Lysophosphatidylcholine Metabolism in the Rabbit Heart CHARACTERIZATION OF METABOLIC PATHWAYS AND PARTIAL PURIFICATION OF MYOCARDIAL LYSOPHOSPHOLIPASE-TRANSACYLASE*

Metabolism of lysophosphatidylcholine (LPC), re- cently implicated in arrhythmogenesis, was characterized in rabbit ventricular homogenates. Activities of four enzymatic pathways were distinguishable after subcellular fractionation and DEAE-Sephacel chromatography including microsomal lysophospholipase, mi- crosomal acyl coenzyme A/LPC acyltransferase, cytosolic lysophospholipase, and cytosolic lysophospholi- pase-transacylase. Microsomal lysophospholipase activity was attenuated 81% by acidosis comparable to that in ischemic myocardium (pH 6.5) and was inhibited by substrate. LPC acyltransferase was identified in the microsomal fraction based on CoA-dependent phosphatidyl choline synthesis, the positional specificity of ac- ylation of LPC, and identical reaction velocities with both of its labeled co-substrates. LPC acyltransferase had a Vmax of 5.1 nmol/mg/min, a broad pH optimum centered at pH 7, and an apparent K, for LPC and palmitoyl-CoA of 14 pi and 7 pi. Cytosolic lysophos- pholipase was separated from lysophospholipase- transacylase by DEAE-Sephacel chromatography and distinguished from microsomal lysophospholipase by its broad pH activity curve, Michaelis-Menten kinetics (Vmax was incubated with p~ of I4C-LPC and p~ palmitoyl-CoA or with L unlabeled palmitoyl-LPC and L for 15 min, extracted, and separated by HPLC. PC was treated with phospholipaseA2, and reaction products separated as described under “Experimental Procedures.” Cyto- solic protein was incubated with 100 ~ L M I4C-LPC alone or with L M I4C-LPC and 100 L M palmitoyl-CoA. Results are expressed as the ratio of radioactivity in fatty acid divided by total released radioactivity expressed as a percentage. Data are means rt S. E. of at least five determinations from a total of three preparations. Palm-CoA = palmitoyl-CoA.

solic lysophospholipase, and cytosolic lysophospholipase-transacylase. Microsomal lysophospholipase activity was attenuated 81% by acidosis comparable to that in ischemic myocardium (pH 6.5) and w a s inhibited by substrate. LPC acyltransferase was identified in the microsomal fraction based on CoA-dependent phosphatidyl choline synthesis, the positional specificity of acylation of LPC, and identical reaction velocities with both of its labeled co-substrates. LPC acyltransferase had a Vmax of 5.1 nmol/mg/min, a broad pH optimum centered at pH 7, and an apparent K , for LPC and palmitoyl-CoA of 14 pi and 7 pi. Cytosolic lysophospholipase was separated from lysophospholipasetransacylase by DEAE-Sephacel chromatography and distinguished from microsomal lysophospholipase by its broad pH activity curve, Michaelis-Menten kinetics (Vmax = 9.5 nmol/mg/min, K,,, = 7.5 p i ) , and lack of substrate inhibition. Lysophospholipase-transacylase was identified in the cytosolic fraction by CoA-independent phosphatidyl choline synthesis and purified 4885-fold from homogenate by ammonium sulfate precipitation, DEAE-Sephacel, hydroxylapatite, gel filtration, and polylysine chromatography. The partially purified enzyme had a transacylase/lysophospholipase activity ratio of 0.6, and transacylation of LPC was prominent at submicellar concentrations of substrate.
Amphiphiles, including long chain acylcarnitine and lysophosphatides, have been implicated as biochemical mediators of electrophysiological derangements in ischemic myocardium potentially contributing to malignant ventricular dysrhythmia (1,2). Such compounds alter electrophysiological behavior of canine Purkinje fibers and of ventricular muscle in a fashion closely analogous to myocardial ischemia when as little as 2% of membrane phospholipid is constituted with amphiphiles acquired from exogenous sources (3). Thus, accumulation of even small amounts of LPC' may alter the biophysical char- The abbreviations used are: LPC, 1-palmitoyl-sn-glycero-3-phos-acteristics of cardiac membranes and give rise to electrophysiological perturbations similar to those seen in ischemic myocardium.
Early after the onset of myocardial ischemia phospholipid metabolism appears to be altered. Lysophosphatidylcholine accumulates in ischemic zones (4, 5) and in venular effluents from ischemic regions (6). Such accumulation may reflect: 1) increased production, 2) decreased metabolism, 3) diminished washout, or 4) combinations of these phenomena. Although rabbit myocardial phospholipase activity is only modest (7), the overall capacity for metabolic clearance of LPC in rabbit myocardium was found to be large in preliminary experiments in the present study. Thus, the amount of LPC potentially capable of interacting with membranes may vary considerably depending on relatively modest changes in the overall rates of metabolic clearance.
Although metabolism of lysophospholipids has been studied extensively in liver (e.g. Refs. 8-10), lung (e.g. Refs. 11 and 12), and brain (e.g. Refs. 13 and 14), little information is available characterizing LPC metabolism in myocardium. It is likely that phospholipid metabolism differs markedly in the heart compared to the liver in view of differences of profiles of products accumulating in the two organs when they are rendered ischemic (4, 15). Accordingly, the present study was undertaken to identify metabolic pathways of LPC in rabbit heart and determine kinetic characteristics of competing and interacting pathways to gain insight into specific metabolic mechanisms potentially responsible for accumulation of lysophosphatides in ischemic myocardium. Four enzymatic pathways of rabbit myocardial LPC metabolism were identified and characterized including: 1) acyl coenzyme A/LPC acyltransferase (EC 2.3.1.23) (Fig. 1, top left) present in the microsomal fraction of rabbit ventricular homogenates, exhibiting specific substrate requirements, distinctive positional specificity of acylation, and identical reaction velocities when either of its co-substrates was labeled 2) lysophospholipase (EC 3.1.1.5) (Fig. 1, bottom left) in the microsomal fraction, characterized by kinetic parameters, pH profile, and substrate inhibition; 3) lysophospholipase in the cytosolic fraction separated from lysophospholipase-transacylase activity by ion exchange chromatography and characterized by its pH profile, maximum velocity, and substrate dependence; and 4) lysophospholipase-transacylase (EC 3.1.1.5) (horizontal reaction Fig. 1) identified in the cytosolic fraction based on CoAindependent phosphatidylcholine synthesis, e o -p d k a t i o n of FA and PC synthetic activities, and characteristic substrate activity relationships.

EXPERIMENTAL PROCEDURES
Preparation of Cytosolic Extracts-New Zealand White rabbits fed ad libitum were killed by cervical dislocation. Hearts were removed promptly and placed in homogenization medium (0.25 M sucrose, 10 mM phosphate, 1 mM EDTA, and 1 m~ dithiothreitol, pH 7.40) at 0-4 "C. Ventricular muscle was washed extensively in buffer, minced to form a paste, passed through a Harvard Apparatus Co. tissue press, and homogenized with a Potter-Elvehjem apparatus (six strokes) to yield a 25% (w/v) homogenate. The homogenate was centrifuged sequentially at 3,000, 10,000, and 14,000 X gmSx for 10 min and the final supernatant centrifuged at 100, OOO X gmaX for 1 h. The mitochondrial fraction (10,000 X gmaX pellet) was washed and resuspended in homogenization buffer. The 100,000 X g, , , supernatant fraction was decanted and used as cytosolic extract. The microsomal pellet was washed and resuspended in homogenization medium. Cytosolic and microsomal preparations contained an average of 8 and 3 mg/ml of protein.
Assay Systems-Microsomal or cytosolic protein was added to 700 $ of 75 mM phosphate and 3 mM MgCL buffer, pH 7.4, containing I4C-LPC (specific activity approximately 10, OOO dpm/nmol) and incubated in a metabolic shaker at 37 "C for 10 min. Butanol, 700 pl (pH 7, 0-4 "C), was added and the mixture vortexed twice for 10 s. Layers were separated by centrifugation and the organic phase stored in a sample vial. Butanol extraction resulted in nearly quantitative recovery of reaction products (FA, 98% PC, 96%; and LPC, 89%) based on partition coefficient analysis, extracted over 90% of the total radioactivity from the aqueous phase (the remaining radioactivity in the aqueous phase was >95% LPC as ascertained by TLC), and gave identical results when compared with conventional CHCL/MeOH extractions. Lipids were separated by HPLC or TLC or both. Each method gave the same results. HPLC separations utilized a cation exchange column (PXS-10/25-SCX, Whatman) as the stationary phase and acetonitrile/methanol/HzO (300:90:70) as the mobile phase. The flow rate was 2.5 ml/min. Column eluates were monitored at 203 nm as previously described (16). Radioactive metabolites of LPC were confvmed by HPLC and two-dimensional TLC (1) in comparisons with standards. Thus, on a silica-based stationary phase utilizing both acidic and basic mobile phases as well as HPLC utilizing an anionic stationary phase and a neutral mobile phase, both radiolabeled FA and PC produced in incubations during this investigation had identical RF values and retention times as authentic FA and PC. Eluates were collected in scintillation vials and evaporated prior to addition of 10 ml of Aquasol I1 (New England Nuclear) and quantification of radioactivity by scintillation spectrometry. Radioactivity was corrected for quenching with the use of an external standard. Components in fractions from column chromatography were identitied by one-dimensional TLC performed on Silica OF plates (Analabs) with a solvent system of CHCb/acetone/MeOH/AcOH/H20 (3:4:1:1:0.5). After elution, the plates were developed by brief iodine staining and regions corresponding to fatty acid (RF = 1) or PC (RF = 0.4) were scraped into scintillation vials and radioactivity quantified after the addition of 10 ml of Aquasol 11. In addition, PC synthesized by partially purified lysophospholipase-transacylase had identical RF values and retention times in comparisons with standards when chromatographed by two-dimensional TLC (1) or HPLC (16). Protein content was determined by the Bradford method (17). Initial reaction rates were calculated by measurement of the flux of radioactivity from substrate (LPC) into fatty acid or PC under conditions in which ~2 0 % of substrate was metabolized during the incubation. PC synthesized by cytosolic protein had a specific activity equal to twice that of substrate, a value taken into account in the calculation of initial reaction velocity.
In experiments requiring determination of the position of labeled acyl groups in PC, snake venom phospholipase AQ (Crotalus adamanteus) was incubated with PC isolated by HPLC. HPLC fractions were evaporated to dryness under NP and resuspended in 1 ml of diethyl ether.
Incubations were conducted with 0.16 mg of phospholipaseA2 (200 units), approximately 50 nmol of PC, and 5 nmol of CaClz in 1 ml of ferrous sulfate-washed diethyl ether for 1 h as described by Wells and Hanahan (18). The reaction mixture was evaporated to dryness under NP, 0.25 ml of Hz0 and 0.25 ml of butanol were added, the mixture vortexed, and 25 pl of the butanol layer were separated by HPLC. Reaction products were quantified by scintillation spectrometry.
Kinetic parameters were determined with Lineweaver-Burk plots by linear regression analysis of data points obtained from three measurements of initial enzyme rate at each concentration from at least two different preparations.
Fatty acid release during incubation of I4C-LPC with partially purified lysophospholipase or partially purified lysophospholipasetransacylase was linear with respect to incubation time and mass of protein. PC synthesis during incubation of 14C-LPC with microsomal protein or partially purified lysophospholipase-transacylase was linear also with respect to incubation time and mass of protein. Fatty acid release resulting from incubation of microsomes with I4C-LPC was dependent on the substrate/protein ratio (described under "Results"), but at a constant ratio was linear during the initial 10 min of incubation. Therefore, lysophospholipase activity was measured within this interval.
Na-K ATPase in both microsomal and cytosolic preparations was assayed conventionally based on inorganic phosphate release from ATP (19). CK was assayed spectrophotometrically (20). In some experiments, microsomes were incubated f i s t in 10% or 20% (v/v) ethanoI/water for 15 min at 37 "C in anticipation of liberating latent enzyme activity trapped within microsomal vesicles. No latent activity was found.
Partial Purification of Lysophospholipase and Lysophospholipase-Transacylase Actiuities-Twelve rabbits were killed by cervical dislocation. Cytosolic extracts were prepared as previously described except that three 30-s bursts of a Polytron homogenizer at one-half maximal setting were used instead of homogenization with an Elvehjem apparatus. To 175 ml of cytosol prepared from ventricular muscle, 65 g of solid ammonium sulfate were added over 2 min and stirred for an additional 9 min. The precipitated protein was peUeted at 15,000 X gmax for 5 min and quickly resuspended in Buffer A (25 mM Po4, pH 7.60, 10% glycerol, and 10 m~ BME). The protein was then dialyzed against 100 volumes of Buffer A for 14 h at 0 "C. A precipitate formed during this interval which did not contain lysophospholipase or lysophospholipase-transacylase activity, and was removed by centrifugation at 10, OOO X g, , , for 15 min. The supernate was then loaded onto a DEAE Sephacel column (2.6 X 26 cm) previously equilibrated with Buffer A. After the UV absorbance had fallen to base-line, the column was developed with IO volumes of a Lysophosphatidylcholine Metabolism linear NaCl gradient (0-600 mM) at a flow of 30 ml/h. The fractions containing lysophospholipase-transacylase activity eluted at approximately 400 mM NaCl and were subsequently pooled, dialyzed against 50 volumes of 28 mM potassium phosphate buffer (pH 7.0) containing 10% glycerol and 10 mM BME, for 12 h at 0 "C, and loaded onto a column (2.6 X 10 cm) of hydroxylapatite previously equilibrated with the same buffer. After elution of the void volume, transacylase activity was eluted with a linear gradient of 28-500 mM potassium phosphate, pH 7.0 (containing 10% glycerol and 10 mM BME), at a flow rate of 45 ml/h. Active fractions eluted at approximately 200 mM PO4 and were concentrated with an Amicon device with a YM-10 filter, from 45 ml-5 ml. The concentrated protein was loaded onto an AcA 44 column (92 X 1.6 cm) and gel filtration was conducted a t a flow of 12 ml/h with 25 mM sodium phosphate buffer, pH 7.0, containing 10% glycerol, 10 mM BME and 0.5 M NaCl. The active fractions were dialyzed against 100 volumes of sodium phosphate buffer (pH 7.0) containing 10% glycerol and 10 mM BME, for 10 h and loaded onto a polylysine agarose column (0.9 X 10 cm). A 0-700 mM NaCl gradient was run and active fractions eluted at approximately 150 mM NaCI.
Source a n d Purity of Materials-LPC was obtained from Sigma.
It yielded only a single spot after two-dimensional TLC (1) and l2 staining. "C-LPC was obtained from New England Nuclear and was >98% pure as ascertained by HPLC (16). DEAE-Sephacel was obtained from Pharmacia, polylysine agarose from Sigma, AcA 44 resin from LKB, and hydroxylapatite from Bio-Rad. C. adamanteus phospholipase A? was obtained from Sigma.

RESULTS
Initial experiments identified the reaction products resulting from incubation of microsomal or cytosolic protein with 14C-LPC. Incubations of microsomes with 14C-LPC resulted in the release of fatty acid and synthesis of phosphatidylcholine with specific activities of 5.9 and 4.2 nmol/mg/min under conditions as shown in Table I (Table I) and consequently was not explored further. Radioactivity in PC synthesized by the microsomal or cytosolic proteins was not present in choline plasmalogens reflected by the observation that mild acid hydrolysis (0.1 N HCI in emulsions of butanol and Hz0 at 22 "C for 15 min) did not result in release of radiolabeled aldehydes and radioactivity was quantitatively recovered in PC. Furthermore, radioactivity was not present in glycerol ethers under these conditions since base-catalyzed methanolysis (1 N NaOH in MeOH at 22 "C for 15 min) of PC synthesized by microsomal or cytosolic protein resulted in quantitative production of radiolabeled FA methyl esters but no radioactivity co-migrating with LPC.
CK was used as a marker for cytosolic enzymes and was found to have a specific activity of 66.0 -t 3 IU/mg of protein (mean k standard error, n = 6) in the cytosolic fraction but only 11.5 -e 0.4 IU/mg of protein ( n = 6) in the microsomal fraction. No latent CK activity was detected in the microsomal fraction after preincubation in ethanol. Na-K ATPase had a specific activity of 7.8 f 0.1 pmol/mg/h in the microsomal fraction but could not be detected in the cytosolic fraction ( n = four of each). Since the microsomal fraction contained the highest enzymatic specific activity for fatty acid release as well as PC synthesis, membrane-bound enzymes in rabbit ventricular homogenates must hydrolyze as well as acylate LPC. The cytosolic fraction contains an enzyme(s) which hydrolyzes LPC and synthesizes PC which cannot be accounted for by microsomal contamination since the membrane marker Na-K ATPase was not detected in the cytosolic fraction. Since the mitochondrial fraction has the lowest specific activity for either PC synthesis or fatty acid release, contamination with mitochondrial components cannot account for any of the higher specific activities observed. These results demonstrate that the cytosol contains enzyme(s) that hydrolyze and acylate LPC which are separate and distinct from their microsomal counterparts. This view is substantiated by independent criteria delineating differences between the cytosolic and microsomal enzyme systems (see below).
Acyl CoA Dependence of P C Synthesis by Microsomal Enzymes and Acyl CoA Independence of Cytosolic Enzymes"Microsoma1 PC synthesis required palmitoyl CoA which could be replaced by ATP and CoA in combination but not by either alone (Table 11). Thus, PC synthesis by microsomes is acyl-CoA dependent requiring the presence of a high energy thioester for enzymatic esterification. This interpretation is substantiated by the similar rates for incorporation of ['4C]palmitoyl CoA and I4C-LPC into PC by microsomes ( Table I). Among potential source(s) for FA in these incubations are: 1) endogenous FA in the microsomal preparation; 2) FA produced by enzymatic hydrolysis of LPC; and 3) FA produced by hydrolysis of endogenous lipids in the microsomal fraction during the incubation. The production of radiolabeled FA decreased by >90% in incubations containing both ATP and CoA, suggesting that possibility 2) is quantitatively most important.
Incubation of I4C-LPC with cytosolic protein resulted in PC synthesis that was independent of the presence of acyl-CoA, CoA, ATP, or combinations of CoA and ATP (Table 11).
Thus, PC synthesis by the cytosolic fraction was acyl-CoA independent with similar rates of PC synthesis in the presence or absence of acyl-CoA or an acyl-CoA-generating system. In support of this interpretation was the marked difference in incorporation of [14C]palmitoyl-CoA or I4C-LPC into PC when incubated with cytosolic protein (Table I). These results demonstrate that the cytosolic PC synthetic pathway is acyl-CoA independent and does not proceed by esterification of a high energy thioester.
Positional Specificity of Acylation-To further define possible differences between the acyl-CoA-dependent and the acyl-CoA-independent pathway, studies were performed to characterize the positional specificity of acylation of PC syn-   In contrast, PC synthesis by cytosolic protein resulted in a 1.1:l distribution of radioactivity at the C-1 and C-2 carbons. Furthermore, this ratio did not change when incubations were conducted with 14C-LPC and 100 p~ unlabeled palmitoyl-CoA (Table 111). This result demonstrates the preferential utilization of the C-1 fatty acid in LPC for esterification at the C-2 carbon of another LPC molecule over that of exogeneous acyl-CoA. Taken together, these results demonstrate that PC synthesis by rabbit myocardial microsomes is acyl-CoA dependent, proceeds by acylation of the C-2 hydroxyl by the cosubstrate acyl-CoA, and is, therefore, mediated by LPC acyltransferase. In contrast, rabbit myocardial cytosolic PC synthesis is acyl-CoA independent, proceeds by trans-esterification of the C1 fatty acid of LPC, and is, therefore, compatible with the putative mechanism of action of lysophospholipasetransacylase (21). Fatty acid-releasing activity of the cytosolic fraction might be accounted for in its entirety by lysophospholipase-transacylase based on published ratios for hydrolytic/transacylase activity for the lung enzyme (11). However, subsequent purification demonstrated the presence of a sep-

Positional specificity of LPC acylation
Microsomal protein (300 p g ) was incubated with 100 p~ of I4C-LPC and 100 p~ palmitoyl-CoA or with 100 ~L M unlabeled palmitoyl-LPC and 100 ~L M ['4C]palmitoyl-CoA for 15 min, extracted, and separated by HPLC. PC was treated with phospholipaseA2, and reaction products separated as described under "Experimental Procedures." Cytosolic protein was incubated with 100 ~L M I4C-LPC alone or with 100 ~L M I4C-LPC and 100 ~L M palmitoyl-CoA. Results are expressed as the ratio of radioactivity in fatty acid divided by total released radioactivity expressed as a percentage. Data are means rt S. E. of at least five determinations from a total of three preparations. Palm-CoA = palmitoyl-CoA. arate and distinct lysophospholipase in the cytosolic fraction of myocardium. Separation of Cytosolic Lysophospholipase from Lysophospholipase-Transacylase-Anion exchange chromatography separated cytosolic lysophospholipase from lysophospholipase-transacylase (Fig. 2). Cytosolic lysophospholipase eluted in the void volume and was completely devoid of PC synthetic activity. Lysophospholipase-transacylase eluted at approximately 400 mM sodium chloride and each active fraction had a 1.1:l ratio of radioactivity in PC compared to FA. The active fractions were purified further by chromatography on hydroxylapatite. Lysophospholipase-transacylase activity eluted at approximately 200 mM potassium phosphate. Again the ratio of PC/FA radioactivity was 1.1:l. Active fractions were concentrated as described under "Experimental Procedures" and loaded onto an AcA 44 column. Lysophospholipase-transacylase activity eluted as a single peak with a 1.1:l distribution of radioactivity in PC/FA. Active fractions were then dialyzed against pH 7 sodium phosphate buffer and loaded onto a polylysine agarose column. Lysophospholipasetransacylase eluted a t 200 mM sodium chloride, again with a 1.l:l ratio of radioactivity in PC/FA. Partially purified lysophospholipase-transacylase synthesized PC at a rate of 3.2 nmol/mg/min a t a substrate concentration of 2 p~ LPC. A summary of the purification is shown in Table IV. Substrate Dependence-LPC acyltransferase had a maximum velocity of 5.1 nmol/mg/min with an apparent K,,, for LPC of 14 p~ and an apparent K,,, for palmitoyl-CoA of 7 p~ (Fig. 3). Microsomal lysophospholipase activity was dependent upon the ratio of substrate to microsomal protein ( Fig. 4) with marked inhibition at high substrate/protein ratios. Additional experiments with 10 pg of microsomal protein revealed an inflection point at 3 p~ LPC, well below published values of the CMC (22). The inflection point had a constant protein/substrate ratio of 3.3 -+ 0.1 pg of protein/nmol of LPC (a k SE). The maximum velocity at each protein concentration tested was similar (5.9 k 0.5 nmol/mg/min ( X k SE)), although it occurred over an order of magnitude variation in substrate concentration. Substrate dependence differed considerably for the cytosolic enzymes. Cytosolic lysophospholipase (partially purified fraction after DEAE-Sephacel chromatography) had a maximum velocity of 9.5 nmol/mg/min and an apparent K , for LPC of 7.5 p~ (Fig. 5). Lysophospholipase-transacylase (partially purified after DEAE-Sephacel chromatography) demonstrated Michaelis-Menten kinetics for fatty acid release with an apparent K , of 14 p~. However, substrate dependence for PC synthesis was nonlinear, especially at low substrate concentrations (Fig. 6).  One unit of activity is defined as the activity necessary to convert 2 nmol of LPC into 1 nmol of PC and 1 nmol of glycerophosphoryl choline in 1 min at 37 "C. Lineweaver-Burk plot of PC synthesis in the microsomal fraction. Incubation conditions: left, 90 p g of microsomal protein were incubated with 100 p~ palmitoyl-CoA and from 15-150 p~ "C-LPC for 10 min at 37 "C in a total volume of 1 ml. Reaction products were extracted in butanol and separated by HPLC as described under "Experimental Procedures." Right, 90 p g of microsomal protein, 100 p~ I4C-LPC, and from 4-37 p~ palmitoyl-CoA were incubated for 10 min at 37 "C, extracted into butanol, and separated by HPLC. Results are means f S. E. of three determinations.
pH Dependence-Fatty acid release during incubations of microsomes with 14C-LPC had maximum activity from pH 7.0-8.0 but was markedly attenuated at pH 6.5 (rate = 19% of maximum, mean of three preparations). In addition, microsomal lysophospholipase activity at pH 6.0 was less than 10% that at pH 7.0 at every substrate and protein concentration shown in Fig. 4. Similar pH profiles were obtained with Tris-C1 buffer substituted for phosphate. In contrast, PC synthesis by microsomal protein showed a flat pH curve in the range of 6-8.5 when measured with either radiolabeled LPC and cold palmitoyl-CoA or radiolabeled palmitoyl-CoA and cold LPC (n = three of each). Cytosolic lysophospholipase had a broad pH optimum between 7 and 8 and was only mildly attenuated (rate = 80% of control) at pH 6.0 (n = 3). Cytosolic lysophospholipase-transacylase demonstrated a bell-shaped pH profiie for both fatty acid release and PC synthesis with an optimum at pH 7.0 (n = 3).

DISCUSSION
This investigation demonstrates the presence of four enzymes in rabbit ventricular homogenates which metabolize LPC. The microsomal fraction contained lysophospholipase activity which demonstrated unusual kinetics. The reaction velocity was not linearly related to the amount of protein in the incubation media. At a constant protein concentration reaction velocity increased with increasing substrate concentration until an inflection point was reached when the inverse became true at substrate concentrations higher than the inflection point. The inflection point occurred at a constant protein/substrate ratio and was associated with a constant maximum velocity. The maximum velocity was independent of the total concentration of substrate in the incubation medium but was related to the substrate/protein ratio. This suggests that the relevant kinetic parameter is the density of LPC in the membrane that is potentially capable of interacting with the membrane-bound enzyme to form a productive binary complex. The ratio of substrate to protein would influence reaction velocity by determining the density of LPC within the membrane. Biphasic kinetics for membrane bound

I4
lysophospholipase activity has previously been noted for rat brain lysophospholipase and has been interpreted as inhibition by micellar aggregates of LPC (22). Subsequently, the possibility that a critical substrate/protein ratio in the membrane is responsible for lysophospholipase inhibition has been suggested (23). Utilizing 16-doxy1 stearate as a probe, we have noted marked alterations in the rotational correlation times of myocardial microsomal membranes exposed to LPC'. Since inhibition of lysophospholipase activity occurs at submicellar concentrations of LPC, the inhibitory nature of "high" substrate concentrations can not be related solely to the presence of micellar aggregates of substrate but rather most likely reflects alterations in the biophysical characteristics of the membrane preparation produced by exceeding a critical mole proportion of LPC.
Myocardial ischemia is accompanied by an abrupt increase in hydrogen ion concentration to values approaching pH 6 (24). It is anticipated, therefore, that membrane bound lysophospholipase activity would be severely attenuated during myocardial ischemia potentially contributing to accumulation of LPC we have previously noted in ischemic tissue (4). The abrupt decline in activity during incubations with mild acidosis can not be related to titration of phosphate ion since a similar profile was obtained with Tris-C1 buffer. Similarly, the decline in activity can not be related solely to altered inflection points with acidosis since the activity at pH 6.0 was 4 0 % maximal activity over a 20-fold variation in substrate/protein ratios at four different protein concentrations.