ATP-dependent Regulation of Rabbit Myocardial Cytosolic Calcium-independent Phospholipase A2*

We report here that rabbit myocardial cytosolic cal- cium-independent phospholipase A2 exists as a high molecular weight complex comprised of catalytic and regulatory polypeptides whose activity and stability are influenced by specific interactions with ATP. Multiple lines of evidence document the functional signif- icance of interactions between the catalytic complex and ATP including: catalytic affinity although the 40- kDa catalytic subunit neither binds to ATP affinity matrices nor is subject to ATP-dependent activation and stabilization. The catalytic and regulatory ele- ments were functionally resolved by differential thermal inactivation and the ATP-regulatable phospholi- pase A2 catalytic complex was reassembled by reconstitution of highly purified catalytic and partially purified regulatory proteins. Thus, alterations in ATP concentration influence the activity and longevity of the myocardial cytosolic calcium-independent phospholipase A2 catalytic complex, thereby potentially modulating the release of lipidic second messengers and facilitating adaptive alterations in membrane

We report here that rabbit myocardial cytosolic calcium-independent phospholipase A2 exists as a high molecular weight complex comprised of catalytic and regulatory polypeptides whose activity and stability are influenced by specific interactions with ATP. Multiple lines of evidence document the functional significance of interactions between the catalytic complex and ATP including: 1) adenine nucleotide triphosphates attenuate the rate of thermal denaturation of native cytosolic phospholipase A2; 2) ATP augments the initial rate of phospholipid hydrolysis in a manner independent of the concentration, interfacial properties, and physical state of aggregated substrate; 3) adenine nucleotide triphosphates attenuate the reactivity of an essential thiol residue to covalent modification by 5,5'-dithiobis(Z-nitrobenzoic acid); and 4) the catalytic complex specifically and reversibly binds to ATP affinity matrices, although the purified

40-
kDa catalytic subunit neither binds to ATP affinity matrices nor is subject to ATP-dependent activation and stabilization. The catalytic and regulatory elements were functionally resolved by differential thermal inactivation and the ATP-regulatable phospholipase A2 catalytic complex was reassembled by reconstitution of highly purified catalytic and partially purified regulatory proteins. Thus, alterations in ATP concentration influence the activity and longevity of the myocardial cytosolic calcium-independent phospholipase A2 catalytic complex, thereby potentially modulating the release of lipidic second messengers and facilitating adaptive alterations in membrane physical properties.
Activation of intracellular phospholipases A, is responsible for the majority of arachidonic acid mass released during cellular stimulation in most systems (e.g. . Accordingly, the biochemical mechanisms underlying the activation of intracellular phospholipases A, have been the subject of intense scrutiny (e.g. . Although the majority of prior studies have emphasized the importance of calcium ion as the primary regulator of intracellular calcium-dependent phospholipases A, (e.g. Refs. 6,[10][11][12][13][14], the role of additional * This work was supported in part by National Institutes of Health Grant HL34839 and the Monsanto Co. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Recipient of an Established Investigator Award from the American Heart Association. T o whom correspondence should be addressed Molecular and Cellular Cardiovascular Biochemistry, Washington University School of Medicine, 660 S. Euclid Ave., Box 8020, St. Louis, MO 63110. Tel.: 314-362-2690;Fax: 314-362-1402. modulatory elements in the regulation of intracellular phospholipases A, has recently attracted considerable interest (e.g. . During the last several years, multiple groups have demonstrated the presence of calcium-independent phospholipase A2 activity in a wide variety of tissues (21)(22)(23)(24)(25)(26)(27). The necessity for regulatory factors distinct from calcium ion in the regulation of intracellular phospholipases A, is clearly most apparent in the case of calcium-independent phospholipases A, whose molecular mechanisms for regulation have yet to be elucidated.
Early studies demonstrated that the majority of measurable phospholipase A, activity in normal canine myocardium was present in the cytosolic compartment, possessed a neutral pH optimum, and was calcium independent (21). Subsequent purification and detailed characterization of this calciumindependent phospholipase A, revealed that the enzyme was a low abundance high specific activity 40-kDa polypeptide which possessed absolute specificity for the hydrolysis of the sn-2 fatty acid of diradyl glycerophospholipids and which preferentially hydrolyzed plasmalogens containing arachidonic acid at the sn-2 position (27). Since the release of arachidonic acid and the concomitant accumulation of amphiphilic products modulate cellular responses to a wide variety of physiologic and pathophysiologic perturbations (e.g. , the identification of the biochemical mechanisms which regulate the activity of this calcium-independent phospholipase A, are of obvious importance. During the course of the purification of canine myocardial cytosolic calcium-independent phospholipase Az, we exploited the unanticipated finding that phospholipase AS activity specifically and reversibly binds to adenine nucleotide triphosphate-agarose affinity matrices (27). To determine if ATP induced alterations in the physical or kinetic properties of calcium-independent phospholipase Al, we examined the effects of ATP (and its nonhydrolyzable analogs) on myocardial cytosolic phospholipase A, thermal stability, kinetics, and susceptibility to chemical inactivation. We now report that rabbit myocardial cytosolic calcium-independent phospholipase A2 exists as a catalytic complex comprised of catalytic and regulatory polypeptides and demonstrate that the catalytic complex is both activated and stabilized by ATP.

EXPERIMENTAL PROCEDURES
Purification of Rabbit Myocardial Cytosolic Calcium-independent Phospholipase A2-New Zealand White rabbits were sacrificed, and hearts were rapidly removed and perfused retrograde (60 mmHg) for 10 min with modified Krebs-Henseleit buffer utilizing a Langendorf perfused heart model (35). Ventricular tissue was subsequently isolated, weighed, and placed in 0 "C homogenization buffer (0.25 M sucrose (grade I), 10 mM imidazole, 10 mM KCl, pH 7.8) a t 25% (w/ v). Myocardium was minced utilizing sharp scissors, homogenized utilizing a Potter-Elvehjem apparatus, and cytosol was subsequently prepared by differential centrifugation as previously described (27). of ea2+-independent Phospholipase Az 14527 Cytosol was routinely dialyzed against two changes of 500 volumes of homogenization buffer (8 h/dialysis) and either utilized directly or stored in liquid nitrogen. No differences in experimental results were observed in comparisons of dialyzed cytosol with dialyzed cytosol stored in liquid nitrogen in individual aliquots for up to 3 months. Cytosolic phospholipase A, from perfused rabbit myocardium was purified by sequential ion exchange, chromatofocusing, ATP affinity and Mono-Q chromatographies by methods similar to those described previously for canine myocardial cytosolic phospholipase A% (27). For reconstitution experiments, either the ATP affinity column eluent (specific activity = 21 pmol/mg.min; 13,500-fold purified) or the Mono-Q fast protein liquid chromatography eluent (specific activity = 59 pmol/mg. min; 37,800-fold purified) was utilized. The purity of the Mono-Q eluent was corroborated by "'I-autoradiography of SDS-PAGE' gels of Bolton Hunter "sI-labeled proteins which demonstrated a predominant band at 40 kDa. Preparation of Synthetic Phospholipids-Synthesis of homogeneous 16:0,[:'H]18:1 plasmenylcholine was performed by dicyclohexylcarbodiimide-mediated synthesis of radiolabeled oleoyl anhydride and subsequent condensation with reverse phase HPLC purified l-O-(Z)hexadec-l'-enyl-sn-glycero-3-phosphocholine (36) utilizing N,N-dimethyl-4-aminopyridine as catalyst as previously described (37). Homogeneous 16:0,18:1 [3HH-(Me)choline]plasmenylcholine (2.0 X lo5 dpm/nmol) was prepared by exhaustive methylation of reverse phase HPLC purified lysoplasmenylethanolamine (36) with ["HICHJ utilizing benzyltrimethylammonium chloride as catalyst (38) and subsequent acylation of [(Me)~holine-~H]lysoplasmenylcholine with oleoyl anhydride as previously described (37).
Phospholipase A, Assay-Phospholipase A, activity was assessed by incubating enzyme with the indicated concentration of radiolabeled phospholipid (introduced by ethanol injection (10 p1)) in assay buffer (final conditions: 100 mM Tris, 4 mM EGTA, 4 mM EDTA, pH 7.0) at 37 "C for 60 s in a final volume of 210 pl (unless otherwise indicated). Reactions were quenched by addition of butanol (100 pl) and radiolabeled products were separated by TLC prior to quantitation by scintillation spectrometry as previously described (27). Thermal denaturation of phospholipase A, was assessed similarly, except enzyme was incubated at 37 "C for the indicated times in the absence or presence of nucleotide prior to quantitation of phospholipase A, activity. Phospholipase A, activity was linear with respect to both time and protein under all conditions examined. The standard error of the phospholipase A4 assays was typically within 3% and always within 10%. All nucleotides utilized in assays were initially dissolved in H,O (0 "C), immediately titrated to neutral pH, and either utilized directly or stored at -20 "C in individual aliquots.
HPLC Analysis of Endogenous Myocardial Nucleotides and Commercially Available A TP-Nucleotides in fresh myocardial cytosol were purified after perchloric acid precipitation of endogenous protein constituents by addition of 2 volumes of 6% HC10, a t 0 "C. Precipitated proteins were removed by centrifugation and the resultant supernatant was immediately neutralized prior to utilization for HPLC (39). Endogenous myocardial nucleotides were resolved with a SAX-HPLC column (Whatman Magnum 9, 9.4 X 25 cm, 10-pm particles) utilizing a discontinuous potassium phosphate gradient (39). HPLC fractions were neutralized (1 N NaOH, 0 " C ) prior to immediate analysis of their effects on the activation and stabilization of myocardial cytosolic phospholipase A,. Utilization of HPLC-purified ATP resulted in identical effects on activation and stabilization of cytosolic phospholipase A, as commercially available non-HPLC purified ATP.
Miscellaneous Procedures and Sources of Materials-Both Bio-Rad and Quanti Gold (Diversified Biotech) protein assay kits were utilized for protein content determinations.
'""IAutoradiography of SDS-PAGE gels of Mono-Q column eluents were performed as described previously (27). Homogeneous 16:O,["H]18:1 plasmenylethanolamine (9.6 X lo5 dpm/nmol) was a generous gift from Paul Glaser (Washington University). ['HH]CH:J was purchased from Amersham C o p . , while all other radiolabeled starting materials were purchased from Du Pont-New England Nuclear. Bovine heart lecithin and bovine heart ethanolamine glycerophospholipids were purchased from Avanti Polar Lipids, while oleic acid was obtained from Nu-Chek Prep. Inc. Nucleosides and nucleotides of the highest quality (ie., vanadate-free where possible) were purchased from either Boehringer Mannheim Biochemicals or Sigma. ATP-agarose (N'-linkage) and most other reagents were obtained from Sigma.

A T P Attenuates the Thermal Denaturation of Rabbit Myocardial Cytosolic Calcium-independent Phospholipase Az-
Thermal denaturation of cytosolic calcium-independent phospholipase A2 activity at 37 "C was a rapid first order process with a t1,> = 2.8 min ( Fig. 1, left). Incubation of cytosol in the presence of either 10 mM ATP or 10 mM AMP-PCP dramatically attenuated the thermal denaturation of cytosolic phospholipase A2 activity. Since myocardium contains an abundance of enzymes which hydrolyze ATP, the magnitude and concentration dependence of ATP-mediated stabilization of cytosolic calcium-independent phospholipase AP activity was examined utilizing the nonhydrolyzable ATP analog, AMP-PCP. At each concentration of AMP-PCP examined, the kinetics of phospholipase AZ thermal inactivation at 37 "C were first order ( Fig. 1, left). Remarkably, although the tI,? of thermal denaturation was only 2.8 min in the absence of nucleotide, it was 250-fold greater (>11 h) in the presence of saturating concentrations of AMP-PCP. Stabilization of myocardial cytosolic calcium-independent phospholipase Az was both concentration dependent and saturable with maximal effects manifest at 2 1 mM AMP-PCP ( Fig. 1, right). Since both ATP and AMP-PCP effectively stabilize myocardial cytosolic calcium-independent phospholipase AZ activity, these results demonstrate that adenine nucleotide triphosphates interact with phospholipase A, (or associated regulatory constituent(s)) in a manner which does not require participation of an intermediary kinase.
A T P Activates Rabbit Myocardial Cytosolic Calciurn-inclependent Phospholipase A2-To examine the effects of adenine nucleotide triphosphates on the initial rates of phospholipid

FIG.
1. AMP-PCP-mediated attenuation of the thermal denaturation of rabbit myocardial cytosolic calcium-independent phospholipase Az. Left, rabbit myocardial cytosol (400 pg of cytosolic protein) was incubated with AMP-PCP (0-10 mM) for the indicated times a t 37 "C, and remaining phospholipase A, activity was subsequently quantified by incubation with 160,[3H] 181 plasmenylcholine substrate (10 p~ final concentration with a specific activity of 1.25 X lo6 dpm/nmol) for 60 s as described under "Experimental Procedures." Values are expressed as the fractional percentage of phospholipase A, activity a t a given AMP-PCP concentration present after the indicated period of incubation at 37 "C compared to that present at t = 0 (i.e. in the absence of a 37 "C preincubation interval). X, no addition; ., 10 p~ AMP-PCP; Q31.6 p~ AMP-PCP; A, 100 p~ AMP-PCP; A, 316 p~ AMP-PCP; 0, 1 mM AMP-PCP 0, 3.16 mM AMP-PCP; 0, 10 mM AMP-PCP. Data points represent the mean of duplicate determinations. Right, the tH of thermal inactivation (at 37 "C) of myocardial cytosolic calciumindependent phospholipase A, (calculated by linear regression analyses of semilogarithmic plots a t each nucleotide concentration) is plotted as a function of nucleotide concentration.  ATP Alters the Susceptibility of Rabbit Myocardial Cytosolic Calcium-independent Phospholipase AB to Chemical Znactiuation By DTNB-To further demonstrate the interaction of adenine nucleotide triphosphate with phospholipase A, (or its regulatory constituents) by an independent technique, the ability of AMP-PCP to attenuate chemical inactivation of phospholipase AS activity by DTNB was examined. Incubation of myocardial cytosol with DTNB (0-10 WM) at 0 "C for 1 min in the absence of AMP-PCP demonstrated that DTNB inactivated myocardial calcium-independent phospholipase A2. In contrast, incubation of cytosol containing 10 mM AMP-PCP with DTNB did not result in substantial inactivation (Fig. 3). These results suggest that the reactivity (i.e. accessibility) of an essential thiol residue on the catalytic complex (see below) is modified after interaction of AMP-PCP with phospholipase A, or its associated regulatory constituents.
ATP-mediated Activation of Myocardial Cytosolic Calciumindependent Phospholipase A, Is Independent of the Physical Properties and Interfacial Characteristics of Aggregated Substrate-Comparisons of the activating effect of adenine nucleotide triphosphates on phospholipase A, activity utilizing substrate in two different physical states (i.e. liquid-crystalline or inverted hexagonal I1 phase) were performed by exploiting the fact that the liquid-crystal to hexagonal I1 phase transition temperature of 16:0,18:1 plasmenylethanolamine is 30 "C (40). Analysis of phospholipase A,-catalyzed 16:0,[:'H] 18:l plasmenylcholine hydrolysis (liquid-crystalline bilayer a t 37 "C) at different substrate concentrations demonstrated that rabbit myocardial cytosolic calcium-independent phospholipase A, possessed an apparent V, , , of either 1.7 or 7.8 nmol/mg.min in the absence or presence of AMP-PCP, respectively (Fig. 4, left). Similarly, analysis of phospholipase concentrations demonstrated an apparent V, , , of either 0.7 or 2.6 nmol/mg. min in the absence or presence of AMP-PCP, respectively (Fig. 4, right). Since 16:0,18:1 plasmenylcholine and 16:0,18:1 plasmenylethanolamine possess substantially different interfacial characteristics (e.g. surface charge, hydration, physical state), these results demonstrate that ATPmediated alterations in the catalytic complex are independent of changes in the macromolecular organization and interfacial characteristics of substrate. Furthermore, phospholipase A, activation by adenine nucleotide triphosphates results almost entirely from augmentation of the measured apparent maximum reaction velocity since substantial changes in the apparent affinity of enzyme for substrate in the presence of nucleotides were not observed.
Specificity of Nucleotide-dependent Modulation of Myocardial Cytosolic Calcium-independent Phospholipase A2-To examine the chemical interactions responsible for modulation of rabbit myocardial cytosolic calcium-independent phospholipase A2 by nucleotide triphosphates, the effects of the nonhydrolyzable analogs AMP-PCP and GMP-PCP were compared (nonhydrolyzable analogs were utilized to minimize alterations in activator concentration during incubations). AMP-PCP was more potent than GMP-PCP in activating myocardial cytosolic calcium-independent phospholipase A, (Fig. 5). The presence of magnesium resulted in a rightward shift of the dose-response profiles of both adenine and guanine nonhydrolyzable nucleotide triphosphates accompanied by a modest overall reduction in the magnitude of nucleotidemediated activation of phospholipase A, activity (Fig. 5). Addition of nonhydrolyzable guanine nucleotide triphosphates to saturating concentrations of GMP-PCP (21 mM) failed to elicit additional increases in phospholipase A, activity.
The magnitude of nucleotide triphosphate-mediated activation of phospholipase At was greater at pH 6 than at pH 8 (4.3 versus 2.5-fold, respectively). The concentration at which 50% of nucleotide-mediated activation was elicited for both adenine and guanine nucleotide triphosphates (in the presence or absence of magnesium) was higher at pH 6 than at pH 8 (e.g. 1.7 and 2.8 mM for Mg . AMP. PCP and Mg . GMP. PCP, respectively, at pH 6 uersus 1.1 and 1.5 mM for Mg. AMP. PCP and Mg. GMP. PCP, respectively, at pH 8). Similar results were obtained with other nonhydrolyzable nucleotide triphosphates (e.g. ATP-/& AMP-PNP, GTP& GMP-PNP). In contrast, addition of nucleosides (i.e. adenosine, guanosine, inosine) had no effect on phospholipid hydrolytic rates. In general, nucleotide di-and monophosphates (e.g. ADP, GDP, NAD(H), AMP, GMP, CAMP, cGMP) as well as phosphate and polyphosphates (e.g. PPi and tripolyphosphate) possessed intermediate or no effects on phospholipase A, activity. Estimates of the relative potencies of nucleotides for stabilization of cytosolic calcium-independent phospholipase A2 were: nucleotide triphosphate > Mg.nucleotide triphosphate > nucleotide diphosphate, NAD(H) >> nucleotide monophosphate, cyclic nucleotide monophosphate, nucleoside.
To ascertain which endogenous myocardial nucleotides were present in sufficient quantities to mediate cytosolic phospholipase A2 activation and stabilization in intact myocardium, endogenous cytosolic metabolites were chromatographed after perchloric acid extraction (neutralized perchloric acid extracts retained their ability to attenuate thermal denaturation and induce activation of phospholipase A, activity). Preincubation (37 "C for 20 min) of dialyzed cytosol with  buffer, pH 6.0, containing either 4 mM EGTA and 4 mM EDTA (A, W) or 1 mM EGTA and 10 mM MgC1, (0, +). Released 3H-labeled fatty acid was subsequently quantified as described under "Experimental Procedures." The only radiolabeled product detected was 3Hlabeled fatty acid without detectable radiolabeled lysophospholipid, diglyceride, monoglyceride, or phosphatidic acid. Data points represent the mean of duplicate determinations. NTP, nucleotide triphosphate.
HPLC chromatographic eluents of perchloric acid extracts of cytosol, prior to quantitation of phospholipase AZ activity, demonstrated that the cytosolic stabilizing factor precisely coeluted with ATP (the major uv absorbing peak) with diminutive amounts of stabilizing activity confined to the region coeluting with ADP (Fig. 6). Direct incubation of cytosolic calcium-independent phospholipase Az with HPLC column chromatographic eluents (i.e. no 37 "C preincubation period) resulted in a 4-fold increase in measurable calcium-independent phospholipase AP activity (in the peak column fraction) FIG. 6. Anion exchange HPLC of endogenous rabbit myocardial cytosolic nucleotides. Rabbit myocardial nucleotides (obtained from neutralized perchloric acid extracts of fresh myocardial cytosol) were fractionated by SAX-HPLC utilizing a discontinuous K[PO,] gradient as described under "Experimental Procedures." HPLC column fractions were subsequently neutralized and assessed for their ability to both activate and stabilize cytosolic calciumindependent phospholipase AP activity by quantitating 3H-labeled fatty acid release from 16:0,['H]18:1 plasmenylcholine substrate (2 PM final concentration with a specific activity of 1.25 X lo6 dpm/ nmol) in co-incubations of myocardial cytosol (400 pg of protein) and aliquots from HPLC column fractions. Phospholipase A, activation was determined from initial rate analyses without preincubation (O), while phospholipase A, stabilization was assessed by preincubating reaction mixtures for 20 min at 37 "C prior to quantitating remaining phospholipase A, activity (A). -, UV absorbance. Data points represent the mean of duplicate determinations. which coeluted with ATP (Fig. 6).

Reconstitution of ATP-dependent Activation and Stabilization of Myocardial Cytosolic Calcium-independent Phospholipase
Az-Highly purified calcium-independent phospholipase Az (dialyzed ATP-agarose eluent as well as near-homogeneous Mono-Q active fractions) could no longer be activated or stabilized by ATP. However, when highly purified phospholipase AP (dialyzed ATP-agarose eluent, specific activity = 21 pmol/mg.min; 13,500-fold purified) was mixed with crude myocardial cytosol, ATP-dependent activation and stabilization of the purified enzyme was restored (Fig. 7). Accordingly, dialyzed ATP-affinity column chromatographic eluent (specific activity = 21 pmol/mg. min) was utilized as a probe to 18:l plasmenylcholine substrate (10 PM final concentration; specific activity = 1.25 X lo6 dpm/nmol). Reactions were performed in the absence (-) or presence (+) of 10 mM ATP for 60 s at 37 "C as described under "Experimental Procedures." Radiolabeled fatty acids were extracted into butanol, isolated by TLC, and quantitated by scintillation spectrometry as described under "Experimental Procedures." Lower panel, ATP-mediated stabilization of phospholipase A, was assessed by preincubatingcytosol(400 pg of protein), heat-treated cytosol (400 pg of protein), highly purified phospholipase A2 (30 ng), or the indicated mixtures in the absence (-) or presence (+) of 10 mM ATP for 20 min at 37 "C. Subsequently, remaining phospholipase A, activity in these fractions or mixtures was assessed by addition of 16:0,[:'H]181 plasmenylcholine substrate (10 ~L M final concentration; specific activity = 1.25 X lo6 dpm/nmol) for 60 s at 37 "C as described under "Experimental Procedures." Released radiolabeled fatty acid was extracted into butanol, separated by TLC, and quantified by scintillation spectrometry as described under "Experimental Procedures." C , cytosol; C,,, cytosol previously incubated at 37 "C for 20 min to ablate endogenous catalytic activity; C , , cytosol previously incubated at 60 "C for 45 min; Cn cytosol previously incubated with trypsin (2%, w/w) for 24 h at 4 "C. pPLA2, highly purified phospholipase AP (dialyzed ATP affinity column eluent; specific activity = 21 pmol/mg.min; 13,500-fold purified). characterize the cytosolic constituent(s) which mediated ATP-dependent activation and stabilization of myocardial cytosolic calcium-independent phospholipase AP.
First, phospholipase Az activity in myocardial cytosol was ablated by preincubation at 37 "C for 20 min in the absence of ATP. Subsequent addition'of ATP to heat-treated cytosol did not result in recovery of phospholipase AZ activity (Fig.  7). However, co-incubation of heat-treated cytosol and highly purified phospholipase AP (dialyzed ATP affinity column eluent; specific activity = 21 gmol/mg-min; 13,500-fold purified) resulted in the reconstitution of ATP-dependent activation and stabilization of highly purified phospholipase AP activity (Fig. 7). Thus, the cytosolic constituent(s) conferring ATP-dependent activation and stabilization can be functionally discriminated from phospholipase Az catalytic activity (i.e. the polypeptide(s) mediating ATP-dependent activation of phospholipase AP is more resistant to thermal denaturation at 37 "C than the 40-kDa polypeptide catalyzing phospholipase Az activity). Finally, preincubation of cytosol at 60 "C, proteolytic digestion of cytosol with trypsin (Fig. 7), or treatment of cytosol with 1 mM DTNB (followed by exhaustive dialysis) completely ablated the ability of the cytosolic constituent(s) to reconstitute ATP-mediated activation and stabilization of highly purified phospholipase Az. Thus, a cytosolic protein which is functionally distinct from endogenous phospholipase A2 catalytic activity is capable of conferring ATP-dependent activation and stabilization to highly purified phospholipase A*.
Gel Filtration Chromatography of the Nucleotide-dependent Regulatory Component-To purify and further characterize the polypeptide(s) responsible for the activation and stabilization of myocardial cytosolic phospholipase AB, crude myocardial cytosol was fractionated by gel filtration chromatography. Although the polypeptide mediating rabbit myocardial cytosolic calcium-independent phospholipase A2 activity has a molecular mass of 40 kDa (similar to the canine enzyme (27)), crude rabbit myocardial cytosolic phospholipase AP activity migrated as a high molecular weight complex exhibiting an apparent molecular mass of 400 kDa (Fig. 8, upper  panel). Crude cytosolic phospholipase A2 activity reproducibly migrated at 400 kDa on gel filtration chromatography even in the presence of supraphysiologic ionic strength buffers (e.g. 400 mM K[PO,]). Multiple attempts at gel filtration chromatography of the nearly homogeneous 40-kDa catalytic polypeptide ( i e . Mono-Q eluent) were unsuccessful reflecting the extreme lability of the homogeneous catalytic polypeptide even in the presence of ATP. The polypeptide(s) mediating ATP-dependent activation of endogenous myocardial cytosolic phospholipase Az reproducibly ( n = 3) eluted at a slightly lower apparent molecular mass (approximately 350 kDa) than the phospholipase AP catalytic complex (i.e. peak catalytic activity eluted 1 tube earlier than activating and stabilizing activities which both co-chromatographed) (Fig. 8, upper  panel).
The apparent molecular weight of the polypeptide(s) mediating nucleotide-dependent stabilization of myocardial cytosolic phospholipase AP was also determined in reconstitution assays utilizing highly purified phospholipase AP (dialyzed ATP affinity column eluent; specific activity = 21 pmol/mg. min; 13,500-fold purified) and heat-treated (37 "C for 20 min) gel filtration column chromatographic fractions (to ablate endogenous phospholipase A2 catalytic activity in the gel filtration eluents). Preincubation (37 "C for 20 min) of highly purified phospholipase A2 alone or in combination with heattreated gel filtration fractions in the absence of ATP did not result in the preservation of highly purified phospholipase Az activity (Fig. 8, upper panel). In stark contrast, incubation of heat-treated gel filtration fractions and highly purified phospholipase Az for 20 min at 37 "C in the presence of ATP resulted in the preservation of phospholipase AP activity only when fractions eluting at 350-400 kDa were utilized (Fig. 8,  lower panel). Trypsin treatment of the 350-400 kDa constituents obtained from gel filtration chromatography resulted in the complete ablation of both their endogenous phospholipase A, catalytic activity as well as their ability to mediate nucleotide-dependent activation and stabilization of highly purified phospholipase A2 (Fig. 9). Thus, myocardial cytosolic phospholipase A2 migrates as a high molecular weight complex on gel filtration chromatography ( i e . 400 kDa) which marginally, but reproducibly, precedes the elution of the polypeptide(s) conferring nucleotide-dependent stabilization and activation (i.e. 350-400 kDa). Gel filtration chromatography of native rabbit myocardial cytosolic phospholipase Az. Upper panel, rabbit myocardial cytosol was fractionated by gel filtration chromatography utilizing tandem Superose 12 columns as described under "Experimental Procedures." Endogenous calcium-independent phospholipase A, activity in each chromatographic fraction was assessed in the absence (0) or presence (0) of 10 mM ATP utilizing 16:0,[3H]181 plasmenylcholine substrate (10 p M final concentration; specific activity = 1.25 X lofi dpm/nmol) as described under "Experimental Procedures." Upper panel (inset), plot of KAv uersus molecular weight for the tandem Superose 12 columns employing ferritin, catalase, lactate dehydrogenase, phosphorylase b, bovine serum albumin, and ovalbumin. Lower panel, first, chromatographic fractions were preincubated for 20 min at 37 "C in the absence of ATP to ablate endogenous phospholipase A, activity in each fraction. Next, 30 ng of highly purified myocardial cytosolic phospholipase A, (dialyzed ATP-affinity column eluent; specific activity = 21 pmol/mg.min) was subsequently mixed with 50 pl of each heat-treated gel filtration column fraction in assay buffer and these mixtures were subsequently incubated at 37 "C for an additional 20 min in the absence (0) or presence (0) of 10 mM ATP. Finally, remaining phospholipase A, activity in these mixtures was quantitated by injection of 16:0,[3H]181 plasmenylcholine substrate (10 p M final concentration; specific activity = 1.25 X lo6 dpm/nmol) and subsequent incubation for 60 s at 37 "C as described under "Experimental Procedures."

Chromatofocusing of Cytosolic Constituents Mediating Nucleotide-dependent Phospholipase Az Activation and Stabilization-
To further characterize the cytosolic constituents mediating nucleotide-dependent activation and stabilization, chromatofocusing of cytosolic proteins was performed. Under the conditions employed (see "Experimental Procedures") modest amounts of phospholipase Az activity (<3,000 dpm/ 50-pl eluent of radiolabeled fatty acid was released from [3H] plasmenylcholine substrate under the assay conditions utilized) were present in the chromatofocusing fractions eluting at or near pH 7.2. To remove this endogenous phospholipase A, catalytic activity, each fraction was first preincubated at 37 "C for 20 min which resulted in the complete loss of phospholipase Az catalytic activity. Subsequent addition of ATP (10 mM) to these heat-treated fractions did not restore phospholipase Az catalytic activity. However, co-incubation of highly purified phospholipase Az (dialyzed ATP affinity eluent; specific activity = 21 pmol/mg-min) with heat-treated chromatofocusing eluents identified a single peak of ATPdependent phospholipase A, activating activity (eluting at pH Reconstitution of ATP-dependent activation and stabilization of highly purified rabbit myocardial cytosolic calcium-independent phospholipase Az (PLA,) with partially purified phospholipase A, regulatory protein(s). Partially purified regulatory protein(s) from gel filtration chromatography ( i e . 50 pg of protein from fraction 37 in Fig. 8 ) was mixed with highly purified phospholipase A, (30 ng of dialyzed ATP affinity column eluent; specific activity = 21 pmol/mg.min) to further characterize the regulatory factor(s) mediating ATP-dependent activation and stabilization of phospholipase A, activity. Upper panel, activation of phospholipase A, activity by ATP was determined by quantifying the initial rate of 16:0,[3H]181 plasmenylcholine hydrolysis (10 p M final concentration; specific activity = 1.25 X lo6 dpm/nmol) in the absence (-) or presence (+) of 10 mM ATP in the indicated mixtures during 60-s incubations at 37 "C. Lower panel, ATP-dependent stabilization of phospholipase A, activity was determined by quantifying remaining phospholipase A, activity in the indicated fractions and mixtures after a 20-min preincubation at 37 "C performed in the presence (+) or absence (-) of 10 mM ATP. pPLA,, highly purified phospholipase A, (dialyzed ATP-affinity column eluent, 21 pmol/mg.min; 13,400fold purified); 350K, gel filtration fraction 37 from Fig. 8; 350K37, gel filtration fraction 37 preincubated at 37 "C for 20 min to ablate endogenous catalytic activity; 3 5 0 K~, gel filtration fraction 37 previously incubated with trypsin (2%, w/w) for 24 h at 4 "C. (Fig. 10, upper panel). Similarly, the factor which mediated ATP-dependent stabilization of highly purified phospholipase AP eluted as a single peak which precisely cochromatographed with the ATP-dependent phospholipase Az activating factor (Fig. 10, lower panel). DISCUSSION Early studies suggested a role for ATP in the regulation of myocardial cytosolic calcium-independent phospholipase A:! since partially purified enzymic activity specifically and reversibly bound to ATP affinity matrices (27). This unique property was exploited during the initial purification of myocardial calcium-independent phospholipase AP, leading to the identification of a low abundance high specific activity 40-kDa polypeptide as the catalytic entity responsible for canine myocardial cytosolic calcium-independent phospholipase AP activity (27). In this report, we demonstrate that native rabbit Chromatofocusing of cytosolic constituents mediating nucleotide-dependent phospholipase AP activation and stabilization. Phospholipase A, regulatory constituents from rabbit myocardial cytosol were purified by chromatofocusing as described under "Experimental Procedures." Endogenous calcium-independent phospholipase A, activity in each chromatographic fraction was first irreversibly ablated by heat-treatment at 37 "C for 20 min prior to use in reconstitution experiments. Upper panel, the elution of protein(s) mediating ATP-dependent activation of highly purified rabbit myocardial cytosolic phospholipase A, (dialyzed ATP-affinity column eluent; specific activity = 21 pmol/mg.min) was assessed by quantifying phospholipase A, activity in mixtures of highly purified phospholipase A, (30 ng) and column chromatographic fractions (50 pl) in the absence (0) or presence (0) of 10 mM ATP as described under "Experimental Procedures." Lowerpanel, the elution of ATP-dependent stabilizing factors of phospholipase A, activity from the chromatofocusing columns was determined by quantitating phospholipase A, activity in mixtures of highly purified myocardial cytosolic phospholipase A, (30 ng of dialyzed ATP-affinity column eluent; specific activity = 21 pmol/mg.min) and each chromatofocusing column fraction (50 p1) following a 20-min preincubation at 37 "C in the absence (0) or presence (0) of 10 mM ATP. Remainingphospholipase A, activity in all mixtures was subsequently assessed by quantitating "H-labeled fatty acid released during incubation of substrate (16:0,[:'H]18:1 plasmenylcholine; 5 p M final concentration; specific activity = 1.25 X lo6 dpm/nmol) with these mixtures for 60 s at 37 "C as described under "Experimental Procedures." myocardial cytosolic calcium-independent phospholipase A, exists as a high molecular weight complex which can be regulated by ATP and is comprised of functionally distinct catalytic and regulatory components. These results are the first to demonstrate that one biochemical motif potentially utilized for regulation of this class of calcium-independent phospholipases A, exploits ligand-regulated protein-protein interactions.

7.2)
Multiple lines of experimental evidence demonstrate an interaction between ATP and the cytosolic phospholipase A, catalytic complex. First, ATP dramatically (250-fold) attenuates the rate of thermal inactivation of crude, but not purified, phospholipase AS. Second, interaction of ATP with the catalytic complex results in a 3to &fold increase in the initial rate of phospholipase A, activity which is independent of the concentration, interfacial characteristics, and physical state of substrate. Third, ATP attenuates the covalent modification of a critical thiol residue by DTNB. Fourth, myocardial cytosolic calcium-independent phospholipase A, specifically and reversibly binds to ATP-affinity matrices even in the presence of excess AMP. These interactions appear to be unique characteristics of cytosolic calcium-independent phospholipase A, since all other calcium-dependent phospholipases AS examined to date (e.g. pancreatic, Naja nu;,, and platelet cytosolic phospholipases AS) neither demonstrate ATP-dependent alterations in their physical and kinetic properties' nor bind to ATP affinity matrices (27).
High performance liquid chromatography of both endogenous nucleotides from perchloric acid extracts of myocardial tissue, as well as commercial ATP, demonstrated that the chemical moiety facilitating cytosolic phospholipase A, activation and stabilization precisely co-chromatographed with ATP. Analysis of endogenous myocardial nucleotides demonstrated that ATP was the only nucleotide present in sufficient quantities to modulate phospholipase AS activity (in the absence of substantial compartmentation). Since multiple synthetic nonhydrolyzable adenine nucleotide triphosphate analogs (e.g. AMP-PCP, ATPTS, AMP-PNP) also possessed similar effects on both activation and stabilization of calciumindependent phospholipase A, activity, the participation of an intermediary kinase in mediating these phenomena can be effectively excluded (41)(42)(43)(44). Likewise, since adenine nucleotide triphosphates are more potent than guanine nucleotide triphosphates, divalent cations are nonobligatory components of ATP-mediated phospholipase AS activation and stabilization, and the phospholipase AS examined in this study is a soluble activity regulated by a soluble protein(s), the data also exclude conventional G-protein-mediated regulation (45)(46)(47).
Since the initial description of phospholipid depletion, the selective release of arachidonic acid and the accumulation of lysophospholipids during liver ischemia (48), multiple independent groups have demonstrated alterations in phospholipid catabolism in ischemic myocardium (e.g. Ref. 49), brain (e.g. Ref. 50), kidney (e.g. Ref. 51), and intestinal mucosa (e.g. Ref. 52). Similarly, reperfusion of ischemic tissues is accompanied by rapid alterations in intracellular ATP and concurrent changes in phospholipid catabolism (53,54). The demonstration that cytosolic phospholipase A, from myocardium exists as a high molecular weight complex whose kinetic characteristics are modulated by ATP suggests a potential biochemical mechanism which can couple alterations in high energy phosphate metabolism with changes in phospholipid catabolism. Recent evidence has suggested that compartmentalized pools of ATP are present in myocytes (39,55). Accordingly, local alterations in ATP concentration potentially modulate calcium-independent phospholipase A, activity at or near critical subcellular loci, leading to adaptive alterations in the physical properties of individual subcellular membranes which can be precisely tailored to a changing metabolic milieu. However, it is important to note that profound pathophysiologic perturbations (e.g. ischemia and reperfusion) are accompanied by a multiplicity of complex biochemical alterations (c.f. Ref. 56) which may modulate the interactions of calcium-independent phospholipase A' (or its regulatory proteins) with nucleotide triphosphates or may initiate alternative regulatory mechanisms.
Collectively, the results of the current study demonstrate that ligand-induced modulation of protein-protein interactions of a soluble phospholipase AS complex is one biochemical mechanism capable of regulating myocardial cytosolic calcium-independent phospholipase A,. Molecular identification of the constituent(s) which comprise the regulatory complex, clarification of the precise alterations induced by ATP, and elucidation of the critical domains and contact points which modulate protein-protein interactions in the catalytic com-' S. L. Hazen and R. W. Gross, unpublished observations. plex should provide detailed insights into the precise molecular mechanisms which regulate this novel class of intracellular calcium-independent phospholipases Az.