ELSEVIER Biochimica et Biophysica Acta 1302 ( 1996} 61-68

BB Biochific~a et Biophysica A~ta

The acylation of lysophosphatidylglycerol in rat heart: evidence for both in vitro and in vivo activities

Philip Cheng a, Vernon Dolinsky ~. Grant M. Hatch a.b., " Department qt'Bi,chemisttT umt Moh'cuh~r Biology. Unirersity o.fManitolm. Room P-122 77t) Bammtyne Acem~e. Winnipeg, Manitoba R3E OW3,

C,m,&l b Dep, lronent of htternal Medicim'. Unit er~'ity of Manito/nt. Room P-122 770 B,mmttyne A remw, Winnipeg, Manitoba R3E OW3. Canada

Received 14 December 1995: accepted 17 January 199h

Abstract

The reacylation of lysophospholipids back to their parent molecules is important for attaining the appropriate fatty acyl composition in many phospholipids and for preventing the accumulation of arrhythmia generating lysophospholipids in the heart. ~n this study, we report the presence of an active acyltransferase activity for lysophosphatidylglycerol reacylation to phosphatidylglycerol in rat heart membrane preparations. The activity of acyl-Coenzyme A:l-acylglycerophosphorylglycerol acyltransferase in rat heart subcellular fractions was in the order of microsomal > mitochondrial > cytosol. The activity in lhe membrane fractions were characterized and found to have a pH optimum in the alkaline range. However, significant enzyme activity was observed at physiological pH. With oleoyi-Coenzyme A as substrate, the microsomal activity had a preference lbr lysophosphatidylglycerol substrates in the order of myristoyl > palmitoyl > oleoyl > stearoyl. The apparent K m values for I-palmitoylglycerophosphorylglycerol and oleoyI-Coenzyme A were 9.4 and 7.1 ~M, respectively. In contrast, the mitochondrial activity had a preference for lysophosphatidylglycerol substrates in the order of oleoyl > myristoyi = stearoyl = palmitoyl. The apparent K m values for I-oleoylglycerophosphorylglycerol and oleoyl-Coenzyme A were 17.8 and 18.0/xM, respectively. Both membrane activities were heat labile as pre-incubation at 55°C for ! min completely abolished the activity. However, pre-incubation at 50°C resulted in different profiles of inactivation in both microsomal and mitochondrial fractions. Both membrane activities were inhibited by high concentrations of lysophosphatidylglycerol and affected to a similar extent by various detergents. To demonstrate whether reacylation of lysophosphatidylglycerol to phosphatidyiglycerol occurred in vivo, isolated rat hearts were perfused for 60 min in the Langendofff mode with 0.1 gM I-palmitoylglycerophoslzhoryl[14C]glycerol bound to albumin. l-Palmitoylglycerophosphoryl[14C]glycerol was readily taken up by the isolated perfused rat heart and significant synthesis of phosphatidyl[14C]glycerol was observed. The findings indicate the presence of an acyl-Coenzyme A:l-acylglycerophosphorylglycerol acyitransferase activity in the rat heart subcellular membranes which is capable of catalyzing lysophosphatidyiglycerol acylation to phosphatidylglycerol in vitro and in vivo.

Keywords: Phosphatidylglycerol; Lysophosphatidylglycerol; Acyltransferase; Arrhythmia; (Heart)

1. In t roduc t ion

Phosphatidylglycerol (PG) is an important polyglyc- erophospholipid in mammal ian membranes (for reviews

Abbreviations: CL, Cardiolipin; PG, phosphatidylglycerol: PA. phos- phatidic acid; PGP, phosphatidylglycerolphosphate; KHB. Krebs- Henseleit buffer; CDP-DG, cytidine-5'-diphosphate-l,2-diacyl-sn- glycerol; CTP, cytidine-5'-triphosphate; LPG, ly:~ophosphatidylglycerol; LPG AT, acyl-coenzyme A:l-acylglycerophosphorylglycerol ac2citrans- ferase; oleoyI-CoA, oleoyi-Coenzyme A.

* Corresponding author. Fax: +1 (204) 7830864: e-mail: ghatch@bldghsc.lan I.umanitoba.ca.

0005-2760/96/$15.00 Published by Elsevier Science B.V. PIi S0005-2760(96)000 ! 2-4

see [!,2]). in addition, PG is required for the activity of several enzymes [3]. In the heart, PG comprises approx. 1.5% of the entire cardiac phospholipid mass [4,5]. The de novo biosynthesis of PG occurs via the CDP-DG pathway [6,7]. Biosynthesis of PG commences with the formation of CDP-DG from PA catalyzed by PA:CTP cytidylyitrans ferase. CDP-DG is then converted to PG by sequential action of PGP synthase and PGP phosphatase. The regula- tory mechanisms which govern PG biosynthesis in the

heart are largely unknown. Membrane phospholipids exist in a dynamic flux in

which continuous biosynthesis is countered by continuous degradation. Phospholipids are known to undergo a rapid

62 P. Cheng et al. / Bim'himi,'a et Biophysiva Avta 1302 f 1996) 61-68

deacylation-reacylation process involving phospholipases and acyltransferases (for review see [8]). The deacylation-reacylation cycle for the modification of fatty acids in phosphatidylcholine was first described by William Lands [9]. This cycle was postulated to be reponsible tbr the introduction of polyunsaturated fatty acids into phos- pholipids (for review see [ 10]). For example, injection of [t4C]linoleic acid into rats followed by rapid isolation of the phospholipids of the liver indicated that 95% of the linoleic acid in l-stearoyl-2-1inoleoylphosphatidylethanol- amine entered the molecule by means of deacylation and reacylation [I I]. In vitro LPG AT activity had been de- scribed in several tissues [12,13]. However, the in viw~ reacylation of LPG to PG in the heart had never been reported.

PG is the immediate precursor of CL in the CDP-DG pathway of CL biosynthesis [14]. in the heart, newly synthesized phosphatidic acid is utilized for new PG and CL biosynthesis [5]. In addition, the newly synthesized PG may be preferentially utilized for CL biosynthesis in rat heart mitochorvdria [15]. it was not known if deacylation- reacylation of PG is an obligatory requirement it,, heart mitochondria given the tact that newly synthesized CL undergoes extensive remodelling [16]. Limited information was available on the resynthesis of PG from LPG in the heart. In fact the presence of LPG AT activity in heart mitochondria had never been reported. Recently, we ob- served the presence of a phospholipase A activity directed towards PG in rat heart microsomes and mitochondria [I 7]. Since the majority of this activity was localized to the mitochondria [17] and the majority of PG in the heart is observed in mitochondria [ 15], it is logical to suspect that the heart should possess the ability to reacylate LPG back to D,,.G in the mitochondria. In this study, the in vitro presence of LPG AT activity in rat heart mitochondria is demonstrated. In addition, evidence for the in vivo reacyla- tion of LPG to PG in the intact rat heart is presented.

2. Materials and methods

[m4C]Glycerol-3-phosphate was obtained from Amer- sham, International, [~4C]Oleoyl-CoA was obtained from Dupont, Mississauga, Ontario. Lysophospholipids were ob- tained from Serdary Research Laboratories, Englewood Cliffs, NJ. Thin-layer plates (silica gel 60, 0.25 mm thick- ness) were obtained from Baxter CanLab, Winnipeg, Man- itoba, Ecolite Scintillant was obtained from ICN Bio- chemicals, Costa Mesa, CA. All other biochemicals were of analytical grade and obtained from either Sigma, St. Louis, MO, or Fisher Scientific, Winnipeg, Manitoba. Male Spmgue-Dawley rats (175-200 g) were used throughout the study. Rats had access to tap water and fed ad libitum in a temperature- and light controlled room. Treatment of animals conformed to the Guidelines of the Canadian Council on Animal Care.

2. !. Preparation of subcelhdar fractions

The animals were sacrificed by decapitation and hearts quickly removed, and perfused tbr 2.5 rain in the Langen- dorff mode [18] with an ice cold buffer containing 0.25 M sucrose, 10 mM Tris-HC! (pH 7.4), 2 mM EDTA (ho- mogenizing buffer), to remove blood. A 10% homogenate was prepared in buffer and centrifuged for 10 min at 500 x g (Beckman J2-H with JA-20 rotor). The resulting supernatant was centrifuged at 12 000 x g for 10 min. The resulting pellet was resuspended in 5 ml of homogenizing buffer by 15 strokes of a hand-held Dounce (loose fitting) tissue grinder and ag~in centrifuged at 10000 × g tbr I0 min. The resulting pellet was resuspended in I ml homoge- nizing buffer with a (tight fitting) Dounce tissue grinder and used as the source of mitochondrial fraction tot" assay of enzymes. The post-mitochondrial supernatant was cen- trifuged for 60 min at 100000 × g (Beckman Ultra cen- trifuge with 70.1 Ti rotor) and the resulting pellet resus- pended in 0.5 ml of homogenizing buffer with a (tight fitting) Dounce tissue grinder and used as the source of microsomal fraction. The supernatant ti"om this centrifuga- tion was used as the source of cytosol. Marker enzyme analysis revealed that the mitochondrial h'action was con- taminated with 8% microsomal particles and microsomal fraction with 5% mitochondrial particles [15].

2.2. Assay of LPG AT

LPG AT activity was determined in rat heart subcellular fractions. In 16 × 100 mm test tubes the assay mixture containe~! 50 mM Tris-HCI (pH 9.0-9.5), 40-65 /~M of LPG, 50 btg o1" protein, 60 /,M of [i-14C]oleoyI-Coen - zyme A (spec. act. 2300 dpm/nmol) and deionized water to a final volume of 0.7 ml. The enzyme assay was incubated for 30 min at 25°C and the reaction was termi- nated by the addition of 3 ml chloroform:methanol (2: I, v/v). 0.8 ml of 0.9% KCI was added to facilitate phase separation and the tubes were vortexed and centrifuged at full speed in a bench top centrifuge. The aqueous fraction was removed by suction and the organic fi'action dried under nitrogen gas and resuspended in 25 /xl of chloro- tbrm:methanol (2:1, v /v ) and 20 /xi placed onto 0.5 N oxalate-treated thin-layer plate with PG standard. The plate wa~ developed in a solvent system containing chloroform:methanol:HC! (87:!3:0.2, v / v / v ) . The plate was stained with iodine vapor and silica gel corresponding to PG was removed and [14C]PG was determined by a Beckman Model LS 3801 liquid scintillation counter. In some experiments with microsomal preparations I- pahnitoylglycerophosphorylglycerol and [I- lac]oleoyl- Coenzyme A were omitted and replaced by 40 /zM l- palm i toylglycerophosphoryl[ i 4 C(U)]glycerol (0.5 tzCi/nM).

P. Cheng et al. / Biochinlica el Bieq~hysica Acta 1302 (1996~ 61-68 63

2.3. Preparation and isolation of I t4 CILPG

Dipaimitoylglycerophosphoryl[ ~4C(U)]glycerol was synthesized from [14C]glycerol-3-phosphate as previously described [5]. I-Paimitoyiglycerophosphoryl[ j~C(U)]gly- cerol was synthesized from dipaimitoylglycerophosphor- yl[14C(U)]glycerol via incubation with phospholipase A_, from bovine pancreas, Sigma P-8913 (St. Louis, MO). The incubation mixture contained 50 mM Tris-HCI (pH 8.0), 3 mM CaCL,, 0.1 nM dipalmitoylglycerophosphoryl- [t4C(U)]glycerol (0.5 p, Ci /nM) and 5 mg of phospho- lipase A 2 in a total volume of 0.5 ml. The mixture was incubated with agitation at 37°C for 2 h then terminated by the addition of 3 ml chloroform:methanol (2:i, v / v ) fol- lowed by the addition of i.5 ml 0.9% NaCI to facilitate phase separation. Conversion to I-palmitoylglycerophos- phoryi[14C(U)]glycerol was essentially complete under these conditions. Tubes were centrifuged at full speed in a bench top centrifuge and the upper aqueous layer removed and washed twice with theoretical lower phase. The or- ganic fractions were combined and placed on a 20 x 20 cm 0.4 M borate-treated thin-layer plate. The thin-layer plates were prepared by spraying the plate lightly with 0.4 M boric acid, blotting dry the excess boric acid and then air drying the plates overnight. The plate was developed in a solvent system containing chloroform:methanol:7 M NH4OH (60:30:4, v/v}. I-Palmitoylglycerophosphoryl- [~4C(U)]glycerol migrated with a relative mobility of 0.06-0.10 in this solvent system. PG migrated with a relative mobility of 0.15-0.18. The silica gel correspond- ing to l-palmitoylglycerophosphoryl[14C(U)]gly cerol was removed and l-palmitoylglycerophosphoryl[~4C(U)]- glycerol was extracted by washing the gel three times with 3 ml of chloroform:methanol (2:!, v/v). The extracts were combined and 4.5 ml of 0.9% NaCI was added to initiate phase separation and partition of the l-palmitoylgly- cerophosphoryl[14C(U)]glycerol into the organic phase. The aqueous phase from these washings was washed twice with 3 ml chloroform:methanol (2:!, v /v ) and all organic phases combined, l-Palmitoylglycerophosphoryl[ 14C(U)]. glycerol isolated in this fashion migrated identically with LPG standard on two-dimensional thin-layer plates [4].

2.4. Pe.rfusion t~ isolated rat hearts with radioactive LPG

The animal was sacrificed by decapitation and the heart quickly removed, cleaned of extraneous tissue, and cannu- lated via the aorta. 5 ml of KHB [19] was perfused through the heart with a syringe to remove blood and the heart was subsequently placed on the perfusion apparatus and per- fused in the Langendorff Mode with KHB (5 ml/min) saturated with 95% 0 , , /5% CO 2 at 37°C for 5 min or until electrical stabilization was achieved. An electrode was placed on the aorta and another electrode immersed in the buffer solution bathing the apex of the heart and the

viability of the heart was monitored by electrocardiac recording. This procedure allows assessmer~t of both ven- tricular and atria activities simultaneously. The Po2 of the buffer was 69.6 + 2.0 kPa. Under these conditions the heart remained electrically stable and therelbre viable for up to 4 h of perfusion [20]. Subsequent to stabilization, the hearts were peffused with 12.5 ml of KHB at a flow rate of 4.5 m l / m i n containing 0.1 ~M i-pa!mitoyigly- cerophosphoryi[14C(U)]glyceroi (0.34/,tCi/heart), 0.3 mM oleic acid and 0.3 mM albumin for up to 60 min. Oleic acid was prepared as described [21]. The presence of I-palmitoylglycerophosphoryi[14C(U)]glycerol bound to albumin did not cause cardiac arrhythmia's. Subsequent to perfusion, I0 ml of air was forced through the heart to remove the residual perl'usate in the vessels.

2.5. Isolation and analysis of radioactire LPG and PG

The heart (typically 0.8 g wet weight) was homoge- nized by a 20 s burst of Polytron ~' homogenizer (Kme- matika, Lucerne, Switzerland) in 5 ml chloroform:methanol (2:1, v / v ) containing butylated hydroxy toluene (0.05 mg/ml). Tissue left on the generator probe was removed and homogen ized in another 5 ml of the chloroform:methanol. The homogenates were combined and centrifuged at 2000 rpm for 20 min in a bench top centrifuge to pellet debris. A 100 /xi aliquot of the ho- mogenate was taken for the determination of total radioac- tivity associated with the heart. The homogenate was transferred to 16 x 125 mm screw cap tubes and 5 ml of 0.73% NaCI was added to the homogenate to initiate phase separation. The tubes were shaken for 5 min and then t~entrifuged at 2000 rpm for 20 min in a bench-top Cen- trifuge (Model TJ-6; Beckman Instruments, Mississauga, Ont., Canada). The aqueous phase was removed and 5 mi of 0.9% NaCI was added to wash the organic phase. The tubes were mixed and centrifuged as described above and the organic phase removed and combined with that of the first centrifugation. The aqueous phase was then re-washed with theoretical lower phase. The organic phases were all combined and dried under a stream of N 2 gas and resus- pended in 100/xi of chloroform:methanol (2:1, v/v) . 5/zl was taken for the determination of radioactivity associated with the organic phase. A 20-25/J,I aliquot of the organic phase was placed on a thin-layer plate and phospholipids separated by two-dimensional thin-layer chromatography on borate-treated plates [4]. LPG and PG standards were placed on the plate prior to chromatography. The lipids were visualized with iodine vapor and removed into 7 ml scintillation vials. Finally, 5 ml of Ecolite scintillation cocktail was added. The radioactivity was determined us- ing a Beckman Model LS 3801 Scintillation Counter with internal standards. In some experiments, l-palmitoylg- lycerophosphoryl[14C(U)]glycerol was added immediately prior to homogenization in chloroform:methanol. Protein

64 P. Cheng et aL/ Biochimica et Biophysica Acta 1302 t 1996) 61-68

Table ! Subcellular distribution of LPG AT activity in rat heart subcellular fractions. Rat hearts were homogenized and subcelhdar tractions prepared from the homogenate as described in Section 2

Subcellular fraction Enzyme activity (nmole/min mg)

Microsomes 0.556 + 0.028 Mitochondria 0.036 + 0.006 Cytosol 0.004 + 0.003

LPG AT activity was assayed in microsomes, mitochondria and cytosol using [l-14C]oleoyi-CoA and I-palmitoylglycerophosphorylglycerol as substrate. Each value represents the mean + standard deviation of three hearts.

was determined by the method of Bradlord [22], Phospho- lipid phosphorus mass was determined as described [23].

3. Results

3. i. Reacylation o f LPG to PG occurs in rat heart mem- branes

Recently, we demonstrated that the heart contains phos- phoiipase A activity directed towards PG in microsomes, mitochondria and cytosol which results in the formation of LPG [17]. Thus, we investigated if LPG could be reacy- lated to PG in these subceilular fractions. With oleoyI-CoA and l-paimitoylglycerophosphoryiglycerol as substrates, LPG AT activity was observed in microsomes, mito- chondria and cytosol (Table I). The activity was greatest in microsomes followed by mitochondria and cytosol. The low cytosolic activity was likely due to contamination with membrane particles. Thus, the cardiac membrane fractions had the ability to synthesize PG from LPG in vitro.

3,2. Characterization o f LPG AT act ivi~ in rat heart

the microsomal fraction significant enzyme activity was observed at pH 7.4 (0.10 + 0.02 nmol /min per mg pro- tein).

It was previously demonstrated that hamster heart I- acyl:Coenzyme A acylglycerophosphorylcholine acyltrans- ferase [24] and rat lung l-palmitoyl-lysophosphatidyl- glycerol acyltransferase [13] exhibited substrate specificity. Thus, we investigated the substrate specificity of the rat heart LPG AT for various l-acyl-LPG's. As seen in Table 2, significant LPG AT activity towards various LPG's was observed. The specificity of microsomal LPG AT for various LPG's was in the order of l-myristoyl > l- palmitoyi > I-oleoyl > I-stearoyi. With concentrations of I-palmitoyiglycerophosphorylglyceroi at 0. i mM or greater present in the assay, LPG AT activity was attenuated in a concentration-dependent manner (Fig. I A). On the other hand the specificity of mitochondrial LPG AT for various LPG's was in the order of I-oleoyl > l-myristoyl = i-pal- mitoyl = l-stearoyl (Table 2). With high concentrations ( > 60 /zM) of I-oleoylglycerophosphorylglycerol in the assay, mitochondrial LPG AT activity was also inhibited (Fig. ! B). Thus, high concentrations of LPG inhibited LPG AT activity.

The apparent affinity's for oleoyl-Coenzyme A and LPG of the membrane activities were determined. The activities of the microsomal enzyme at different l- palmitoylglycerophosphorylglycerol concentrations in the presence of a fixed amount of oleoyI-Coenzyme A were determined and the results were depicted in a Eadie- Hofstee plot. From the slope of this line the apparent K m of m i c r o s o m a l LPG AT for I - p a l m i t o y l g l y - cerophosphorylglycerol was determined to be 9.4 p M (Fig. 2A). The activities of the microsomal enzyme at different oleoyI-Coenzyme A concentrations in the pres- ence of a fixed amount of l-palmitoylglycerophosphorylg- lycerol were determined and the results were depicted in a Eadie-Hofstee plot. From the slope of this line the appar-

L ~ AT activity had never been characterized in the heart. ~ u s , we characterized the ability of microsomal and mitochondrial fractions to reacylate LPG to PG. The microsomal LPG AT exhibited maximum activity towards alkaline pH. At the pH optimum of 9.0 the activity was 0.55 +0.03 n m o i / m i n . m g protein. A gradual loss of activity was observed when the pH was decreased towards neutrality. However, significant enzymatic activity (0. I 1 + 0.01 nmol /min , mg protein) was still observed at pH 7.4 indicating that the reacylation reaction may occur at physiological pH. The pH optimum in alkaline condition was similar to that previously reported for rat liver micro- somal LPG AT [12]. The mitochondrial LPG AT exhibited maximum activity towards alkaline pH. At the pH opti- mum of 9.5 the activity was 0.22 + 0.05 n m o l / m i n , mg protein. A gradual loss of activity was observed when the pH was decreased towards neutrality. However, similar to

Table 2 Substrate specificity of LPG AT towards various LPG's in rat heart microsomes and mitochondria

I-Acylglycerophosphorylglycerol Enzyme activity (nmol/min mg protein)

microsomes mitochondria I-Myristoylglycerophosphorylglycerol (14:0) I-Palmitoylglycerophosphorylglyceroi (16:0) I -Oleoylglycerophosphorylglycerol (18:!) I -Stearoyiglycerophosphorylglycerol (18:0)

0.641 -I- 0.037 0.045 4- 0.029

0.546 4- 0.028 0.039 4- 0.028

0.414-t-0.140 0.191 4-0.022

0.094 ::1:0.019 0.036 + 0.018

LPG AT activity in rat heart microsomes and mitochondria were assayed using [1-14C]oleoyI-CoA in the piesence of various LPG's as described in Section 2. Values represent the mean of three hearts.

P. Cheng et al. / Biochimica et Biophysica Acta 1302 (1996161-68 65

0.6

0.6

~'~o~ 0.4 s_ z.

0.2

0.0

!

0.66 .1 .4 1 LPG (raM)

0.3

0.2

| i SF

0.1

0.0

T B

.04 .00 .08 .1 LPG (raM)

Fig. I. LPG AT activity in rat heart membrane fractions in the presence of various concentrations of LPG. LPG AT activity was assayed in rat heart microsomes (A) in the presence of 0.065-1.0 mM concentrations of l-palmitoylglycerophosphorylglycerol or mitochondria (B) in the pres- ence of 0.04-0.1 mM concentrations of I-oleoyiglycerophosphorylg- lycerol. Values represent the mean + standard deviation of three hearts.

mitochondrial and microsomai activities (Fig. 4). By l0 min of pre-incubation at 50°C 80% of the mitochondrial and 55% of the microsomal activities were abolished. We examined the effect of various detergents on LPG AT activity. LPG AT activity was assayed in mitochondrial and microsomal fractions in the absence or presence of 0.05% tyloxapol, CHAPS, Tween 20, Miranol H2M and Triton X-100. Tyioxapol did not markedly affect enzyme activity (Table 3), whereas CHAPS, Tween 20, Miranoi H2M and Triton X-100 were inhibitory. The general in- hibitory effect of detergents was similar to that reported for the cardiac I-acylglycerophosphorylcholine acyltrans- ferase [24]. The microsomal and mitochondrial activities were affected similarly by all detergents tested (Table 3). To determine whether LPG was acylated to PG by transacylase activities, microsomal fractions were incu- bated with ['4C]LPG and the radioactivity incorporated into PG determined. Under our incubation conditions sig- nificant radioactive PG was not formed.

1.0

0.8 J

°6i . 0.4 -~ • ~ ®

0.2

@ 0.0 , , , ,

0.00 0.01 0 .02 0 .03 0.04

v/IS]

ent K m of microsomal LPG AT for oleoyl-Coenzyme A was determined to be 7.1 /xM (Fig. 2B). The apparent K m of mitochondrial LPG AT for I -o leoylg ly- cerophosphorylglycerol was determined to be 17.8 /zM (Fig. 3A) and the apparent K m of mitochondrial LPG AT for oleoyl-Coenzyme A was determined to be 18.0 p,M (Fig. 3B).

We examined the heat inactivation profile of LPG AT. Rat heart microsomal or mitochondrial fractions were pre- incubated at 55°C for up to 10 min and subsequently LPG AT activity was determined. Enzyme activity was rapidly lost by pre-incubation at 55°C in the membrane fractions when assayed with either l-palmitoylglycerophosphorylg- lycerol or 1-oleoylglycerophosphoryiglycerol. By I min of pre-incubation at 55°C LPG AT activity was almost com- pletely abolished (data not shown). Thus, cardiac mem- brane LPG AT activity was heat labile. Milder pre-incuba- tion of the membrane fractions at 40-45°C resulted in little inactivation (data not shown). However, pre-incubation at 50°C resulted in a different rate of inactivation of the

1.0 13

0.8

> 0 . 6

0.4

0.2

@

0.0 , ~ , 0.00 0.02 0.04 0.06 0.08

v/[s]

Fig. 2. LPG AT activity in rat heart microsomes in the presence of various concentrations of oleoyl-CoA and I-paimitoylgly- cerophosphorylglycerol. (A) LPG AT activity was assayed in rat heart microsomes in the presence of 60 /zM oleoyI-Coenzyme A and i -60 /~M I-palmitoylglycerophosphorylglycerol and a graph of activity versus activity/substrate plotted. (B) LPG AT activity was assayed in rat heart microsumes i , the ple,~el~ of 60 /zM I-paimitoylglyccrophosphorylg- lycerol and 1-100 /xM oleoyI-CoA and a graph of activity versus activity/substrate plotted. Values represent the mean of at least two

hearts.

66 P. Cheng et al. / Biochimica et Biophysica Acta 1302 (1996) 61-68

0.4

0,3

> 0,2

0.1

0.0 0.000

A

I I I

0.005 0.010 0 .015 0.020

v/ [S]

0,20

0,16

0.12

0.08

0.04

0.00

0.000

B

O

• •

0

0.004 0.008 0.012

v/[s]

Fig. 3. LPG AT activity in rat heart mitochondria in the presence of various concentrations of oleoyi-CoA and I-oleoylglycerophosphorylg- lycerol. (A) LPG AT activity was assayed in rat heart mitochondria in the presence of 60 g,M oleoyI-Coenzyme A and 0.5-50 /.tM l-oleoylg- lycerophosphorylglycerol and a graph of activity versus activity/sub- strate plotted. (B) LPG AT activity was assayed in rat heart mitochondria in the presence of 40/zM l-oleoylglycerophosphorylglycerol and 0.5-100 ttM oleoyI-CoA and a graph of activity versus activity/suhstrate plotted, Values represent the mean of at least two hearts.

3,3, LPG is reacylated to PG in vivo

The above studies indicated that LPG could be reacy- lated to laG in cardiac membranes in vitro. To determine if the heart had the ability to reacylate LPG to PG in vivo, rat hearts were peffused for 60 min with 0.1 /zM [14C]LPG bound to albumin. Perfusion of the heart with 0.1 p.M [ t 4 C ] L P G bound to 0.3 mM albumin did not result in cardiac arrhythmia. Subsequent to peffusion the hearts were homogenized in chloroform:methanol and the ho- mogenate was separated into aqueous and organic phases and two-dimensional thin-layer chromatography performed on an aliquot of the organic phase and the radioactivity incorporated into LPG and PG determined. As seen in Table 4, approx. 40% of the radioactivity incorporated into the organic phase was observed in PG and the rest in LPG. No formation of radioactive PG was observed when ra- dioactive LPG was added to the heart just prior to homoge- nization. These studies suggested that LPG entered into the heart during peffusion and was converted to PG indicating

i-

1001~

75 ' "~

,_~ u 50 ,c

25

\

\ \ \

0 . . . . . . . . . . J - . . . . . . . . . . . . . . . . . . ] . . . . . . . . L . . . . . . i .........

0 2 4 6 e 10

Time (min)

Fig. 4. Heat inactivation profile of LPG AT activity in rat heart micro- somes and mitochondria. Membrane fractions were pre-incubated for up to 10 min at 50°C and the LPG AT activity subsequently determined in microsomes (closed symbols) and mitochondria (open symbols). Values represent the mean of at least two hearts. The activity at time 0 in microsomes was 0.55 nmol /min per mg protein and in mitochondria 0.19 nmol /min per mg when assayed with I-palmitoylglycerophosphorylg- lycerol and I-oleoylglycerophosphorylglycerol, respectively.

Table 3 The effect of various detergents on the activity of LPG AT in mito- chondria and microsomes

Detergent Percent activity

microsomes mitochondria

Control 100 100 Tyloxapol I i 6 80 CHAPS 59 65 Tween 20 54 39 Miranol H2M 38 28 Triton X- 100 35 26

LPG AT activities were determined in microsomes with 40 ~ M l- palmitoylglycerophosphorylglycerol as substrate and in mitochondria with 40 /~M l-oleoylglycerophosphorylglycerol as substrate. The assay mix- ture contained 0.05% detergent solutions dissolved in assay incubation buffer. Values represent the mean of two hearts, Control activities were 0.5 nmol /min per mg for microsomes and 0.2 nmol /min per mg for mitochondria.

Table 4 Radioactivity inco~)orated into laG in hearts perfused with [14C]LPG

Radioactivity

d p m / g heart % of organic

Lysophosphatidyl[ 14 C]glycerol 16 705 58 Phosphatidyl[ 14 C]glycerol I 15 ! 8 40

Rat hearts were p effused for 60 min with 0.1 /zM [14C]LPG (0,34 p, Ci/he~trt) as described in Section 2 and the radioactivity incorporated into LPG and PG in the heart were subsequently determined by lipid extraction followed by two-dimensional thin-layer chromatography of the organic fraction. Values represent the mean of two hearts.

P. Cheng et al. / Biochimica et Biophysica A cta 1302 (1996) 61-68 67

the inherent ability of the heart to reacylate LPG to PG in vivo. Some radioactivity was found associated with PG in the heart perfusate after 60 min. However, this accounted for only 0.2% of the total organic radioactivity of the perfusate (data not shown). The heart likely contributes minimally to this extracellular pool of PG, since no de- tectable PG phospholipid phosphorus mass was observed in the perfusate of hearts perfused for 60 min.

4. Discussion

The objective of this study was to investigate the pres- ence of LPG AT activity in heart mitochondria and whether the isolated intact heart could resynthesize PG from LPG. The results confirm the presence of an in vitro LPG AT activity in cardiac microsomal fractions. In addition, the presence of an in vitro LPG AT activity in rat hean mitochondria was observed. Finally, the heart perfusion studies with radioactive LPG provide strong evidence for the ability of the heart to reacylate LPG to PG in vivo.

The presence of a microsomal LPG AT activity in rat liver, kidney, lung and heart had been previously reported [12]. The activity was several orders of magnitude lower than the activity directed towards lysophosphatidylcholine reacylation to phosphatidylcholine. The microsomal activ- ity reported in the present study was somewhat higher than previously reported [12] but this may be due to differences in substrate solubility (see below). In contrast, the cardiac enzyme activities observed in our study were several fold lower than that reported in sonicated type II cells [13]. This may possibly be due to the greater requirement for LPG AT activity in these cells, since PG represents 10% of the total phospholipid mass in type II alveolar cells [25] but only 1.5% in the rat heart [4,5].

The microsomal LPG AT activity displayed a prefer- ence for acyl groups of the LPG substrate. It would appear that shorter chain length LPG substrates were preferred over the longer chain LPG's. However, I-oleoylgly- cerophosphorylglycerol proved to be a more preferred substrate than l-stearoylglycerophosphorylglyceroi indicat- ing that specificity of LPG AT for its LPG substrates may be dictated by both chain length and degree of unsatura- tion. In contrast to the microsomal enzyme, the mitochon- drial LPG AT displayed high activity towards l-oleoylg- lycerophosphorylglycerol but limited activity towards the saturated LPG substrates. We cannot, however, exclude the possiblity that the specificity of the membrane LPG AT activities resulted from the different solubility of the vari- ous LPG's in these membranes.

In hearts perfused with radioactive LPG the radioactiv- ity incorporated into PG accounted for 40% of the total radioactivity observed in the organic fraction indicating that LPG was readily converted to PG. This was not surprising since the in vitro studies indicated significant enzyme activity in the physiological pH range and a high

affinity for the LPG substrate. Lysophospholipids are known to promote cardiac an'hythmia's and LPG was the most potent of those iysophospholipids examined [26]. ?G was recently shown to be released from the liver during perfusion [27]. Thus, a serum source of LPG is quite conceivable given the fact that serum also contains signifi- cant Group-If phospholipase A 2 activity [28]. This serum activity may be dramatically increased in a variety of clinical disorders including septic shock, adult respiratory distress syndrome, psoriasis, renal failure, gout and arthri- tis [29]. Indeed, liver peffusion studies have indicated that phospholipase A 2 may be released under both physio- logical and pathophysiological conditions and that this phospholipase A 2 has a preference for PG substrates [30]. Interestingly, renal failure, septic shock, rheumatoid arthri- tis and infection have all been associated with cardiac arrhythmia's [31-34]. Cardiac microsomal membranes are a combination of sarcolemma, Golgi and sarcoplasmic reticulum and significant PG is observed in rat heart microsomes [15]. LPG AT activity was greatest in the microsomal fi'action. This makes sense since the sar- colemma would be the first membrane exposed to LPG generated from outside the cell and the LPG AT activity in this membrane could attenuate the membrane lytic effect of such compounds by resynthesizing them back to PG. The fact that the isolated heart had the ability to reacylate LPG to PG in vivo support this speculation.

The subcellular localization of cardiac acyltransferases has been the subject of considerable debate. In rabbit heart acyitransferase activity directed towards i-acyigly- cerophosphorylcholine was reported to occur exclusively in microsomes [35] or distributed among all subcellular fractions [36]. In myocytes this activity was observed in microsomal and cytosolic fractions [37]. The ability to reacylate both lysophosphatidylcholine and lysophosphati- dylethanolamine had been reported in guinea pig heart mitochondria [38,39]. Cardiac microsomes contain only approx. 27% of the total cardiac PG content [15]. Given the fact that the majority of cardiac PG is localized to the mitochondria the ability to reacylate LPG to PG would be expected to occur in the mitochondrial fraction. In addi- tion, since the rat liver mitochondriai enzymes involved in the synthesis of PG exhibit no acyl selectivity for their substrates [40], the presence of a mitochondrial LPG AT activity would be highly warrented. Indeed, the substrate specificity differences between mitochondriai and microso- real LPG AT activities argue for the presence of an LPG AT activity in the mitochondria. When assayed with l- oleoylglycerophosphorylglycerol, the mitochondrial activ- ity was approx. 50% of th'~' observed in microsomes. In addition, the heat inactivation profiles of the mitochondrial and microsomal enzymes at 50°C differed significantly. Thus, the mitochondrial activity was likely not due to contamination with microsomes.

PG is the immediate precursor of CL in the mito- chondria [14] and the molecular species of CL in bovine

68 P. Cheng et ai. / Biochimica et Biophysica Acta 1302 (1996) 61-68

heart mitochondria contain almost exclusively unsaturated fatty acid [41,42]. At first glance these results might suggest that unsaturated species of LPG are preferred substrates for the mitochondrial LPG AT which could provide PG for the synthesis of CL. However, studies with rat liver mitochondrial membranes indicate that newly synthesized CL undergoes extensive remodelling to attain its appropriate molecular species composition [16]. In addi- tion, extra-mitochondrial PG may be used for CL biosyn- thesis in the heart [I 7]. Thus, the role of this mitochondrial LPG AT activity in the heart may simply be to generate the appropriate acyl composition of PG observed in that organelle.

in summary, the in vitro presence of LPG AT activity in rat heart microsomes is confirmed and now demon- strated in rat heart mitochondria. These enzyme activities appear to have similar kinetic properties with respect to their affinty for substrates and inhibition by detergents. The activities seem to differ in their substrate specificity and their rate of inactivation at 50°C. Finally, evidence is presented for the in vivo reacylation of LPG to PG in the intact rat heart studies. We hypothesize that the cardiac LPG AT activities may play an important role in the metabolism of LPG in the mammalian heart under physio- logical and pathophysiological conditions.

Acknowledgements

The authors wish to thank Dr. Shu Guang Cao for helpful discussions and synthesis of radioactive substrates. This work was supported by a grant from the Heart and Stroke Foundation of Manitoba. G.M.H. is a Heart and Stroke Foundation of Canada Scholar.

References

[1] Hostetler, K,Y, (1982) in Phospholipids (Hawthorne, J.N. and Ansell, G,B,, eds,), pp, 215-61, Elsevier, Amsterdam.

[2] Daum, G, (1985) Biochim. Biophys, Acta 822, 1-42. [3] Sandermann, H. (1978)Biochim, Biophys, Acta 515, 209-237. [4] Poorthuis, B.J,, Yazaki, P,J, and Hostetler, K.Y. (1976) J. Lipid.

Res, 17, 433-437, [5] Hatch, G.M, (1994) Biochem. ], 297, 201-208, [6] Batenburg, J.J,, Klazinga, W. and Van Golde, L.M.G. (1985)

Biochim. Biophys, Acta 833, 17-24. [7] Kiyasu, J.Y,, Pieringer, R.A., Paulus, H. and Kennedy E.P. (1963) J.

Biol, Chem. 238, 2293-2298. [8] Van Den Bosch, H,, Van Golde, L.M.G. and Van Deenen, L.L.M.

(1972) Physiol, Rev. 66, 14-145.

[9] Lands, W.E.M. (1960) J. Biol. Chem. 235, 2233-2237. [10] Irvine, R.F. (1982) Biochem. J. 204, 3-16. [! !] Akesson, B. (1970) Biochim. Biophys. Acta 218, 57-7e [12] Wittels, B. (1973} J. Biol. Chem. 248, 2906-291 !. [13] Funkhouser, J.D., Batenburg, J.J. and Van Golde, L.M.G. (1981)

Biochim. Biophys. Acta 666, i-6. [14] Hostetler, K.Y., Van Den Bosch, H. and Van Deenen, L.L.M.

(1971) Biochim. Biophys. Acta 239, 113-119. [15] Hatch, G.M., Cao, S.G. and Angel, A. (1995) Biochem. J. 306,

759-764. [16] Schlame, M, and Rustow, B. (1990) Biochem J. 272, 589-595. [17] Cao, S.G. and Hatch, G.M. (1995) Biochim. Biophys. Acta 1256,

97-102. [18] Langendorff, O. (1895) Pfluegers Arch. 61,291-332. [19] Krebs, H.A. and Henseleit, K. (1932) Hoppe-Seylers Z. Physiol.

Chem. 210, 33-66. [20] Arthur, G. and Choy. P.C, (1984) Biochim. Biophys. Acta 795,

221 - 229. [21] Cao, S.G. and Hatch, G.M. (1994) Lipids 29, 475-480. [22] Bradtord, M.M. (1976) Anal. Biochem. 72, 248-254. [23] Rouser, G., Siakotos, A.N. and Fleisher, S. (1966) Lipids I, 85-86. [24] Arthur, G. and Choy, P.C. t 1986) Biochem. J. 236, 481-487. [25] Mason, R.J,, Dobbs, L.G,, Greenleaf. R.D. and Williams, M.C.

(1977) Fed Proc, 36, "~697 '~7~'~ [26] Man, R.Y.K., Wong, T. and Choy, P.C. 11983) Life Sciences 32,

1325-1330. [27] Lekka, M., Tokumura, A., Tsuji, H. and Hanahan, D.J. (1993) Arch.

Biochem. Biophys. 302, 380-384. [28] Vadas, P. and Pruzanski, W. (1989)in Phospholipase A~: Role and

Function in Inflammation (Wong, P.Y.-K. and Dennis, E.A. eds.). pp. 83-101, Plenum Press, New York.

[29] Pruzanski, W. and Vadas, P. (1990) in Biochemistry, Molecular Biology, and Physiology of PLA, and its Regulatory Factors (Mukhetjee, A.B., ed.), pp. 209-251, Plenum Press, New York.

[30] Hatch, G.M., Vance, D.E. and Wilton, D.C. (1993) Biochem. J. 293, 143-15(I.

[31] Armstrong, D.S. and Menahem, S. (1993)J. Paediatr. Child Health 29, 309-31 I.

[32] Rokas, S., Mavrikakis, M., lliopoulou. A. and Moulopoulos, S. (1990) Int. J. Cardiol. 26, 75-82.

[33] Odeh, M. (1993) Int. J. Cardiol. 42, 231-238. [34] Charmes, J.P., Ton-That, H., Rince, M., Dupre-Goudable, C. and

Durand, D. (1992} Rev. Pratt. 42, 427-431. [35] Gross, R.W. and Sobel, B.E. (1982)J. Biol. Chem. 257, 6702-6708. [36] Needleman, P. Wyche, A., Spreche, H., Elliot, W.J. and Evers, A.

(1985) Biochim. Biophys. Acta 836, 267-273. [37] Severson, D.L. and Fletcher, T. (1985) Can. J. Physiol. Pharmacol.

63, 944-95 !. [38] Arthur, G., Page, L.L., Zaborniak, C.L. and Choy, P,C. (1987)

Biochem. J. 242, 171-175. [39] Arthur, G., Zaborniak, C.L. and Choy, P.C. (1987} Biochem. Cell

Biol. 65, 1016-1021. [40] Rustow, B., Schlame, M., Rabe, H., Reichmann. G. and Kunze, D.

(1989) Biochim. Biophys. Acta 1002, 261-263. [41] Schlame, M., Horvath, L. and Vigh, L. (1990) Biochem. J. 265,

79-85. [42] Schlame, M., Brody, S. and Hostetler. K.Y. (1993) Eur. J. Biochem.

212, 727-735.