Acyl-CoA synthetase activity in Plasmodium knowlesi-infected erythrocytes displays peculiar substrate specificities

Biochimica er Biophysics Acra 958 (1988) 1-9 Elsevier

BBA 52705

Acyl-CoA synthetase activity in Plasmodium knowlesi-infected

erythrocytes

displays peculiar substrate specificities Bruno D. Beaumelle UA 530 CNRS, INSERM (Received

Key words:

and Henri J. Vial U.58, Monipellier (France)

8 July 1987)

Acyl-CoA synthetase; Acyl-CoA; Fatty acid: Phospholipid; (Plasmodium knowlesi infection); (Simian erythrocyte)

Lipid metabolism;

In its blood stages the malaria parasite, Plasmodium, displays very high lipid metabolism. We present evidence for an abundant long-chain acyl-CoA synthetase (EC 6.2.1.3) activity in Plasmodium knowfesi-infected simian erythrocytes. The activity was found to be 20-fold higher in the schizont-infected (the last parasite stage) than in control erythrocytes. The cosubstrate requirements of the enzyme were similar to those previously reported for acyl-CoA synthetases from other sources. Among the separated reaction products of oleyl-CoA synthetase, only PPi and oleyl-CoA were inhibitory, with Ki over 350 PM. The fatty acid specificity of the parasite acyl-CoA synthetase activity was fairly marked and depended on the unsaturation state of the substrate. The tested fatty acids displayed similar V,,,, whereas their K, ranged from 11 (palmitate) to 59 PM (arachidonate). Finally, experiments involving heat inactivation and separation on hydroxyapatite excluded the presence of a specific arachidonyl-CoA synthetase identical to those present in other cells. On the other hand, fatty acid competition experiments evidenced the existence of at least two distinct enzymatic sites for fatty acid activation in P. know&-infected simian erythrocytes: one is specific for saturated fatty acids and the other for polyunsaturated species, whereas oleate could be activated at both sites.

Introduction During intraerythrocytic development of the malaria parasite Plasmodium, the lipid metabolism of the red cell, which is usually restricted to exchange/acylation/ transacylation reactions [l-3] increases considerably and new phospholipid biosynthetic pathways appear [4,5]. At the end of Plasmodium maturation, the phospholipid and

Abbreviation: sulfonic acid.

Hepes,

4-(2-hydroxyethyl)-l-piperazineethane-

Correspondence: H.J. Vial, CNRS UA 530, INSERM Rue de Navacelles, 34100 Montpellier, France. 0005-2760/88/$03.50

U.58. 60

0 1988 Elsevier Science Publishers

fatty acid contents of the erythrocyte can be as much as 6-times higher [6-lo]. Exogenous fatty acids are probably indispensable for intraerythrocytic Plasmodium development in vivo since they are required for its growth in vitro [6,11]. The fatty acid incorporation capacity of infected cells is more than lo-times higher than that of normal erythrocytes [4,12]. Since they are few biochemical studies of fatty acid metabolism in the Plasmodium-infected erythrocyte [6,11,13], we studied the enzyme that first metabolizes fatty acids, i.e., acyl-CoA synthetase (EC 6.2.1.3). This activation step is necessary for most of the cellular reactions involving fatty acids [14-181. Metabolism such as oxidation and desaturation/

B.V. (Biomedical

Division)

2

elongation of fatty acids has not been detected until now in Plasmodium species 16,131 and, thus, fatty acids incorporated in infected cells are mainly recovered in complex lipids [12]. Most human-cell acyl-CoA synthetase activities examined to date have been found to be the result of the action of both a nonspecific fatty-acyl-CoA synthetase and an arachidonic-specific acyl-CoA synthetase [19-211, except for the red cell, which has no specific arachidonyl-CoA synthetase [19]. Our results show that Plasmodium knowlesi possesses a high acyl-CoA synthetase activity (about 20-times more active than that in control erythrocytes). This enzymatic activity was characterized by kinetic parameters, fatty acid competitions, cosubstrate requirements and pH dependence. The results of the competition experiments are consistent with the existence of a two-site (one for saturated fatty acids, the other for polyunsaturated ones) enzyme or of two enzymes displaying identical behavior during experiments of heatinactivation and separation on hydroxyapatite. The latter experiments rule out the presence of a specific arachidonyl-CoA synthetase similar to those observed in other cells [19-211. Materials and Methods Chemicals

All chemical products were of the highest purity available. Percoll was from Pharmacia (Uppsala, Sweden). All radiolabeled fatty acids, [9,103H]palmitate, [9,10-3H]oleate, [1-‘4C]stearate, [l“C]linoleate, [1-r4C]arachidonate and [1-r4C]oleyl-CoA, were from Amersham. Acyl-CoA standards for thin-layer chromatography were synthesized and purified according to established procedures [22,23]. Oleyl-CoA for product inhibition experiments was from Sigma (St. Louis, MO). Enzyme preparation

Monkeys (Macuca mulatta or Macaca fasciculUris) were infected by P. knowlesi and schizont-infected blood was collected as previously described [7,24]. White blood cells were removed by passage through a cellulose powder column (CFll, Whatman) [25]. Pure schizont-infected erythrocytes were isolated on a discontinuous Percoll gradient [26]. The parasitemia of the preparations was always

greater than 7OW, as measured by giemsa-stained smears. The infected erythrocytes, at a hematocrit of 20% in a phosphate buffer (6.8 mM potassium phosphate, 26.2 mM sodium phosphate, 118 mM sodium chloride and 5.6 mM glucose (pH 6.5), designated here as isolated buffer) were lysed by adding 1 vol. of bidistihed water, followed by sonication (30 W, 30 s) using a Branson sonifier. Control cells were treated in the same way. Membranes were then isolated by centrifugation at 100000 X g for 1 h at 4’ C. The latter step was repeated after addition of fresh buffer, and homogenisation of the membrane pellet. Membrane extracts were then stored at a 20% equivalent hematocrit in isolation buffer at - 80” C and used within 1 month. Long-chain acyl-CoA synthetase assay

Acyl-CoA synthetase (EC 6.2.1.3) activity was determined using the isotopic assay described by Wilson et al. [20]. Briefly, infected-erythrocyte membranes (3 * lo6 cells) or uninfected-erythrocyte membranes (10’ cells) were incubated for 3 or 6 min, respectively, in 0.15 ml of a standard reaction mixture containing 15 pmol Tris-HCl (pH 8), 1 pmol ATP, 100 nmol CoA, 750 nmol ditbiothreitol, 3 pmol MgCl,, and 0.04 ml of radiolabeled fatty acid (200000 dpm) solution in 50 mM NaHCOJ7.5 mM Triton X-100. The reaction was terminated by adding 2.25 ml of stop solution, i.e., isopropanol/heptane/2 M sulfuric acid (40: 10: 1, v/v), followed by 1.5 ml of heptane, 1 ml of H,O, and vigorous vortexing. After centrifugation for 5 min at 2000 x g, the upper phase was discarded. The lower aqueous phase was then washed twice with 2 ml of heptane containing 4 mg/ml palmitate and finally with 2 ml of heptane. An aliquot (1 ml) of the aqueous phase was counted in 4 ml of scintillation fluid (Packard). The blanks consisted of tubes without ATP or CoA, i.e., those giving rise to minimum activity compared to other controls, such as when boiled cells were used as the enzyme source or stop solution was added at the beginning of the incubation. Identification of labeled products

After neutralization of the aqueous phase to pH 6 with 1 N NaOH, and reaction with neutral

3

hydroxylamine [27], the hydroxamate derivatives were then chromatographed on silica gel G plates (Merck, Darmstadt, F.R.G.). Chloroform/methanol/ water mixtures in volumetric proportions of 95: 5.6:0.5 1281 or 68:25:4 [29] were used as soIvents. More than 80% of the radioactivity of the aqueous phase comigrated with authentic standards derived from the corresponding acyl-CoA in either system. We also checked that neutral hydroxylamine did not attack the ester bonds of lipids, as previously reported [29]. Since P. knowZesi has a very active phospholipid metabolism f4], we verified that no radiolabeled phospholipids were formed during the reaction, even after a 6 min incubation. In this case, the reaction was stopped by chloroform/methanol (2 : 1, v/v), followed by lipid extraction [30]. Aql-CoA hydrolase assay Hydrolysis of [l-‘*C]oleyl-CoA was measured as the disappearance of heptane-insoluble 14C, using the same procedure as for the acyl-CoA synthetase assay. Heat inactivation of acyI-CoA synthetase activities

Membranes were heated for various times at 45 “C in isolation buffer supplemented with 6.67 mM ATP, 20 mM MgCl, and 20 mM 2mercaptoethanol 11201. After cooling on ice, palmitoyl-, oleyl- and arachidonyl-CoA synthetase activities were measured using 0.15 mM concentration of [ 3H]palmitate, [ 3H]oleate and [ 14C]arachidonate, respectively, as substrates. Separation of acyi-CoA synthetases membrane solubi~zation and hydroxyapatite chromatographies were performed exactly as previously described [19], except that hydroxyapatite chromatography was performed using phosphate buffers (pH 6.5) at + 4’ C. Other methods Proteins were determined according to the method of Lowry et al. 1311. All kinetics corresponded to a Michaelis-Menten model. Enzymatic kinetic parameters were deduced from Lineweaver-Burk plots. The best-fit lines were determined using the least-squares method. When the substrate concentration was lowered by more

than 10% by the reaction, the modified Lineweaver-Burk equation using the average substrate concentration was used; this calculation could be done [32], since we verified that the reaction was not significantly in~bited by products, nor reversible, and that all cosubstrates were at saturating concentrations under the conditions used (see Results). The extent of the substrate transformation never exceeded 40%, so that the error in the K, determination was less than 2% [32]. All calculations were done on an ‘Apple IIe’ computer using Basic programs. Results Comparison of the acyl-CoA synthetase activities in simian erythrocytes before and ufter infection by P. knowlesi; linearity of the assay Fig. la shows that the linoleyl-CoA synthetase activity varied linearly with the cell number until the numbers reached 2.5 . 107 infected cells or at least 25 - lo7 control cells. After correction for assay time, the acyl-CoA synthetase activity of P. know/es&infected erythrocytes was calculated to be about 20-times that of control erythrocytes. Hence, it can reasonably be concluded that such an increase in acyl-CoA synthetase activity after infection of the erythrocyte is due to parasite enzyme appearance, and that the enzymatic activity displayed by infected erythrocytes is nearly totally provided by parasite acyl-CoA synthetases. Similar results were obtained using either palmitate or arachidonate as the substrate (data not shown). Additionally, Fig. lb shows that the linoleyl-CoA synthetase activity in infected cells was linear with time for at least 12 min. Ofey(-CoA hydrolase activity; reversibility of the activation reaction Less than 5% of the aqueous soluble [ “C]oleylCoA disappeared over the concentration range used (from 10 to 50 PM) in 3 min. Addition of PP, or AMP (1 mM) did not lead to a disappearance of significant amounts of aqueous soluble radioactivity either. Hence, acyl-CoA hydrolase activity as well as the reverse activity of acyl-CoA synthetase were negligible in membrane extracts from schizont-stage P. knowlesi-infected erythrocytes, under the conditions used.

A

TABLE

I

EFFECTOR AND COSUBSTRATE EFFECTS ON THE ACTIVITY OF THE LINOLEYL-CoA SYNTHETASE FROM P. KNOWLESI-INFECTED SIMIAN ERYTHROCYTES Assays were performed using as the substrate 0.15 mM [‘4C]iinoleate (2~~ dpm). as described in Materials and Methods, except when otherwise indicated. Values are mean& SE. (n = 3).

Time

W

-so-

-0

0

5

(min)

I

10

Equivalent

!

15 cell

I

20 number

I

System

Activity (% of control)

Complete - ATP, - CoA - MgCl z - Dithiothreital - Triton X-100 - Pahnitate during extraction

100+ 5 0 23k 3 116k12 173i 6 125k19

25 x IO“)

Fig. 1. Linearity of the assay of the linoleyl-CoA synthetase from control and P. knowlesi-infected monkey erythrocytes. The enzyme activity was measured using [i4C]linoleate as the substrate (0.15 mM, 200000 dpmjassay) as described in Materials and Methods. (A) Linearity of enzyme activity versus the number of cells per assay (for 6 or 4 min for control or infected erythrocytes, respectively). 0, Control erythrocytes; O, P. know/&-infected erythrocytes (P > 85%). (B) Linearity of enzyme activity versus time of the assay of enzyme from pure schizont-infected erythrocytes (5. lo6 cells).

1554 t 73 PM, respectively. Among all the substrates and cosubstrates of the enzyme, only ATP had an in~bito~ effect, at concentrations over 13.3 mM (data not shown). In Fig. 2 it can be seen that the enzyme displayed quite a sharp peak of maximum activity at a rather basic pH of 7.5-8.5.

Inhibitory effects of the reaction products on oleylCoA synthetase acriuity Only AMP (at concentrations

as high as 1 mM)

Influence of effecters and pH-dependence of longchain acyl-CoA synthetase activity The substrate requirements of the enzyme in schizont-infected cell were measured using [‘“Cllinoleate as the substrate (Table I). Linoleyl-CoA synthetase activity was totally dependent on the presence of ATP and CoA, but only partially on was used in the presence of Mg’+. Dithiothreitol the assay to ensure that all the CoA was available in its reduced form. This effect, as well as that of palmitate added during the extraction procedure, appeared to increase the reproducibility of the assay. Triton X-100 solubilized the fatty acid used as the substrate, although it had an inhibitory effect on enzyme activity. We did not test other detergents, but we verified that all reaction cosubstrates were at saturating concentrations in the assay. The apparent K, of the enzyme for CoA, ATP and Mg” were 48 + 5, 730 t_ 60 and

PH Fig. 2. pH dependence of oleyl-CoA synthetase from P. knowf&-infected monkey erythrocytes. Activity was measured using 0.15 mM 13H]oleate (400000 dpm) as the substrate, as described in Materials and Methods, and using buffers (sodium acetate, sodium maleate, Hepes, Tris and sodium borate) within the pH interval of their pK, + 1.

5

had no inhibitory effect on oleyl-CoA synthetase activity in P. knowlesi-infected simian erythrocytes, as shown in Table II. The most inhibitory product was oleyl-CoA, acting by a competitive mechanism with a K, of about 0.4 mM. PPi displayed a rather low inhibiting activity, with a Ki near 1 mM for the competitive component, but it was the onty product with a rather poor noncompetitive inhibitory effect. The inhibitory effect was greatly enhanced when the products were mixed (in a mot to mol ratio), insofar as the Ki observed under these conditions was far lower than the K; of any single reaction product at the same concentration. It should be noted that since the product concentrations at the end of all the assays never reached 5 PM, product inhibition could not alter the measurement of kinetic parameters. It can thus be assumed that acyl-CoA synthetase activity in P. k~o~~~e~~-infected erythrocytes is not inhibited by its products, under these conditions.

length of the carbon chain, from 11 PM for palmitate to 59 PM for arachidonate, whereas all the fatty acids displayed similar apparent V,,,. The in~bition of fatty acyl-CoA synthetase

Apparent kinetic parameters and fatty acid specificity of acyl-CoA syrtthetase activities Table III shows the Michaelis kinetic parameters of all the fatty acids tested. The apparent K, increases with the extent of unsaturation and the

TABLE

II

INHIBITORY EFFECT OF THE REACTION PRODUCTS ON THE ACTIVITY OF THE OLEYL-CoA SYNTHETASE FROM P. K,VOWLESf-INFECTED ERYTHROCYTES Conversion of [ 3H]oieate to ~3~~]oieyl-CoA was monitored as described in Materials and Methods. The inhibition constants of the linear mixed-type inhibition (reduced to a competitive mechanism when a = m) were calculated from apparent V,,, and apparent K, as described by Segei f32]. -, no inhibition observed until 1 mM. Product

AMP PP, Oieyl-CoA AMP + PP, + oleyl-CoA ’ (0.33 : 0.33 : 0.33, mol/mol) a Concentration of each product total concentration.

Inhibitory

effect

competitive

noncompetitive

(K, (FM))

(aKi

(mM))

_ 814144 36O_i77

5.4+0.7 -

163_+ 41

0.8 rt 0.03

in the mixture

was 1/3 of the

Fig. 3. lnhibition by paimitate and arachidonate arachidonyl-CoA synthetase and palmitoyl-CoA synthetase from P. k~~~/~sj-infected monkey erythrocytes. The conversion of (‘HJpalmitate and [14C]arachidonate into [3H]peimitoyl-CoA and [‘4C]arachidonyl-CoA, respectively, was assayed as described in Materials and Methods. (A) l, No inhibitor; 0, 15 pM palmitate; l. 120 PM arachidonate. (B) l, No inhibitor: n, 20 PM arachidonate; 0, 100 PM palm&ate.

6

activities by their substrates was always competitive; the K, had about the same values as the K,, except for linoleate and especially arachidonate, whose Ki is 45times lower than its K,. Pahnitoyl- and stearoyl-CoA synthetases displayed the same inhibition pattern by other fatty acids: they were highly (i.e., with a low K,) and competitively inhibited only by the other saturated fatty acid and poorly inhibited by the unsaturated species (except for the inhibition of palmitoyl-CoA synthetase by oleate, although there was a noncompetitive component). The least effective inhibitor was arachidonate, with K, consistently over 200 PM. Oleyl- and linoleyl-CoA synthetase activities in P. knowlesi-infected erythrocytes displayed about the same substrate specificities; they were similarly inhibited by all the other fatty acids and their apparent V,, and K, had about the same val-

TABLE

ues. Nevertheless, linoleyl-CoA synthetase activity was more effectively inhibited by stearate than was oleyl-CoA synthetase (but with a noncompetitive component). Oleate was the only substrate whose K, was similar to its Ki for inhibiting the other fatty acyl-CoA synthetase activities. Arachidonyl-CoA synthetase activity showed the most particular inhibition pattern. It was the only one of all the tested acyl-CoA synthetase activities that was not at all inhibited by saturated fatty acids, as can be seen in Fig. 3. On the other hand, oleate and especially linoleate were good inhibitors, the latter inhibiting arachidonyl-CoA synthetase as effectively as arachidonate, (or as 20:3( n - 6) whose K, was 18 + 2 PM). The results showing that linoleate and oleate are rather good inhibitors of arachidonyl-CoA synthetase activity suggest that the schizont stage of P. knowlesi-infected erythrocytes does not have an

III

APPARENT KINETIC KVOWLESI-INFECTED

PARAMETERS AND ERYTHROCYTES

FATTY

ACID

CROSS-INHIBITION

OF ACYL-CoA

SYNTHETASE

FROM

P.

Apparent kinetic parameters and inhibition constants are expressed as mean+S.E. of at least four or two separate experiments, respectively, performed using at least five different substrate concentrations and three inhibitor concentrations. The inhibition constants (K, for the competitive and aK, for the noncompetitive component) of the linear-mixed type inhibition were calculated as described elsewhere [32]. -, no inhibition observed until 0.5 mM. Substrate palmitate Apparent

oleate

stearate

linoleate

arachidonate

kinetic parameter 11,

3

28?

5

30*

5

41+

I

59*

9

29+

6

21+

1

38+

6

461

9

28k

5

7* _

2

13i_ _

3

20+ 1 77+_20

K, aK, Oleate

12* _

3

22* _

9

K, aK1 Linoleate

19k 8 96k20

K, (PM) Vm,. (nmoI/ min per lo9 cells) Inhibitor (pM) Palmitate K, aK, Stearate

K, UK, Arachidonate K, aK,

16Oi50 z 500 315 f 80 > 500

42113 54*21

_ _

98+ 2 126+43

_

163+ _

6

78+23 224 f 70

34+

5

24& _

2

78k21 _

138520 230 * 15

51i

7

175 _

1

15+ _

2

13+ _

5

232 f 21 > 500

99+18 2500

79+23 _

7

0

IO

30

20 Time

40

(mln)

Fig. 4. Heat inactivation of acyl-CoA synthetases from P. knowlesi-infected monkey erythrocytes. Membranes were heated at 45OC for various times and assayed for either palmitoyl- (0) oleyl- (A), or arachidonyl(0) CoA synthetase activities, using as substrates 0.15 mM concentrations of [‘Hlpalmitate, (aH]oleate or [‘4C]arachidonate, respectively. Results are expressed as the percentage of the activity present in ice-kept membrane preparations.

arachidonate-specific acyl-CoA synthetase identical to that described previously in other cells [19-211. This was further confirmed by experiments involving heat inactivation or separation of the enzymatic activity on hydroxyapatite. Heat-inactivation of acyl-CoA synthetase activities Fig. 4 shows the results of a typical experiment in which the arachidonyl-CoA synthetase activity was more heat stable than the oleyl- or palmitoylCoA synthetase activities, whereas the behaviors of the three enzymatic activities were very similar. Effectively, half-inactivation times of 18 + 3, 22 + 4 and 49 _+ 10 min (not significantly different) were deduced from triplicate independent experiments for palmitoyl-, oleyl- and arachidonyl-CoA synthetase activities, respectively. Further evidence that P. knowlesi-infected erythrocytes are devoid of the arachidonyl-specific acyl-CoA synthetase found in other cells was provided by chromatographic separation experiments performed on hydroxyapatite as described by Laposata et al. [19]. The same peak of activity was observed using either [‘4C]arachidonate or [‘Hloleate as the substrate (data not shown). Discussion In the present study we characterized the acylCoA synthetase activity present in the schizont

stage of P. knowlesi-infected simian erythrocytes. The activity of this key enzyme was found to be about 20-times greater than that of simian control erythrocytes assayed under the same conditions. Hence, the general properties of the acyl-CoA P. knowlesi-infected synthetase activity in erythrocytes probably result mainly from parasitic enzyme action. This increase in activity is greater than the increase in the total fatty acid content of erythrocytes at the end of parasite maturation, which we have found to be about 6-fold [7]. Compared to other cell types, this activity is very high. Taking into account the fact that our membrane preparations contained about 22 mg of protein per lOI* infected erythrocytes, the apparent kinetic parameters shown in Table III can be used to calculate that the parasite oleyl-CoA synthetase would have a specific activity of 12 nmol/min per mg protein under the conditions used by Laposata et al. [19], which amounts to about 1.5times the activity recovered in human cells with the highest specific activity [19]. The linoleyl-CoA synthetase activity in P. knowfesi (schizont form) has about the same cosubstrate requirements as acyl-CoA synthetases from other sources. The apparent K, for CoA (48 PM) is similar to that of palmitoyl-CoA synthetase for rat liver microsomes [33-361 and mitochondria [37]. This value is in the range observed for arachidonyl-CoA synthetase in other cells [20,21,38]. On the other hand, the apparent K, of the parasitic enzyme for ATP (730 FM) is close to that of arachidonyl-CoA synthetase for platelets [20]. Lastly, the apparent K, for MgClz is in the range (1-5 mM) observed with enzymes from other sources [20,21,34,36,37]. This is also the case of the optimal pH range (7.5-8.5) of the parasite oleyl-CoA synthetase, which is similar to those reported for other cell types [33,37-391. Triton X-100 was found to be inhibitory in the present study, like all detergents tested elsewhere on rat liver [34,35] or brain [38] enzymes. In the product inhibition experiments, the parasite enzyme was inhibited only by PP, and oleyl-CoA. The apparent K, of the enzyme for the different tested substrates were of the same orders of magnitude as those reported by others [20,21,33, 36,38-431.

8

The parasite acyl-CoA synthetases were inhibited in different ways by saturated or polyunsaturated fatty acids, the members of each class being rather poor inhibitors of the enzyme activity related to the other class. This effect was maximal when ~ac~donyl- and pal~toyl-CoA synthetases were tested with palm&ate and arachidonate as inhibitor, respectively. As shown in Fig. 3, 100 /.LM palmitate (i.e., lo-times its K,) did not at all inhibit arachidonyl-CoA synthetase, whereas arachidonate inhibited palmitoyl-CoA synthetase with a Ki of only about 400 FM (i.e., &fold its K,). If palmitate and arachidonate have been activated at the same site, the K, of each fatty acid as a substrate would have had a value similar to its Ki in inhibiting the activation of the other fatty acid. These data thus suggest that P. knowlesi has at least two long-chain acyl-CoA synthetases or, alternatively, a two-sites enzyme. In the latter case, linoleate, which behaved like arachidonate, would be activated at the same site, specific for polyunsaturated species, whereas palmitate and stearate would be activated at the other site. Oleate would be capable of being activated at both sites, since it was the only fatty acid displaying similar values of K, as a substrate and K, as an inhibitor. This phenomenon can be related to the position of inco~oration on the glycerophospholipid backbone we found in P. knowlesi, i.e., saturated and polyunsaturated fatty acids are incorporated exclusively at positions 1 and 2, respectively, while oleate is incorporated at both positions (unpublished data). Although the kinetic data are consistent with the existence of two enzymatic sites responsible of fatty acid activation, numerous results provide evidence for the absence of an arachidonyl-specific acyl-CoA synthetase in P. knowlesi resembling those previously reported in other cells [19--211: (i) Heat denaturation experiments showed that the half-denaturation times do not differ significantly between oleyl- or palmitoyl-CoA synthetase and arachidonyl-CoA synthetase activities. Experiments of this kind have usually provided clearly different denaturation kinetics when different acyl-CoA synthetases were present, e.g., one nonspecific, and the other specific for either arachidonate [20] or lignocerate [44].

(ii) In our work, Iinoleate rather effectively inhibited the parasite arachidonyl-CoA synthetase activity, which has not been observed in cells containing significant amounts of arachidonicspecific acyl-CoA synthetase (20,211. (iii) A specific arac~donyl-CoA synthetase differing from the nonspecific enzyme was not observed after chromatography on hydroxyapatite

D91. In conclusion, in this study we showed that acyl-CoA synthetase activity is present at a very high level in the schizont stage of P. knowlesi-infected simian erythrocytes. Many characteristics of the enzyme, i.e., cosubstrate requirements, optimum pH and product inhibition pattern. were found to be similar to those reported for acyl-CoA synthetases from other sources [20,21,33-411. Nevertheless, its fatty acid specificity appears to be largely dependent on the extent of unsaturation of the substrate; the cross-inhibition experiments between palmitate and aracbidonate provide evidence for the existence in P. knon~I~~~of at least two different enzymatic sites for activating longchain fatty acids. This substrate specificity does not appear to arise from the specific arachidonylCoA synthetase activity previously described [19-211. The high acyl-CoA synthetase activity in Plas~odiu~-infected erythrocytes is yet another component of the very active lipid metabolism of this parasite. Since fatty acids are necessary for the in vitro development of Plasmodium 16,111, fatty acid metabolism, especially the first step characterized here, is a potential target for a chemotherapeutic approach, for example, by the use of specific inhibitors of acyl-CoA synthetase ]45]. Acknowledgements This work was supported by the UNDP/ World Bank/WHO special program for Research and Training in Tropical Diseases (grant T16-181-MZ15B). We greatly appreciated the help of M.H. Authier and M. Cayzac, who performed the protein determinations. We are also indebted to J. Sainte Marie and J. Broussal for technical assistance, and to Professor A. Bienveniie for fruitful discussion.

9

References I Mulder, E. and Van Deenen. L.L.M. (1965) Biochim. Biophys. Acta 106, 106-117. 2 Mulder, E., Van Den Berg, J. W.O. and Van Deenen, L.L.M. (1965) Biochim. Biophys. Acta 106, 118-127. 3 Renooij, W., Van Golde, L.M.G., Zwaal, R.F.A. and Van Deenen, L.L.M. (1976) Eur. J. B&hem. 61, 53-58. 4 Vial, H.J., Thuet, M.J., Broussal. J.L. and Philippot, J.R. (1982) J. Parasitol. 68, 379-391. 5 Vial, H.J., Thuet, M.J. and Philippot, J.R. (1982) J. Protozool. 29, 258-263. 6 Holz, G.G. (1977) Bull WHO 55, 2377248. ‘7 Beaumelle. B.D. and Vial, H.J. (1986) Biochim. Biophys. Acta 877, 262-270. 8 Angus, G.N., Fletcher. K.A. and Maegreth, B.G. (1971) Ann. Trop. Med. Parasitol. 6.5, 429-439. 9 Rock. R.C., Standefer, J.C., Cook, R.T., Little, W. and Sprinz, H. (1971) Comp. B&hem. Physiol. 38B, 425-437. 10 Vial, H.J., Philippot, J.R. and Wallach, D.F.H. (1984) Mol. B&hem. Parasitol. 13, 53-65. 11 Sherman, I.W. (1983) in Malaria and the Red Cell (Ciba Foundation Symposium 94), pp. 206-221, Pitman, London. 12 Rock, R.C. (1971) Comp. Biochem. Physiol. 40 B, 893-906. 13 Haldar, K., Ferguson, M.A.J. and Cross. G.A.M. (1985) J. Biol. Chem. 260.4969-4974. 14 Oliveira, M.M. and Vaughan, M. (1964) J. Lipid Res. 5, 156-162. 15 Riendeau, D. and Guertin, D. (1986) J. Biol. Chem. 261, 976-981. 16 Marinetti, G.V. and Cattieu, K. (1982) Biochim. Biophys. Acta 685, 109-116. 17 Clark, D.L. and Queener, S.F. (1985) B&hem. Pharmacol. 34,4305-4310. 18 Bremer, J. (1983) Physiol. Rev. 63, 1420-1480. 19 Laposata, M., Reich, E.L. and Majerus, P.W. (1985) J. Biol. Chem. 260, 11016-11020. 20 Wilson, D.B., Prescott, S.M. and Majerus, P.W. (1982) J. Biol. Chem. 257, 3510-3515. 21 Taylor, A.S., Sprecher, H. and Russel, J.H. (1985) Biochim. Biophys. Acta 833, 229-238. 22 Reitz. CR., Lands, W.E.M., Christie, W.W. and Holman, R.T. (1968) J. Biol. Chem. 243, 2241-2246. 23 Okuyama, H., Lands. W.E.M., Christie, W.W. and Gunstone, F.D. (1969) J. Biol. Chem. 244, 6514-6519.

24 Vial, H.J., Thuet, M.J. and Philippot, J.R. (1984) Biochim. Biophys. Acta 795. 372-383. 25 Homewood, CA. and Neame, K.D. (1976) Ann. Trop. Med. Parasitol. 70, 249-251. 26 Saul, A., Myler, P., Elliot, T. and Kidson, C. (1982) Bull. WHO 60, 755-759. 27 Rosenfeld, I.S., D’Agnolo, G. and Vagelos, P.R. (1975) Anal. B&hem. 64,221-228. 28 Mishina, M., Kamiryo, T., Tashiro, S. and Numa, S. (1978) Eur. J. Biochem. 93, 347-354. 29 Kaufman, J.F., Krangel, MS. and Strominger, J.L. (1984) J. Biol. Chem. 259, 7230-7238. 30 Folch, J., Lees, M. and Sloane-Stanley, G.H. (1957) J. Biol. Chem. 226, 497-509. 31 Lowry, O.H., Rosebrough, N.J., Fan, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 32 Segel, I.H. (1975) Enzyme Kinetics, John Wiley and Sons, New York. 33 Bar-Tana, J., Rose, G. and Shapiro, B. (1971) B&hem. J. 122, 353-362. 34 Pande, S.V. and Mead, J.F. (1968) J. Biol. Chem. 243, 352-361. 35 Tanaka, T., Hosaka, K., Hoshimaru, M. and Numa, S. (1979) Eur. J. B&hem. 98, 165-172. 36 Normann, P.T., Thomassen, MS., Christiansen, E.N. and Flatmark, T. (1981) B&him. Biophys. Acta 664, 416-42’7. 37 Philipp, D.P. and Parson, P. (1979) J. Biol. Chem. 254, 10785-10790. 38 Reddy, T.S. and Bazan, N.G. (1983) Arch. B&hem. Biophys. 226, 125-133. 39 McLean Grogan, W. and Huth, E.G. (1986) Lipids 21, 11-16. 40 Morisaki, N., Saito, Y. and Kumagai, A. j1983) Biochim. Biophys. Acta 752, 301-306. 41 Suzue, G. and Marcel, Y.L. (1972) J. Biol. Chem. 247, 6781-6783. 42 Kameda, K. and Nunn, W.D. (1981) J. Biol. Chem. 256, 5702-5707. 43 Vaswani, K.K. and Ledeen, R.W. (1987) J. Neurosci. Res. 17, 65-70. 44 Bhushan, A., Singh, R.P. and Singh, I. (1986) Arch, Biothem. Biophys. 246, 374-380. 45 Omura, S., Tomoda, H., Xu, Q.M., Takahashi, Y. and Iwai, Y. (1986) J. Antibiot. 39, 1211-1218.