Studies on enzyme-substrate interactions of cholinephosphotransferase from rat liver

BBA 52006

Studies on enzyme-substrate

interactions of cholinephosphotransferase

from

rat liver Gabriele Pontoni, Caterina Manna, Antonio Salluzzo, Luisa de1 Piano, Patrizia Galletti, Mario De Rosa and Vincenzo Zappia *

(Received

Key words:

Cholinephosphotranaferaae:

May 9th. 19X5)

Enzyme-whatrate Diacylglycerol:

interactlon; CDPcholine: (Rat liver)

Phoaphatidylcholine

synthesis:

In order to elucidate the reaction mechanism and the substrate-binding sites, CDPcholine:1,2-diacylglycerol cholinephosphotransferase (EC 2.7X2), prepared from rat liver microsomal fraction, has been subjected to kinetic analysis and substrate specificity studies. Kinetic evidence supports the hypothesis of a Bi-Bi sequential mechanism, involving a direct nucleophilic attack of diacylglycerol on CDPcholine during the reaction. To investigate the substrate requirements for recognition and catalysis, several CDPcholine analogs, modified in the nitrogen base or in the sugar or in the pyrophosphate bridge, have been synthesized, characterized and assayed as substrates and/or inhibitors of the reaction. The amino group on the pyrimidine ring, the 2’-alcoholic function of the ribose moiety as well as the pyrophosphate bridge have been identified as critical sites for enzyme-substrates interactions.

Introduction phosphatidylcholine The biosynthesis of through Kennedy’s pathway involves, as the final step. the transfer of the cholinephosphate moiety from CDPcholine to 1,2-diacylglycerol, catalyzed by 1,2-diacylglycerol:CDPcholine cholinephosphotransferase (cholinephosphotransferase) [l-7]. The enzyme is located at the cytoplasmic site of the endoplasmic reticulum [8] where diacylglycerol is also selectively located. Studies on enzyme specificity towards diacylglycerol support a dependency of the activity on the chain length [6,7] and degree of unsaturation of fatty acids [6]. Moreover, early reports on enzyme specificity towards CDPcholine suggest a * To whom correspondence 0005.2760,‘85,‘$03.30

should

he addressed.

(1, 1985 Elsewer Science

Publishers

critical role of the pyrimidine moiety of the molecule in recognition [3]. This study represents the first effort to elucidate the mechanism of catalysis of cholinephosphotransferase by means of two complementary approaches: kinetic analysis and substrate specificity studies. Kinetic studies were undertaken in order to elucidate whether a single or a double displacement mechanism is operative, a phosphorylcholine enzyme covalent intermediate being predictable only in a double displacement process. Substrate specificity studies, carried out using several analogs of CDPcholine, allowed us to elucidate some enzyme-substrate interactions. Most of the CDPcholine analogs required by the latter approach have been synthesized in this laboratory by means of new synthetic procedures of wide applicability for further analog designs.

B.V. (Biomedical

Division)

223

Preliminary been reported

results 191.

of this work

have already

Materials and Methods Chemicals. CDPcholine monosodium salt was a generous gift from Cyanamid Italy, Catania, Italy. Cytidine, phosphorylcholine calcium salt, AMP, dTMP, dCMP, cytosine-l-/%D-arabinofuranosyl5’-monophosphate (araCMP) and CMP-free acids, UMP and 5-iodoCMP disodium salts, egg yolk phosphatidylcholine phospholipase C from Bacillus cereus, AGlX8 resin and CDPdiacylglycerol (dipalmityl) were from Sigma Chemical Co., St. Louis, MO, U.S.A. Affi-Gel601 boronic resin was from Bio-Rad Lab., Richmond, CA, U.S.A. Valerie p-toluenesulphonyl carbonildiimidazole, acid, chloride (tosylchloride) and all other chemicals at the purest grade available were from EGA Chemie, Steinheim, F.R.G. CDP[methyl-‘4C]choline (50 mCi/mmol) and phosphoryl[methyl-‘4C]choline (50 mCi/mmol) were from Amersham International, Amersham, Bucks., U.K. Preparation of microsomes. Male rats, weigh@ 200-250 g, were killed by decapitation after an overnight fast. Livers were perfused with ice-cold 0.9% NaCl and homogenized with 4 vol. of 0.25 M sucrose/l mM EDTA (pH 7.4). After centrifugation of the homogenate at 10000 X g, the supernatant was further centrifuged 1 h at 105 000 X g. The pellet was washed twice with 0.1 M Tris-HCl (pH 7.4) and was suspended in the same buffer containing 1 mM EDTA and 0.5 mM dithiothreitol, yielding a final protein concentration of 5 mg/ml. Enzymatic

assay for

cholinephosphotransferase.

The enzymatic assay was performed at 30°C according to Kanoh [2]. The standard reaction mixture contained 25 pmol Tris-HCl (pH S), 0.5 pmol freshly prepared deoxycholate, 50-100 nmol 1,2-diacylglycerols, 5-25 nmol (0.25-1.25 PCi) CDP[methyl-‘4C]choline and 50-100 pg microsomal proteins in a final volume of 250 ~1. 1,2-diacylglycerol was prepared from egg yolk phosphatidylcholine by enzymatic hydrolysis with phospholipase C from B. cereus [2]. The diacylglycerol was added in the assay mixture as a 20% Tween-20 emulsion. When experiments with increasing 1,2-diacylglycerol concentration were

performed, appropriate amounts of Tween-20 were added to reach the same final Tween-20 concentration. Control samples were prepared lacking either 1,2-diacylglycerol or the enzyme. The reaction rate is expressed as pmol of phosphatidylcholine formed per second. Synthesis of analogs of CDPcholine Compounds 3 to 8 and 10 (Table III). Dry phosphorylcholine-free acid was prepared from its calcium salt by oxalic acid precipitation of calcium oxalate solution followed by lyophilization of the aqueous supernatant [lo]. UMP and 5-iodoCMPfree acids were prepared from their sodium salts by means of a Dowex 50 column (7.5 cm X 2.5 cm i.d.), as previously described [ll]. The dry compounds were prepared by lyophilization. To 0.70nmol aliquots of dry phosphorylcholine a 2 M anhydrous solution of tosylchloride in dimethylformamide was added to obtain a final phosphorylcholine/tosylchloride equimolar ratio. In about 5-10 min, the phosphorylcholine was dissolved by means of very vigorous shaking, yielding a highly viscous solution, 0.2 mmol of the dry nucleotide or cytidine were then added, followed by 1-2 h of vigorous shaking. The reaction mixtures containing compounds 3 to 8 were then dissolved in water and loaded onto a Dowex AGlX8 column (7 cm X 4 cm i.d.), formate form; 1 1 linear gradient O-0.04 M formic acid was employed to elute compounds 3, 5, 7 and 8; the elution volumes were 420 ml, 540 ml, 775 ml and 775 ml, respectively. For purification of compounds 4 and 6, the column was washed with 1 1 linear gradient O-O.15 M formic acid, followed by 600 ml of 0.15 M formic acid which eluted both compounds. In order to purify compound 10, the reaction mixture was dissolved in water, brought to pH 8.8 by diluted ammonia and loaded onto an Affi-Gel601 column (8 cm X 1.1 cm i.d.) previously equilibrated with 250 mM ammonium formate (pH 8.8). The column was then washed with 250 ml of the same buffer and compound 10 was eluted with 100 ml of 250 mM ammonium formate (pH 4.4). The fractions containing the above-mentioned compounds were then pooled and lyophilized. and the various compounds were characterized by ultraviolet, nuclear magnetic resonance (NMR) and high-pressure liquid chromatography (HPLC)

224

techniques as described below. The yields for the synthesized compounds were as follows: compound 3, 47%; compound 4, 62%: compound 5, 20%; compound 6, 66%; compound 7. 78%; compound 8, 51%; compound 10. 40%. Compound 2 (Table III). 0.9 mmol valeric acid were dissolved in 1 ml of anhydrous dimethylformamide; 1.8 mmol carbonyldiimidazole were then added to yield valeric acid imidazolide [12]. In order to obtain CDPcholine sodium tetrabutylammonium, 1 ml of 0.1 M aqueous solution of CDPcholine monosodic salt was treated with tetrabutylammonium hydroxide in a 1 : 1 molar ratio. The lyophilized salt was then dissolved in 4 ml of anhydrous dimethylformamide and 5 mg of dimethylaminopyridine (as a catalyst) and the freshly prepared imidazolide solution were then added. The reaction mixture was kept at 50°C in an oil bath, and 24 h later a new addition of imidazolide was made. The reaction was monitored by thin-layer chromatography (TLC) (system e. Table 1). After 40 h, CDPcholine was quantitatively derivatized yielding equal amounts of N4-valerylCDPcholine (N-ValCDPcholine) and 2’,3’-di-OValery], N4-valerylCDPcholine. The reaction was stopped by adding water and the valeryl derivatives were then extracted with 3 x 1 vol. ethyl acetate. The combined organic layers were dried under reduced pressure and the residue, dissolved in methanol, was loaded onto a silica gel column (30 cm x 1.5 cm i.d.) previously equilibrated with methanol. TrivalerylCDPcholine was eluted by 80’% aqueous methanol. while N-ValCDPcholine was eluted by a 70%’ methanol solution. The latter purified compound was finally lyophilized and analyzed by ultraviolet, NMR and HPLC techniques as described below. The yield was 48%. Compound 9 (T&e II). 3 mmol CDPcholine were dissolved in 5 ml of acetic acid, then 10 ml of acetic anhydride and 1 ml of pyridine were added and the solution was kept under vigorous shaking at 0°C for 40 min. These experimental conditions promote a selective acetylation of the two alcoholic functions of ribose moiety, yielding 2’,3’di-0-acetylCDPcholine (diAcCDPcholine). The product was precipitated from the reaction mixture upon addition of tetrahydrofurane and purified by repeated precipitations from methanol solution upon addition of 4 ~01s. of tetrahydrofurane.

Radiochemical synthesis (compounds 3 to 8 und 10. Tuhie ZZZ). 0.7 pmol (8. IO’ dpm) phosphoryl[meth>~l-‘4C]choline ammonium salt were dissolved in 400 ~1 of 0.1 M formic acid and lyophilized. 100 ~1 of a 2 M solution of tosylchloride in dimethylformamide were then added and, after 15 min shaking, 0.1 mmol of the various nucleotides or cytidine were added and the solution was stirred for 2 h. The reaction was stopped by adding 1 vol. water and the mixture was partitioned three times with 1 vol. ethyl acetate. The labelled compounds were finally purified from the water layer by HPLC on a DEAE column according to conditions b of Table I. The radioactive and ultraviolet absorbing

TABLE

I

CHROMATOGRAPHIC CHOLINE ANALOGS

CHARACTERIZATION

OF CDP-

C‘hromatographic conditions: (a) A Whatman Partisil SCXlO strong cation exchange HPLC column (25 cmX4.6 mm i.d.) was employed with a flow rate of 1 ml/min; 250 mM sodium phosphate (pH 6.8) was used as eluent. (b) An LKB Ultropac TSK 545 DEAE weak anion exchange semipreparative column (15 cmx7.5 mm i.d.) was employed with a flow, rate of 1 ml/mm: 25 mM ammonium formate (pH 4.4) w’as used as eluent. (c) A Macherey-Nagel nucleosil C-IX reversed-phase column (20 cm x 4.0 mm i.d.) was employed with a flow rate of 1 ml/min: a mixture containing 85% of 250 mM ammonium formate (pH 4.0) and 15% methanol was used as eluent. (d) Glass-supported Merck 60 F254 silica gel thin-layer plates were eluted with methanol/water/acetic acid (60: 25 : 15. v/v). (e) The same plates as in(d) were eluted with butanol/water/acetic acid (60: 25 : 15. v/v). Compound

1 CDPcholine 2 Ai-ValCDPcholme 3 5.iodoCDPcholine 4 U DPcholine 5 ADPcholine 6 dTDPcholine 7 dCDPcholine 8 araCDPcholine 9 diAcCDPcholine 10 CMPcholine

HPLC retention (min) 7.0 14.0 20.0 14.5 13.5 5.0 16.0 6.0 5.0 16.0 6.0 13.0 6.0 13.0 20.0 11.0

(a) (b) (a) (c) (a) (a) (b) (a) (a) (h) (a) (h) (a) (b) (a) (a)

TLC time

R, (S) 3 (d) 0 (e) 12 (e)

23 (d)

peak was collected and lyophilized. The yields of the various radioactive compounds were always above 90%. Chromatographic characterization. All the synthesized analogs were characterized by HPLC and/or TLC as illustrated in Table I. HPLC analyses were performed either on a Perkin Elmer Series II HPLC chromatograph, equipped with a Perkin Elmer LC-15 ultraviolet detector or on a Varian Vista 5000 HPLC chromatograph, equipped with an LKB Uvicord S II ultraviolet detector and a Varian Series RI-3 refractive index detector. Ultraviolet spectroscopic measurements. Ultraviolet spectra were carried out on a Varian DMS 100 ultraviolet-visible spectrophotometer interfaced with a Varian techtron DS-15 data station. Nuclear magnetic resonance spectra. NMR spectra were run on a Bruker 500 MHz NMR spectrometer, equipped with a Fourier transform accessory. The samples were analyzed using 2H20 as a solvent. All peaks have been assigned as shown in Table II. Radioactivity measurements. The radioactivity was measured in a Beckman LS 7800 liquid scintillation counter, equipped with an automatic quench correction system, using Insta-gel (Packard) as scintillation liquid. Results Kinetic analysis In order to meet the initial velocity conditions required by a kinetic analysis, well purified enzyme preparations are usually employed [12]. Nonetheless, the lack of procedures for obtaining homogeneous preparations of cholinephosphotransferase prompted us to verify the initial velocity conditions using a partially purified microsomal fraction as enzyme source. A good linearity of enzymatic activity with respect to protein concentration and reaction time was indeed obtained [9] (up to 100 pg of protein and 10 minof incubation time) thus permitting the selection of initial velocity conditions acceptable for a kinetic analysis. The kinetic results are illustrated in Fig. 1: the double reciprocal plot A, with diacylglycerol as variable substrate at three fixed concentrations of CDPcholine, shows an intersecting pattern with

Fig. 1. (A) Double reciprocal plot of initial velocity of cholinephosphotransferase at variable concentrations of diacylglycerol and the following fixed levels of CDPcholine: n . 34 PM; 0, 54 PM; 0, 102 PM. The replots of both intercepts and slopes are also reported, showing a linear pattern. (B) Double reciprocal plot of initial velocity of cholinephosphotransferase at variable concentrations of CDPcholine and the following fixed levels of diacylglycerol: n, 187 PM; 0, 312 pM; and 0, 436 PM. The intercept and slope replots are linear. The enzymatic assays were run for 10 mm at 30°C as described in Materials and Methods using 50 pg of protein. K, values. extrapolated by means of the reported replots are 320 PM for diacylglycerol and 33 I.IM for CDPcholine.

linear intercepts and slope replots. In plot B, the same linear intersecting pattern is observable when CDPcholine is the variable substrate. The K,, values for diacylglycerol and CDPcholine, extrapolated by means of the replots in Fig. 1 are 320 PM and 33 PM, respectively, in fairly good agreement with reports from the literature [4,13]. The reported intersecting patterns suggest, according to Cleland [14], a sequential mechanism for cholinephosphotransferase. Product inhibition experiments have been carried out with phosphatidylcholine up to 2 mM, which represents 8-times the endogenous amount of phosphatidylcholine

226

221

22x

TABLE

III

CDP-CHOLINE

AND

ITS ANALOGS

AS SUBSTRATES

AND

INHIBITORS

OF CHOLINEPHOSPHOTRANSFERASE

The enzymatic assay was run for 10 min at 30°C as described in Materials and Methods “C-labelled substrates and/or of inhibitors. 436 PM diacylglycerol. 100 pg of enzyme protein. expressed as percentage of the activity measured at 40 PM CDPcholine.

using the reported The relative activity

concentrations and inhibition

0

-c)yr

R“’

II

p-0

A-

0

R

n

R” R’

R

Compound 1 CDPcholine

cytosine

2 Y-ValCDPcholine

,Y-Valcytosine

3 5iodoCDPcholine

5-iodocytosine

4 UDPcholine

uracyl

5 ADPcholine

adenine

6 7 8 9 10 11 I?

dTDPcholine dCDPcholine araCDPcholine diAcCDPcholine CMPcholine CDP CDPdiacylglycerol

I) i

Substrate concn.

No.

f!JMl 40 140 40 80 40 40 40 40 100 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40

R”

R

a-OH

a-OH

choline

2

-H ;HOH

n-OH a-OH cu-OH ol-OCOCH, a-OH a-OH cu-OH

choline choline choline choline choline -OH -diacylglycerol

7 2 2 2

II

!

thymine cytosine cytosine cytosine cytosine cytoaine cytosine

Substrate

I?,

R’

u-OCOCH, a-OH a-OH wOH

lnhibttor compound No.

2 2 3 4 4 5 5 9 9 10 10 11 11 11 12 12

Inhibitor concn. (PM)

40 400 40 140 40 400 40 400 40 400 40 400 40 400 1000 40 200

Phosphatidylcholine formed (pmol/a per mg) 88.3 0.7 < 0.3 < 0.3 1.7 17.7 1.5 < 0.3 < 0.3 XX.3 69.X 12.5 47.x 8X.3 37.2 XX.3 69.0 XX.3 70.7 X8.3 71.6 XX.3 X8.3 X8.3 XX.3 XX.3

1 7 7

Activity (“cl 100 0.x 0 0 I .9 20 1.7 0 0

100 80 x2 54 100 42 100 7X 100 80 100 Xl 100 100 100 100 100

of are

229

added with microsomes [15]. Under such conditions, no detectable inhibition of the enzyme was observed.

{=o

0

+

CHj-KH&C?

fl

OH

.P-0 p

25 cl

Chemical synthesis of CDPcholine analogs In order to investigate the substrate specificity, several analogs of CDPcholine have been designed and synthesized with modifications in three structural regions of the molecule, namely the nitrogen base, the sugar moiety and the pyrophosphate bridge (compounds 2 to 10 in Table III). Three procedures for the synthesis of the designed analogs have been developed. As a first synthetic tool, the procedure by Kikugawa and Ichino [lo] with major modifications was employed. As reported in Scheme I, the two-step reaction is based on the activation of phosphorylcholine (compound 14) by tosylchloride (compound 13) in dimethylformamide under strict anhydrous conditions, followed by the in situ addition of the suitable nucleotide (compounds 16 to 21) to the 4-fold excess of phosphorylcholine tosyl derivative (compound 15). The reaction is complete in 90 min as opposed to the 9 days required by dicyclohexylcarbodiimide condensation [17], with higher yields ranging from 50 to 90%. The method is suitable for radiochemical synthesis in that an almost quantitative utilization of any labelled precursor is achievable. The versatility of the reaction is higher, since phosphodiesters, as well as pyrophosphodiesters, can be synthesized. N-ValCDPcholine (compound 2) was synthesized by a second procedure, implying the direct acylation of CDPcholine (Scheme II), with the imidazolide of valeric acid as acylating agent (compound 27) [12]. The CDPcholine

ROH=

%doct.v dCMP araCMP

0

CH,

0 +

CH30b6(-0-CH,-CH:ii-CH, ”

Scheme I.

:

dH

15

UMP dTMP

16 17 s 19 20

AMP

21

Cytldme

22

OH OH

k

lnBu~-N+

1

LiH bH

Scheme II.

was solubilized in dimethylformamide in its tetrabutylammonium salt form. Finally, the acetylation of both 2’ and 3’ alcoholic groups of ribose was performed by a standard procedure with acetic anhydride [18], in glacial acetic acid. The procedures for direct acylation of CDPcholine are of interest in that conditions for group-selective acylation of alcoholic and amino groups have been obtained [19]. Substrate specificity To further investigate the reaction mechanism with respect to CDPcholine requirements for enzyme recognition and catalysis, the analogs reported in Table III have been tested as substrates and/or inhibitors of cholinephosphotransferase. Four structural domains of CDPcholine have been screened by this approach as possible enzyme-interacting sites. The methyl-‘4C-labelled CDPcholine analogs reported in Table III were tested for their ability to act as cholinephosphate donors in the phosphatidylcholine synthesis. None of the synthesized analogs becomes as active as the natural cholinephosphate donor, confirming the narrow specificity of the enzyme towards CDPcholine, previously suggested by Kennedy and Weiss [3]. Changes in the pyrimidine ring structure, as in UDPcholine and 5-iodoCDPcholine, are not compatible with substrate activity. Conversely, dCDPcholine, endowed with 20% of the activity exerted by CDPcholine, is the most active among the analogs assayed. This result confirms the physio-

230

Fig. 2. Double reciprocal plot of cholinephosphotransferase activity using CDPcholine (0) dCDPcholine (0) and araCDPcholine (W) as cholinephosphate donors. The assay was run for 10 min at 30°C as described in Materials and Methods using 436 PM diacylglycerol and 100 pg protein. The K, values are 57, 130 and 120 PM, while Vmnxvalues are 3.66, 1.23 and 0.112 pmol/s for CDPcholine. dCDPcholine and araCDPcholine. respectively.

relevance of such a naturally occurring compound [20,21]. The other 2’-deoxy derivative tested, i.e., dTDP choline, is also active as substrate, suggesting a lower enzyme specificity towards 2’-deoxy substrates. Nevertheless, the inversion of chirality at 2’ carbon, as in araCDPcholine, leads to a compound endowed with a quite poor activity. In order to more closely examine the involvement of 2’-OH group in the catalytic process, the affinity of the natural substrate was compared with that of 2’-modified analogs. Fig. 2 illustrates the double reciprocal plots of CDPcholine and its deoxy and arabinosyl analogs. Under the reported conditions the apparent K, (V,,,) values for CDPcholine, dCDPcholine and araCDPcholine, were 57 PM (3.66), 130 (1.23) and 120 (0.112), respectively. Inhibition studies The analogs reported in Table III have also or higher been tested as inhibitors at equimolar concentrations with respect to the physiological substrate.

Fig. 3. Inhibition curve of cholinephosphotransferase activity in the presence of increasing amounts of either 5-iodoCDPcholine (0) or UDPcholine (0). The assay was run for 10 min at 30°C as described in Materials and Methods using 40 PM CDPcholine, 436 pM added diacylglycerol and 100 pg of protein. Ii,, values were 170 pM for 5AodoCDPcholine and 300 PM for UDPcholine.

Among the analogs modified in the cytosine moiety, only 5’-iodoCDPcholine and UDPcholine are endowed with significant inhibitory activity. As reported in Fig. 3. the 5-iodo derivative, exerts the most significant inhibition (I,, = 170 PM). while UDPcholine inhibits only at higher concentrations (I,, = 300 PM). All the other cholinecontaining analogs (compounds 2. 5. 9 and 10) exert a very poor inhibition. The inhibition of CMPcholine, when kinetically analyzed (Fig. 4). was found to be uncompetitive (K, = 555 I_IM vs. K,, = 57 PM) [13]. In the light of this result, the

Fig. 4. Double reciprocal plot of uncompetitive inhibition of cholinephosphotransferase activity by different concentrations of CMPcholine. The assay was carried out for 10 min at 30°C as described in Material and Methods using 436 nM added diacylglycerol, 100 pg of proteins and the following concentrations of CMPcholine: l , none; 0. 0.7 mM; n . 1.5 mM. Under these conditions, K, for CMPcholine is 555 PM, while K,, for CDPcholine is 57 PM.

231

previously reported absence of substrate activity of CMPcholine (Table III) can be ascribed to a lack of recognition at the CDPcholine binding site. Consequently, it is still questionable whether this phosphodiester is endowed with sufficient free energy content to warrant the transfer of the phosphorylcholine moiety. Finally, CDP and CDPdiacylglycerol were found to be devoid of any detectable inhibitory effect, even at high concentrations. Discussion Few data on the enzymatic transfer of phosphoalcoholic groups are reported in the literature. Among them, the studies by Dowhan on the transfer of a phosphodiacylglycerol moiety from a cytidine-activated substrate to glycerophosphate or serine support a sequential mechanism for phosphatidylglycerophosphate synthase [22] and a ping-pong mechanism for phosphatidylserine synthase 1231. Our data allow us to propose a sequential mechanism for rat liver cholinephosphotransferase. Consistent with the demonstrated mechanism, a direct nucleophilic attack of the alcoholic function of diacylglycerols on the /?-phosphorus of CDPcholine is predictable, leading to the cleavage of the phosphorus-oxygen bond. The kinetic and substrate specificity results permit us to draw some inferences on the overall enzymatic process. At least three binding sites involved in the enzyme-CDPcholine recognition can be proposed, namely the cytosine amino group, the 2’ alcoholic function of the ribose and the positively charged quaternary ammonium pole. The key role of the cytosine amino group in the catalytic process is also shown by the lack of substrate activity of the analogs modified at this site (compounds 2, 3 and 4). Moreover, the appropriate stereochemical configuration of the alcoholic group at the 2’ carbon is essential for catalysis, as shown by the poor substrate activity of araCDPcholine. However, the removal of this group, as in dCDPcholine, affects the substrate activity less than the inversion of its configuration, thus suggesting a steric hindrance of the P-OH group. The relevance of the choline moiety in enzyme recognition can be inferred by the absence of any

inhibitory effect by CDP; The finding that no detectable inhibition is exerted by CDPdiacylglycerol, which can be envisioned as a transition state analog lacking the choline moiety [24], leads to the same conclusion. Finally, the length of the pyrophosphate bridge is also critical in the recognition process, as CMPcholine, being uncompetitive inhibitor, does not interact with the enzyme at the CDPcholine binding site. Acknowledgement This work was supported by grant 83.02019.04 from the Italian National Research Council. References 1 Weiss, S.B., Wagner Smith, S. and Kennedy, E.P. (1958) J. Biol. Chem. 231, 53-64 2 Kanoh, H. and Ohno, K. (1981) Methods Enzymol. 71, 536-546 3 Kennedy, E.P. and Weiss, S.B. (1956) J. Biol. Chem. 222, 193-214 4 Kanoh, H. and Ohno, K. (1976) Eur. J. Biochem. 66, 201-210 5 Kanoh, H. (1970) Biochim. Biophys. Acta 218, 249-258 K. and Kanoh, H. (1978) J. Biol. Chem. 253, 6 Morimoto, 5056-5060 R. and Bell, R.M. (1977) J. Biol. Chem. 282, 7 Coleman, 3050-3056 S.B., Lim, P.H. and 8 Vance, D.E., Choy, P.C., Farren, Schneider, W.J. (1977) Nature (London) 270, 268. 9 Zappia, V., De Rosa, M., Pontoni, G., Manna, C., SalIuuo, A., de1 Piano, L. and Galletti, P. (1985) in Novel Biochemical, Pharmacological and Clinical Aspects of Citidinediphosphatecholine (Zappia, V., Kennedy, E.P., Nilsson, B. and Galletti, P., eds.), pp. 41-44, Elsevier, New York 10 Kikugawa, K. and Ichino, M. (1971) Chem. Pharm. Bull. 19 (5) 1011-1016 11 Romeo, G., Cocchiara, G., De Rosa, M., Giordano, F. and Zappia, V. (1981) Boll. Sot. It. Biol. Sper. 57 (ll), 1175-1181 12 Staab H.A. and Rohr, W. (1968) in Newer Methods of Preparative Organic Chemistry (Foerst, W., ed.), Vol. 5, pp. 61-108, Academic Press, New York 13 Morimoto, K. and Kanoh, H. (1978) Biochim. Biophys. Acta 531, 16-24 14 Cleland, W.W. (1970) in The Enzymes (Boyer, P.D., ed.), Vol. 2, pp. 1-61, Academic Press, New York 15 Gem, G.S., Bartley, W., Stirpe, F., Notton, B.M. and Renshaw, A. (1962) Biochem. J. 83, 181-191 16 Cleland, W.W. (1963) B&him. Biophys. Acta 67, 104-137 17 Khorana, H.G. (1961) in Some Recent Developments in the Chemistry of Phosphate Esters of Biological Interest, John Wiley and Sons, New York

232 18 Martinez, A.P.. Lee. W.W. and Goodman. L. (1966) J. Med. Chem. 9, 268 19 Pontoni. G., Salluzzo. A., Russo, G.L., Palumbo. P. and Galletti. P. (1984) 30th Meeting It. Biochem. Sot. Ischia. October 1984, abstr. p. 325 20 Sugino, Y. (1960) Biochim. Biophys. Acta 40, 425-434 21 Kennedy, E.P., Fencil Borkenhagen. L. and Wagner Smith. S. (1959) J. Biol. Chem. 234. 199882000

22 Hirabayashi, T.. Larson. T.J. and Dowhan, W. (1976) Biochemistry 15, 5205-5211 23 Larson. T.J. and Dowhan. W. (1976) Biochemistry 15, 5212-5217 24 Lashmet. P.. Tang, K.C. and hedron Let. 24. 1121. 7124

Coward.

J.K. (1982)

Tetra-