.J Mol

Cell

Cardiol

24, 1101--l

Metabolism

111 (1992)

of Platelet-activating Massimo

Division

of Clinical

Factor

Triggiani*

and Floyd

in the Guinea-pig H. Chilton

Immunology, The Johns Hopkins University School of Medicine, Boulevard, Baltimore, MD 21224, USA

(Received 28 Mq

Heart

1991, accepted in revised form 23 Apil

301 Bayview

1992)

M. TRIGCIANI AND F. H. CHILTON. Metabolism of Platelet-activating Factor in the Guinea-pig Heart. Journal ?/ Molecular and Cellular Cardiology (1992) 24, 1101-l 111. Platelet-activating factor (PAF; 1-alkyl-2-acetyl-glycerophosphocholine) is a biologically active phospholipid which is synthesized by a variety of blood cells and organ systems. PAF exerts many effects on the cardiovascular system including hypotension, depression of myocardial contractility and coronary constriction. The present study has examined the capacity of the guinea-pig heart to regulate the levels of exogenous PAF in two different models: isolated perfused heart and isolated ventricular myocytes. In the first model, isolated hearts were perfused with labeled PAF (IO-“‘M) in a recirculating manner at How rates of 15 ml/min (normal flow perfusion; NFP) and 2 ml/min (low flow perfusion, LFP). ExogenousI) provided PAF appeared in the tissue in a time-dependent manner. The rate of extraction of PAF was higher during LFP than during NFP. PAF was metabolized by the heart to two major products, lyso-PAF and I-alkyl2-acyl-sn-glycero-3-phosphocholine (1-alkyl-2-acyl-GPC). Lyso-PAF was found primarily in the perfusion buffer while both lyso-PAF and 1-alkyl-P-acyl-GPC were detected in the tissue. No qualitative difference in the metabolic products derived from PAF catabolism was observed between hearts undergoing NFP and LFP. Aretyl hydrolase activity was detected in the perfusion fluid at both flow rates, probably accounting for the formation of lyso-PAF in the perfusate. However, perfusion fluid from LFP contained a higher acetyl hydrolase activity per pg of protein as compared to fluid from NFP. Isolated ventricular myocytes incubated with labeled PAF (3 x 10m9 M) also converted it to 1-alkyl-P-acyl-GPC. Kinetic experiments suggested that PAF was initially deacetylated to form lyso-PAF and that this intermediate was then rapidly reacylated with a fatty acyl moirt) at the sn-2 position. HPLC analysis of the fatty acids inserted at the sn-2 posItion 01 1-alkyl-P-acyl-GPC revealed that the myocytes reacylated lyso-PAF predominantly with arachidonic acid. These data indicate that the guinea-pig heart may regulate PAF levels by at least two mechanisms: ( 1 ! it may release acetyl hydrolase into the vascular compartment, particularly under low flow conditions; and (21 thr ventricular myocyte has the capacity to take up PAF and catabolize it to inactive products. KEY WORDS: Acetyl

hydrolase;

Arachidonic

acid;

Platelet-activating

Introduction Platelet-activating factor (PAF) is a phospholipid molecule that possesses a variety of biological activities. PAF was first described as a factor produced by immunologically activated basophils which was able to induce the aggregation and degranulation of platelets [l-3]. Since then, several other biological activities, both on isolated cell preparations and isolated perfused organs, have been reported. For example, PAF has the capacity to stimulate chemotaxis, aggregation and *Present address: Division 8013 I Naples, Italy.

of Clinical

Please address all correspondence Medicine, 300 S. Hawthorne Road, 0022-

2828/92/101101+

I1 $08.00/O

Immunology,

factor;

Ventricular

myocytes.

degranulation of inflammatory cells such as the neutrophil [4], eosinophil [.5j and basophil [q, suggesting that this molecule has a primary role in inflammatory processes. In addition, PAF influences the functions of different organ systems such as the heart, the kidney, the liver and the brain, implicating its significance in a number of pathophysiological states [7, 81. The effects of PAF on the cardiovascular system are complex and, sometimes, speciesspecific. One of the most reproducible conse-

University

of Naples

II School

to: Floyd H. Chilron, Division of Pulmonary Winston-Salem, NC 27103. USA.

of Medicine,

Medicine,

101992

Via S. Pan&i

Bowman

Academic

Gray

School

Press Limited

5, of

1102

M. Triggiani

and F. H. Chilton

quence of intravenous administration of PAF is a systemic hypotension that has been variably attributed to a reduced cardiac output, a decrease in peripheral resistances and a reduction of plasma volume [9, 101. The reduced cardiac output induced by PAF appears to be directly related to a negative inotropic effect [II, 121; however, it may also be due, at least in part, to a concomitant coronary constrictor effect [13]. The variations of coronary blood flow induced by PAF have also been related to the activation of blood cells, such as platelets and neutrophils, circulating in the coronary vessels and to the release of secondary vasoactive mediators [9, 10, 141. Several potential sources of PAF are present within the cardiovascular system. Myocardial cells appear to produce PAF when irreversibly injured with ATP-depleting agents [15]. Endothelial cells obtained from human umbilical vein produce PAF in response to several stimuli such as thrombin, histamine, bradykinin and leukotriene C, [16, 171. The cardiovascular system also contains other resident cells such as the mast cell and the macrophage which, in studies conducted with cells of non-cardiac origin, have been shown to synthesize PAF [28, 291. Within the vascular compartment, cells like the neutrophi1 [20], eosinophil [21] and, in certain species, platelet [22], may also generate large amounts of PAF. The release of PAF in isolated perfused hearts has been demonstrated during anaphylaxis [23] and during ischemia-reperfusion [24, 251. This last observation, together with the ability of exogenous PAF to reproduce or to aggravate myocardial ischemia suggest a role for this molecule in the mechanism of ischemia-reperfusion injury [2S, 271. This hypothesis, however, is far to be conclusively proven, although some beneficial effects of PAF receptor antagonists have been reported in models of ischemia-reperfusion damage [28]. Most of the effects of PAF, both on isolated cells and organ systems, can be observed at very low concentrations (i.e. nanomolar). Therefore, the capacity of the cardiovascular system to regulate the levels of this mediator by converting it to inactive metabolites is a key element in limiting its potentially harmful effect in vivo. In this regard, the present study was designed to explore the uptake and the

metabolism of PAF by the guinea-pig perfused heart and isolated ventricular myocyte.

Methods Isolated perfused heart Guinea-pigs were killed by cervical dislocation and the hearts were isolated and perfused with Krebs solution which contained (mM): NaCl (118); KC1 (5.4); NaH2P0, (1); MgSO, (1.2); CaCl, (1.9); NaHCO, (25); and glucose (11.1). The solution was constantly oxygenated with 95% O,-5% CO,. The hearts were perfused at a constant flow rate using a peristaltic pump and the flow rate was adjusted as indicated in each experiment. After 20 min of equilibration, the hearts were perfused (15 ml/min) for 10 min with Krebs containing bovine serum albumin (BSA; 0.25 mg/ml) to remove the serum-derived acetyl hydrolase. r3H]-PAF ( 1-[3H]-alkyl-2-acetyl-GPC; New England Nuclear, Boston, MA, USA; 60 Ci/mmole) was dissolved in an appropriate volume of Krebs containing 0.05 mg/ml BSA (80 ml for NFP and 15 ml for LFP) to a final concentration of lo-” M (1.5 and 0.2 pmoles/ min for NFP and LFP, respectively). These concentrations of PAF are reported to induce a 10-40% reduction of contractility in isolated guinea-pig hearts perfused at constant flow [ 13, 231. The hearts were perfused ( 15 or 2 ml/min, as indicated) for 30 min with the solution containing PAF in a recirculating manner for a total of 6 passes (5 min each pass). The total dose of PAF circulated through the heart within 5 min in both NFP and LFP conditions. No catabolism of PAF occurred during the circulation in the perfusion apparatus if the heart was excluded. Aliquots (1 ml) of the perfusate were collected after each pass. A portion of these aliquots (100 ~1) was used to quantify the amount of radioactivity in the perfusate. Lipids were then extracted from the remaining portion (900 ,ul) by the method of Bligh and Dyer [29]. After 30 min, the hearts were perfused for 5 min with Krebs containing BSA (0.25 mg/ml) to remove the PAF that had not been incorporated into the tissue. Subsequently, the hearts were homogenized in methanol/water (2/ 1; v/v), acidified (pH 3) with formic acid and the lipids were extracted

PAF

Metabolism

by the method of Bligh and Dyer [29]. This procedure extracted > 93% of the radioactivity from the tissue. Preliminary experiments in which labeled PAF was added to the tissue immediately before the homogenization indicated that ~5% of PAF was degraded during the homogenization and extraction procedure. The radioactive compounds recovered from the extraction were separated by thin-layer chromatography (TLC) as described below and quantified by liquid scintillation counting. The sum of the radioactivity recovered in the perfusate and that recovered in the tissue at the end of the experiment was 89f 3% (NFP experiments) and 84f4% ( LFP experiments) of the initial radioactivity. The remaining lo-15% probably represents labeled PAF adhering to the tubes of the perfusion system or removed from the tissue during the final wash with BSA. Preparation

of ventricular myocytes

Myocytes were obtained by a modified procedure of Kamp et al. [30]. Briefly, guinea-pig hearts were perfused (15 ml/min) with Ca2+free Krebs solution containing collagenase ( 1 mg/ml). This perfusion was continued in a recirculating manner for 20 min. Collagenase was removed by perfusion with Ca2+-free Krebs for 10 min. The ventricles were then cut into pieces and placed in a 50 ml conical tube with 20 ml of Ca2+-free Krebs. The tissue was incubated at 37°C with gentle shaking for 10 min to disperse the isolated cells. The tissue was then filtered through a gauze and the isolated cells were allowed to sediment 1’30 min) at room temperature. The supernatant fluid was removed and the cell pellet was resuspended in Ca’+-free Krebs solution. The percentage of rod-shaped cells used in these experiments was >85% and the viability, assessed by Trypan blue exclusion, was >80%. In these experiments, the myocytes were resuspended ( lo6 cells/tube) in 1 ml of Krebs and incubated at 37°C with 3 pmoles ;3 x IO-"M) of [3H]-PAF or [3H]-lyso-PAF complexed to BSA (final concentration: 0.05 mg/ml). The incubations were terminated by the addition of 3 ml of methanol/ chloroform lipids were (2/l; v/v) and extracted from the whole incubation mixture : cells + supernatant) by a modified method of

in

the

1103

Heart

Bligh and Dyer [29] in which acid was added to lower aqueous phase to 3.0. Chromatography

sufficient the pH

formic of the

of lipids

In order to identify the products derived from the catabolism of PAF, lipids extracted from the perfusion fluid, from the homogenized tissue or from the myocytes were separated on layers of Silica Gel G developed in chloroform/methanol/glacial acetic acid/water (50/ 251814; v/v). The standards (PAF, lyso-PAF and 1-0-alkyl-P-acyl-GPC) were visualized with I, vapors and the distribution of label in different products was determined by scanning the plate for radioactivity using a Bioscan System 200 Imaging Scanner (Washington, DC, USA). Various products were then isolated from the TLC plate and the radioactivity of these products was determined by liquid scintillation counting. The predominant molecular species of I[3H]-alkyl-2-acyl-GPC produced during PAF metabolism were determined b>- a reverse phase HPLC system which separates individual molecular species as benzoate derivatives. l-[3H]-alkyl-2-acyl-GPC was purified by scraping the corresponding area of the TLC plate and extracting the lipids 13 X ) from the silica gel with methanol/water ; 3/ 1; v/v). The benzoate derivatives of I-[‘HIalkyl-2-acyl-GPC were synthesized as prcviously described [31]. The various labeled benzoate cliglyceride molecular species were isolated by reverse phase HPLC using a 250 X 4.6 mm analytical column (Ultrasphere ODS, Beckman,, rluted with acetonitrile/2-propanol (80/20; v/\? at a flow rate of 1 ml/min. The eluate was c~)llected and the label in the fractions was determined by liquid scintillation counting. Radioactive peaks were identified by their elution with synthetic 1-hexadec-yl-2-ac)-IGPC containing specific acyl chains i 18: 1, 18:2, 18:3, 20:4, 20:5 or 22:6) at the sn-2 position. Recovery of total radioactivity with this procedure is > 75% [31]. Ace@ h_ydrolase assay Isolated 10 min

hearts were perfused (15 ml/min) for with Krebs containing 0.25 mg/ml

1104

M.

Triggiani

and

BSA to remove serum-derived acetyl hydrolase. Subsequently, the hearts were perfused with Krebs without BSA for 45 min at NFP, followed by 30 min at LFP and then reperfused for 5 min at NFP. The perfusion fluid was collected at various time points during the experiment. This fluid was centrifuged (2 x , 400 g, 8 min) to assure that the perfusate was cell-free. Acetyl hydrolase activity in the perfusion fluid was determined utilizing a previously described procedure [31]. The perfusate (1 ml, -3-7 ,ug of protein) was then incubated (120 min) with [3H]-PAF (1 pmole) added as a complex with BSA at a final concentration of 0.05 mg/ml. Preliminary experiments demonstrated that this assay was linear in time up to 120 min. At the end of the incubation the lipids were extracted and analyzed as described above. The protein concentration in each sample was determined by the method of Bradford [32]. The specific activity of acetyl hydrolase was expressed as fmoles of [3H]-PAF converted to [3H]-lyso-PAF/min normalized per pg of protein.

LDH assay LDH activity (expressed as mIU/ml) in the perfusion fluid was determined spectrophotometrically according to the method of Rosalki and Wilkinson [33]. Results

Metabolismof PAF in the isolatedperfusedheart Initial experiments were performed to determine whether PAF was metabolized by the isolated guinea-pig heart. Guinea-pig hearts were perfused in a recirculating manner with Krebs containing [3H]-PAF (IO-” M). Figure 1 shows the uptake of labeled PAF by the guinea-pig heart at two different flow rates (15 and 2 ml/min). The data indicate that exogenously provided PAF was removed from the coronary perfusate in a time- and flowdependent manner. Approximately 60% of PAF was removed from the perfusate within 30 min (6 passes) during NFP. In the case of LFP, more than 70% of PAF was removed from the perfusion buffer during the first pass. These data indicate that the isolated heart can remove PAF from the coronary circulation

F. H. Chilton

I 0

1 20

IO Time

I 30

(min)

FIGURE 1. Uptake of [3H]-PAF by isolated perfused guinea-pig hearts. Isolated hearts were initially perfused (15 ml/min for 10 min) with Krebs buffer containing BSA (0.25 mg/ml) to remove serum-derived acetyl hydrolase. The hearts were then perfused in a recirculating manner with [‘HI-PAF complexed to albumin at flow rates of 15 ml/min (Cl) and 2 ml/min (m). At different time intervals, aliquots of the perfusion buffer were removed and the radioactivity was determined by liquid scintillation counting. These data are the means f S.E.M. of three experiments.

and that the flow rate is an important factor in determining the extraction of exogenous PAF. In these experiments, we also characterized the metabolites derived from the catabolism of PAF and their relative distribution in the tissue and in the pet&sate. Table 1 shows that after 30 min of perfusion with labeled PAF at both flow rates (NFP and LFP), the majority (approximately 80%) of the radioactive products in the tissue consisted of unmetabolized PAF. However, small amounts of two metabolites, lyso-PAF and 1-alkyl-2-acyl-GPC, were found in the tissue. No significant difference in the relative percentages of the radioactive products was observed between NFP and LFP. Figure 2 shows the distribution of PAF and its metabolites (lyso-PAF and 1-alkyl-2-acylGPC) in the perfusion buffer as a function of perfusion time. These data indicate that there is progressive appearance of lyso-PAF in the perfusion buffer. This metabolite accounted for 35% of the total radioactivity in the perfusion buffer after 30 min of perfusion (NFP). In contrast, no significant formation of 1-alkyl-P-acyl-GPC was detected in this compartment at any time point tested. Similar

PAF TABLE

1. Labeled

Metabolism

products

in the

in the heart

Heart

perfused

% of radioactivity Product

NFP

(n=4)

78.8 f 4.8 7.7f2.a

PAF Lyso-PAF I-alkyl-P-acyl-GPC

9.2*

1.1

1105 with

[sH]-PAF in the tissue LFP

(n=3)

a1.7k9.4 7.0f4.5 6.3A3.2

Isolated guinea-pig hearts were perfused with [“HI-PAF as described in Figure 1 at flow rates of 15 ml/min (NFP) or 2 ml/min (LFP). After 30 min. the hearts were flushed with Krebs buffer containing BSA (0.25 mg/ml) and immediately homogenized as described in Methods. Lipids were extracted [29] and analyzed by TLC. Tissue uptake of radioactivity (expressed as a percentage of the total radioactivity perfused in the heart) was 47.1 f 2.0% for NFP and 64.3 f 4.2% for LFP.

0

IO

20 Time

30

(mm)

FIGURE 2. Time-course of formation of lyso-PAF in the perfusion buffer. Isolated hearts were perfused (15 ml/min) as described in Figure 1. Aliquots of the perfusion buffer were obtained at different times and lipids were extracted by the method of Bligh and Dyer [29]. Lipids were then separated by TLC as described in Methods. The radioactivity migrating in the area of PAF, lyso-PAF and 1-alkyl-P-acyl-GPC was determined by liquid scintillation counting. The data are expressed as percentages of the total radioactivity in the perfusion buffer. These data are the meansfs.E.n. of four experiments.

data were observed when were used (data not shown).

LFP

conditions

Release of acetyl hydrolase in the coronary perfisate There are several potential explanations for the appearance of labeled lyso-PAF in the coronary perfusate during the perfusion with labeled PAF. For example, the accumulation of lyso-PAF could be due to the release of acetyl hydrolase from the cardiac tissue into

the perfusion buffer. This enzyme, present in the serum [34] and in several cell types [31, 3.5, 34, has been shown to convert PAF to lyso-PAF in a highly selective manner. Alternatively, lyso-PAF could be formed during the catabolism of PAF by the tissue and subsequently released into the perfusion buffer. In order to better understand this phenomenon, we measured the acetyl hydrolase activity in the cell-free perfusion fluid. In these experiments, the hearts were initially perfused (15 ml/min) with BSA (0.25 mg/ml) to remove the serum-derived acetyl hydrolase. The data in Table 2 show that acetyl hydrolase activity is detectable in the perfusion buffer after NFP (15 ml/min for 30 and 45 min). During LFP (2 ml/min for 5, 15 and 30 min) both the acetyl hydrolase activity and the protein of the perfusion fluid content increased as compared to NFP. However, acetyl hydrolase activity increased more than the protein concentration. This resulted in a significant increase (up to three-fold) in the specific activity of the enzyme (normalized to pg of protein) during LFP as compared to basal conditions. The maximum increase in the specific activity of the enzyme was observed during the.first 5 min of LFP. Restoration of the initial perfusion flow (reperfusion) resulted in the return of the levels of the enzyme to the basal levels. Table 2 also shows that, under these experimental conditions, no significant increase of LDH release was observed during LFP as compared to NFP. To investigate the possibility that PAF was deacetylated in the perfusion buffer by a non-

1106

M. Triggiani

TABLE

2. Acetyl

Flow rate (ml/min)

hydrolase

Time (min)

30 45 5 15 30

15

15 2 2 2 15

1

activity

Acetyl activity

and F. H. Chilton

in the perfusion

hydrolase (Units’/ml)

2.0* 1.4 1.8f 1.0 5.8zt2.6 3.9zt l.82 3.8f l.42 1.4f

1.0

fluid Acetyl hydrolase specific activity (Units’/pg of protein)

Protein (,eW)

3.3f 2.1% 4.6z!c 5.2f 4.8% 3.2h

1.0 1.0

1.9

0.30 f 0.09 0.37f0.12 1.01*0.13*

1.1 1.1

0.49 zk0.08?

1.6

0.33zkO.18

0.65+x0.13

LDH activity (mIU/ml)

N.D. 2.6 1.6 2.6 3.7 4.2

Isolated hearts were perfused (15 ml/min, 10 min) with Krebs buffer containing BSA (0.25 mg/ml) to remove serumderived acetyl hydrolase. The hearts were then perfused with Krebs at constant flow (15 ml/min: NFP). Aliquots (I ml) of the perfusion buffer were collected after 30 min and 45 min to establish a baseline of acetyl hydrolase activity. Subsequently, the hearts were perfused at 2 ml/min (LFP) and the perfusion buffer was collected after 5, 15 and 30 min. Finally, the hearts were reperfused at 15 ml/mitt and the perfusion buffer was collected after I min. The perfusion fluid collected at each time point was centrifuged and incubated (120 min) with radiolabeled PAF complexed to BSA. The acetyl hydrolase activity, protein content and LDH activity were monitored as described in Methods. Data on acetyl hydrolase activity and protein content are expressed as the mean& S.E.M. of five experiments. Data on LDH activity are expressed as the mean of two experiments. ’ 1 Unit = 1 fmole of sH-PAF catabolized/min at 37°C; N.D.: not done; yP< 0.05 when compared to basal values (15 ml/min, 30 min).

specific phospholipase A, (PLA,), we performed two types of experiments. The first is based on the observation that acetyl hydrolase is a calcium-independent enzyme whereas most of PLA, are calcium-dependent [34]. Therefore, we measured acetyl hydrolase activity in four samples of perfusate (two from NFP and two from LFP) with or without the addition of EDTA (5 mM). No difference was observed between control and EDTA-treated samples (3.5 f 2.7 and 3.6f 2.8 fmoles/min, respectively). In the second group of experiments, we incubated 1 ,2-dipalmitoyl-n-j3Hjmethyl-GPC (81 Ci/mmole, New England Nuclear) with perfusion fluids from two different heart preparations in the same conditions used for acetyl hydrolase assay. In this case, the conversion of the labeled substrate to l-palmitoyl-2-lyso-n-[3H]-melhyf-GPC was 0.29 fmoles/min. This value was more than IO-fold lower than that obtained in the same samples using C3H]-PAF as a substrate. These two types of experiments suggest that the enzyme released by cardiac tissue is acetyl hydrolase and not a PLA,. Metabolism Many

different

of PAF by ventricular myoqtes cells in the cardiac

tissue could

participate in the metabolism of PAF. A primary candidate would be the myocyte, since, on a mass basis, this is the predominant cell type in the cardiac tissue [371. In addition, the accumulation of PAF in the isolated heart (Table 1) raised the question of whether isolated myocardial cells have the capacity to metabolize exogenous PAF. Therefore, in the next series of experiments, ventricular myocytes were incubated with [3H]-PAF for various periods of time and the labeled metabolites were examined by TLC. Figure 3 shows that PAF is metabolized by ventricular myocytes to one major product, 1-alkyl-Zacyl-GPC. The formation of very small quantities (-5% of total radioactivity) of lyso-PAF was also observed at the earliest time point (15 min). In order to investigate the nature of the acyl chain at the sn-2 position of the 1-[3H]alkyl-2-acyl-GPC formed during the catabolism of [3H]-PAF, this product was isolated, converted to benzoate derivative and analyzed by HPLC. Figure 4 shows the HPLC profile of diradylglycerobenzoate derivatives of the labeled 1-alkyl-P-acyl-GPC obtained from ventricular myocytes incubated with [3H]-PAF. The major acylation product obtained from the myocyte eluted with

PAF Metabolism l

PAF

A Lyso-PAF --

80

0

30

60 Time

90

(mln)

FIGURE 3. Catabolism of [‘HI-PAF in isolated ventricular myocytes. Ventricular myocytes (106/tube) were incubated in 1 ml volume with [3H]-PAF (3 pmoles) complexed to BSA. At different time points the reactions were stopped and the lipids were extracted and analyzed by TLC as described in Methods. The data are expressed as percentages of the total radioactivity in the lipid extract. These data are the means* S.E.M. of four rxperimerits.

in the Heart

1 107

GPC, 1-alkyl-Z-oleoyl-GPC and 1-alkyl-2linoleoyl-GPC. Lyso-PAF has been shown to be an intermediate of the conversion of PAF to I -alkyl-2acyl-GPC in a number of inflammatory cells [38, 391. The data shown in Figure 3 suggest that lyso-PAF does not accumulate within the myocytes but is rapidly and completely acylated by the transfer of a fatty acyl chain to the sn-2 position of the molecule to form I-alkyl-2-acyl-GPC. In order to support this conclusion, we explored the metabolic fate of lyso-PAF in parallel with that of PAP in the myocytes. These experiments showed that lyso-PAF was converted to 1-alkyl-2-acylGPC and that the reacylation profile / > 80% converted to 1 -alkyl-2-arachidonoyl-GPC:) was identical to that of PAF (data not shown). Discussion

1-alkyl-2-arachidonoyl-GPC. This product accounted for >80% of the radioactivity in I-alkyl-2-acyl-GPC. In addition, minor peaks of radioactivity eluted with I-alkyl-2-eicosapentaenoyl-GPC, I-alkyl-2-docosahexaenoyl-

The present study revealed that the level of PAF within the isolated guinea-pig heart can be efficiently regulated by at least two mechanisms. First, PAF can be partially inactivated by the release of acetyl hydrolase into

40 60 Fraction (min) FIGURE 4. Selective reacylation oflyso-PAF with arachidonic acid in isolated ventricular myocytes. I-( ‘HI-alkyl2-acyl-GPC obtained from [3H]-PAF catabolism in the myocytes was isolated, converted to its benzoate derivatives and analyzed by a reverse-phase HPLC system. The retention times of standards containing eicosapcntarnoic acid! docosahexaenoic acid (20: 5/22: 6)) arachidonic acid (20:4~, linoleic acid (18:2) or oleic acid (18: 1) at the sn-2 position are indicated. These data are representative of three separate experiments.

1108

M. Triggiani

and F. H. Chilton

vascular compartments. Second, it can be taken up by the myocytes and converted to inactive products. These two tissue-specific mechanisms may operate in vivo together with others such as the acetyl hydrolases present in serum and in blood cells to maintain low levels of circulating PAF. Although the relative importance of PAF-degrading enzymes in the blood and in the heart is not clear, there may be circumstances in which the catabolic capacity of the tissue may become important. This may be the case, for example, when PAF is produced within the heart (by myocytes, macrophages, etc.) or by blood cells (neutrophils, eosinophils) invading the tissue in different pathological conditions. The flow rate of perfusion appears to be an important factor in the regulation of coronary PAF levels by regulating both the specific activity of acetyl hydrolase released into the vascular compartments and the uptake of PAF from the coronary vascular compartment. In particular, the majority of PAF provided to the heart is taken up in a single pass at a perfusion rate of 2 ml/min. Subsequent passes do not apparently result in further PAF removal from the perfusate (Fig. 1) . There may be several potential explanations for this phenomenon. The first is that PAF is partially converted to lyso-PAF in the perfusion fluid by acetyl hydrolase (as shown in Fig. 2). In this case we should assume that the uptake of lyso-PAF is slower than that of PAF. A second explanation is that the extraction of PAF is concentration-dependent so that, at low concentrations, PAF tends to remain bound to its carrier (albumin). Finally, we cannot exclude that, although PAF uptake is rapid in the first 5 min of LFP, the extraction of the phospholipid may be impaired by prolonging the duration of low flow perfusion. The simultaneous presence of lyso-PAF and acetyl hydrolase in the perfusion buffer suggests that lyso-PAF may be formed by a direct deacetylation of PAF within the coronary vascular compartment. Our data, however, do not exclude the possibility of an intravascular release of lyso-PAF initially formed in the tissue. The increase in acetyl hydrolase activity during LFP is not simply due to a concentration effect of low flow since it exceeds the increase in protein concentra-

tion. Furthermore, the release of acetyl hydrolase is not expression of a non-specific leakage of cytoplasmic enzyme since no significant increase in LDH release was observed at the low flow conditions used in this study. The presence of acetyl hydrolase in the perfusion fluid has a number of potentially important implications. First, the heart may have the potential to inactivate PAF by a mechanism that does not necessarily require the interaction of this mediator with target cells. This “extracellular” inactivation of PAF may be an important step in limiting the amount of PAF available for the cellular effects. In addition, the specific activity of acetyl hydrolase in the per&sate increases during LFP. It is tempting to speculate that the higher acetyl hydrolase specific activity may contribute to the inactivation of PAF generated during myocardial ischemia within the coronary circulation. In addition, the acetyl hydrolase may be responsible for the generation of lysoPAF and, presumably, of other lysophospholipids in the coronary circulation. Although lyso-PAF is considered to be inactive in most biological systems when compared to PAF, lysophospholipid molecules have long been known for their detergent and amphiphilic properties. This is particularly the case of the heart where the accumulation of lysophospholipids is thought to be partially responsible for cell membrane damage and generation of arrhythmias occurring during myocardial ischemia [40]. From a methodological point of view, the presence of acetyl hydrolase in the perfusate may affect the measurement of endogenously synthesized PAF in isolated heart models by artificially lowering the amount of this mediator. Appropriate steps to inactivate the enzyme, such as the acidification of the samples or pretreatment of the tissue with agents that block this enzyme (e.g. phenylmethylsulfonyl fluoride) 1 may be necessary when measuring the synthesis of PAF in this model. It should be pointed out that the amount of acetyl hydrolase released from the isolated perfused heart is relatively low as compared to that of acetyl hydrolase found in plasma. Although a direct comparison of PAF-degrading activity in plasma and perfusion fluid may be difficult because of the different composition of the two media (albumin concentra-

PAF

Metabolism

tion, lipid composition, etc.), acetyl hydrolase activity in the perfusion fluid is approximately loo-fold lower than that in plasma. Our data do not identify the cell type(s) responsible for the release of acetyl hydrolase. Several cell types, including platelets [351, mast cells [32] and macrophages [36J, have been shown to release acetyl hydrolase spontaneously and after appropriate activation. A reasonable source of this enzyme in our model is the endothelial cell, which is directly exposed to the perfusion fluid in the isolated heart model. It is interesting to note that human umbilical vein endothelial cells have a large capacity to catabolize PAF. During this catabolic process large amounts of lyso-PAF are formed, indicating the presence of a high acetyl hydrolase activity in these cells [41]. Experiments are presently under way in our laboratory to verify if endothelial cells release acetyl hydrolase. Once PAF reaches the tissue, a primary target in the heart is presumably the myocyte. Several studies have shown that PAF has direct effects on the isolated human and guinea pig papillary muscle, such as a decreased contractility and modulation of intracellular Na+ and Ca2+ [11, 12, 421. These effects are thought to be exerted via the activation of specific PAF receptors on the myocyte. Isolated myocytes metabolize PAF predominantly to 1-alkyl-2-acyl-GPC. While our data do not address the question of the existence of PAF receptors on myocytes, they suggest that these cells have the capacity to take up PAF and convert it to metabolic products. It should be noted, however, that kinetic data obtained in vitro, in a static system, cannot be directly extrapolated to the in rliz~o situation in which a dynamic interaction between PAF and myocytes occurs. The first step in the catabolism of PAF in the myocyte appears to be the removal of the acetate at the sn-2 position of the molecule generating lyso-PAF. This intermediate does not accumulate and is converted to 1-alkyl-2acyl-GPC by the transfer of a long chain fatty acid at the sn-2 position. This metabolic pathway in the myocyte is similar to that described in a variety of inflammatory cells such as the neutrophil [43], platelet [38], macrophage [.?9] and mast cell [31]. Although our data do not exclude a direct transacylation of PAF to

in the

Heart

1109

1-alkyl-P-acyl-GPC, the observation that lysoPAF has a metabolic fate similar to that of PAF in the myocytes support the conclusion that lyso-PAF is an intermediate in the conversion of PAF to I-alkyl-2-acyl-GPC. The acylation of lyso-PAF in the myocytes shows a striking selectivity for arachidonic acid (conversion of >80% of lyso-PAF to 1 -alkyl-2arachidonoyl-GPC). The selectivity of the reacylation of lyso-PAF is evident when considering that arachidonic acid constitutes only 5-27% of the total fatty acids in phosphatidylcholine in the heart of several species including rats, dogs, pigs and humans [44471. The specificity of this reaction observed in other cell systems has been attributed to an arachidonate-specific coenzyme A-independent transacylase [38, 391. Our data suggest that this biochemical pathway, originally described in inflammatory cells, may also be functional in the myocytes. The ability of the myocyte to accumulate 1 -alkyl-2-arachidonoyl-GPC may also be important in view of the activation of PLA, occurring during myocardial ischemia [48]. In inflammatory cells, 1-alkyl-2-arachidonoyl-GPC has been shown to be acted upon by PLA, with the simultaneous generation of lyso-PAF and free arachidonic acid [499]. If this pathway is functional in the cardiac myocyte, it would provide a mechanism for the simultaneous generation of PAF and arachidonic acid metabolites during ischemia. Taken together, our study demonstrates that PAF in the coronary vascular space can be degraded by acetyl hydrolase released from the tissue, particularly under conditions of reduced perfusion flow. In addition, PAF can be bound by target cells, such as the myocytr, and rapidly metabolized to inactive products. Both these and other non-specific mechanisms, such as the metabolic capacity of serum acetyl hydrolase and various blood cells, are likely to participate in the regulation of PAF levels within the cardiovascular system. Acknowledgements The authors gratefully acknowledge Dr Myron L. Weisfeldt (New York, NY, USA) and Dr Giuseppe Ambrosio (Naples, Italy) for helpful discussion. This work was supported by NIH grants AI 24985 and

1110

M. Triggiani

and F. H. Chilton

AI 26771. M.T. is the recipient of a Fogarty International Fellowship Award (NIH, Bethesda). A preliminary report of this work was

presented at the 62nd Scientific Sessions of the American Heart Association [Circulation 80, II-449 (1989)].

References 1 HENSON, P. M. Release of vasoactive amines from rabbit platelets induced by sensitized mononuclear leukocytes and antigen. J Exp Med 131, 287-304 (1970). 2 BENVENISTE, J., HENSON, P. M., COCHRANE, C. G. Leukocyte-dependent histamine release from rabbit platelets. The role of IgE, basophils and a platelet-activating factor. J Exp Med 136, 13561377 (1972). 3 DEMOPOULOS, C. A., PINCKARD, R. N., HANAHAN, D. J. Platelet-activating factor (PAF). Evidence for I-0-alkyl-2acetyl-sn-glyceryl-3-phosphorylcholine as the active component. A new class of lipid chemical mediators. J Biol Chem 254, 9355-9358 (1980). 4 SHAW, J. O., PINCKARD, R. N., FERRIGNI, K. S., MCMANUS, L. M., HANAHAN, D. J. Activation of human neutrophils with l-O-hexadecyl/octadecyl-2-acetyl-sn-glyceryl-3-phosphorylcholine (platelet-activating factor). J Immunol 127, 1250-1255 (1981). 5 WARDLAW, A. J., MOQBEL, R., CROMWELL, O., KAY, A. B. Platelet-activating factor: a potent chemotactic and chemokinetic factor for human eosinophils. J Clin Invest 78, 1701-1706 (1986). 6 COLUMBO, M., CASOLARO, V., WARNER, J. A., MACGLASHAN, D. W. JR., KAGEY-SOBOTKA, A., LICHTENSTEIN, L. M. The mechanism of mediator release from human basophils induced by platelet-activating factor. J Immunol 145, 3855-386 1 ( 1990). 7 HANAHAN, D. J. Platelet activating factor: a biologically active phosphoglyceride. Annu Rev Biochem 55, 4833509 (1986). 8 BRAQUET, P. L., TOUQUI, L., SHEN, Y., VARGAFTIG, B. B. Perspectives in platelet-activating factor research. Pharmacol Rev 39, 97-145 (1987). 9 KENZORA, J. L., PEREZ, J. E., BERGMANN, S. R., LANGE, L. G. Effects of acetyl glyceryl ether of phosphorylcholine (platelet activating factor) on ventricular preload, afterload, and contractility in dogs. J Clin Invest 74, 1193-1203 (1984). 10 LAURINDO, F. R. M., GOLDSTEIN, R. E., DAVENPORT, N. J., EZRA, D., FEUERSTEIN, G. Z. Mechanism of hypotension produced by platelet-activating factor. J Appl Physiol 66, 2681-2690 (1989). 11 ROBERTSON, D. A., GENOVESE, A., LEVI, R. Negative inotropic effect of platelet-activating factor on human myocardium: a pharmacological study. J Pharmacol Exp Ther 243, 834-839 (1987). 12 ROBERTSON, D. A., WANG, D.-Y., LEE, C. O., LEVI, R. Negative inotropic effect of platelet-activating factor: association with a decrease in intracellular sodium activity. J Pharmacol Exp Ther 245, 124128 (1988). 13 FELIX, S. B., BAUMANN, G., AHMAD, Z., HASHEMI, T., NIEMCZYK, M., BERDEL, W. E. Effects of platelet-activating factor on myocardial contraction and myocardial relaxation of isolated perfused guinea pig hearts. J Cardiovasc Pharmacol 16, 7508756 (1990). 14 JACKSON, C. V., SCHUMACHER, W. A., KUNKEL, S. L., DRISCOLL, E. M., LUCCHESI, B. R. Platelet-activating factor and the release of a platelet-derived coronary artery vasodilator substance in the canine. Circ Res 58, 218229 (1986). 15 JANERO, D. R., BURGHARDT, C. Production and release of platelet-activating factor by the injured heart-muscle cell (cardiomyocyte). Res Comm Chem Pathol Pharmacol 67, 201-218 (1990). 16 PRESCOTT, S. M., ZIMMERMAN, G. A., MCINTYRE, T. M. Human endothehal cells in culture produce plateletactivating factor (I-alkyl-2-acetyl-sn-glycero-3-phosphocholinc) when stimulated with thrombin. Proc Nat1 Acad Sci USA 81, 3534-3538 (1984). 17 MCINTYRE, T. M., ZIMMERMAN, G. A., PRESCO~, S. M. Leukotrienes C, and D, stimulates human endothelial cells to synthesize platelet-activating factor and bind neutrophils. Proc Nat1 Acad Sci USA 83, 2204-2208 (1986). 18 TRIGGIANI, M., HUBBARD, W. C., CHILTON, F. H. Synthesis of 1-acyl-2-acetyl-sn-glycero-3-phosphocholine by an enriched preparation of the human lung mast cell. J Immunol 144, 47734780 (1990). 19 TRIGGIANI, M., SCHLEIMER, R. P., WARNER, J. A., CHILTON, F. H. Differential synthesis of I-acyl-2-acetyl-snglycero-3-phosphocholine and platelet-activating factor by human inflammatory cells. J Immunol 147, 668666 (1991). 20 LOTNER, G. Z., LYNCH, J. M., BETZ, S. J., HENSON, P. M. Human neutrophil-derived platelet-activating factor. J Immunol 124, 676684 (1980). 21 CROMWELL, O., WARDLAW, A. J., CHAMPION, A., MOQBEL, R., OSEI, D., KAY, A. B. IgG-dependent generation of platelet-activating factor by normal and low density human eosinophils. J Immunol 145, 3862-3868 (1990). 22 TOUQUI, L., HATMI, M., VARCAFTIG, B. B. Human platelets stimulated by thrombin produce platelet-activating factor (I-0-alkyl-2-acetyl-sn-glycero-3-phosphocholine) when the degrading enzyme acetyl hydrolase is blocked. Biochem J 229, 811816 (1985). 23 LEVI, R., BURKE, J. A., Guo, Z.-G., HATTORI, Y., HOPPENS, C. M., MCMANUS, L. M., HANAHAN, D. J., PINCKARD, R. N. Acetyl glyceryl ether phosphorylcholine (AGEPC). A putative mediator of cardiac anaphylaxis in the guinea-pig. Circ Res 54, 117-124 ( 1984).

PAF 24

25

26 27 28 29 30 31 32 33 34

35 36 37 38

39

40 41

42 43 44 45

46 47 48 49

Metabolism

in the

Heart

1111

MONTRUCCHIO, G., ALLOATTI, G., TETTA, C., DE LUCA, R., SAUNDERS, R. N., EMMANUELLI, G., CAMUSSI, G. Release of platelet-activating factor (PAF) from ischemic-reperfused rabbit heart. Am J Physiol 256, Hl236-H1246 (1989). BERTI, F., MAGNI, F., RO~~ONI, G., DE ANGELIS, L., GALLI, G. Production and biologic interactions of prostacyclin and platelet-activating factor in acute myocardial ischemia in the perfused rabbit heart. J Cardiovasc Pharmacol 16, 727-732 (1990). STAHL. G. L., TERASHITA, Z.-I., LEFER, A. M. Role of platelet-activating factor in propagation of cardiac damage during myocardial ischemia. J Pharmacol Exp Ther 244, 898-904 (1988). MXKELSON, J. K., SIMPSON, P. J., LUCCHESI, B. R. Myocardial dysfunction and coronary vasoconstriction induced by platelet-activating factor in the post-infarcted rabbit isolated heart. J Mol Cell Cardiol 20, 547-561 , 1988). BRAQUET, P., PAUBERT-BRAQUET, M., KOL+AI, M., BOURGAIN, R., BUSSOLINO, F., HOSFORD, D. Is there a case t& PAF antagonists in the treatment of ischemic states? Trends Pharmacol Sci 10, 23-30 (1989). BLIGH. E. G., DYER. W. J. A rapid method oftotal lipid extraction and purification. Can J Biochem Physiol37.91 I917 ‘1959). KAMP, 1‘. J.. SANGUINETTI, M. C., MILLER, R. J. Voltage-dependent binding of dihydropyridine calcium r.hannel blockers to guinea pig ventricular myocytes. J Pharmacol Exp Ther 247, 124@1247 (1988). TRIGGIANI, .M., CHILTON, F. H. Influence of immunologic activation and cellular fatty acid levels on the catabolism of platelet-activating factor within the murine mast cell (PT-18). Biochim Biophys Acta 1006. 41-51 11989). BRADFORD, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteinAye binding. Anal Biochem 72, 248-254 (1976). ROSALKI. S. B., WILKINSON, J. Reduction of ketobutyrate by human serum. Nature 188, I1 l&l 111 ,196W. WARDLOW, M. L., Cox, C. P., MENG, K. E., GREENE, D. E., FARR, R. S. Substrate specificity and partial ~&aracterization of the PAF-acylhydrolase in human serum that rapidly inactivates platelet-activating factor. .J Immunol 136, 3441-3446 (1986). SUZI:KI, Y., MIWA, M., HARADA, M., MATSUMOTO, M. Release of aretylhydrolase from platelets on aggrrqaticm with platelet-activating factor. Eur J Biochem 172, 1177120 (1988). ELST.~D. M. R.. STAFFORINI, D. M., MCINTYRE, T. M., PRESCOTT, S. M., ZIMMERMAN, G. A. Platelet-activating factor acetylhydrolase increases during macrophage differentiation. J Biol Chem 264, 8467~8470 ( 1989:. ANV~RSA, P.. LEVXKY. V., BEGHI, C., MCDONALD, S. L., KIKKANA, Y. Morphometry of rxrrrise-indwrd right ventricular hypertrophy in the rat. Circ Res 52, 57-64 (1983). KRAMER, R. M., PATTON, G. M., PRITZKER. C. R., DEYKIN, D. Metabolism of platelet-activating factor ill human platelets. Transacylase-mediated synthesis of 1-O-alkyl-2-arachidonoyl-sn-glycer~~-3-phosphocholinc. ,J Biol Chcm 259, 13316-13320 (1984). ROBINSON, M., SNYDER, F. Metabolism of platelet-activating factor by rat alveolar macrophages: lyso-PAF as an obligatory intermediate in the formation ofalkyarachidonoyl glycerophosphorholinr species. Biochim Biophys Acra 837, 52-56 (1985). CORR, P. B., GROSS, R. W., SOBEL, B. E. Amphipatic mctabolitrs and membrane dysfunction in whemic myocardium. Circ Res 55, 1355154 (1984). BLANK, .M. L., SPECTOR, A. A., KADUCE, T. L., LEE, T.-C., SNYDER, F. Metabolism of platelet-activating factor ( Ialkyl-2-acetyl-sn-glycero-3-phosphorholine) and 1-alkyl-2-acetyl-sn-glycerol by human endothelial cells &whim Biophys Acta 876, 373-378 (1986). TAMARGO, .J., TEJERINA, T., DELGADO, C., BARRIGON, S. Electrophysiological effects of platrlet-activatirlg factor {PAI’-acether) in guinea-pig papillary muscles. Eur J Pharmacol 109, 219-227 [ 1985j. CHILTON, F. H., O’FLAHERTY, J. T., ELLIS, J. M., SWENDSEN, C. L., ~YYKLE, R. L. Selective acylation uf Iyw platelet-activating factor by arachidonate in human neutrophils. J Biol Chem 258, 7268-7271 (19831. SHAIKH, N. .4., DOWNAR, E. Time course of changes in porcine myocardial phospholipid levels during ia(.hemia. .\ reassessment of the lysolipid hypothesis. Circ Res 49, 316-325 (1981). CHIEN, K. R., HAN, A., SEN, A., BUJA, L. M., WILLERSON, J. T. Accumulation of unesterified arachidonic acid in ischemic canine myocardium. Relationship to a phosphatidylcholinr deacylation~reacylation cycle and thr depletion of membrane phospholipids. Circ Res 54, 313-322 (1984). BENEDIKTSDOTTIR, v. E., GUDBJARNASON, S. Modification of the fatty acid composition of rat heart wrwlrmm,~ with dietary cod liver oil, corn oil or butter. J Mel Cell Cardiol 20, 141-147 (19881. ROQVELIN, G., GUENOT, L., ASTORG, P. O., DAVID, M. Phospholipid conwnt and fatty acid composition (11‘ human heart. Lipids 24, 775-780 (1989). KAWACXXHI, H., YASUDA, H. Prostacyclin biosynthesis and phospholipasr activity in hypoxic rat myocardium.