11
Eur. J. Biochem. 233, 62-72 (1995) 0 FEBS 1995 Molecular cloning of CTP : phosphocholine cytidylyltransferase from Plasmodium falci~arum Hye-Jeong YE0 I, Joannes SRI WIDADA’, Odile MERCEREAU-PUIJALON’ and Henri Joseph VIAL‘ Laboratoire de Dynamique Moltculaire des Interactions Membranaires, CNRS URA 1856, Dtpartement Biologie-Santt, * Laboratoire de Microbiologie et Pathologie Cellulaire Infectieuse, INSERM U-431, Dtpartement Biologie-Santt, Unite de Parasitologie Exptrimentale, Institut Pasteur, Paris, France (Received 31 July 1995) - EJB 95 1156/1 Universite Montpellier 11, case 107, France Universitt Montpellier 11, case 100, France CTP: phosphocholine cytidylyltransferase (CCT) is the rate-limiting and regulatory enzyme in the synthesis of phosphatidylcholine, the major membrane phospholipid, in Plasmodium. The structural gene encoding CCT was isolated from the human malaria parasite Plasmodium faleiparurn. This was achieved using the PCR to amplify genomic DNA with degenerate primers constructed on the basis of conserved regions identified within yeast and rat liver CCT molecules, and using the PCR product to screen a genomic library. The 19 ,fulcipurum CCT gene encodes a protein of 370 amino acids (42. 6 kDa) and displays 41 -43% similarity (28-29% identity) to CCT molecules of the other organisms cloned to date. The central domain of CCT, proposed as the catalytic domain of the CTP-transfer reaction, shows 68- 72% similarity and 48-55% identity among I-1 Julcipurum, human, rat and yeast enzymes. This gene is present in a single copy, as determined by Southern-blotting of genomic DNA, and located on chromo- some 13 of I-1 ,fakiparum. Large transcripts were detected by Northern analysis and indicate that this gene is expressed in the asexual intraerythrocytic stages. The coding region of the I-1 jalciparum CCT gene was inserted into an Esclzerichiu coli expression vector to confirm the function of the CCT product. The recombinant CCT expressed in E. coli is catalytically active, as evidenced by the conversion of phosphocholine to CDP-choline. Keywords: Plasmodiunz ,fakiparum ; phospholipid : cytidylyltransferase ; mRNA expression ; chromosome localisation. The phospholipid metabolic pathway has a number of fea- tures suggesting that it is particularly important to the most pa- thogenic human malaria parasite, Pla.smndiunz,fulciparum, mak- ing it a very attractive candidate for rational drug design. Asex- ual intraerythrocytic proliferation of P 1a.smodium is accompa- nied by the synthesis of a considerable number of phospholipids needed for the production of parasite membranes. The phospho- lipid composition of the parasite membranes is quite distinct from that of host human cells [I, 21. Adapted phospholipid bio- synthesis is required for the membrane biogenesis supporting malarial parasite growth [3, 41. The parasite directs the synthesis of its own phospholipid synthetic enzymes, some of which have Correspondence to H.-J. Yeo, Laboratoire de Dynamique Moltcu- laire des Interactions Membranaires, CNRS URA 1856, Dtpartement Biologie-Sante, Universite Montpellier 11, case 107, F-34095 Montpel- her Cedex 5, France Ahhreviutions. CCT, CTP:phosphocholine cytidylyltransferase; PtdCho, phosphatidylcholine; PFGE, pulsed-field gel electrophoresis; IPTG, isopropyl 1-thio-P-D-galactopyranoside. Enzymes. CTP:phosphocholine cytidylyltransferase (EC 2.7.7.1 5); CTP:glycerol-3-phosphate cytidylyltransferase (EC 2.7.7.39); choline kinase (EC 2.7.1 32) ; choline phosphotransferase (EC 2.7.8.2) ; CDP- diacylglycerol pyrophosphatase (EC 3.6.1.26); DNA polymerase (EC 2.7.7.7); 3-deoxy-manno-octulosonate cytidylyltransferase (EC 2.7.7. 38); T4 DNA ligase (EC 6.5.1.1). Note. The novel nucleotide sequence data reported here have been submitted to the EMBLKenBank sequence data bank and are available under accession number X84041, been shown to possess catalytic properties quite distinct from those of their human homologues [5, 61. This has given rise to the idea of blocking the development of the parasite using a drug which selectively interferes with this metabolism [7, 81. Currently, a pharmacological model, through effectors of the choline carrier [which provides the intracellular parasite with choline, the precursor of phosphatidylcholine (PtdCho), the ma- jor phospholipid of the parasite] is under development in our laboratory. Potent in vitro antimalarial activity is associated with the complete cure of highly infected mice and of I? falciparum- infected Aotus monkeys [9]. A potential problem with using a parasite biochemical path- way as a drug target is that many parasites have the ability to use alternate biochemical pathways, conferring resistance and rendering the drug ineffective. Therefore, it is necessary to eluci- date the mechanism of regulation of the phospholipid biosynthe- sis pathway. Molecular cloning of key enzymes of Plasmodium phospholipid metabolism is an obvious step in this process. PtdCho is the major phospholipid in most eukaryotic cells [3, 10, 111 and functions not only as an important structural compo- nent but also as a major source of second messengers for signal- transduction molecules in many cell types [ 12- 141. In higher eukaryotic cells, PtdCho is synthesized mainly by the CDP-cho- line pathway [lo]. In the yeast, Succhuromyces cerevisiae, a lower eukaryote, PtdCho is synthesized via three main pathways [ 15, 161 : (a) the decarboxylation of phosphatidylserine to phos-

Molecular Cloning of CTP: Phosphocholine Cytidylyltransferase from Plasmodium falciparum

Embed Size (px)

Citation preview

Eur. J. Biochem. 233, 62-72 (1995) 0 FEBS 1995

Molecular cloning of CTP : phosphocholine cytidylyltransferase from Plasmodium falci~arum Hye-Jeong YE0 I, Joannes SRI WIDADA’, Odile MERCEREAU-PUIJALON’ and Henri Joseph VIAL‘

’ Laboratoire de Dynamique Moltculaire des Interactions Membranaires, CNRS URA 1856, Dtpartement Biologie-Santt,

* Laboratoire de Microbiologie et Pathologie Cellulaire Infectieuse, INSERM U-431, Dtpartement Biologie-Santt,

’ Unite de Parasitologie Exptrimentale, Institut Pasteur, Paris, France

(Received 31 Ju ly 1995) - EJB 95 1156/1

Universite Montpellier 11, case 107, France

Universitt Montpellier 11, case 100, France

CTP: phosphocholine cytidylyltransferase (CCT) is the rate-limiting and regulatory enzyme in the synthesis of phosphatidylcholine, the major membrane phospholipid, in Plasmodium. The structural gene encoding CCT was isolated from the human malaria parasite Plasmodium faleiparurn. This was achieved using the PCR to amplify genomic DNA with degenerate primers constructed on the basis of conserved regions identified within yeast and rat liver CCT molecules, and using the PCR product to screen a genomic library. The 19 ,fulcipurum CCT gene encodes a protein of 370 amino acids (42. 6 kDa) and displays 41 -43% similarity (28-29% identity) to CCT molecules of the other organisms cloned to date. The central domain of CCT, proposed as the catalytic domain of the CTP-transfer reaction, shows 68- 72% similarity and 48-55% identity among I-1 Julcipurum, human, rat and yeast enzymes. This gene is present in a single copy, as determined by Southern-blotting of genomic DNA, and located on chromo- some 13 of I-1 ,fakiparum. Large transcripts were detected by Northern analysis and indicate that this gene is expressed in the asexual intraerythrocytic stages. The coding region of the I-1 jalciparum CCT gene was inserted into an Esclzerichiu coli expression vector to confirm the function of the CCT product. The recombinant CCT expressed in E. coli is catalytically active, as evidenced by the conversion of phosphocholine to CDP-choline.

Keywords: Plasmodiunz ,fakiparum ; phospholipid : cytidylyltransferase ; mRNA expression ; chromosome localisation.

The phospholipid metabolic pathway has a number of fea- tures suggesting that it is particularly important to the most pa- thogenic human malaria parasite, Pla.smndiunz,fulciparum, mak- ing it a very attractive candidate for rational drug design. Asex- ual intraerythrocytic proliferation of P 1a.smodium is accompa- nied by the synthesis of a considerable number of phospholipids needed for the production of parasite membranes. The phospho- lipid composition of the parasite membranes is quite distinct from that of host human cells [ I , 21. Adapted phospholipid bio- synthesis is required for the membrane biogenesis supporting malarial parasite growth [3, 41. The parasite directs the synthesis of its own phospholipid synthetic enzymes, some of which have

Correspondence to H.-J. Yeo, Laboratoire de Dynamique Moltcu- laire des Interactions Membranaires, CNRS URA 1856, Dtpartement Biologie-Sante, Universite Montpellier 11, case 107, F-34095 Montpel- her Cedex 5 , France

Ahhreviutions. CCT, CTP:phosphocholine cytidylyltransferase; PtdCho, phosphatidylcholine; PFGE, pulsed-field gel electrophoresis; IPTG, isopropyl 1-thio-P-D-galactopyranoside.

Enzymes. CTP:phosphocholine cytidylyltransferase (EC 2.7.7.1 5 ) ; CTP:glycerol-3-phosphate cytidylyltransferase (EC 2.7.7.39); choline kinase (EC 2.7.1 3 2 ) ; choline phosphotransferase (EC 2.7.8.2) ; CDP- diacylglycerol pyrophosphatase (EC 3.6.1.26); DNA polymerase (EC 2.7.7.7); 3-deoxy-manno-octulosonate cytidylyltransferase (EC 2.7.7. 38); T4 DNA ligase (EC 6.5.1.1).

Note. The novel nucleotide sequence data reported here have been submitted to the EMBLKenBank sequence data bank and are available under accession number X84041,

been shown to possess catalytic properties quite distinct from those of their human homologues [ 5 , 61. This has given rise to the idea of blocking the development of the parasite using a drug which selectively interferes with this metabolism [7, 81. Currently, a pharmacological model, through effectors of the choline carrier [which provides the intracellular parasite with choline, the precursor of phosphatidylcholine (PtdCho), the ma- jor phospholipid of the parasite] is under development in our laboratory. Potent in vitro antimalarial activity is associated with the complete cure of highly infected mice and of I? falciparum- infected Aotus monkeys [9].

A potential problem with using a parasite biochemical path- way as a drug target is that many parasites have the ability to use alternate biochemical pathways, conferring resistance and rendering the drug ineffective. Therefore, i t is necessary to eluci- date the mechanism of regulation of the phospholipid biosynthe- sis pathway. Molecular cloning of key enzymes of Plasmodium phospholipid metabolism is an obvious step in this process. PtdCho is the major phospholipid in most eukaryotic cells [3, 10, 111 and functions not only as an important structural compo- nent but also as a major source of second messengers for signal- transduction molecules in many cell types [ 12- 141. In higher eukaryotic cells, PtdCho is synthesized mainly by the CDP-cho- line pathway [lo]. In the yeast, Succhuromyces cerevisiae, a lower eukaryote, PtdCho is synthesized via three main pathways [ 15, 161 : (a) the decarboxylation of phosphatidylserine to phos-

Ye0 et al. ( E m .I. Biochem. 233) 63

phatidylethanolamine, which is sequentially methylated to form PtdCho, (b) the methylation of de now biosynthesized phospha- tidylethanolamine via the CDP-ethanolamine pathway, and (c) the de novo biosynthesis of PtdCho by the CDP-choline path- way. In Plasmodium, it is now considered very likely that PtdCho can be synthesized through these three pathways 131. The CDP-choline pathway that is common to most eukaryotes involves three distinct enzymic reactions; (a) the phosphoryla- tion of choline by choline kinase, (b) the production of CDP- choline from phosphocholine by CCT, and (c) condensation of CDP-choline with diacylglycerol to form PtdCho by the choline phosphotransferase. In most cell systems, the step catalyzed by CCT is rate limiting and the regulated step in PtdCho biosynthe- sis [ lo , 15, 171.

Over the last few years, in response to the understanding of the importance of the CCT in the CDP-choline pathway, a number of reports have described mechanisms by which CCT activity is regulated. In most cells, the dominant mechanism for the regulation of this enzyme involves its interconversion be- tween an inactive cytosolic form and an active membrane-bound form [lo, 181. Free fatty acids, diacylglycerol and membrane PtdCho contents, and the phosphorylation state of the CCT regu- late the reversible translocation of the enzyme between cytosol and membranes [19-23). The recent cloning of the gene or cDNA encoding CCT in several organisms has allowed the in- vestigation of the regulatory mechanisms. In rat liver, a putative membrane-interaction domain of CCT has been identified, and i t has been proposed that the mechanism of rat liver CCT activa- tion involves lipid-specific stabilization of an amphipathic heli- cal structure in the C-terminal region of the protein [24-261. In the yeast, substantial differences in lipid regulation have been shown in comparison to the rat CCT [27].

To elucidate the regulatory mechanism of CCT in P. falci- parum and provide information for evaluating the phospholipid metabolic pathway as a potent antimalarial target, we have initi- ated studies of genes encoding key phospholipid metabolism en- zymes. This paper reports, to our knowledge, for the first time, the cloning of a gene-encoding a phospholipid-metabolism en- zyme, CTP:phosphocholine cytidylyltransferase, in P. ,fulci- parum.

MATERIALS AND METHODS

Malaria strain. The Nigerian strain (Richard) of P. faki- parum was used for PCR. genomic Southern-blot and Northern- blot analyses. The parasites were maintained in a human erythro- cyte culture containing Hepes-buffered RPMI 1640 (Gibco BRL) and 10% human serum as described by Trager and Jensen [28]. The cultures were kept in 5-cm Petri plates at 7 % hemato- crit and at 37°C.

Bacterial strain, cultures and molecular biological mate- rials. The Escherichia coli strain DH5n was used for all trans- formations and plasmid maintenance. Standard procedures were used for bacterial culture and isolation of plasmid DNA [29]. Essentially, pBluescript I1 KS (Stratagene) was used for most of the subcloning. Restriction enzymes and DNA-modifying en- zymes were from Boehringer Mannheim, Gibco BRL and Euro- gentec.

DNA isolation from parasites. The parasites were isolated by lysis of the infected erythrocytes with 0.025% saponine. DNA was isolated from non-synchronized parasites by protein- ase K treatment at 37°C for 3 h in the presence of detergents. After several extractions with phenol/chloroform, plasmodia1 DNA was isolated, following overnight precipitation and centrif- ugation.

PCR and cloning of pCT1. Degenerate oligonucleotide primers, constructed on the basis of conserved regions deduced within the yeast and rat liver CCT, were used to amplify the appropriate fragments by PCR utilising 2 U Taq polymerase (Promega), and an automated thermocycler (Biometra). 0.2 pg genomic DNA template were denatured at 94°C for 5 min, fol- lowed by 30 cycles of amplification (denaturation at 94°C for 90 s, annealing at 45°C for 80 s, and extension at 72°C for 70 s). Primers used in this study were synthesized by Eurogentec and had the following sequences :

5’-GCAAGCTTATCC(A/G)GATGG( Am) ATATT( T/ C)GA-3’, based on the conserved amino acid motif YADGVFD (amino acids 107-1 13 in yeast CCT numbering, HindIII site underlined) ; P3, 5’-CCAAGCTTTGT(A/C)GATGGGT(A/T)- GA(T/C)GA(A/G)GT-3’, based on the amino acid motif HCRWVDEV (amino acids 165 - 172, HindIII site underlined) ;

TG-3’, based on the motif AHDDIPY (complementary strand, amino acids 194-200, BurnHI site underlined); P6, 5’-CC=

based on the amino acid motif DD(V/I)YK(H/P)IKE (comple- mentary strand, amino acids 206-214, BumHI site underlined).

The restriction enzyme sites were included to facilitate sub- sequent cloning of PCR products and the amino acid numberings are derived from the yeast CCT amino acid sequence. PCR with primers P3 and P4 resulted in an intense band of product of approximately 110 bp. The 110-bp product was purified on DEAE-cellulose paper, cut with restriction enzymes of cloning sites BamHI and HindIII, and cloned into pBluescript I1 KS (named pCT1).

Screening of genomic DNA library. A ANMI149 genomic EcoRl DNA library of P. fulciparum strain K1 was kindly pro- vided by Dr G. Lansley (Pasteur Institut, Paris). About I00000 recombinant phages of this genomic library were screened with CT1 fragment (1 10 bp), ”P-labelled with [rr-”P]dATP (Amers- ham Life Science) using Prime-It Random primer kit (Strata- gene). In total, approximately 40 positive clones were obtained from first screening. Five of these clones were chosen and each clone was followed by second-round and third-round screening. Once confirmed as 100% pure clone at the third screening, phage DNA was purified and the inserts containing the 19 ,fa&- parum CCT gene were released by EcoRI digestion, resulting in 4.3-kb DNA fragment.

Subcloning, DNA sequencing and sequence analysis. The EcoRI-cut 4.3-kb insert was subcloned into pBluescript IT KS (clone named pCTRR8) for mapping and sequencing. Defined restriction fragments of pCTRR8 were ligated into pBluescript I1 KS. Among the different subclones constructed for sequencing, unidirectional deletions were generated on pCTBBI and pCTBBII containing 2.3-kb inserts (Fig. 1 A, pCTBBI has the opposite orientation to pCTBBII) by exonuclease 111 (Phar- macia) digestions in order to obtain overlapping clones 1301. The nucleotide sequences were determined using the Sequenase Ver- sion 2.0 (U.S. Biochemical Corp.) according to the dideoxy chain termination method [31] on double-stranded plasmid. Primers used in the sequencing were reverse, T7, T3, KS and SK sequences within pBluescript I1 KS. Amino acid sequences were compared against the non-redundant protein data bases by using the National Center for Biotechnology Information net- work Basic Local Alignment Search Tool (BLAST) server. Pro- tein sequence alignments were carried out using CLUSTAL W program [32]. Secondary structures and hydrophobic moment analysis [33] were provided from the Bisance facilities.

Southern blotting and hybridization. 4 pg high-molecular- mass genomic DNA of P. falciparum (Nigerian strain) was di- gested to completion with several restriction enzymes, fraction-

P 1,

P4, 5’-CCGGATCCATA(T/A)GG(A/T)ATATC(A/G)TC(A/G)-

m T T ( A/T)ATTTGTTTATA(T/A)A(C/T)ATC(A/G)TC-3’,

64 Ye0 et al. ( E M J. Biochern. 233)

ated on a 0.7 % agarose gel and blotted using 0.4 M NaOH onto HybondrM-N+ (Amersham Life Science) membrane according to the protocol provided by the manufacturer. The membrane was probed with the 1.1-kb DNA fragment spanning the entire I? fulcipurum CCT-coding region (generated by PCR), labelled using the sandom-primer method. Prehybridization was per- formed at 42°C in a solution containing 5xSSPE (20XSSPE = 3 M NaCl, 0.2 M sodium phosphate, 0.02 M EDTA), 5XDen- hardt's solution, 1 % SDS, 50 mM sodium phosphate, 100 pg/ml salmon sperm DNA and 25% formamide, for 4 h. Denatured DNA probe was added at lo6 cpmiml, and hybridization was performed at 42°C for 18 h. The membrane was then washed twice at 37°C for 15min in 2XSSPE and once at 56°C for 30 min in 0.1 XSSPE.

RNA isolation and Northern-blot analysis. I? $ilciparum- infected erythrocytes (parasitemia 5-10%) from six culture dishes were washed twice in 137 mM NaC1, 2.7 mM KCI, 10.1 mM Na,HPO,, 1.8 mM KH,PO,, pH 7.4 (NaCVP,) contain- ing 5-10 mM vanadyl ribonucleoside complex (Gibco BRL). The asynchronized parasites were released by treatment with NaCIIP, containing 10 mM vanadyl ribonucleoside complex and 0.1 % saponin at 4°C for 5 min and washed once with NaCI/P, containing 1 0 mM vanadyl ribonucleoside complex. Total RNA samples were then isolated from the asynchronized parasites using TrizolTM (Gibco BRL). The poly(A)-rich and poly (A)- depleted RNAs were isolated by Quick prepTM Micro mRNA Purification Kit (Pharmacia). Total RNAs, from human leukemic monocyte-like U937 cells, were used as a control. Total RNAs, poly (A)-rich and poly (A)-depleted RNAs were electrophoresed in a 2.0% agarose/formaldehyde gel and blotted onto Hybond- N' membrane in 3 M NaC1, 0.3 M sodium citrate (20xNaCli Cit). All membranes were prehybridized at 42°C in a solution containing 5 XNaCl/Cit, 5 XDenhardt's solution, 50 mM sodium phosphate, pH 7, 25% or 50% formamide, 100 pg/ml yeast tRNA and 100 pg/ml salmon sperm DNA for 4 h, then hybrid- ized for 18 h following the addition of the ."P-labelled 1.1-kb DNA probe as described above. The membrane was then washed twice at 37°C for 15 min in 2xNaCI/Cit, and once at 60°C for 30 min in 0.2XNaCI/Cit.

Chromosomal localization of R falciparum CCT. Chromo- somal-sized DNA of I? fulcipurum 3D7 was isolated in agarose blocks as described (341. Pulsed-field gel electrophoresis (PFGE) were performed on a CHEF apparatus (351. Chromo- somes were separated in a 0.7% agarose gel (Bio-Rad) in 67 mM Tris/borate, 1.5 mM EDTA chamber buffer. The gel was photographed and transferred onto Hybond-N ' membrane. Hy- bridization was overnight at 65°C in a solution containing 6X N a C K i t , 2.5% non-fat powdered milk and 100 pg/ml herring sperm DNA. The 1 . l -kb CCT-coding-region fragment, radio- labelled by nick-translation (Boehringer Mannheim), was used as a probe. The membrane was washed at 65°C several times in 6XNaCI/Cit, and once in 2XNaCKit .

Cloning of the R falciparum CCT gene in the expression vector PET. For expression of I? fulciparum CCT without a fusion component, the coding region of the P fulciparurn CCT gene was inserted into the expression vector pET3d (Novagen). Amplification of the entire coding region (1 .I kb; restriction en- zyme sites underlined) was achieved by PCR using Pexl (5'- AATATTCTACCATGGATAGTTC-3', corresponding to bases -11 to 11 with the base changes in positions -2 and -1, result- ing in a NcoI site which contains the start codon); Pex2 (5'- CAAGAATTCTGGAGTTACAACCCAAGGGC-3', comple- mentary to bases 491 -519 with base changes in positions 495 and 516, resulting in the elimination of the NcoI site in order to protect the cloning site, and the introduction of an EcoRI site for the further utility) ; Pex3 (5 ' -CTTGGGTTGTAACTCCAa

ATTCTTGG-3', Corresponding to bases 494- 520 with the base changes in positions 495 and 516, with the same aim as in Pex2, as they are complementary, resulting in the elimination of the NcoI site and the introduction of an EcoRI site); Pex4 (5'- GACGGATCCCATTTTAACTACTGG-3', complementary to bases 1103-1126 with base changes in positions 1118, 1119, 1122 and 1123, resulting in a BumHI site downstream of the stop codon). Amino acid sequences were not modified by the base changes. We effected two rounds of PCR (Fig. 8A). The first two PCRs were accomplished with Pexl and Pex2 and with Pex3 and Pex4, separately. The second PCR was achieved with Pexl and Pex4 by mixing an aliquot of 0.1 ng each first PCR product as template DNA. PCR was carried out as described above, except for amplification conditions (denaturation at 94°C for 90 s, annealing at 55°C for 70 s, and extension at 72°C for 70 s). The first PCR products were single bands of 530 bp and 640 bp in each reaction. The second PCR resulted in a major band of 1 150 bp. The PCR-generated fragment was introduced into NcoI and BumHI sites of the bacterial expression vector. The bacterial host DH5a was used for initial cloning of the en- tire CCT gene into the pET3d vector and for maintaining plas- mid yields. For protein production, the recombinant PET P. fulci- purum CCT is transferred to the host E. coli strain containing a chromosomal copy of the gene for T7 RNA polymerase, BL21 (DE3)pLysS (Novagen).

Expression of R falciparum CCT in E. coli. Single colo- nies from E. coli BL21 (DE3)pLysS cells harboring PET R fulcr- parum CCT were picked from cells plated on Luria-Bertani me- dium and used to inoculate into 5 ml cultures. The 5-ml culture was grown overnight at 37"C, then used to inoculate a large culture (to a dilution of 1 : 100) of fresh Luria-Bertani medium supplemented with 100 pg/ml ampicillin and 34 pg/ml chloram- phenicol. The large culture was grown at 37°C until the absor- bance at 550 nm reached a value of about 0.4-0.8. Isopropyl- thio-p-D-galactopyranoside (IPTG) was then added to a concen- tration of 0.4 mM, and the culture was grown for 3 h at 37"C, harvested, and resuspended in lysis buffer (10 mM Tris, pH 7.4, 50 mM NaCI, 1 mM EDTA, 2 mM dithiothreitol, 0.05% Triton X-100). A cell extract was then prepared by three cycles of soni- cation of 10 s each, followed by centrifugation at 12 OOOXg for 15 min at 4°C. The supernatant fraction was used for a CCT activity assay.

Assay of CCT activity. The CCT activity was assayed as previously described [36] with minor modifications. The reac- tion mixture containing 50 mM sodium phosphate, pH 7.0, 5 mM CTP, 25 mM MgCl,, 0.1 mM [methyl-'4C]cholinephos- phate (56 Ci/mol), and the cell extract in a total volume of 20 pl was incubated at 37°C for 20 min. The reaction product was fractionated by TLC on a silica gel 60 plate with ethanol/2% ammonia (1 :1, by vol.). The plate was exposed to X-OMAT film (Kodak). The reaction product was quantified by liquid scintillation counting.

RESULTS

Cloning strategy and isolation of the f? falciparum CCT gene. Only two CCT genes, from yeast [37] and rat liver [38], had been cloned when this study was initiated. Our strategy for cloning the P ,fulcipurum CCT gene was to obtain a specific DNA probe rather than relying on heterologous hybridization, which was unlikely, due to the biased codon usage of l? falci- parurn 139, 401. To generate a DNA probe, we first performed the PCR on R ,fulciparum genomic DNA, assuming that the gene did not contain an intron or contained an intron of a size not precluding efficient amplification. Primers were derived from

Ye0 et al. (Eur: J . Biochern. 233) 65

A

B

-11

50

1 1 0

170

230

2 9 0

350

4 1 0

470

530

5 9 0

650

710

770

830

890

950

1010

1 0 7 0

*. -.

pCTBBI I

AA AGTATGATCA TAATTATTAT TATTATAATA ATACATATGA TAAATTAAGT AACGTAACTT TTAATTTAAA CGGAAAAAAA AAAAAAAGTT ATGATAATGA ATCTTCCTAT ATCAAAACAG CTTATCATGA TATAATTGAT GAAGATAAAA ATTTATTAAA TAAAGAAAAA AGTTTTAATA AAGATGAAAA AAAAAAAAAA AGTATTCTTG AGACAGATCA ATATAAAAAT TGTGAGGATC AACAATTTGA TAATAAATAT AATCATATAA TAATTGGAAA GAAAAAAAGT GATAGCACTT ATGATAATAG TTTGGATAAA GAATCATCAA ATGAAAAAAA ATTATCAATT GATAAAGGAC

AATATTCTAA TEGATAGT TCAAATTATT TTCATGATTG TAAAACCATG CTAAGTGAAC

ATAATGAATC TATTGAATCT AGTAATAATG ATATAAATGG AAAGCAAAAG GAACACATTA K G N S E N Q D V D P D T N P D A V P D

RAAAAGGAAA TTCTGAGAAT CAAGATGTAG ATCCAGATAC AAACCCAGAT GCAGTTCCAG D D D D D D D N S N D E S E Y E S S Q M

ATGATGATGA TGATGATGAT GATAATAGTA ATGATGAAAG TGAATATGAA AGTAGTCAAA D S E K N K G S I K N S K N V V I Y A D

TGGATAGTGA AAAAAATAAG GGATCAATAA AAAATTCAAA AAATGTTGTT ATATATGCTG

ACGGAGTATA TGATATGCTT CATTTAGGGC ATATGAAACA ATTGGAGCAA GCCAAAAAAC F E N T T L I V G V T S D N E T K L F K

TTTTTGAAAA TACTACTTTA ATAGTAGGTG TAACTAGCGA TAATGAAACC AAATTATTTA G Q V V Q T L E E R T E T L K H I R W Y

AAGGTCAAGT TGTTCAAACT CTTGAAGAAA GAACAGAGAC TTTAAAACAT ATACGATGGG D E I I S P C P W V V T P E F L E K Y 5

TAGATGAAAT AATTTCTCCA TGCCCATGGG TTGTAACTCC AGAATTTTTG GAAAAATATA

AAATAGATTA TGTTGCACAT GATGATATAC CATATGCAAA TAATCAGAAA AAGAAAAAGA K K S K G K S F S ! ? D E E N E D I Y A W

AAAAAAAATC TAAAGGGAAG TCATTTAGTT TTGATGAAGA AAATGAAGAT ATATATGCTT L K R A G K F K A T Q R T E G V S T T D

GGTTAAAAAG AGCAGGCAAA TTTAAAGCAA CACAAAGAAC AGAAGGTGTA TCAACTACAG L I V R I L K N Y E D Y I E R S L Q R G

ATTTAATAGT AAGAATATTA AAAAATTATG AGGATTATAT TGAAAGGTCA TTACAACGTG

GTATACATCC TAATGAATTA AATATTGGTG TAACCAAAGC TCAGTCAATA AAAATGAAAA N L I R W G E K V T D E L T K V T L T D

AAAATTTAAT AAGATGGGGA GAAAAAGTGA CAGATGAATT AACTAAAGTT ACTTTAACAG

ACAAGCCATT AGGTACGGAT TTTGATCAAG GAGTTGAAAA TCTTCAAGTC AAATTTAAAG L F K I W K N A S N K L I T D ! ? T R K L

AACTTTTTAA AATATGGAAA AATGCATCGA ATAAATTGAT AACTGATTTT ACAAGAAAAC E A T S Y L T S I Q N I I D Y E I E N D

TTGAAGCAAC ATCTTATTTA ACATCTATTC AAAATATCAT AGATTATGAA ATAGAAAATG D Y A S S N F D D E T S S *

ATGATTATGC TAGTAGTAAT TTTGATGATG AAACCAGTAG T W T G T A ATTTGTCTTT TATTTATATT TGACATGATA TAATAAaATA AAAAATAATA TAAAATATAA TAAGAAGAAA AATAAAAAGA ATTGTACATA TAAATCATAT ACAGATTATA TACATATAAT GTGTATAAAA TATTTTTTTT ATCTTGTCAT AGTTTTTAAT ATATTAATTT ATATAAAATT TTTTTTTTTT GTTTTTATTA TATCTGCCTA AATTTCTCAA TTTATCAATT TTTATGAAGC ATTCTCTCAT CTTTATCTTT TCCCCTTCTT TTTATTTCAT TTCTCTTTTT CCTTTTATTT CTTCTTTTTT ATATATAGTT TTTATGTTTA TGTAAATAAT TAAATAATCT TCTTTTTTAA AGGTAATTCT TTTTTTTCTT TTTTTTT

M D S S N Y F H D C K T M L S E H

N E S I E S S N N D I N G K Q K E H I K

G V Y D M L H L G H M K Q L E Q A K K L

I D Y V A H D D T P Y A N N Q K K K K K

I H P N E L N I G V T K A Q S I K M K K

K P L G T D F D Q G V E N L Q V K F K E

1 7

37

57

77

97

117

137

157

177

197

217

237

2 57

277

297

317

337

357

370

Fig. 1. Restriction map, DNA sequencing strategy and nucleotide sequence of I? fakiparum CCT gene. (A) Restriction map for the clone pCTRRX (4.3 Kb). The abbreviations used for the restriction sites are: A, AluI; B, BamHI; Bs, B.spH1; E, EcoRI; Nc, N c d ; Ns, N,siI; P, P.rtI; R, R.su1; S, Sspl. The arrows in the lower part of the figure represent the length and direction of individual sequencing reactions. The direction of translation is from left to right. The open reading frame of the coding region is shown by hatched box. The positions of the initiation codon ATG and termination codon TAA are also indicated. (B) Nucleotide sequence of CCT gene and predicted amino acid sequence (one-letter code). The sequence of the DNA fragment (CTI ) obtained following PCR is underlined. Sequences related to potential polyadenylation signals at 3' downstream region are underlined. The asterisk indicates the termination codon.

the highly conserved (up to 75 % identity) central domain of CCT, corresponding to residues 75-235 of the rat protein or residues 102-262 of the yeast protein. The primers were de- fined according to the /? jdciparum codon usage [39, 40) and redundancy was limited so as to increase the possibility of base pair matching at the 3' end, required for PCR priming f41j. PCR amplification using degenerate oligonucleotides P3 and P4 (see

Materials and Methods) resulted in an abundant 110-bp product, which was of the expected size, equivalent to the distance be- tween oligonucleotides P3 and P4 in the rat and yeast proteins. The amplified DNA was cloned and sequencing showed that the pCTl clone encoded 35 amino acids, presenting 70% identity with the sequences from the yeast and rat CCTs (residues 166- 200 and 139-173, respectively). This CT1 region had an A+T

66 Ye0 et al. ( E m J . Biochem. 233)

PfCCT 1 yCCT 1 hCCT 1 rCCT 1

PfCCT 60 yCCT 58 hCCT 39 rCCT 39

PfCCT 1 0 6 yCCT 118 hCCT 9 1 rCCT 9 1

PfCCT 1 6 6 yCCT 178 hCCT 1 5 1 rCCT 1 5 1

PfCCT 226 yCCT 220 hCCT 193 rCCT 1 9 3

I I -MDSSNYFHDCKTMLSEHNESIESSNNDINGKQKEHIKKGNSENQDVDPDTNPDAVPDDD MANPTTGKSSIRAKLSNSSLSNLFKK"KRQREETEE--QDNED~ES~QDENKDTQ ______----------- MDAQSSAKVN--SRKRRKEVPG---PNGATEEDGIPSKVQRCA ----__----------- MDAQCSAKVN--ARKRRKEAPG---PNGATEEDGVPSKVQRCA ... f .

DDDDDNSNDESEYE------SSQMDSEKNKGSIK---------NS~IYADG~DMLHL L T P R K K R R L T K E F E E K E A R Y T N E L P K E L R K Y R P K G F R F N L I R I Y A D G V F D L F H L VGLRQPAPFSDEIEVDFSKPYVRVTMEEACRGTP--------CE~~~ADGIFDLFHS VGLRQL'APFSDEIEVDFSKPYVRVTMEEASRGTP---------CE~~~ADGIFDLFHS * * * * * . . * . * . . . * "

GHMKQLEQCKKAFPNVTLIVGVPSDKITHKLKGLTVLTDKQRCETLTHCRWVDEVVPNAP GHARALMQAKNLFPNTYLIVGVCSDELTHNFKGFTVMNENERYDAVQHCRYVDEWRNAP G H A R A I S . I Q A K N L F P N T Y L I V G V C S D E L T H N F K G F T V M N "I . * * . * * *I*** * * * . * 1 x . . * . . . * * * * * . * . *

WCVTPEFLiEHKIDtlVA~~IPWSAD---------------------SD~IYKP~KEMGKFI, WTLTPEFLAEHRIDFVAHDDIPYSSAG------------------SDD~KHIKEAG~A WTLTPEFLAEHRIDFVAHDDIPYSSAG------------------SDD~KHIKEAGMFA * ..la.* , * * *I**"*** . * . * . * * *

~\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\y 1 ATQRTEGVSTTDLIVRILKNYEDYIERSLQRGIHPNELNIGVTKAQSIKMKKNLIRWGEK TTQRTNGVSTSDIITKIIRDYDKYLMRNFARGATRQELNVSWLKKNELEFKKHINEFRSY PTQRTEGISTSDIITRIVRDYDWARRNLQRGYTAKEL"-----EKKYHLQER~K PTQRTEGISTSDIITRIVRDYDVYARRNLQRGYTAKELNVSFIN----EKKYHLQER~K

* .. **** * - * * A * m * - * * . * * * * ** * * * *

pfCCT 2 8 6 yCCT 2 8 0 hCCT 2 4 9 rCCT 249

~~

I VTDELTKVTLTDKPLGTDFDQGVENLQVKFKELFKI~ASN------------------ FKKNQTNLNNASRDLYFEVREILLKKTLGKKLYSKLIGNELKKQNQRQRKQNFLDDPFTR VKKKVKDVEEKSKEFVQKVEEKSIDLIQKWEEKSREFIGSFLEMFGPEGALKHMLKEGKG VKEKVKDVEEKSKEFVQKVEEKSIDLIQKWEEKSREFIGSFLEMFGPEGALKHMLKEGKG

. .

PfCCT 328 yCCT 340 hCCT 309 rCCT 309

~

I J KLITDFTRKLEATSYLTSIQNIIDYEIEN-------DD--------YASSNFDDETSS*- KLIREASPATEFANEFTGENSTAKSPDDNGNLFSQEDDEDTNSNNTNTNSDSDSNTNSTP RMLQAISPKQSPSSSPTHERSPSPSFRWPFSGKTSPSS---------SPASLSRC~VTCD RMLQAISPKQSPSSSPTRERSPSPSFRWPFSGKTSPPC--------SP~LSRHK~YD * . . ... .

PfCCT yCCT 400 hCCT 3 6 1 rCCT 3 6 1

59 57 38 38

1 0 5 117 90 90

1 6 5 177 150 150

225 219 1 9 2 1 9 2

285 279 248 248

327 339 308 308

370 399 360 360

Fig. 2. Comparison of R fakiparum (PfCCT), human (hCCT), rat liver (rCCT) and yeast (yCCT) CCT amino acid sequences. The EMBL/ GenBank Database accession numbers are as follows: human, L28957; rat liver, L13245; yeast, X06598. Sequences were aligned with the CLUSTAL W multiple sequence alignment program. First, profile alignment was used to align PfCCT and yeast CCT sequences before adding the set of human and rat CCT sequences, using the Blosum 62 matrices. Similar results were obtained when the alignment was performed with the Dayhoff PZAM 250 matrix. Asterisks indicate identical matches and dots indicate conservative changes. Gaps in alignment were introduced to optimize similarities. Domain structure of CCTs is schematized on the P. ,fulciparum CCT in the upper part of the figure. The corresponding regions are indicated on the sequence alignment.

content of 66%, close to the average value of the fl falcipurum coding sequence, and codon usage was typically that of this or- ganism. This confirmed that the pCTl was derived from a gene encoding the fl falcipurum CCT. A genomic library, EcnRI, con- structed in iNM1149 phage vector was screened with the CTl probe. The partial restriction map of the positive clone (pCTRR8) was established (Fig. 1 A) and several subclones were constructed for sequencing. As the nucleotide sequence compo-

sition in fl ,fulciparurn resulted in the frequent absence of usual restriction enzyme sites for further subcloning, we performed an exonuclease 111 deletion on the pCTBBI and pCTBBII contain- ing 2.3-kb inserts in order to complete sequencing. In most parts of the 2.3-kb fragment, both strands of DNA were sequenced according to the strategy shown (Fig. 1 A). Sequences deter- mined in the same orientation were confirmed by independent clones and different sequencing primers.

Ye0 et al. (ELII: J . Biochern. 233) 67

Nucleotide sequence analysis of the l? fakiparum CCT gene. The P ,fakiparum CCT gene is apparently intronless, encoding a protein of 370 amino acids whose length is closer to that of rat liver CCT (367 amino acids) than that of the yeast CCT (424 amino acids). The predicted protein molecular mass is about 42.6 kDa. The start codon ATG (TAATAaG) is in an ideal context for translation [ 421. The overall nucleotide composition is typical for a P ,fakiparum sequence since (a) the coding re- gion (1110 bp) is 72% A + T and (b) the A + T content in the 5’ upstream (373 bp) and the 3’ downstream (396 bp) non-coding regions are 82% and 85% A+T, respectively, higher than that of the coding region. To ensure that the f? falciparum CCT se- quence had been cloned without rearrangement, we controlled the organization of the gene and a cDNA by PCR using specific primers designed to flank several different regions in the coding sequence. The PCR products generated from the cDNA were identical to those obtained from genomic DNA (unpublished re- sult), confirming the integrity of the clone and the lack of intron in the P jalciparum CCT gene.

Properties of the predicted l? falciparum CCT and similarity to CCTs from other organisms. The P falciparum CCT gene has an open reading frame encoding a protein of 370 amino acids. The protein sequence composition revealed few Pro, Arg and Cys residues and abundant Lys, Asp and Asn residues when compared with the yeast or mammalian CCT sequences. The predominance of Lys over Arg is remarkable in the I? ,fakiparum CCT protein sequence (43 Lys and nine Arg residues) compared to 47 Lys and 28 Arg residues in yeast CCT, and 28(29) Lys and 27(26) Arg residues in human(rat) CCT. This could reflect the high A +T nucleotide sequence composition in Plasmodiuwz. The plasmodia1 enzyme is predicted to be the most acidic among these CCTs, carrying a net charge of -19, compared to + 6 and -2 in the yeast and mammalian enzymes, respectively. In agreement with this, the predicted isoelectrical points are 5.2 for P ,falciparitm CCT, but 8.8 and 7.3 for the yeast and rat en- zymes, respectively. Putative consensus motifs present in the P julciparurn CCT amino acid sequence were searched for using the MOTIFS program and the PROSITE data base. Potential phosphorylation sites for the protein kinase C (amino acid posi- tions 79, 85, 150, 227, 272, 296, 326 and 334) and the casein kinase I1 (24, 50, 66, 70, 75, 129, 143, 206, 234, 294, 302 and 362) were detected.

Comparison of the four CCT sequences shows that the central domain has very strong similarity, whereas N-terminal and C-terminal domains have low but significant similarities (Fig. 2). The overall sequence of the predicted I? ,faleiparum CCT shows 41 -43% similarity and 28-29% identity to the overall sequences of yeast and mammalian CCTs, in comparison to the 48% similarity and 34% identity between yeast and mam- malian CCTs, and 97% similarity and 93% identity between hu- man and rat CCTs. The central domain ( R ,fakiparum CCT resi- dues 90-268), presumed to contain the catalytic site, shares 72% similarity (55% identity) with yeast protein residues 102- 262 and 6 8 % similarity (48% identity) with mammalian protein residues 75-235, not taking into account the 18-amino-acid in- sertion. In each CCT, the adjacent downstream region from the well-conserved central domain shows a highly charged domain corresponding to residues 274-323 of P ,faleiparum CCT, resi- dues 263 -322 of yeast CCT, and residues 237 -284 of rat CCT (38%, 38% and 58% charged amino acids, respectively) with the predominance of basic residues.

The hydropathy profile of the CCT gene product, as analysed by the method of Kyte and Doolittle [43] revealed that P ,fulci- parum CCT is highly hydrophilic (data not shown). The overall

322 311

Fig. 3. Helical-wheel projections of residues Lys (K274) to Thr (T294) and residues Val (V308) to Asn (N327) of l? fakiparum CCT. Charged polar residues are darkly shaded and neutral polar residues are hatched. The first and the last residues are marked by the directional arrows.

sequence of the P ,faleiparum CCT showed no segment suffi- ciently extended and sufficiently hydrophobic for a membrane- spanning helix, as also described in CCT from other organisms. Interestingly, in mammalian CCTs, the conserved central domain is followed by a long and unbroken 58-amino-acid a-helix (resi- dues 236-293) which includes a potential amphipathic a-helix of 33 amino acids (residues 256-288) composed of three 11- nucleotide repeats in tandem [38, 441. This domain has been reported to be involved in the membrane-binding domain [24, 271. Although the CCT sequence from P falciparum does not predict such a long and unbroken a-helix structure, the second- ary-structure analysis using Chou-Fasman Garnier and Robson and Nakashima algorithms and the GOR method revealed poten- tial a-helices in the regions of residues 274-294 and 308-324 in the P falciparum CCT sequence. Amphipathic distributions between polar and hydrophobic residues of these a-helices were evident when these regions were plotted on a helical wheel pro- jection (Fig. 3). The hydrophobic moment p values [33] were high throughout these regions, with average values of 0.46 for a-helix of residues 274-294 and of 0.33 for a-helix of residues 308 - 327.

Sequence similarity with other cytidylyltransferases. Tsuka- goshi and his colleagues [37] have already shown that signifi- cant degrees of local similarity exist among the yeast CCT, E. coli CDP-diacylglycerol pyrophosphatase, E. coli DNA poly- merase, E. coli 3-deoxy-manno-octulosonate cytidylyltransfer- ase and T4 DNA ligase. It should be noted that all these en- zymes are involved in a nucleotidyl transfer reaction. We, there- fore, compared the amino acid sequences in the functionally re- lated enzymes, limiting investigations to the CTP-cytidy- lyltransferases, including P ,fakiparum CCT, yeast CCT, human CCT, rat CCT and Bacillus subtilis CTP :glycerol-3-phosphate cytidylyltransferase [45]. The region comprising the central do- main of the four CCT homologues displayed about 37% identity with the sequence of B. suhtilis CTP:glycerol-3-phosphate cyti- dylyltransferase (Fig. 4). The significant conservation of the amino acid sequences in these functionally related enzymes strengthens the prediction of this domain as a catalytic core of the CTP-transfer reaction and probably suggests that they share a common evolutionary ancestor.

Gene copy number (Southern-blot). To determine the number of CCT genes in I? falciparum, we performed a Southern-blot analysis using genomic DNA (Nigerian strain) digested with dif- ferent restriction enzymes. One band was detected in the diges- tion with Pstl, EcoRI, BglII, Hind111 and BamHI, and two bands were observed in the sample treated with Ncol (Fig. 5) . This result is fully consistent with the cloning results (see restriction

68

7 - 6 - 5 - 4 -

3 -

2 -

1.6 - 1 -

0.5 -

Ye0 et al. ( E L M J . Biocliem. 233)

PfCCT 91 TKLFKGQWQTL 144

hCCT 76 ELTHNFKGFTVMNE 129 rCCT 76 P ELTHNFKGFTVMNE 12 9 yCCT 103 P KITHKLKGFTVLTD 156

BsGCT 1 MK EFNLQKQKKAYHSY 54

PfCCT 198 KKSKGKSFS 245 212 212 239

BsGCT 95 _._________-- 12 9

hCCT 174 _ _ _ _ _ _ _ _ _ rCCT 174 _ _ _ _ _ _ _ _ _ yCCT 201 _ _ _ _ _ _ _ _ _

Fig.4. Homology domain in the amino acid sequences of I? falciparum CCT (PfCCT) human CCT (hCCT), rat liver CCT (rCCT), yeast CCT (yCCT) and B. subtiZis CTP: glycerol-3-phosphate cytidylyltransferase (BsGCT). Identical residues (shadowed) and con5ervative changes are shown with solid boxes Gaps in alignment dre introduced to optimize \imilarities

A B

Fig. 6. Chromosomal localization. Parasite chromosomes from the 3D7 strain were separated by PFGE and blotted onto the Hybond-N+ mem- brane. Chromosomal-sized DNAs were loaded in different quantities for each lane. (A) The ethidium-bromide-stained gel. (B) Autoradiograph of a blot. The membrane was probed with "P-labelled 1.1-kb DNA frag- ments spanning the whole coding region of the R fulciparurn CCT gene. The number indicates the chromosomal assignment of the P ,fakiparum CCT gene; w denotes the position of the wells.

.W

- 1 3

Fig. 5. Southern-blot analysis. High-molecular-mass DNA was isolated from R ,fulciparurn (Nigerian strain) and samples of 4 pg each were digested with the restriction enzymes Pstl (lane l), Ncol (lane 2), EcoRl (lane 3 ) , Hind111 (lane 4), BglII (lane 5 ) and BamHI (lane 6) . The sizes of the molecular-mass markers (M) are given in kilobase pairs; w indi- cates the position of the wells. Filter was probed with "P-labelled coding region DNA fragments (1.1 kb).

map in Fig. 1 A) and indicates the presence of one gene/haploid genome. The result of EcoRI digestion, showing a band of 4.3 kb aided the choice of genomic library EcoRI for screening (see above), since this length would be of an appropriate size for accomodation in vector iLNM1149.

Chromosome localization. Hybridization of the F1 fulcipurum CCT gene probe to Southern blots of chromosomal DNA sepa-

rated by PFGE shows a single signal in chromosome 13. This was performed with the reference clone 3D7 at several DNA concentrations (Fig. 6). To confirm this result, we performed an- other hybridization using the same probe with a membrane con- taining three additional parasite strains (data not shown). In all the strains tested, a single band was detected in the region of chromosome 13. This result is consistent with a single-copy gene of the above Southern-blot analysis. It is noteworthy that all other metabolic pathway genes cloned to date, e.g. glycolytic pathway genes [46-513 from P fulciparum which have been defined for the chromosomal location, have also been found to be single copies.

R faleiparum CCT mRNA. The F1 falciparum CCT gene ex- pression was confirmed by Northern-blot analysis of erythro- cytic-stage mRNA. This showed a major 3-kb RNA transcript with a minor band at 4 kb. To confirm the specificity of our

Ye0 et al. (Eur: J . Biochern. 233) 69

--

285

I8S

SF

4 CDP-cho

4 P-cho

-0

Fig. 7. Northern-blot analysis of CCT expression. (A) About 20 pg total RNAs extracted from asynchronous cultures of erythrocytic-stage .? f'lciparunz were loaded i n lane 1. Total RNAs from human leukemia monocyte-like U937 cells were loaded in an equivalent amount (lane 2), as a control. Left, a photograph of the bromide-stained gel. Right, autoradiograph of a blot of the same gel. The membrane was hybridized with the '2P-labelled 1.1-kb DNA probe encompassing the entire coding region. Closed and open triangles indicate the positions of the ribosomal RNAs (28s and 18s) of plasmodial and human leukemia monocyte-like U937 cells, respectively. The arrows indicate the position of the tran- scripts. (B) To determine the specificity of CCT transcripts, about 20 pg total RNAs of P. ,fnki/mrunz were loaded in each lane. Left, photograph of the bromide-stained gel showing plasmodia1 rRNA (closed triangles) ; right, autoradiograph of blots. Following the blotting, the membrane was cut into three pieces. The filters were probed as follows: lane 1, hybrid- ization with the 5' specific probe spanning the 5' non translated region (370 bp), released from one of the clones obtained following exolll dele- tion; lane 2, hybridization with the 1.1-kb DNA probe of entire coding region; lane 3, hybridization with the 3' probe of the BwnHI-P.st1 frag- ment (about 1.2 kb) flanking the 3' of the gene, obtained from the clone pCTRR8.

hybridization signals, we performed several blottings and hy- bridizations under different stringencies. A Northern-blot of po- ly(A)-rich RNA confirmed the presence of both transcripts with the absence of the signal in poly(A)-depleted RNA (data not shown). There was no hybridization signal using total RNAs from human cells (monocyte-like U937 cells), even after pro- longed exposure of the autoradiogram (Fig. 7A). These results were confirmed using probes encompassing different regions of the gene, i.e. (a) 5' upstream fragment (370 bp, released from one of the clones following the ex0111 deletion), (b) coding re- gion (1.1 kb) and (c) 3' downstream fragment from the gene (about 1.2 kb, generated by the BumHIiPstI digestion of plasmid pCTRRX), as a non-specific probe (Fig. 7 B),

Expression of P. fakiparum CCT in E. coli and CCT activity. Expression of P ,fulciparum CCT in E. coli cells that are origi- nally deficient in this enzyme has enabled us to establish that this gene encodes a protein with the expected CCT activity. The coding region of P fulcipurum CCT was inserted downstream of bacteriophage T7 gene-1 0 promoter in the pET3d expression vector (named pETPfCCT), such that unfused, native P ,fulci- purum CCT is produced upon induction with IPTG. The expres- sion of P ,fdciparurn CCT in E. coli BLZl(DE3)pLysS trans- formed with pETPfCCT was analyzed by SDS/PAGE. Following induction with IPTG, overexpression of a protein cor- responding to 48 kDa was observed. Cells transformed with the pET3d plasmid without insert did not show significant overex- pression of any polypeptide following induction. The Western- blot using a rat anti-(CCT peptide) IgG [221 (kind gift of D. E. Vance, Edmonton, Canada) confirmed the identity of the overex- pressed protein (data not shown). The molecular mass of the expressed P. ,fakiparum CCT, estimated by SDS/PAGE, is slightly different from the predicted molecular mass (42.6 kDa).

A

terndale HCCT in wCTBBlI

Pexl Pex2 & a PCR product 1 E3zIEB

Pex3 Pex4 it

PCR product 2

recombinant PfCCT - PCR product 3

R

Fig. 8. Expression off! fakiparum CCT (PfCCT) in E. coli and CCT activity. (A) Scheme of the PfCCT coding region generated by PCR. For expression of the PfCCT without the fusion region, the PfCCT gene insert was obtained by two rounds of PCR. The arrows above the dif- ferently patterned boxes indicate the positions and the orientations of the primers used in P. ,fcdciparum CCT coding region amplification. (B) TLC of the reaction product. Extracts were prepared from E. coli harboring pETPfCCT and about 3 pg protein, except for in lane 2 (1.5 pg), were incubated with the CCT activity assay mixture as described in Materials and Methods. The reaction mixture was fractionated by TLC and the plate was exposed to an autoradiography film. Lane 1, no cell extract; lanes 2 and 3, complete assay systems containing 1.5 pg and 3 pg pro- tein, respectively; lane 4, minus MgC12; lane 5 , minus CTP; lane 6, with the extract from BL21 (DE3)pLysS cells haboring PET instead of pETPfCCT. SF, solvent front; 0, origine; CDP-cho, CDP-choline; P- cho, phosphocholine.

This is probably due to the nature of the protein sequence, possi- bly due to the presence of the poly(Lys) residues in the central domain. The reaction product of a CCT activity assay (Fig. 8 B) was identified as CDP-choline. CDP-choline was not formed when the supernatant of uninfected cells or cells infected with pET3d was used. The activity was stimulated by MgZ ' (Fig. 8B, lane 4) and dependent on CTP (lane 5 ) , as was the activity of other CCT molecules. The specific activity was calculated to be 10 pmol . mg ' . min-'. This result supports the notion that our cloned CCT gene is the structural gene for P. fulcipurum CCT.

70 Ye0 et al. (Eur: 3. Biochrm. 233)

DISCUSSION This is the first report of the molecular cloning of a phospho-

lipid metabolic pathway enzyme gene in I? ,faleiparum. The gene is intronless, encoding a protein of 370 amino acids. The nucleo- tide sequence compositions of the non-coding regions are 82% A + T (5’ upstream region) and 85% A + T (3’ downstream re- gion), higher than that of the coding sequence composition, 72% A+T. These are typical values in F! falciparurn [39, 521. The identity of our clone as a genuine l? fulciparum CCT gene i s based on its strong similarity (48-55% identity) to previously isolated yeast 1371, human [44] and rat protein [38] sequences in the central domain, presumably the catalytic core of this enzyme providing the CTP-transfer specificity. This criterion is strength- ened by extensive similarity within these four CCT homologues, and low but significant similarity to other CTP-cytidylyltransfer- ases, including CTP:glycerol-3-phosphate cytidylyltransferase from B. subtilis [45]. Recently, molecular cloning of CCT genes has been achieved in some other mammalian cell types, such as mouse (Rutherford, M., Tessner, T., Jackowski, S. and Rock, C. 0. (1992) Genbank accession number Zl2302), Chinese hamster 1531 and fetal rat type-I1 cells [54]. These mammalian CCTs appeared remarkably conserved. However, only one CCT, from the lower eukaryote S. cerevisiae, has, to our knowledge, been cloned to date [37], and our l? falciparum CCT is the second cloned enzyme from lower eukaryote. Regarding the C-terminal region of F! fulciparum CCT, which seemed to be incomplete when compared with the other homologues, we suspected that the gene was interrupted by intron(s). We have sequenced more than 800 bp downstream from the termination codon and no sig- nificant in-frame amino acid sequences have been determined. As intron sizes in Plasmodi~im are known to be relatively short, we excluded the possibility of an intron and conclude that the predicted protein sequence reported here is the complete se- quence.

One of the striking features emerging from our cloning i s the presence of an insertion of highly charged 18 amino acids in l? faleiparum CCT within the highly conserved central domain. Very probably, this region is not an intron, since no boundaries that confirmed to the AG/GT rule for intron-exon junctions [52] were observed within an appropriate range. A control experi- ment by PCR also confirmed the absence of the intron. Inser- tions have frequently been reported within the P. ,fulciparum se- quences, e.g. two 92-amino-acid insertions of dihydro-6-hy- droxy methylpterin pyrophosphokinase 15.5 I , one 62-amino-acid insertion of the 6-phosphate dehydrogenase gene 1561, and one 5-amino-acid insertion of lactate dehydrogenase within the con- served mobile loop region [461. More strikingly, recent cloning of the I? faleiparum enolase gene has described a pentapeptide insertion found uniquely in the higher plant homologues [51]. Unfortunately, a sufficient number of CCT sequences from the various species were not available to obtain informative points for the meaning of the insertion.

The results of hydropathy analysis indicated that the pre- dicted translation product of the CCT gene is a highly hydrophi- lic protein, as expected for a soluble cytosolic protein. The l? jblcipurum CCT shares an extended hydrophilic region within the N-terminal region with yeast CCT in the corresponding re- gion. The conservation of the catalytic site and of the adjacent highly charged domain between the various CCTs strengthens the prediction that this domain is involved in a substrate and co- factor binding domain. A number of consensus sequence motifs were found in R fiilciparum CCT and it would be interesting to study whether or not these sites are applied in vivo. The dupli- cated motifs, Ser-Pro, for the Pro-directed protein kinase de- scribed in mammalian CCT [571, were not found in CCT from l? falciparum.

CCT has been shown to be the principal rate-limiting en- zyme in the CDP-choline pathway in Plasmodium [I71 as well as in mammalian cells [ 101, whereas CCT is not the sole regula- tory enzyme in the CDP-choline pathway of yeast [%I. Con- cerning the regulation in mammalian cells, it is well established that the binding of the enzyme to membranes is mediated by an amphipathic a-helical domain [24-261. Interestingly, two re- gions are predicted to be a-helical structures i n the correspond- ing regions of l? ,faleiparum CCT. Furthermore, each of these domains are highly amphipathic, with a-helical hydrophobic mo- ment values 0.33 and 0.46. A cluster of charged residues, with the predominance of cationic residues, is present in each amphi- pathic a-helical structure. This suggests an interaction of this domain with the negatively charged membrane surface. Thus, l? falciparum CCT could interact with the membrane in the same way as the rat CCT. Although two a-helical structure portions also exist in the corresponding regions of yeast CCT, including the high density of charged residues, the helices are not amphi- pathic throughout the portions with hydrophobic moment value lower than 0.2.

Analysis of CCT gene expression has shown two large tran- scripts, a major band of 3 kb and a minor band of 4 kb. Admit- ting the presence of a 100-200-residue poly(A) tail, it could be suggested that the mature I? fulciparum CCT transcripts contain more than 1.5 kb of 5’ or/and 3‘ untranslated sequences. Al- though the presence and the function of such a long untranslated region in Plasmodium has not been well established, long tran- scripts have already been described for some other malarial genes, i.e. a protein kinase gene [60J, the glucose-6-phosphate dehydrogenase gene [56]. Two CCT transcripts have been re- ported from several rat tissues (1.5-1.8 kb and 4-7.5 kb) [38, 541. Interestingly, during myelination of the mouse nervous sys- tem, the CCT messenger levels are quite important during the first two postnatal weeks and decline thereafter [61 J , signifying that gene expression is developmentally regulated at the tran- scriptional level. Although the large transcript (4kb) of R fulci- parum CCT cannot be unambigously defined as a real transcript, since it is present as a minor signal, we could not exclude the possibility that multiple transcripts reflect specific regulation of l? fulciparum CCT gene transcription.

The in v i m activity of the recombinant F! fulciparum CCT clearly shows that the cloned CCT encodes a functional CCT. Our expressed protein i n E. coli will be important for enzymo- logical studies, especially those regarding the interaction of CCT and lipids. Since different regulatory modes between yeast CCT and rat CCT have already been observed, it will now be interest- ing to investigate regulatory modes of P. fakiparum CCT. The availability of the recombinant protein should certainly facilitate more rapid developments on the structure and enzymic function of CCT as well as its role in regulating PtdCho biosynthesis in Plasmodium.

In conclusion, we have established the presence of the F! falciparum CCT gene, which is, to our knowledge, the first cloning of a phospholipid metabolic-pathway enzyme in Plas- modium at the molecular level. This gene is present in a single copy and located on chromosome 13. This confirms that the intraerythrocytic Plusmodium possesses its own phospholipid biosynthetic machinery for manufacturing its membranes. The analysis of the predicted protein sequence led us to divide l? fulciparum CCT into three domains; (a) a N-terminal region, more similar to yeast CCT than mammalian CCTs, characterized by a long hydrophilic stretch, (b) a very conserved central do- main present among all the CCT homologues, indicating the ca- talytic core of this enzyme, and (c) a C-terminal region, more similar to the mammalian CCTs than yeast CCT, residing the two amphipathic a-helices that might be involved in the interac-

Ye0 et al. (Eul: J . Biochem. 233) 71

tion with t h e membrane . A t present, promising advance in o u r pharmacological model has been achieved using chol ine ana- logues that inhibit plasmodia1 PtdCho biosynthesis [9]. CCT cat- a lyzes the rate-limiting s tep in this pathway. The isolation of t h e CCT g e n e and its expression in E. coli o p e n u p n e w possibilities for s tudying the regulatory role of CCT in this important P t d C h o synthet ic pathway, a novel drug target, in F! falciparum.

We are very grateful to Dr G. Langsley for providing genomic librar- ies, and S. Bonnefoy for the kind gift of the chromosome blot. This work was supported by the UNDPiWorld B a n W H O Special Programme for Research and Training in Tropical Diseases (grant 920556), and the Commission of the European Communities (N TS3”-CT92. 0084). We thank Dr M. L. Ancelin for helpful discussions, Dr C. Roy for a critical reading of the manuscript, Dr D. E. Vance for his kind gift of rat CCTpep antibody and Dr J.-L Nikawa for providing the yeast CCT expression vector, having allowed precious control experiments.

REFERENCES 1. Holz, G. G. (1977) Lipids and the malaria parasite, Bull. WHO 55,

2. Sherman, L. (1979) Biochemistry of Plasmodium (malarial para- sites), Microhiol. Rev. 43, 453-495.

3. Vial, H. J., Ancelin, M. L., Philippot, J. R. & Thuet, M. J. (1990) Biosynthesis and Dynamics of Lipids in Plnsmodiurr2-infected mature mainrnalian erythrocytes, Blood Cells 16, 531 -555.

4. Vial, H. J. & Ancelin, M. L. (1992) Malarial Lipids, an overview, in Subcellular hiochemisrq (Avila, J . L. & Harris, J. R., eds) pp. 259-306, Plenum Press, NY.

5 . Ancelin, M. L., Vial, H. J . & Philippot, J. R. (1986) Characterization of choline and ethanolamine kinase activities in P lasnzodium-in- fected erythrocytes, in En7yme.s of lipid metabolism I1 (Freysz, L., Dreyfus, H., Massarelli, R. & Gatt, S., eds) pp. 59-64, Ple- num Press, NY.

6. Ancelin, M. L. & Vial, H. J . (1986) Several lines of evidence dem- onstrating that P lasmodium ,fo/cipcirum, a parasitic organism, has distinct enzymes for the phosphorylation of choline and ethanola- mine, FEBS Lett. 202, 217-223.

7. Ancelin, M. L. & Vial, H. J. (1986) Quaternary ammonium com- pounds efficiently inhibit Plasmodium jalcipcirum growth in vivo by impairment of choline transport, Antimicrobial Agents Chem- other: 29, 814-820.

8. Vial, H. J., Thuet, M. J., Ancelin, M. L., Philippot, .I. R. & Chavis, C. (1984) Phospholipid metabolism as a new target for malaria chemotherapy. Mechanism of action of D-2-amino-l -butanol, Biochem. Pharmacol. 33, 2761 -2770.

9. Vial, H. J., Calas, M., Bourguignon, J. J., Ancelin, M. L. & Giral, L. (1 994) Bis-2-aminopyridine: leur proctdt de prtparation et leur application, Patent No F941230, Group VIRBAC.

10. Vance, D. E. (1989) Regulatory and functional aspects of phosphati- dylcholine metabolism, in Phosphatidylcholiize metabolism (Vance, D. E., ed.) pp. 225-239, CRC press, Inc., Boca Raton, n.

11. Zinser, E.. Sperka-Gottlieb, C. D. M., Fasch, E.-V., Kohlwein, S. D., Paltauf, F. & Daum, G. (1991) Phospholipid synthesis and lipid composition of subcellular membranes in the unicellular eu- karyote Sacchuromyce.r cerevisiae, J . Bacteriol. 173, 2026-2034.

12. Billah, M. M. & Anthes, J . C. (1990) The regulation and cellular functions of phosphatidylcholine hydrolysis, Biochem. J. 269, 281 -291.

13. Exton, .I. H. (1990) Signaling through phosphatidylcholine break- down, J . Bid. Chem. 265, 1-4.

14. Pelech, S. & Vance, D. E. (1989) Signal transduction via phosphati- dylcholine cycles, Trends Biochem. Sci. 14. 28-30.

15. Carman, G. M. & Henry, S. A. (1989) Phospholipid biosynthesis in yeast, Annu. Rev. Biochem. 58, 635-669.

16. Nikoloff, D. M. & Henry, S. A. (1991) Genetic analysis of yeast phospholipid biosynthesis, Annu. Rev. Genet. 25, 559 -583.

17. Ancelin, M. L. & Vial, H. J. (1989) Regulation of phosphatidylcho- line biosynthesis in Plasmodictm-infected erythrocytes, Biochim. Biophys. Actci 1001, 82-89.

237 -248.

18. Kent, C. (1990) Regulation of Phosphatidylcholine Biosynthesis, P rog. Lipid Res. 29, 87 - 105.

19. Weinhold, P. A., Rounsifer, M. E., Williams, S. E., Brubaker, P. G . & Feldman, D. A. (1984) CTP:phosphorylcholine cytidy- lyltransferase in rat lung. The effect of free acids on the transloca- tion of activity between microsomes and cytosol, 1. Biol. Chem. 259, 10315- 10321.

20. Cornell, R. & Vance, D. E. (1987) Binding of CTP:phosphocholine cytidylykransferase to large unilamellar vesicles, Biochim. Bio- phjs. Acra 919, 37-48.

21. Jamil, H., Yao, 2. & Vance, D. E. (1990) Feedback regulation of CTP:phosphocholine cytidylyltransferase translocation between cytosol and endoplasmic reticulum by phosphatidylcholine, J . Biol. Chem. 265, 4332-4339.

22. Jamil, H., Utal, A. K. & Vance, D. E. (1992) Evidence that cyclic AMP-induced inhibition of phosphatidylcholine biosynthesis is caused by a decrease in cellular diacylglycerol levels in cultured rat hepatocytes, J . Bid. Chem. 267, 1752-1760.

23. Kolesnick, R. N. & Hemer, M. R. (1990) Physiologic 1,2-diacyl- glycerol levels induce protein kinase C-independent translocation of a regulatory enzyme, J. Biol. Chern. 265, 10900-10904.

24. Craig, L., Johnson, J. E. & Cornell, R. (1994) Identification of the membrane-binding domain of rat liver CTP:phosphocholine cyti- dylyltransferase using chymotrypsin proteolysis, J . Biol. Chem. 269, 3311-3317.

25. Johnson, J. E. & Cornell, R. B. (1994) Membrane-binding amphi- pathic a-helical peptide derived from CTP:phosphocholine cyti- dylyltransferase, Biochemistry 33, 4327 -4335.

26. Wieder, T., Geilen, C. C., Wieprecht, M., Becker, A. & Orfanos, C. E. (1994) Identification of a putative membrane-interacting do- main of CTP :phosphocholine cytidylyltransferase, FEBS Lett. 345, 207-210.

27. Johnson, J. E., Kalmar, G. B., Sohal, P. S., Walkey, C. J., Yamashita, S. & Cornell, R. B. (1992) Comparision of the lipid regulation of yeast and rat CTP:phosphocholine cytidylyltransferase expressed in COS cells, Biochem. J. 285, 815-820.

28. Trager, W. & Jensen, J. B. (1976) Human malaria parasites in con- tinuous culture, Science 193, 673-675.

29. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular clon- ing: u laboratory manual, 2nd edn, Cold Spring Harbor Labora- tory, Cold Spring Harbor, NY.

30. Henikoff, S. (1984) Unidirectional digestion with exonuclease Ill creates targeted breakpoints for DNA sequencing, Gene (Amst.)

31. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) DNA sequencing with chain terminating inhibitors, Proc. Nut1 Acud. Sci. USA 74,

32. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) CLUSTAL W : improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap pe- nalties and weight matrix choice, Nucleic Acids Res. 22, 4673- 4680.

33. Eisenberg, D., Weiss, R. M. & Terwilliger, T. C. (1984) The hy- drophobic moment detects periodicity in protein hydrophobicity, Proc. Narl Acad. Sci. USA 81, 140-144.

34. Bonnefoy, S., Guillotte, M., Langsley, G. & Mercereau-Puijalon, 0. (1 992) Plasmodiumfalciparum : characterization of gene R45 en- coding a trophozoite antigen containing a central block of six amino acid repeats, Exp. Parasirol. 74, 441 -451.

35. Hinterberg, K. & Scherf, A. (1994) PFGE: Improved conditions for rapid and high-resolution separation of Plasmodium falciparum chromosomes, Purasirol. Toduy 10, 225.

36. Tsukagoshi, Y., Nikawa, I.-I., Hosaka, K. & Yamashida, S. (1991) Expression in Escherichia coli of the Saccharomyces cerevisiae CCT gene encoding cholinephosphate cytidylyltransferase, J. Bacteriol. 173, 2134-2136.

37. Tsukagoshi, Y., Nikawa, J. -I . & Yamashita, S. (1987) Molecular cloning and characterization of the gene encoding cholinephos- phate cytidylyltransferase in Saccharomyces cerevisiae, Eul: J . Biochem. 169, 417-486.

38. Kalmar, G., Kay, R. J., Lachance, A., Aebersold, R. & Cornell, R. B. (1 990) Cloning and expression of rat liver CTP:phosphocholine cytidylyltransferase: An amphipathic protein that controls phos-

28, 351 -359.

5463 -5467.

12 Ye0 et al. (Eur: J. Biochem. 233)

phatidylcholine synthesis, Proc. Nut1 Acad. Sci. USA 87, 6029- 6033.

39. Hyde, J. E., Kelly, S. L., Holloway, S. P., Snewin, V. A. & Sims, P. F. G. (1 989) A general approach to isolating Plu.smodiumfalcipariim genes using non-redundant oligonucleotides inferred from protein sequences of other organisms, Mol. Biochem. Purusitol. 32, 247- 262.

40. Saul, A. & Battistutta, D. (1988) Codon usage in Plusmodiumfulci- parum, Mol. Biochem. Purusitol. 27, 35-42.

41. Sommer, R. & Tautz, D. (1989) Minimal homology requirements for PCR primers, Nucleic Acids Res. 17, 6749.

42. Kozak, M. (1986) Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ri- bosomes, Cell 44, 283-292.

43. Kyte, J. & Doolittle, R. F. (1982) A simple method for displaying the hydropathic character of a protein, J. Mol. Biol. 157, 105-

44.

45.

132. Kalmar, G. B., Kay, R. J., Lachance, A. C. & Cornell, R. B. (1994)

Primary structure and expression of a human CTP:phosphocho- line cytidylyltransferase, Biochim. Biophys. Acfa 1219, 328-334.

Manuel, C., Young, M. & Karamata, D. (1991) Genes concerned with synthesis of poly(g1ycerol phosphate), the essential teichoic acid in Bacillus suhtilis strain 168, are organized in two divergent transcription units, J. Gen. Microhiol. (37, 929-941.

46. Bzik, D., Fox, B. A. & Gonyer, K. (1993) Expression of Pln.smo- dium falciparum lactate dehydrogenase in Escherichia coli, Mol. Biochem. Parusitol. 59, 155 - 166.

47. Hicks, K. E., Read, M., Holloway, S. P., Sims, P. F. G. & Hyde, J. E. (1991) Glycolytic pathway of the human malaria parasite Plasmodium julcipurum : primary sequence analysis of the gene encoding 3-phosphoglycerate kinase and chromosomal mapping studies, Gene (Amst.) 100, 123- 129.

48. Kaslow, D. C. & Hill, S. (1990) Cloning metabolic pathway genes by complementation in Esclzerichiu coli, J. Biol. Chem. 265, 12337- 12 341.

49. Olafsson, P., Matile, H. & Certa, U. (1992) Molecular analysis of Plasmodium fulciparum hexokinase, Mol. Biochem. Pnrasitol. 56, 89- 102.

50. Ranie, J., Kumar, V. P. & Balaram, H. (1993) Cloning of the triosephosphate isomerase gene of Plusmodium fafciparum and

expression in Escherichia coli, Mol. Biochem. Parurifol. 61 159- 170.

51. Read, M., Hicks, K. E., Sims, P. F. G. & Hyde, J. E. (1994) Molecu- lar characterisation of the enolase gene from the human malaria parasite ,? fulczparum. Evidence for ancestry within photosyn- thetic lineage, EUK J. Biochern. 220, 513-520.

52. Weber, J. L. (1988) Molecular Biology of Malaria Parasites, Exp. Parusitol. 66, 143-170.

53. Sweitzer, T. D. & Kent, C. (1994) Expression of wild-type and mu- tant rat liver CTP:phosphocholine cytidylyltransferase in a cyti- dylyltransferas-deficient Chinese Hamster Ovary cell line, Arch. Biochem. Biophys. 31 1, 107 - 11 6.

54. Hogan, M., Zimmerman, L. J. , Wang, I., Kuliszewski, M., Liu, J. & Post, M. (1994) Increased expression of CTP:phosphocholine cytidylyltransferase in maturing type I1 cells. Am. J. Physiol. 267, L25-L32.

55.

56.

Triglia, T. & Cowman, F. (1994) Primary structure and expression of the dihydropteroate synthetase gene of Plasmodium falci- parum, Proc. Nut/ Acad. Sci. USA 91, 7149-7153.

O’Brien, E., Kurdi-Haidar, B., Wanachiwanawin, W., Carvajal, J.- L., Vulliamy, T. J. , Cappadoro, M., Mason, P. J. & Luzzatto, L. (1 994) Cloning of the glucose 6-phosphate dehydrogenase gene from Pla.smodium falciparum, Mol. Biochem. Parnsitol. 64, 313 - 29L J L V .

57. MacDonald, J. I. S. & Kent, C. (1994) Identification of phosphoryla- tion sites in rat liver CTP:phosphocholine cytidylyltransferase, J. Biol. Chem. 26’9, 10529-10537.

58. McMaster, C. R. & Bell, R. M. (1994) Phosphatidylcholine biosyn- thesis via the CDP-choline pathway in Sacchuromyces cerevisiae, J . B i d . Cheni. 269, 14776-14783.

59. Zhao, Y., Kappes, B., Yang, J. & Franklin, R. M. (1992) Molecular cloning, stage-specific expression and cellular distribution of a putative protein kinase from Pla.smodiumf~lcipurum, Eur: J. Bio- chem. 207, 305-313.

60. Garbay, B. & Cassagne, C. (1 994) Expression of the ceramide galac- tosyltransferase gene during myelination of the mouse nervous system. Comparison with the genes encoding myelin basic pro- teins, choline kinase and CTP:phosphocholine cytidylyltransfer- ase, Dev. Bruin Rex 83, 119-124.