Purification and Characterization of Canine Myocardial Cytosolic Phospholipase A2 A CALCIUM-INDEPENDENT PHOSPHOLIPASE WITH ABSOLUTE w-2 REGIOSPECIFICITY FOR DIRADYL GLYCEROPHOSPHOLIPIDS*

Recently, we identified a novel calcium-independent, plasmalogen-selective phospholipase A2 activity in canine myocardial cytosol which represents the major measurable phospholipase A2 activity in myocardial homogenates (Wolf, R. A., and Gross, R. W. (1985) J. Biol. Chem. 260, 7295-7303). We now report the 154,000-fold purification of this phospholipase A2 to homogeneity through utilization of sequential anion exchange, chromatofocusing, affinity, Mono Q, and hydroxylapatite chromatographies. The purified enzyme had a molecular mass of 40 kDa, possessed a specific activity of 227 mumol/mg min, had a pH optimum of 6.4, and catalyzed the regiospecific cleavage of the sn-2 fatty acid from diradyl glycerophospholipids. The purified polypeptide was remarkable for its ability to selectively hydrolyze plasmenylcholine in homogeneous vesicles (subclass rank order: plasmenylcholine greater than alkyl-ether choline glycerophospholipid greater than phosphatidylcholine) as well as in mixed bilayers comprised of equimolar plasmenylcholine/phosphatidylcholine. Purified myocardial phospholipase A2 also possessed selectivity for hydrolysis of phospholipids containing arachidonic acid at the sn-2 position in comparison to oleic or palmitic acid. Taken together, these results constitute the first purification of a calcium-independent phospholipase with absolute regiospecificity for cleavage of the sn-2 acyl linkage in diradyl glycerophospholipids and demonstrate that myocardial phospholipase A2 has kinetic characteristics which are anticipated to result in the selective hydrolysis of sarcolemmal phospholipids during myocardial ischemia.


Lysophospholipids
are potent membrane perturbing metabolites which alter the dynamics of myocardial sarcolemmal membranes (9) and precipitate electrophysiologic alterations in vitro which are indistinguishable from those present during myocardial ischemia in uiuo (10, 11). Accordingly, we and others have suggested that activation of phospholipase A2 and the resultant accumulation of lysophospholipids is intimately related to the development of electrophysiologic dysfunction in ischemic myocardium.
Myocardial sarcolemma is predominantly comprised of plasmalogen molecular species (12,13), and the sarcolemmal membrane is the primary target of accelerated phospholipid catabolism in myocytes subjected to simulated ischemia (14). In previous studies we demonstrated that the major measurable phospholipase AP activity in canine myocardium is calcium-independent and has direct physical access to the sarcolemmal membrane (15). Since accelerated sarcolemmal phospholipid catabolism has been implicated as the biochemical mechanism underlying electrophysiologic dysfunction and myocytic cell death during myocardial ischemia, the purification and characterization of this calcium-independent phospholipase AZ is of obvious importance. We now report the 154,000-fold purification of canine myocardial cytosolic phospholipase AZ to homogeneity and demonstrate that the purified enzyme has kinetic properties which make it the likely enzymic mediator of accelerated sarcolemmal phospholipid catabolism during myocardial ischemia. Ai! Activity to Chemical Modification-The ATP affinity column eluate (5 pa) was incubated with either 1 mM dithiobisnitcobenzoic acid, 1 rni pacabcomophenacylbcomide, or 10 FM phenylmethylsulfonyl fluoride, dialyzed against buffer 3 and subsequently assayed as described above. Thermal Denaturation-The purified protein was incubated at 37 "C oc at 60 "C in assay buffer for various times prior to addition of neat saturating concentrations of substrate (3 X K,). After an additional 1 min incubation, products were extracted with butanol, separated by thin layer chromatography and quantified as described above. Zodination

Characterization
of Crude Myocardial Cytosolic Phospholipase AP Activity-As previously demonstrated (15), the major measurable phospholipase AZ activity in canine myocardium was present in the cytosolic fraction and manifest maximal enzymic activity in the presence of the calcium chelator EGTA. No calcium-independent hydrolysis of plasmenylcholine substrate could be detected in homogenates of whole blood or plasma. The release of fatty acid from the sn-2 position of plasmalogen substrate by the cytosolic enzyme was catalyzed by phospholipase AZ since inclusion of an excess of lysophospholipid, diacylglycerol, 1-0-alk-l'-enyl-2-acyl-snglycerol or phosphatidic acid did not significantly attenuate the rate of fatty acid release from radiolabeled plasmenylcholine substrate. Kinetic analyses of the cytosolic fraction utilizing several synthetic sn-2-radiolabeled diacyl, alkyl-acyl, and vinyl-ether choline glycerophospholipid molecular species (Table I) confirm and extend our previous report (15) that the major phospholipase AZ activity in myocardium selectively hydrolyzes ether-linked choline glycerophospholipids. Furthermore, the present results indicate that cytosol contains a calcium-independent phospholipase AZ activity which preferentially hydrolyzes choline glycerophospholipids containing arachidonic acid at the sn-2 position.
Purification of Canine Myocardial Cytosolic Calcium-independent Phospholipase AZ-To characterize the polypeptide(s) responsible for the observed calcium-independent phospholipase AZ activity, canine myocardial cytosolic phospholipase AZ was purified to homogeneity by sequential anion exchange, chromatofocusing, affinity, FPLC-anion exchange, and HPLC-hydroxylapatite chromatographies. First, dialyzed cytosol was applied to a DEAE-Sephacel column, and phospholipase AP activity was quantitatively eluted by application of a 100 mM NaCl stepwise gradient. The active fractions were pooled, dialyzed, and loaded onto a previously equilibrated chromatofocusing column as described under "Experimental Procedures." Phospholipase AZ activity was eluted by the generation of a shallow pH gradient which resulted in a sharply focused peak of activity with an apparent isoelectric point of 7.55 (Fig. 1). This step typically resulted in a 70-100fold purification of myocardial phospholipase AP activity (Table II).
Since initial studies identified the potential association of ATP with myocardial phospholipase A2 (34), further purification was accomplished by exploiting the interaction of myocardial phospholipase AZ with an ATP-agarose affinity matrix. When active fractions from the chromatofocusing column were applied to an ATP-agarose affinity column, phospholipase A, activity was quantitatively and selectively adsorbed (over 99% of other proteins present in the load eluted in the void volume which was devoid of phospholipase activity). The specificity of the interaction between myocardial phospholipase A, and the ATP matrix was further exploited through utilization of sequential washes of the affinity matrix with 10 mM adenosine and 10 mM AMP (which removed the majority of bound protein but did not elute substantive phospholipase A2 activity). Enzymic activity was quantitatively eluted from the ATP-agarose matrix with 1 mM ATP (Fig. 2). Use of this nucleotide affinity matrix resulted in a 150-fold purification of myocardial phospholipase AZ in quantitative yield accompanied by a 50-fold reduction in volume. Thus, this 3-day procedure results in a 52,000fold purification of myocardial phospholipase AY activity in 86% yield which is moderately stable when stored at O-4 "C (ts = 5-7 d).
FIG. 1. Chromatofocusing of myocardial cytosolic phospholipase AZ. The eluate from the DEAE-Sephacel column was dialyzed, applied to a chromatofocusing column, and phospholipase A, activity was focused by development of a shallow pH gradient as described under "Experimental Procedures." Aliquots of column eluates were assayed by quantifying radiolabeled fatty acid release from l-O-(Z)hexadec-l'-enyl-2-[9,10-3H]octadec-9'-enoyl-GPC (0) as described under "Experimental Procedures." -, ultraviolet absorbance at 280 nm; W, pH. As. Active fractions from chromatofocusing were immediately applied to a previously equilibrated ATP-agarose column. After loading, the column was washed with equilibration Phospholipase AZ was further purified by application of the ATP-agarose eluate onto an FPLC-Mono & anion exchange column which was subsequently eluted utilizing a shallow discontinuous NaCl gradient (Fig. 3). Mono Q active fractions were directly loaded onto an HPLC-hydroxylapatite column, and phospholipase Ar activity was eluted with a nonlinear K[PO,] gradient as described under "Experimental Procedures" (Fig. 4). Since the purified enzyme was extremely labile (t!,, g 30 min at 4 "C), assays of enzymic activity following hydroxylapatite chromatography were performed directly after elution of each fraction. Collectively, this series of column chromatographic steps resulted in a 154,000-fold purification of canine myocardial cytosolic phospholipase AP to a specific activity of 227 pmol/mg min with an overall yield of 19% (Table II).
Purity of Myocardial Phospholipase AZ after Column Chiomatography-To assess the purity of myocardial phospholipase A2 after sequential column chromatographies, the active fractions from the hydroxylapatite column were iodinated with Bolton-Hunter reagent, separated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and protein was visualized by autoradiography.
Only a single intense band at 40 kDa was observed in the most active fraction (Fig. 5 phospholipase Al activity from multiple a&amide-based gel electrophoresis systems utilizing either pulverized, extracted, or electroeluted gel slices have failed. In fact, incubation of enzyme with even minute amounts of polymerized acrylamide results in complete and unrecoverable loss of all enzymic activity.

Characterization of Myocardial Phaspholipase AP Binding to
Nucleotide Affinity Matrices-The specificity of the interaction responsible for the adsorption of calcium-independent phospholipase A:! to immobilized nucleotide affinity matrices was examined to gain insight into the chemical interactions contributing to the association of ATP with this phospholipase. Of the three ATP resins tested (see "Experimental Procedures"), coupling via the M-amino group provided the highest yield. Attachment through the C-8 or the ribose hydroxyl groups resulted in recovery of 60-80% of loaded enzymic activity in the ATP wash with the majority of remaining activity present in the void volume. Other matrices such as GTP-agarose, UTP-agarose, ADP-agarose, CoA-agarose as well as AMP-agarose all bound myocardial phospholipase to varying extents in the specified rank order (strongest-weakest, 60-10% binding). In contrast, D-ribose-5'-phosphate-agarose did not bind canine myocardial phospholipase A2 activity. Although Blue Sepharose (CL-6B) quantitatively adsorbed enzymic activity (no activity was present in the void volume), recovery of phospholipase activity after elution with buffer containing ATP, ATP and 1 M NaCl or ATP, and 1 M K[POJ was poor (C5%). With the exception of Blue Sepharose (which nonspecifically adsorbed approximately 50% of the loaded proteins), greater than 99% of loaded proteins did not bind to these affinity matrices under the conditions employed. Furthermore, although classic calcium-dependent, low molecular weight phospholipases AZ are known to bind to the nucleotide analog dye Cibacron Blue FBGA (35), none of the phospholipases A, examined (i.e. Naja naja, pancreatic, bee venom, platelet cytosolic) adsorbed to the ATP resins used.

Kinetic Analyses of Purified Myocardial
Phospholipase AZ-- The homogeneous polypeptide exhibited maximal enzymic activity in the presence of EGTA and possessed a pH optimum of 6.4 for each phospholipid substrate examined. Incubation of the purified enzyme with sn-2-radiolabeled phospholipid (e.g. plasmenylcholine, phosphatidylcholine, or phosphatidylethanolamine molecular species) resulted in the release of radiolabeled fatty acid with no observable radioactivity in lysophospholipid, diradylglycerol, or phosphatidic acid. The possibility that the release of sn-2 fatty acid from diradyl glycerophospholipids occurred by sequential phospholipase A1 and lysophospholipase activities was eliminated by multiple independent techniques. First, myocardial phospholipase AP was incubated with l-30 gM [3H-Me]choline-labeled DPPC, and the reaction products were isolated and quantified as described under "Experimental Procedures." For each concentration of substrate examined, the loss of PC and the accumulation of LPC were stoichiometric (Fig. 6) with no detectable radiolabel in GPC. Second, when sn-2-3H-labeled DPPC was utilized as substrate under identical assay conditions, the resultant increase in 3H-fatty acid equalled (*3%, n = 2) the increase in t3H-Me]LPC at each concentration examined (Fig.   6). Third, incubation of l-[ l-'4C]palmitoyl-2-palmitoyl-GPE with purified enzyme resulted in the production of l-[l-i4C] palmitoyl-LPE without measurable amounts of radiolabeled palmitic acid, and the mass of phosphatidylethanolamine hydrolyzed was quantitatively accounted for by the mass of l-acyl LPE produced (Fig. 6). Furthermore, no [1-14C]palmitate was released from l-[l-'4C]palmitoyl-2-palmitoyl-GPE in the presence of several detergents (i.e. Triton X-100, n-octyl glucoside, Lubrol-PX, or Tween-20). Finally, loo-fold molar excesses of LPC, diacylglycerol, and PA did not substantially diminish release of 3H-fatty acid from l-O-(Z)-hexadec-l'enyl-2-[9,10-3H]octadec-9'-enoyl-GPC.
Thus, myocardial cytosolic phospholipase Az is specific for hydrolysis of the sn-2 ester linkage in choline and ethanolamine diradyl glycerophospholipids and is devoid of measurable phospholipase Al, C, or D activities. Attempts to demonstrate significant reversibility of the reaction by incubation of purified enzyme with lysophospholipid and radiolabeled fatty acid (in the absence or presence of CoA) were unsuccessful.
Characterization of the phospholipid substrate specificity of purified myocardial cytosolic phospholipase A:! was performed by kinetic analyses of the ATP eluent (52,000-fold purified, specific activity = 76 pmol/mg min) since the marked lability of Mono Q or hydroxylapatite eluents precluded their use. Examination of the choline glycerophospholipid subclass specificity of the 52,000-fold purified enzyme revealed that hydrolysis of plasmenylcholine substrate was more rapid than hydrolysis of alkyl-ether choline glycerophospholipid or phosphatidylcholine (Fig. 7, Table III). Comparisons of phospholipase AS activity utilizing phosphatidylcholine molecular species containing palmitate at the sn-1 position and either palmitic, oleic, or arachidonic acid at the sn-2 position as substrates demonstrated a rank order preference for cleavage of arachidonate > oleate > palmitate (Fig. 7, Table III). Furthermore, substantial enzymic activity required the presence of a long chain acyl group at the sn-2 position since PAF  was hydrolyzed three orders of magnitude more slowly than 1-0-hexadecyl-2-arachidonyl-GPC.
Since previous work has demonstrated that plasmenylcholine and phosphatidylcholine bilayers possess distinct molecular dynamics (36), additional experiments were performed to examine the substrate specificity of myocardial phospholipase A2 in systems which minimize differences in the physical properties of aggregated substrate. In initial experiments, we prepared mixed micelles of phospholipids with selected detergents (e.g. Triton X-100, Tween-20, n-octyl glucoside, Nonidet P-40, CHAPS, Lubrol-PX, Brij-35, deoxycholate, and taurocholate) to compare hydrolytic rates for each choline glycerophospholipid subclass in identical microenvironments. Unfortunately, myocardial phospholipase A2 activity was completely abolished by each of these detergents. To circumvent this difficulty, additional experiments employing binary mixtures of plasmenylcholine and phosphatidylcholine in mixed bilayers were performed (Table IV). Incubation of vesicles comprised of 50 mol% plasmenylcholine and 50 mol% phosphatidylcholine with purified enzyme resulted in the  -enyl-2-[5,6,8,9,11,12,-14,15-3H]arachidonyl-GPC; phosphatidylcholine = 1-palmitoyl-2arachidonyl-GPC; [3H]phosphatidylcholine = 1-palmitoyl-2- [5,6, 8,9,11,12,14,15-3H] (Table IV) demonstrating that the observed subclass selectivity of myocardial phospholipase AZ is independent of alterations in the physical properties and interfacial characteristics of aggregated substrate. To compare hydrolysis of each phospholipid subclass in a microenvironment possessing physical properties and interfacial characteristics of its phospholipid subclass counterpart, binary mixtures comprised of 10 mol% [3H]plasmenylcholine in phosphatidylcholine bilayers or 10 mol% [3H] phosphatidylcholine in plasmenylcholine bilayers were prepared. Purified myocardial phospholipase AP efficiently catalyzed the hydrolysis of plasmenylcholine when the physical characteristics of the vesicles were largely those of phosphatidylcholine.
In contrast, phosphatidylcholine was not substantially hydrolyzed even when present in vesicles possessing the physical properties of the preferred substrate in homogeneous systems (i.e. plasmenylcholine) (Fig. 7). Since the purified enzyme selectively hydrolyzed plasmenylcholine in 1) homogeneous systems, 2) equimolar mixtures of plasmenylcholine/phosphatidylcholine, and 3) vesicles whose physical properties resemble those of phosphatidylcholine, these results demonstrate that myocardial phospholipase AP selectively hydrolyzes arachidonylated plasmenylcholine in physiologically relevant matrices.
To further investigate the diversity of the substrate specificity of purified cytosolic myocardial phospholipase A*, a battery of lipids was examined.
To examine the potential physiologic relevance of lysophosphatidylcholine hydrolysis catalyzed by myocardial cy- tosolic phospholipase Aa, additional studies were performed. When bilayers containing 9 mol% lysophosphatidylcholine (5 pM l-[l-'4C]palmitoyl-LPC in 50 pM unlabeled l-O-(Z)-hexadec-l'-enyl-2-octadec-9'-enoyl-GPC) were incubated with purified myocardial cytosolic phospholipase AZ, no radiolabeled fatty acid was released from LPC even though over 10% of plasmenylcholine was hydrolyzed. Similarly, since the loss of DPPE and DPPC and the accumulation of LPE and LPC were stoichiometric (Fig. 6), measurable amounts of lysophospholipid hydrolysis did not occur. Thus, under physiologically relevant conditions, myocardial cytosolic phospholipase A:! hydrolyzes endogenous phospholipids to 1-acyl lysophospholipids and does not act effectively as a lysophospholipase.

DISCUSSION
The results contained herein constitute the first purification of a calcium-independent phospholipase activity which has absolute regiospecificity for cleavage of the sn-2 acyl linkage in diradyl glycerophospholipids.
Although other calcium-independent phospholipases have previously been , detailed kinetic analyses have demonstrated that these phospholipases either specifically catalyze hydrolysis at the sn-1 position or indiscriminately hydrolyze acyl groups at both the sn-1 and sn-2 positions. Since phospholipase A, activity was not present utilizing multiple diradyl glycerophospholipid substrates in different physical states, these results demonstrate the absolute regiospecificity of myocardial cytosolic phospholipase AP and identify this phospholipase as the first regiospecific calcium-independent phospholipase AP purified to date. Myocardial cytosolic calcium-independent phospholipase A2 is the major measurable phospholipase activity in myocardium and is a low abundance, high specific activity polypeptide which required a 154,000-fold purification to reach homogeneity. This degree of purification was facilitated by the unique, highly selective, and reversible adsorption of myocardial cytosolic phospholipase AP to ATP-agarose resin. The purity of the preparation was demonstrated by the presence of a single 40-kDa protein band visualized by the highly sensitive method of lZ51 autoradiography.
Although attempts at obtaining phospholipase activity after polyacrylamide gel electrophoresis have failed (the enzyme is irreversibly inactivated by acrylamide), the high sensitivity and dynamic range of the visualization method employed, the high specific activity of the purified polypeptide (230 wmol/mg.min), as well as the concordant appearance and disappearance of 40-kDa mass with phospholipase activity, collectively demonstrate that the 40-kDa polypeptide catalyzes phospholipase Al activity. Kinetic analyses demonstrated several novel features of the purified protein.
Myocardial phospholipase AZ is the first purified calcium-independent phospholipase AZ which selectively hydrolyzes plasmalogen substrate and arachidonylated glycerophospholipids.
Remarkably, the purified polypeptide also contained intrinsic lysophospholipase and palmitoyl-CoA hydrolase activities, albeit at rates two to three orders of magnitude less than its phospholipase AZ activity. The conclusion that phospholipase AS, lysophospholipase, and palmitoyl-CoA hydrolase activities are catalyzed by a single polypeptide is substantiated by the coelution of each activity through multiple chromatographic steps to a single polypeptide, similar sensitivities of each activity to divalent cations and thiol oxidizing agents, and identical thermal denaturation profiles of each activity at different temperatures.
The possibility that phospholipase AZ, lysophospholipase, and palmitoyl-CoA hydrolase activities are catalyzed by highly homologous yet distinct polypeptides of nearly identical molecular mass which copurify over 154,000-fold cannot be definitively excluded but seems unlikely. Parenthetically, we note that venom phospholipase AZ (the paradigm of sn-2 regiospecificity) possesses minute levels of lysophospholipase activity (12). The highly regiospecific phospholipolysis catalyzed by the venom phospholipase A2 and myocardial cytosolic phospholipase AZ are in stark contrast to the lack of regiospecificity of the previously isolated 98" kDa calcium-independent phospholipase in guinea pig intestinal mucosa which possessed nearly identical phospholipase Al, API and lysophospholipase activities (40). Although the 40-kDa polypeptide is the major measurable phospholipase Az in myocardium, its lysophospholipase and palmitoyl-CoA hydrolase activities comprise only a small fraction of the total lysophospholipase and palmitoyl-CoA hydrolase activities in myocardium (4,19,21,43,44). Accordingly, based upon in vitro kinetic measurements with the purified protein as well as measurements of activities present in myocardial homogenates, it appears likely that this protein functions as a phospholipase AZ and does not make substantial contributions to lysophospholipid or palmitoyl-CoA hydrolysis in intact tissue.
The phospholipase AZ purified in the present study is easily distinguished from other previously described myocardial cytosolic phospholipase activities. A calcium-independent phospholipase A1 activity is present in rat myocardial cytosol but specifically cleaves the sn-1 acyl linkage (41). A phospholipase B activity was reported in Syrian hamster myocardial cytosol (42) but differs from the enzyme purified in the present study by the following features: 1) it does not hydrolyze plasmenylcholine substrate;' 2) its specific activity is three to four orders of magnitude less than the polypeptide purified herein; 3) it has a molecular weight of only 14 kDa; and 4) the regiospecificity of the hamster phospholipase B is predominantly directed toward the sn-1 position while the polypeptide purified in this report has absolute specificity for hydrolysis of the acyl group at the sn-2 position. It is important to note that these cytosolic phospholipase A, and B activities comprise less than 10% of the phospholipase AZ activity present in myocardial cytosol (Table I) utilizing optimal homogenization methods and substrates for each activity (41,42). Thus, cytosolic calcium-independent phospholipase AZ is the major measurable phospholipase in myocardium and possesses separate and distinct physical characteristics and kinetic properties from other myocardial cytosolic phospholipase activities previously described.
Early experiments demonstrated that calcium-independent phospholipase AZ was not present in serum or whole blood and that comparable levels of calcium-independent phospholipase A2 activity were present in perfused and nonperfused hearts. However, comparisons of other calcium-independent lipases (which are predominantly localized in plasma) to the myocardial enzyme merit brief consideration. First, PAF acetyl-hydrolase possesses different chromatographic characteristics (binds to DEAE-Sephacel resin at pH 6.8), thermal stability (stable overnight at room temperature), detergent sensitivity (measurable activity in Triton X-100 or Tween-20), and a substantially different pH optimum (pH 7.8) (45) than the myocardial enzyme. Most importantly, PAF acetylhydrolase is highly specific for hydrolysis of alkyl-ether choline glycerophospholipids containing acetyl groups at the sn-2 position (45). In contrast, myocardial phospholipase A2 hydrolyzes alkyl-ether choline glycerophospholipids with long chain sn-2 aliphatic constituents three orders of magnitude more rapidly than PAF. Second, phospholipase activity me-' Hazen, S. L., and Gross, R. W., unpublished observation. diated by 1ecithin:cholesterol acyltransferase is distinguished from myocardial phospholipase AZ since cholesterol acyltransferase is catalyzed by a 68-kDa polypeptide, requires a serum protein cofactor for expression of phospholipase activity (in its pure form), and exhibits no strict regiospecificity for phospholipid hydrolysis (46,47). Third, endothelial cell-derived lipoprotein lipase is easily distinguished from myocardial phospholipase A2 since myocardial lipoprotein lipase is a 34-kDa polypeptide, avidly binds to Heparin-Sepharose resin (unlike myocardial cytosolic phospholipase AZ), and tolerates acetone precipitation as well as homogenization in detergents (48), both of which completely ablate myocardial phospholipase AZ activity. Fourth, plasma carboxylesterase possesses a different substrate selectivity, thermal stability profile, and molecular weight than myocardial phospholipase AP (49). Finally, cholesterol esterase has a different substrate specificity, a larger molecular mass (68 kDa), and has an absolute requirement for cofactors for lipolysis (29). Taken together, these results demonstrate that the cytosolic calcium-independent myocardial phospholipase AS purified in this report has physical and kinetic characteristics which discriminate it from other calcium-independent lipase activities previously studied.
We have recently demonstrated that myocardial sarcolemma (the electrophysiologically active membrane in myocytes) is the primary target of accelerated phospholipid hydrolysis in myocytes subjected to simulated ischemia (14) and that myocardial sarcolemma is predominantly comprised of plasmenylcholine and plasmenylethanolamine molecular species which are highly enriched in arachidonic acid (12). Since the myocardial phospholipase AP purified herein has direct physical access to the sarcolemmal membrane and selectively hydrolyzes both plasmalogen substrate and arachidonylated glycerophospholipids, this phospholipase has the catalytic potential to selectively hydrolyze the predominant phospholipid constituents present in myocardial sarcolemma (i.e. arachidonylated plasmalogens). Accordingly, activation of this polypeptide is anticipated to result in the selective release of arachidonic acid and the catabolism of sarcolemmal membrane phospholipids similar to that seen during myocardial ischemia (4,14). Although regulation of intracellular phospholipases activated by physiologic increments in calcium ion is now accepted (e.g. Refs. 50 and 51), the biochemical mechanisms responsible for regulation of calcium-independent phospholipases AZ are unknown. Accordingly, future efforts directed toward identification of the molecular mechanisms responsible for the activation of this calcium-independent phospholipase AZ should provide direct insight into the biochemical mechanisms precipitating electrophysiologic dysfunction during myocardial ischemia.