Biosynthesis of Phosphatidylglycerol in Isolated Mitochondria of ...

Plant Physiol. (1994) 105: 1269-1274

Biosynthesis of Phosphatidylglycerol in lsolated Mitochondria of Etiolated Mung Bean (Vigna radiizta 1.) Seedlings' Regina Criebau and Margrit Frentzen*

lnstitut für Allgemeine Botanik, Universitat Hamburg, Ohnhorststrasse 18, D-22609 Hamburg, Cermany

castor bean endospeh have been detennined (Moore, 1974), whereas PGPase activities have not been measured independently of PGPS activities in plant organelles. Furthermore, direct evidence is still missing about whether the plant cardiolipin synthase, as for the enzyme of other eukaryotes (Daum, 1985; Tamai and Greenberg, 1990; Schlame et al., 1993), utilizes CDP-acy12Gro in the synthesis of cardiolipin (Fig. 1).In addition, the mechanisms that cause the different fatty acid pattems of acylzGroPGro and cardiolipin from plant mitochondria, as well as the origin of polyunsaturated Cls fatty acids, which are esterified in a distinctly higher proportion in cardiolipin than in acy12GroPGro(Bligny and Douce, 1980; Fuchs et al., 1981; Edman and Ericson, 1987; Dome and Heinz, 1989), have not been elucidated. To approach these problems, we have continued our experiments on de novo biosynthesis of glycerolipids in plant mitochondria. In this paper the properties of both PGPS and PGPase in isolated mitochondria from etiolated mung bean (Vigna radiata L.) seedlings are presented and their substrate specificities are discussed with respect to the fatty acid composition of the glycerolipidssynthesized in mitochondria.

Phosphatidylglycerophosphate synthase (sn-glycerol-3-phosphate:CDP-diacylglycerol phosphatidyltransferase) and phosphatidylglycerophosphate phosphatase were characterized in mung bean (Vigna radiafa 1.) mitochondria. The synthase has a rather broad p H optimum between 7 and 9, whereas the phosphatase has one of about 7. 60th enzymic activities are stimulated by Triton X100 and require divalent cations but differ in their cation specificities. The synthase shows apparent K,,, values of 9 and 3 WM for sn-glycerol-3-phosphate and CDP-diacylglycerol, respectively. Phosphatidylglycerophosphate, in contrast to lysophosphatidicand phosphatidicacid, is effectively dephosphorylated by the phosphatase, which exhibits an apparent K,,, value of 12 p~ for its substrate. Each enzyme shows higher activities with the dipalmitoyl species of its substrate than with the dioleoyl species. These substrate specificities of both enzymes are predominantly based on differences in apparent V,,, values.

In plant cells not only plastids and ER but also mitochondria are capable of de novo biosynthesis of glycerolipids. Mitochondria, like plastids, are semiautonomous with regard to the formation of their membrane lipids and have to import phospholipids from the ER. The biosynthesis capacity of mitochondria is largely confined to acy12GroPGro and diphosphatidylglycerol (cardiolipin), the characteristic and functionally important glycerolipids of these organelles (Moore, 1982; Kinney, 1993; Robinson, 1993). Depending on plant species and organ, acy12GroPis formed either in both mitochondrial membranes or in the outer one (Sparace and Moore, 1979; Frentzen et al., 1990). On the other hand, a11 enzymic activities that catalyze the reaction sequence from acylzGroP to acy12GroPGro(Fig. l), namely CTP:phosphatidate cytidyltransferase, PGPS, and F'GPase, are located in the inner membrane (Douce et al., 1972; Sparace and Moore, 1979). Within this membrane acylzGroPGro presumably serves as a substrate for the biosynthesis of cardiolipin (Fig. 1).Recently, the conversion of acy12GroPGrointo cardiolipin was clearly demonstrated by Schlame et al. (1993) in isolated plant mitochondria for the first time. Little is known about the properties of the PGPS, PGPase, and cardiolipin synthase of plant mitochondria (Douce and Dupont, 1969; Moore, 1974; Schlame et al., 1993). So far, kinetics and requirements of the PGPS from mitochondria of

MATERIALS A N D M E T H O D S

Chemicals

[U-14C]GroP(5.88 kBq nmol-I) was purchased from Amersham Buchler and CDP-decanoyl2Gr0, CDP-palmitoylzGro, CDP-oleoy12Gro,and CDP-acylzGro from egg lecithin were obtained from Serdary Research Laboratories (London, Ontario, Canada) or Sigma. Palmitoy12-and ~leoyl~GroP[U-'~C]GroP were synthesized enzymically from the respective species of CDP-acylzGroand [U-14C]GroPin a way similar to that described by Tamai and Greenberg (1990). The assay contained 0.1 M Tricine-NaOH, pH 8.2, 0.1 M MgClz, 0.5% (w/v) Triton X-100, 300 PM CDPpalmitoy12Gro or 300 PM CDP-oleoy12Gro, 130 I ~ M[U-"C]GroP (1.4kBq nmol-'), and 20 pg of crude membrane protein of Escherichia coli strain HW55/pPGL3008, which overproduces PGPS, in a total volume of 2 mL. After a 30-min ~~

~

Abbreviations: acylGroP, 1-acyl-sn-glycerol-3-phosphate(lysophosphatidicacid); acylzGroP, diacyl-sn-glycerol-3-phosphate(phosphatidic acid); acy12GroPGro,phosphatidylglycerol; acylZGroPGroP,

phosphatidylglycerophosphate; CDP-acy12Gro, CDP-diacylglycerol; GroP, sn-glycerol-3-phosphate;PGPase, phosphatidylglycerophosphate phosphatase; PGPS, phosphatidylglycerophosphate synthase (sn-glycerol-3-phosphate:CDP-diacylglycerolphosphati-

This work was financially supported by the Deutsche Forschungsgemeinschaft. * Correspondingauthor; fax 49-40-82282-254.

dylbansferase). 1269

1270

Criebau and Frentzen

ocyl2 GroP

COP-acyl2Gro

::I1

P-Gro-P

acyl2 GroPGroP

Plant Physiol. Vol.. 105, 1994

of total radioactivity and for the analysis of the reaction products, 1 mL of the chloroform layer was used. To determine PGPase activity, standard incubations were performed for 10 min at 3OoC in the presence of 0.1 M MopsKOH, pH 7, 0.2% (w/v) Triton X-100, 1 m~ M&, 20 p~ acyl~GroP[U-'~C]GroP (1.4 kBq nmol-'), and up to 20 pg of mitochondrial protein in a total volume of 50 pL. After incubation lipids were extracted with 240 pL of chloroform: methanol (l:l,v/v) containing 50 pg of acylZGroPGroand 100 pL of 0.2 M H3P04 and 1 M KCl. After phasc? separation 100 pL of the chloroform layer were used for separating substrate and reaction product by TLC. Lipid Analysis

P-Gro

P-Gro-P

o c y l z GroPGro

cardiolipin

Figure 1. Biosynthetic pathway of acyI&roPGro and cardiolipin from acylzCroP within the inner mitochondrial membrane (1, CTP:phosphatidate cytidyltransferase; 2, PGPS; 3, PGPase; 4, cardiolipin synthase).

incubation at 37OC, lipids were extracted and acylzGroPGroP was purified by TLC (Kelly and Greenberg, 1990). The yield was about 220 kBq of a~yl~GroP[U-'~C]GroP.

Reaction products of the enzyme assays were routinely separated by TLC on Silica-gel 60 plates in chloroform:methanol:glacial acetic acid (65:25:8, v/v). To confirm the identity of the products, they were rechroniatographed in difjferent solvent systems, such as chloroform: methano1:glacial acetic acid:water (50:25:4:8, v/v) or chloroform:methanol:concentrated ammonia (65:25:4, v/v). In addition t:heir water-soluble products obtained by deacylation with 0.1 M sodium methoxide were separated on cellulose plates in isopropano1:concentratedammonia:wai:er (7:3:1, v/ v) or in ethano1:l M ammonium acetate (pH 7.4) (65:35, v/v) (Chang and Kennedy, 1967a; Poorthuis and Hostetler, 1975). Labeling within the polar head groups of the reaction products was confirmed by analyzing their hydroly sis products after phLospholipase C digestion (Zwaal and Roelofsen, 1974). Radioactivity on plates was detected by scannirig the plates with an automatic TLC linear analyzer and quantified by scintillation counting. RESULTS A N D DISCUSSION

Purification of Mitochondria

Assay Conditions of PGPS

Mitochondria were isolated from 3-d-old mung bean (Vigna radiata L.) seedlings germinated at 3OoC in the dark according to the procedure of Neuburger et al. (1982), but instead of phosphate buffer Mops-NaOH was used in a11 media. This procedure resulted in an efficient separation of mung bean mitochondria from plastidial and ER membranes. Purified mitochondria were finally resuspended in 20 m~ Pipes-KOH, pH 7.2, and 5 m~ DTT, mixed with glycerol to a final concentration of 50% (v/v), and either used directly for enzyme assays or stored at -2OOC. The protein concentration in mitochondrial fractions was determined according to the method of Bradford (1976).

Assa:y conditions of PGPS from mung bean rnitochondria were optimized using [U-'4C]GroP as tracer molwules. Incorporation of GroP into lipophilic reaction products was strictly dependent on exogenously added CDP-acy12Gro (Table I), suggesling that the endogenous CDP-acylzGro pool within the mitochondrial membranes was negligibly lovu. The analysis of the reaction products revealed that acylzGroPGroPand acylzGroPGrowere labeled (Table I) and that both products carried the labeling in their polar head groups. 'These results

Enzyme Assays

Table 1. Labeling rates of total lipophilic products antl the individual ones by mung bean mitochondria from [U-4C]CroP and

PGPS was determined by monitoring the incorporation of [U-14C]GroPinto chloroform-soluble products in the following standard assay unless otherwise stated. The 1OO-pL assay consisted of 0.2 M Pipes-KOH, pH 7.3, 3 m~ MnClZ,0.1% (w/v) Triton X-100, 8 mM DTT, 30 p~ CDP-palmitoylzGro, 65 p~ [U-'4C]GroP (1.4 kBq nmol-I), and up to 80 pg of mitochondrial protein. The reaction was terminated and lipophilic products were extracted after 30 min at 3OoC according to the method of Hajra (1974). For the determination

CDP-palmitoylzCro Standard Assay

Total

Acy1,CroPCroP

Acy1,CroPGro

pKat mg-' proteiri

Complete Minus CDP-acyl,Gro Minus DTT P~US Hg,CIz

50 PM 500 p~

0.63 0.01 0.53

0.08

0.48 0.47

0.42 0.47

0.1 1

,

0.55 0.01 0.42

0.06

Mitochondrial Lipid Synthesis

are consistent with the biosynthesis of acylzGroPGro from GroP and CDP-acylzGrovia PGPS and PGPase activity (Fig. 1). Under standard assay conditions acyl2GroPGro was formed as the main product (Table I). These results suggested that mung bean mitochondria displayed distinctly higher PGPase than PGPS activity as described for the organelles of castor bean endosperm (Moore, 1974). But the proportion of the two products and, thus, the relative activities of the two enzymes varied in dependence on the assay conditions (Table I; Fig. 2). Total incorporation rates of GroP as a function of the pH of the reaction mixture showed that the mitochondrial PGPS has a rather broad pH optimum of about 7.5 (Fig. 2A). Acy1,GroPGro was, however, formed as the main product at a pH value of about 7 only (Fig. 2A). These results indicate that the PGPase has a sharper pH optimum than the PGPS (see below). Highest PGPS activity was measured in PipesKOH buffer, whereas its activity was slightly lower in MopsKOH and distinctly lower in phosphate buffer. On the other hand, Pipes-KOH concentrations from 50 to 300 mM hardly altered the PGPS activity. As depicted in Figure 2C, PGPS activity was appreciably stimulated by Triton X-100, and maximal activity was measured with 0.1% (w/v) Triton X-100. Furthermore, the PGPS activity required divalent cations (Fig. 2B). Highest activities were obtained in the presence of Mn2+in concentrations of about 3 mM. Stimulation of PGPS activity by Mg” was distinctly lower than with Mn2+(Fig. 2B). Even at high Mg2+ concentrations of up to 100 mM, the activity was about 3fold lower than in the presence of 3 mM Mn2+.Concentrations of Ca2+up to 10 mM did not stimulate the enzymic activity. DTT within the reaction mixture affected the incorporation rates of GroP only slightly (Table I), but it was found to stabilize PGPS activity during storage at -2OOC. Maximal incorporation rates of GroP were attained at 60 PM GroP and 20 p~ CDP-palmitoylzGro. Under these conditions the incorporation rates were constant for at least 30 min and proportional to the mitochondrial protein added up to 80 r g . First kinetic analyses were camed out by determining the PGPS activity at various concentrations of CDPpalmitoyhGro at different fixed levels of GroP. The initial velocity pattem thus obtained indicates a sequential reaction mechanism as described for the PGPS of E. coli (Hirabayashi

et al., 1976). Apparent K, values of 9 and 3 PM were calculated for GroP and CDP-palmitoyl,GroP, respectively. These values are very similar to those reported for the mitochondrial PGPS from castor bean endosperm, which also displayed highest activities at pH 7.5 in the presence of Mn2+andTriton X-100 (Moore, 1974). Thus, the mitochondrial PGPS from plant tissue investigated so far have very similar properties. With regard to their strict dependence on divalent cations they resemble the respective enzymes from plastids and microsomes (Marshall and Kates, 1972; Moore, 1974; Andrews and Mudd, 1984),as well as from bacterial cells (Chang and Kennedy, 1967a; Hirabayashi et al., 1976; Carman and Wieczorek, 1980). Their properties, however, differ from the mitochondrial PGPS from yeast and mammalian cells, the activities of which are hardly influenced by divalent cations (Carman and Belunis, 1983; Daum, 1985; KarkhoffSchweizer et al., 1991). Assay Conditions of PCPase

As mentioned above, the proportion of the two reaction products acylzGroPand acylzGroPGroformed by mung bean mitochondria from [U-14C]GroPand CDP-acy1,Gro varied in dependence on the assay conditions (Fig. 2 ) and, thus, indicates differences in the properties of PGPS and PGPase. To determine the PGPase activity independently of the PGPS activity, mitochondrial fractions of mung bean seedlings were incubated with exogenously added palmit~yl,GroP[U-’~C]GroP. The analyses of the reaction products showed that palmitoylzGroP[U-’4C]GroPwas converted to labeled palmitoylzGroPGro only. As depicted in Figure 3A, the PGPase displayed a pH optimum of about 7 as deduced from experiments with in situ synthesized acylzGroPGroP (Fig. 2A). Similar to PGPS, PGPase was stimulated by divalent cations (Fig. 3B) and by Triton X-100 (Fig. 3C). In contrast to PGPS, PGPase showed highest activities in the presence of 1 RIM Mg2+, whereas MnZ+in concentrations higher than 1 ~lu even l inhibited the enzymic activity (Fig. 3B). Substituting Co2+or Ca2+for Mnz+gave results similar to those obtained with Mn2+(Fig. 3B). A dependence on Mg2+ has also been reported for the PGPase activity in pea chloroplasts (Andrews and Mudd, 1984,1985). Similarly to the plastidial enzyme (Andrews and

Figure 2. Effect of p H (A), divalent cation (B), and Triton X-100 (C) on PGPS activity in isolated mung bean mitochondria. lncorporation O], rates of [U-’4C]CroP (total [O, acyl2GroPGro [O], and acyl&roPGroP [A]) were measured in A: a mixture of Mes, Mops, Tricine, 2-(N-cyclohexylamino)ethanesulfonic acid and 3-(cyclohexylamino)-1-propanesulfonic acid each at 40 mM; in B: the presence of the indicated concentrations of MnCI2 (O) and MgCI, (O); and in C: the presence of the indicated concentrations of Triton X-100. Other conditions were the standard assay conditions.

0.6 c

.-c u

’c

K

r

1271

44

c

a

1 a.

o. 2

6

10

8

PH

2

b

6

Cationr [mWl

8

0.2

0.b

0.6

Triton X-100 [%I

0.0

-~

~

1272

Griebau and Frentzen

Mudd, 1985), as well as that of cauliflower mitochondria (Douce and Dupont, 1969), but unlike that of spinach leaf microsomes (Marshall and Kates, 1972), 0.5 m HgC12completely inhibited the PGPase activity of mung bean mitochondria but hardly affected the PGPS activity (Table I). With regard to its pH optimum, its dependence on Mg2+, and its stimulation by Triton X-100, the PGPase of mung bean mitochondria resembles more closely the respective enzyme of E. coli (Chang and Kennedy, 1967b) than those of yeast and mammalian mitochondria (MacDonald and McMurray, 1980; Kelly and Greenberg, 1990). The dephosphorylation rates of palmitoylzGroPGroP by mung bean mitochondria increased with increasing substrate concentrations up to 60 PM giving maximal rates of 12 pKat mg-’ mitochondrial protein. The apparent K, for , is closer to the value palmitoylzGroPGroPwas 12 p ~ which reported for the enzyme from yeast mitochondria (Kelly and Greenberg, 1990) than to that for the enzyme from E. coli (Chang and Kennedy, 1967b) and mammalian mitochondria (MacDonald and McMurray, 1980). To save substrate, PGPase assays were routinely carried out with 20 p~ palmitoylZGroPGroP, giving dephosphorylation rates that were proportional to the amount of protein added up to 25 pg and constant for at least 15 min. In summary, optimal assay conditions for PGPase from mung bean mitochondria differ in certain aspects from those for PGPS. The two enzymic activities can be affected differently, especially by divalent cations and pH of the reaction mixture (Figs. 2 and 3). Under optimal conditions PGPase displayed distinctly higher activities than PGPS. This appears to be also true under in vivo conditions since acy12GroPGroP is not found in mitochondrial membranes of mung beans (Bligny and Douce, 1980). A comparison of the presented results with those reported for PGPS and PGPase of other organisms reveals that both enzymic activities from mung bean mitochondria resemble more closely the respective enzymes of E. coli (Chang and Kennedy, 1967a, 1967b) than those of mitochondria from mammalian and yeast cells (MacDonald and McMurray, 1980; Carman and Belunis, 1983; Daum, 1985; Kelly and Greenberg, 1990; Karkhoff-Schweizer et al., 1991).

Plant Physiol. Vol 105, 1994

Table II. Substrate specificity of PCPase from mung bem mitochondria Dephosphorylation rates were determined under standard PCPase c:onditions with t h e given substrates (15 PM eatzh). Reaction products of acyl,CroP and acylGroP incubations were separated by TLC in ch1oroform:methanol:water (65:25:4, v/v). PCPase Activity

Substrate

pkat

mg-’ protein

Palmit~yl~GroP[U-’~C]GroP Oleoy12C~roP[ U-’4C]GroP

6.1 3.1

Oleoy12[U-’4ClCroP Palmitoyl[ U-’4C]CroP Oleoyl[CJ-’4C]GroP

0.09 0.06 0.1 1

%

1O0 51 1 1

2

Substrate Specificities of PGPase and PGPS

Under conditions giving high PGPase activity, mitochondrial fractions of mung bean seedlings hardly dephosphorylated acyl-GroP or acylzGroP (Table 11). Hence, the PGPase from plant mitochondria shows the same substrate specificity as the phosphatase of E. coli encoded by the pgpA gene but differs from the E. coli phosphatase encoded by the pgpB gene (Icho and Raetz, 1983; Icho, 1988). PGPase of mung bean mitochondria dephosphorylated palmitcly12GroPGroP with distinctly higher rates than oleoy1,GroPGroP (Table 11; Fig. 4B). This subs trate species specificity of the enzyme was due to differences in the apparent V,,, values. As shown in Figure 4A, ths2 PGPS from mung bean mitochondria displayed the same, biit even more pronounced, species specificity than the PGPase. Highest incorporation rates of [U-I4C]GroPinto acy12GroPGroPand acy1,GroPGro were obtained with CDP-palmito ylzGro,lower ones with CDP-acylzGro from egg lecithin, which predomi(Hostetler et nantly consists of CDP-1-palmitoyl-2-oleoylGrc~ al., 1975), and very low ones with CDP-oleoylzC;roand CDPdecanoly12Gro (Fig. 4A). Again, the differencej in enzymic activities determined with the various substrate species were predominantly based on differences in the apparent V,,, values. With regard to its specificities, the PGPS from mung bean nitochondria differs from the respective mzyme from

Figure 3. PCPase activity in isolated mung

bean mitochondria in dependence on pH (A), divalent cation concentrations (B), and Triton X-100 concentrations (C). Dephosphorylation

rates of palmit~yl~GroP[U-’~C]GroP were measured in A, t h e buffer mixture given in Figure 2; in B, the presence of the given concentration of MgClz (O) and MnCI, (O); and in C, the presence of the given concentration of Triton X-100. All other conditions were the standard assay conditions. 1

6

.

1

‘

e PH

10

1

2 Cotions

3

[mHl

4

-

0.1 o 2 0.3 Triton I(-100 [%I

O,&

1273

Mitochondrial Lipid Synthesis LITERATURE CITED

c IE

.-

g

O,&

L

a

F

c

r

O.?

10

20

30

LO

CDP-acyl2Gro h M 1

I a

acyl2GroPGroP b M ]

Figure 4. Substrate species specificities of t h e mitochondrial PCPS (A) and PCPase (B) from mung bean mitochondria. A, Total incorporation rates of [U-14C]CroP into the lipophilic products (acy12GroPCroplus acy12GroPCroP)are given as a function of the concentrations of CDP-palmit~yl~Gro (O), CDP-oleoy12Gro (O), CDP-decanoyl&ro (O),and CDP-acy12Grofrom egg lecithin (A). B, Dephosphorylation rates are given as a function of the concentra(O) and oleoy12GroPtions of palmit~yl~CroP[U-'~CICroP [U-'4C]CroP (O).

mammalian cells (Hostetler et al., 1975; Stuhne-Sekalec and Stanacev, 1989). The results shown in Figure 4 correlate very well with the fatty acid composition of acylzGroPGro from mung bean mitochondria, which is predominantly esterified with palmitoyl groups (Bligny and Douce, 1980). They, however, do not correspond to the fatty acid composition of cardiolipin, which is formed from acylzGroPGro (Schlame et al., 1993) and which contained almost exclusively unsaturated C18acyl groups (Bligny and Douce, 1980). To approach the problem via which reaction sequences the typical fatty acid pattem of cardiolipin is established, experiments are in progress to determine the properties of the cardiolipin synthase, especially its substrate specificities. ACKNOWLEDCMENTS

We thank Dr. Burton Tropp for providing E. cozi strain H W 5 5 and Dr. William Dowhan for plasmid pPGL3008. Received January 10, 1994;accepted April 10, 1994. Copyright Clearance Center: 0032-0889/94/105/1269/06.

Andrews J, Mudd JB (1984)Characterization of CDP-DG and PG synthesis in pea chloroplast envelope membranes. In PA Siegenthaler, W Eichenberger, eds, Structure, Function and Metabolism of Plant Lipids. Elsevier North Holland Biomedical Press, Amsterdam, The Netherlands, pp 124-131 Andrews J, Mudd JB (1985)Phosphatidylglycerol synthesis in pea chloroplasts, pathway and localization. Plant Physiol 79: 259-265 Bligny R, Douce R (1980)A precise localization of cardiolipin in plant cells. Biochim Biophys Acta 617: 254-263 Bradford MM (1976)A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Ana1 Biochem 7 2 248-254 Carman GM, Belunis CJ (1983)Phosphatidylglycerophosphatesynthase activity in Saccharomyces cerevisiae. Can J Microbiol 2 9 1452-1457 Carman GM, Wieczorek DS (1980)Phosphatidylglycerophosphate synthase and phosphatidylserine synthase activities in Clostridium perfringens. J Bacterioll42 262-267 Chang YY, Kennedy EP (1967a)Biosynthesis of phosphatidyl glycerophosphate in Escherichia coli. J Lipid Res 8: 447-455 Chang YY, Kennedy EP (1967b)Phosphatidyl glycerophosphate phosphatase. J Lipid Res 8: 456-462 Daum G (1985)Lipids of mitochondria. Biochim Biophys Acta 822 1-42 Dome AJ, Heinz E (1989)Position and pairing of fatty acids in phosphatidylglycerolfrom pea leaf chloroplasts and mitochondria. Plant Sci 6 0 39-46 Douce R, Dupont J (1969)Biosynthbe du phosphatidylglyckol dans les mitochondries végétales isolées: mise en évidence du phosphatidylglycérophosphate. CR Acad Sci Ser D 268: 1657-1660 Douce R, Manella CA, Bonner WD (1972)Site of the biosynthesis of CDP-diglyceride in plant mitochondria. Biochem Biophys Res Commun 4 9 1504-1509 Edman K, Ericson I(1987) Phospholipid and fatty acid composition in mitochondria from spinach (Spinacia oleracea) leaves and petioles. Biochem J 243 575-578 Frentzen M, Neuburger M, Joyard J, Douce R (1990)Intraorganelle localization and substrate specificities of the mitochondrial acylCoA:sn-glycerol-3-phosphateO-acyltransferase and acyl-CoA:lacyl-sn-glycerol-3-phosphate O-acyltransferasefrom potato tubers and pea leaves. Eur J Biochem 187: 395-402 Fuchs R, Haas R, Wrage K, Heinz E (1981)Phospholipid composition of chlorophyll-free mitochondria isolated via protoplasts from oat mesophyll cells. Hoppe-Seylers Z Physiol Chem 362 1069-1078 Hajra AK (1974)On extraction of acyl and alkyl dihydroxyacetone phosphate from incubation mixtures. Lipids 9 502-505 Hirabayashi T, Larson TJ, Dowhan W (1976)Membrane-associated phosphatidylglycerophosphatesynthetase from Escherichia cozi: purification by substrate affinity chromatography on cytidine 5'diphospho-1,2-diacyl-sn-glycerolSepharose. Biochemistry 15: 5205-5211 Hostetler KY, Galesloot JM, Boer P, van den Bosch H (1975)Further studies on the formation of cardiolipin and phosphatidylglycerol in rat liver mitochondria, effect of divalent cations and the fatty acid composition of CDP-diglyceride. Biochim Biophys Acta 380 382-389 Icho T (1988)Membrane-bound phosphatases in Escherichia coli: sequence of the pgpA gene. J Bacterioll70 5110-5116 Icho T, Raetz CRH (1983)Multiple genes for membrane-bound phosphatases in Escherichia coli and their action on phospholipid precursors. J Bacterioll53 722-730 Karkhoff-Schweizer RR, Kelly BL, Greenberg ML (1991)Phosphatidylglycerolphosphatesynthase expression in Schizosaccharomyces pombe is regulated by the phospholipid precursors inositol and choline. J Bacterioll73 6132-6138 Kelly BL, Greenberg ML (1990)Characterization and regulation of phosphatidylglycerolphosphate phosphatase in Saccharomyces cerevisiae. Biochim Biophys Acta 1046 144-150 Kinney AJ (1993)Phospholipid head groups. In TS Moore, ed, Lipid Metabolism in Plants. CRC Press, Boca Raton, FL, pp 259-284

1274

Griebau and Frentzen

MacDonald PM, McMurray WC (1980) Partia1 purification and properties of mammalian phosphatidylglycerophosphate. Biochim Biophys Acta 620 80-89 Marshall MO, Kates M (1972) Biosynthesis of phosphatidylglycerol by cell-free preparations from spinach leaves. Biochim Biophys Acta 260 558-570 Moore TS (1974) Phosphatidylglycerol synthesis in castor bean endosperm, kinetics, requirements, and intracellular localization. Plant Physiol54 164-168 Moore TS (1982) Phospholipid biosynthesis. Annu Rev Plant Physiol 33: 235-259 Neuburger M, Journet EP, Bligny R, Carde JP, Douce R (1982) Purification of plant mitochondria by isopycnic centrifugation in density gradients of Percoll. Arch Biochem Biophys 217 312-323 Poorthuis BJHM, Hostetler KY (1975) Biosynthesis of bis(monoacylglycery1)phosphate and acylphosphatidylglycerol in rat liver mitochondria. J Biol Chem 250 3297-3302

Plant Physiol. Vol. 105, 1994

RobinsoriNC (1993)Functional binding of cardiolipin to cytochrome c oxidase. J Bioenerg Biomembr 2 5 153-163 Schlame M, Brody S, Hostetler KY (1993) Mitochondrhl cardiolipin in diverse eukaryotes, comparison of biosynthesis reactions and molecular acyl species. Eur J Biochem 2 1 2 727-735 Sparace SA, Moore TS (1979) Phospholipid metabolism in plant mitochondria. Submitochondrial sites of synthesis. I’lant Physiol 6 3 963-972 Stuhne-Sekalec L, Stanacev NZ (1989) Modification of the biosynthesis ,md composition of polyglycerophosphatides in outer and inner mitochondrial membranes by cytidine liponucleotides. Membr Biochem 8: 165-1 75 Tamai KT, Greenberg ML (1990) Biochemical characterization and regulation of cardiolipin synthase in Snccharomyces cerevisiae. Biochim Biophys Acta 1046 214-222 Zwaal RIFA, Roelofsen B (1974) Phospholipase C (phosphatidylcholine cholinephosphohydrolase, EC 3.1.4.3) from Bircillus cereus. Methods Enzymol32 154-161