer I).
Heparin-Sepharose Chromatography (Step 8)-The dialysate was applied to a column of heparin-Sepharose (HiTrap heparin, 1 ml; Amersham) preequilibrated in buffer I at a flow rate of 5 ml/h. Unbound protein was eluted by 15 ml of buffer I. Then the column was rinsed by 15 ml of buffer B containing 77.3 M Triton X-100 (buffer BЈ). The bound protein was eluted by a 60-ml linear gradient of NaCl from 0 to 0.15 M in buffer BЈ. Table I summarizes the various steps of the protocol developed for transacylase purification. An eight-step procedure was employed, and the transacylase activity was monitored by the acylation of NAS throughout the purification.

Purification of Transacylase from Calf Brain
In the present study, calf brain was chosen as a starting material. The post-mitochondrial supernatant was precipitated by 20 -70% ammonium sulfate. After dialyzing the suspension of the precipitates, the 100,000 ϫ g supernatant of the dialysate was applied to the DEAE-Sephacel column, and most of the enzyme activity was recovered in the unbound fraction. The recovery rebounded after the DEAE-Sephacel elution. This step gave rise to about 19-fold purification. The unbound fraction was adjusted to 3 M NaCl and applied to the phenyl-Sepharose column. The enzyme activity was recovered by the buffer containing 50% ethylene glycol (Fig. 1). This step gave rise to an additional 7.6-fold purification.
In the next S-Sepharose chromatography, the enzyme activity was eluted in fractions between 60 and 130 mM NaCl as two peaks (Fig. 2). There was no enrichment in the purification of the S-Sepharose fast fraction. On the other hand the S-Sepharose slow fraction was purified 4.1-fold.
Before step 6 (Sephadex G-75 gel filtration), both S-Sepharose fast and S-Sepharose slow fractions were concentrated using Centriprep-10 without loss of enzyme activity. Only a small increase in purity was produced by gel filtration. For S-Sepharose slow, a peak of the enzyme activity appeared at Ve/Vo ϭ 1.41 (Fig. 3B). On the other hand, for S-Sepharose fast two small peaks of the enzyme activity were observed at Ve/ Vo ϭ 1.12 and 1.41 (Fig. 3A). The values of Ve/Vo, 1.12 and 1.41, corresponded to molecular masses of 77.5 kDa and 38.6 kDa, respectively. The active fractions shown in the Fig. 3B were collected and used for further purification.
Most proteins passed through the Con A agarose column (Fig. 4). Buffer containing 500 mM ␣-methyl-D-mannopyranoside was required to elute the enzyme (Fig. 4). This step gave rise to an additional 55-fold purification. Further recovery of the enzyme was accomplished by maintaining the column in buffer H overnight without elution and then rinsing with the same buffer. Because the protein concentration in the active fraction was extremely low, the protein solution was stabilized with 77.3 M Triton X-100. At this stage, the purified transacylase was separated by acidic native gel electrophoresis. The extraction of gel slices and assay of transacylase activity corresponded to a protein of a molecular mass 40 kDa (Fig. 5).
The active fraction obtained from the Con A column was applied to the heparin-Sepharose column and eluted by a linear gradient of 0 -150 mM NaCl. No Con A and Con A fragments were contaminated in the sample after heparin-Sepharose chromatography. The enzyme activity appeared as two peaks between 20 and 50 mM NaCl (Fig. 6). In this single experiment the elution profile of the enzyme activity corresponded to that of protein of a molecular mass 40 kDa, estimated by SDS-PAGE. The lack of a protein band in lane 30 was due to the accidental loss of deoxycholate-trichloroacetic acid precipitate during aspiration. The highly purified enzyme found in the second peak was purified 193,000-fold. The protein appeared as a single band by SDS-PAGE and was used for further characterization of the enzyme. The yield from 10 calf brains (2,400 ϫ g) was 4.5 g with a specific activity of 25 mol/min/mg of protein. The overall recovery of the activity was 1.6%. Transacylase activity was always accompanied with phospholipase A activity throughout the purification steps as determined by thin layer chromatography and charring (for example, Fig. 5).  a Transacylase activity was measured as described under "Experimental Procedures" and as originally described (5). b In order to obtain the enzyme activity, the postmitochondrial supernatant was centrifuged for 60 min at 100,000 ϫ g, and the specific activity was determined using the membrane-free supernatant.
c Fraction S-Sepharose fast was used. d Fraction S-Sepharose slow was used. The active fractions from this Sephadex G-75 chromatography step was pursued for further purification.
Characterization of Transacylase pH Optimum-The purified transacylase was optimally active at pH 4.5 (Fig. 7). Unlike the homogenate obtained from MDCK cells, the purified enzyme still retained the activity at pH 5-6.5 (5). The enzyme activity was totally abolished at pH greater than 7.6.
Effect of Ca 2ϩ -The enzyme activity was slightly inhibited in the presence of 5 mM EDTA. One and 10 mM Ca 2ϩ enhanced the activity by 48 and 27%, respectively (Table II). A similar effect was observed in the presence of Mg 2ϩ , in which 1 and 10 mM Mg 2ϩ enhanced the activity by 32% and 43%, respectively.
Effect of Other Agents-Nucleotides, in particular ATP, have been reported to stabilize and activate a cytosolic Ca 2ϩ -independent PLA 2 (10, 11). The transacylase activity was slightly inhibited in the presence of 2 mM ATP (Table II). Neither the thiol compound, DTT, nor the thiol reagent, NEM, had significant effects on the enzyme activity (Table II). The transacylase activity was slightly inhibited by the pretreatment of the enzyme with 1 mM PMSF (Table II). Both BEL, a potent inhibitor of cytosolic Ca 2ϩ -independent PLA 2 (iPLA 2 ), and AACOCF 3 , a potent inhibitor of cytosolic PLA 2 (cPLA 2 ), slightly inhibited the enzyme activity (Table II and Fig. 8).
In the last purification step, 77.3 M Triton X-100 was used to stabilize the enzyme. Using the active fraction of Sephadex G-75 (from S-Sepharose slow), it was found that there was no significant difference in the enzyme activity between the fractions in the presence and in the absence of 77.3 M Triton X-100 if the final concentration of Triton X-100 in the assay system was reduced less than 773 nM. However, the transacylase activity was strongly affected in the presence of higher concentrations of Triton X-100. The activity completely disappeared in the presence of greater than 773 M Triton X-100 (Fig. 9).
In a preliminary test, it was found that the enzyme activity in the active fraction of Sephadex G-75 eluate (from S-Sepharose slow) was partially inhibited by 100 g/ml heparin. As already shown in Fig. 6, the enzyme was bound to heparin-Sepharose resin. Fifty percent inhibition of the enzyme activity was observed in the presence of 10 -100 g/ml heparin (Table II).
Dual Enzyme Activity, Transacylase and Phospholipase A 2 -As noted above, the transacylase was always accompanied with phospholipase A activity. To confirm the existence of both transacylase and phospholipase A activities, L-␣-1-acyl-2-[1-14 C]arachidonoyl-PE was employed as an acyl donor. The enzyme acted as not only a transacylase but also as a PLA 2 in the presence of NAS and sn-2-radiolabeled PE (Fig. 10). On the other hand, the enzyme showed only PLA 2 activity in the absence of NAS as an acceptor. Interestingly, the total activity of the transacylase and PLA 2 in the presence of NAS was exactly equal to PLA 2 activity in the absence of NAS. Only 7% of total radiolabeled products was 2-[1-14 C]arachidonoyllysophosphatidylethanolamine after 30 min incubation in the presence or absence of NAS. These data are consistent with a high degree of specificity for the enzyme in using the sn-2 fatty acid of PE. 2

DISCUSSION
Our previous study demonstrated that the transacylase activity of brain is higher than that of other tissues (5). Therefore brain was chosen as a starting material. In a preliminary study, it was confirmed that the transacylase activity is efficiently released by sonication from MDCK cell homogenate prepared in isotonic buffer, consistent with its localization in an intracellular compartment. In the present study, fresh calf brains were frozen once before use. The thawed tissues were homogenized with a hypotonic buffer using a Polytron. This procedure presumably lysed intracellular compartments efficiently and resulted in an increased release of the enzyme. However, a very low recovery (20 -30%) of enzyme activity was initially obtained after ammonium sulfate fractionation. Recovery was markedly improved by diluting the dialysate before the 100,000 ϫ g centrifugation. A high protein concentration might have caused a nonspecific coprecipitation of the enzyme with insoluble materials during ultracentrifugation. The observed rebound of enzyme activity after DEAE-Sephacel chromatog-raphy might have been due to separation of the enzyme from an endogenous inhibitor.
The concanavalin A-agarose chromatography step was one of the most successful purification steps. The enzyme was tightly bound to concanavalin A-agarose resin and specifically released from the resin with 500 mM ␣-methyl-D-mannopyranoside (Fig. 4). At this stage, a major band of molecular mass about 40 kDa and some minor bands of that between 43 and 60 kDa were observed in the active fractions from the concanavalin A column by SDS-PAGE (Fig. 5). Further purification by the heparin-Sepharose column showed an exact coincidence between the transacylase activity and a stained protein of a molecular mass of about 40 kDa as determined by SDS-PAGE (Fig. 6), consistent with the molecular mass assignment by Sephadex G-75 chromatography (Fig. 3B). These data are consistent with the purified transacylase being a 40-kDa oligomannose-containing glycoprotein with a single polypeptide chain. This enzyme has a pH optimum of 4.5. Although suggestive of a lysosomal origin for the transacylase, the enzyme activity was still observed at pH 5-6.5 (Fig. 6). Thus the enzyme may function outside the acidic environment of lysosomes.
Secreted PLA 2 s, group I, II, and III, contain 5-8 disulfide bonds essential for the enzyme activity (13,14). Reducing these disulfide bonds by thiol reagents results in inhibition of the lipase activity. The calf brain transacylase activity was insensitive to DTT when preincubated at pH 7.4 in Tris buffer or when added directly to the assay mixture (Table II). Using a partially purified enzyme (step 7), it was found that the enzyme was tightly bound to organomercurial-agarose resin (Affi-Gel 501; Bio-Rad) and eluted by 200 mM 2-mercaptoethanol (data not shown). This indicates that the enzyme has free thiol group(s) exposed on its molecular surface. Pretreatment of the purified enzyme with NEM had no effect on the enzyme activity (Table II). These results suggest that the free thiol group(s) of the enzyme that is thiol reagent-sensitive is neither functionally important nor found in the catalytic site.
Recently human cPLA 2 was found to have a transacylase activity that is CoA-independent, raising the possibility that cPLA 2 forms an acyl-enzyme intermediate (15). Site-directed mutagenesis of cPLA 2 demonstrated that the serine 228 essential for the catalytic activity of the 85-kDa cPLA 2 may be the  active site nucleophile. (17). Additionally, it was reported that AACOCF 3 is both a slow and tight binding inhibitor of cPLA 2 that may form a hemiketal enzyme-inhibitor complex (16,18). The reaction mechanism of cPLA 2 thus might be similar to that of serine protease, esterase, or lipase (15)(16)(17)(18). In general, these types of enzymes are also sensitive to PMSF and diisopropylfluoride. In the present study, neither 1 mM PMSF nor 28 M AACOCF 3 significantly inhibited the activity of the purified transacylase (Table II and Fig. 8). The action of these inhibitors on the enzyme in 100,000 ϫ g supernatant from MDCK cells homogenate was also not significant (5). These results suggest that neither the calf brain transacylase nor MDCK cell transacylase has an essential serine residue for the transacylase reaction.
Although the purified transacylase did not require Ca 2ϩ or Mg 2ϩ for activity, these cations enhanced the activity (Table  II). In the previous study, it was observed that 1 mM Ca 2ϩ slightly enhanced the transacylase activity in the 100,000 ϫ g supernatant obtained from MDCK cells homogenate (5). It is known that divalent cations affect the physical state of lipid assembly (12). The enhancement of the enzyme activity is thought to be due to formation of ionic bridges by these divalent cations on liposomes, thereby modulating molecular packing and availability of substrate.
Recently, it has been reported that myocardial cytosolic iPLA 2 and macrophage cytosolic iPLA 2 interact with ATP, acting on the enzyme as an activator or a stabilizer (10,11). The purified calf brain transacylase was not activated in the presence of ATP (Table II). It has also been reported that BEL is a potent covalently bound inhibitor of these iPLA 2 s but not other PLA 2 s (secreted Ca 2ϩ -dependent PLA 2 or cytosolic Ca 2ϩ -dependent PLA 2 ) (19,20). Myocardial iPLA 2 activity was completely abolished by pre-incubation of the iPLA 2 with 1 M BEL (19), and half-maximal inhibition of macrophage iPLA 2 activity was observed at 60 nM after a 5 min preincubation (20). Preincubation of the calf brain transacylase with 3 M BEL had no significant effect on the enzyme activity (Table II). These results suggest that a functional group sensitive to BEL inhibitors does not exist around the active site and/or lipid recognition site of the calf brain transacylase. Bartolf and Franson (22) reported that a lysosomal iPLA 2 from bovine adrenal medulla was stabilized by adding Triton X-100, although the lipase activity was reduced in a dose-dependent manner at higher Triton X-100 concentrations (22). A similar effect was observed on the purified calf brain transacylase, where activity was completely abolished in the presence of more than 773 M Triton X-100 (Fig. 9). On the contrary, iPLA 2 (39 kDa) partially purified from bovine brain has the optimal concentration of Triton X-100 at 773-1546 M, suggesting that the transacylase is distinct from the 39-kDa iPLA 2 (27).
Using iPLA 2 (39 kDa) partially purified from bovine brain, Yang et al. (21) reported that heparin at either 100 or 200 g/ml produced 25% inhibition of the iPLA 2 activity. The calf brain transacylase was bound to heparin-Sepharose resin under a low ionic condition and eluted by a low concentration of NaCl consistent with weak ionic binding (Fig. 6). In the present study, 50% inhibition of the transacylase activity occurred at 10 -100 g/ml heparin (Table II). The interaction between the enzyme and heparin might be very weak in the assay system because of the ionic strength. Alternatively, heparin is thought to interact with liposomes and to change the surface properties and the molecular packing of liposomes. As a result, an apparent activity of the enzyme might be partially suppressed. iPLA 2 s have been found in many mammalian tissues and cells (23), and some successful purification of iPLA 2 s has been reported, including canine myocardium (10), P388D 1 macro-phages (11), and rat lung (24), brain (25), and kidney (26). Some biochemical properties of the calf brain transacylase, including insensitivity to DTT and Ca 2ϩ independence for enzyme activity, are found among iPLA 2 s. However, many of iPLA 2 s have a neutral pH optimum. For example, although two kinds of cytosolic iPLA 2 , 110 kDa and 39 kDa, were partially purified from bovine brain, the optimal pH values of 110-kDa and 39-kDa enzymes were 7.4 and 8.0, respectively (27). The purified rat kidney soluble iPLA 2 has a molecular mass of 28 kDa and a neutral pH optimum (26). Although a soluble acidic pH-optimum iPLA 2 was purified from rat brain by a protocol similar to ours, the enzyme displayed a molecular mass of 58 kDa and was highly phosphatidic acid-selective (25). Recently Wang et al. (24) purified an acidic-pH optimum iPLA 2 from rat lung. This enzyme displayed some properties similar to the calf brain transacylase, including insensitivity to 2-mercaptoethanol, ATP, AACOCF 3 , and the binding ability to heparin-Sepharose resin. However, the molecular mass of the lung iPLA 2 was 15 kDa (24). A soluble, lysosomal iPLA 2 partially purified from bovine adrenal medulla showed a basic pI, Triton X-100 inhibition, and binding to concanavalin A-agarose resin (22). These profiles are common to the calf brain transacylase. This enzyme, however, gave a molecular mass of 30.6 kDa.
More recently, a calcium-independent lysosomal PLA 2 has been cloned. This enzyme displayed a similar pharmacologic profile to the 15-kDa protein but has a molecular mass of 25.8 kDa (28).
In the previous study, it was strongly suggested that MDCK cell transacylase has dual enzyme activity, i.e. transacylase and PLA 2 (5). In the present study, both transacylase and phospholipase A activities were observed in the purified protein. The assay in which DOPC/L-␣-1-acyl-2-[1-14 C]arachidonoyl-PE/dicetylphosphate liposomes was used as an acyl donor demonstrated that the calf brain transacylase possesses dual enzyme functions, viz. transacylase and phospholipase A activities (Fig. 10). The enzyme also produced 2-[ 14 C]arachidonyllyso-PE but only as a small fraction of the total radioactive products. This supports the view that the phospholipase activity of the transacylase has sn-2 fatty acid specificity. These data are consistent with the interpretation that 1-O-acylceramide synthase is a novel Ca 2ϩ -independent PLA 2 .