BACKGROUND OF THE INVENTION Cholinephosphotransferase catalyzes the final step in the synthesis of PtdCho through the Kennedy (CDP-choline) pathway via the transfer of a phosphocholine moiety from CDP-choline to diacylglycerol with the release of CMP and the formation of PtdCho (Vance, 1996 in Biochemistry of Lipids, Lipoproteins and Membranes, Vance, DE and Vance, JE, Elsevier: Amsterdam, pp 153-182; Kennedy and Weiss, 1956, J Biol Chem 222: 193-214; Weiss et al, 1958, J Biol Chem 231: 53-64; McMaster and Bell, 1997, BiochimBiophysActa 1348: 100-110; Hjelmstad and Bell, 1991, J Biol Chem 266: 4357-4365; McMaster and Bell, 1994, J Biol Chem 269: 28010-28016; and Cornell, 1989 in Phosphatidylcholine Metabolism, Vance, DE, Boca Raton: Florida, pp. 47-64). The fatty acyl composition of the diacylglycerol molecule utilized by cholinephosphotransferase determines the fatty acyl array for de novo synthesized PtdCho. In addition, the intracellular location of cholinephosphotransferase identifies the site of de novo PtdCho synthesis for subsequent transfer to other organelles, or assembly with proteins and other lipids for secretion during the synthesis of lung surfactant, lipoproteins, and bile (Jobe, 1993, N Engl J Med 328: 861-868; Steinberg, 1997, J Biol Chem 272: 20963-20966; Ruetz and Gros, 1994, Cell 77: 1071-1081).

Properties of cholinephosphotransferase activities present in membrane preparations have identified an activity capable of de novo synthesis of the PtdCho structural analogues, platelet activating factor (PAF), a major mediator of inflammatory processes, as well as PAF precursor (Snyder, 1997, Biochim Biophys Acta 1348: 111-116; Leslie, 1997, J Biol Chem 272: 16709-16712; Venable et al, 1993, J Lipid Res 34: 691-702). However, the biological and biochemical roles of a de novo synthesized PAF pathway, and indeed the roles of the various proposed isoforms of cholinephosphotransferase in the regulation of the partitioning of lipid biosynthetic pathways, have yet to be determined as a mammalian cholinephosphotransferase has not been cloned or purified.

Genetic approaches led to the isolation of two genes encoding cholinephosphotransferase activities from the yeast Saccharomyces cerevisiae.

The yeast CPT1 gene product encodes a cholinephosphotransferase (McMaster and Bell, 1994; Hjelmstad and Bell, 1990, J Biol Chem 265: 1755-1764; Hjelmstad and Bell, 1987, J Biol Chem 262: 3909-3917) specific for the synthesis of PtdCho in vitro and in vivo, while the EPT1 gene product codes for a dual specificity choline/ethanolaminephosphotransferase capable of synthesizing PtdCho and phosphatidylethanolamine (PtdEtn) in vitro, but which synthesizes primarily PtdEtn in vivo (McMaster and Bell, 1994; Hjelmstad and Bell, 1991, J Biol Chem 266: 5094-5103). Analysis of chimeric CPT1/EPT1 enzymes (McMaster and Bell, 1994; Hjelmstad and Bell, 1991; Hjelmstad and Bell, 1988, J Biol Chem 263: 19748-19757) mapped the active site domain, and site-directed mutagenesis identified a diagnostic catalytic motif (Hjelmstad et al, 1994, J Biol Chem 269: 20095-21002). It was hypothesized that active site residues would be conserved between Genera. This rationale was used as a basis to isolate human choline/ethanolaminephosphotransferase cDNAs (hCEPT1 and hCEPT2) for subsequent expression and characterization.

SUMMARY OF THE INVENTION According to a first aspect of the invention, there is provided an isolated native, cloned, recombinant or synthetic DNA sequence encoding hCEPT1 protein comprising the DNA sequence of SEQ ID NO: 1 or sub-fragments thereof.

According to a second aspect of the invention, there is provided an hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2.

According to a third aspect of the invention, there is provided a DNA molecule encoding hCEPT1 protein, said DNA deduced from the amino acid sequence of SEQ ID NO: 2.

According to a fourth aspect of the invention, there is provided an isolated native, cloned, recombinant, or synthetic DNA sequence encoding hCEPT2 protein comprising the DNA sequence of SEQ ID NO: 3 or sub-fragments thereof.

According to a fifth aspect of the invention, there is provided hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4.

According to a sixth aspect of the invention, there is provided a DNA molecule encoding hCEPT1 protein, said DNA deduced from the amino acid sequence according to SEQ ID NO: 4.

According to a seventh aspect of the invention, there is provided a recombinant expression system, capable, when transformed into a host cell, of expressing a DNA sequence encoding hCEPT1 (SEQ ID NO: 2) or hCEPT2 (SEQ ID NO: 3) which system comprises control sequences effective in said host cell operably linked to said DNA sequence.

According to an eighth aspect of the invention, there is provided a host cell transformed with the expression system described above.

The host cell may be selected from the group consisting of: a plant cell; a yeast cell; a bacterial cell; and a mammalian cell.

According to a ninth aspect of the invention, there is provided mutant hCEPT1 protein having an amino acid sequence substantially as shown in Figure 2A (SEQ ID NO: 2) and having a missense mutation at glycine 156.

According to a tenth aspect of the invention, there is provided a DNA molecule encoding mutant hCEPT1 protein, said DNA deduced from the amino acid sequence described above.

According to an eleventh aspect of the invention, there is provided mutant hCEPT2 protein having an amino acid sequence substantially as shown in Figure 2B (SEQ ID NO: 4) and having a missense mutation at glycine 156.

According to a twelfth aspect of the invention, there is provided a DNA molecule encoding mutant hCEPT2 protein, said DNA deduced from the amino acid sequence described above.

According to a thirteenth aspect of the invention, there are provided antibodies directed against hCEPT, said hCEPT selected from the group consisting of: hCEPT1 (SEQ ID NO: 2); hCEPT2 (SEQ ID NO: 4); mutant hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2 and having a missense mutation at glycine 156; mutant hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4 and having a missense mutation at glycine 156; and immunoreactive fragments thereof.

The antibodies described above may be used to identify proteins related to hCEPT1 and hCEPT2 or for diagnostic use.

According to a fourteenth aspect of the invention, there is provided a method of synthesizing lipids containing a given fatty acid composition comprising: providing hCEPT protein, selected from the group consisting of: hCEPT1 (SEQ ID NO: 2); hCEPT2 (SEQ ID NO: 4); mutant hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2 and having a missense mutation at glycine 156; mutant hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4 and having a missense mutation at glycine 156; and combinations thereof; providing substrates required for lipid biosynthesis; combining the hCEPT protein and the substrates; incubating the hCEPT protein and the substrates under conditions promoting lipid biosynthesis; and harvesting the lipids.

According to a fifteenth aspect of the invention, there is provided lipids prepared according to the above-described method.

The lipids may be used as a food additive.

According to a sixteenth aspect of the invention, there is provided a method of assaying a compound for modulation of lipid metabolism comprising: providing hCEPT protein, selected from the group consisting of: hCEPT1 (SEQ ID NO: 2); hCEPT2 (SEQ ID NO: 4); mutant hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2 and having a missense mutation at glycine 156; mutant hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4 and having a missense mutation at glycine 156; and combinations thereof; providing substrates required for lipid biosynthesis; providing a compound proposed to modulate lipid metabolism; combining the compound, the hCEPT protein and the substrates; incubating the compound, the hCEPT protein and the substrates under conditions promoting lipid biosynthesis; harvesting the iipids; and characterizing the lipids, thereby determining the effect of the compound on lipid metabolism.

According to a seventeenth aspect of the invention, there is provided a reagent for use in disease diagnosis or genotyping comprising antibodies directed against hCEPT, said hCEPT selected from the group consisting of: hCEPT1 (SEQ ID NO: 2); hCEPT2 (SEQ ID NO: 4); mutant hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2 and having a missense mutation at glycine 156; mutant hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4 and having a missense mutation at glycine 156; and immunoreactive fragments thereof.

According to an eighteenth aspect of the invention, there is provided a reagent for use in identifying proteins related to hCEPT comprising antibodies directed against hCEPT, said hCEPT selected from the group consisting of: hCEPT1 (SEQ ID NO: 2); hCEPT2 (SEQ ID NO: 4); mutant hCEPT1 protein having an amino acid sequence substantially as shown in SEQ ID NO: 2 and having a missense mutation at glycine 156; mutant hCEPT2 protein having an amino acid sequence substantially as shown in SEQ ID NO: 4 and having a missense mutation at glycine 156; and immunoreactive fragments thereof.

According to a nineteenth aspect of the invention, there is provided a nucleotide probe selected from the group consisting of: the DNA sequence of SEQ ID NO: 1; the DNA sequence of SEQ ID NO: 3; and fragments thereof.

According to a twentieth aspect of the invention, there is provided an oligonucleotide for use in identifying genes related to hCEPT, said oligonucleotide selected from the group consisting of: the DNA sequence of SEQ ID NO: 1; the DNA sequence of SEQ ID NO: 3; and fragments thereof.

According to a twenty-first aspect of the invention, there is provided an antisense probe for use in treating lipid metabolic disorders, said antisense probe being complementary to the DNA sequence of SEQ ID NO: 1; the DNA sequence of SEQ ID NO: 3; or fragments thereof.

One embodiment of the invention will now be described in conjunction with the accompanying figures in which: BRIEF DESCRIPTION OF THE DRAWINGS Figure 1A is the nucleotide sequence of the isolated cDNA molecule hCEPT1 (SEQ ID NO: 1); Figure 1B is the nucleotide sequence of the isolated cDNA molecule hCEPT2 (SEQ ID NO: 3).

Figure 2A is the amino acid sequence of hCEPT1 (SEQ ID NO: 2); Figure 2B is the amino acid sequence of hCEPT2 (SEQ ID NO: 4).

Figure 3 is a Western Blot of hCEPT1 protein expressed in S. cerevisiae and E. coli.

Figure 4A is a Northern blot analysis of hCEPT1 and ß-actin transcripts in human cell types; Figure 4B is a Northern blot analysis of hCEPT2 transcripts in human cell types Figure 5 is an alignment of hCEPT1 protein with known choline- ethanolamine-phosphotransferases.

Figure 6 is a predicted secondary structure of hCEPT1 protein.

Figure 7 is a bar graph of in vitro cholinephosphotransferase and ethanolaminephosphotransferase activity in hCEPT1 mutants.

Figure 8 is a bar graph of metabolic labelling in hCEPT1 mutants.

Figure 9 is a schematic diagram showing the role of choline/ethanolaminephosphotransferases in lipid formation.

Table 1 summarizes the diradylglycerol and CDP-aminoalcohol specificities of hCEPT1 protein.

Table 2 summarizes the results of overexpression of hCEPT1 and hCEPT2 in human HEK-293s cells.

DETAILED DESCRIPTION Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belons. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred methods and materials are now described. All publications mentioned hereunder are incorporated herein by reference.

Described herein is the first cloning and expression from a mammalian source of proteins capable of catalyzing choline-and ethanolaminephosphotransferase reactions (hCEPT1 and hCEPT2). A CDP- alcohol phosphotransferase motif has been identified in both hCEPT1 and hCEPT2 and site-directed mutagenesis was subsequently used to determine the residue responsible for choline-versus ethanolamine-phosphotransferase activity.

Specifically, mutation of glycine 156 of hCEPT1 abolished ethanolaminephosphotransferase activity without abolishing cholinephosphotransferase activity. Furthermore, this residue is conserved in hCEPT2. hCEPT1 and hCEPT2 can be used for transferring phosphocholine or phosphoethanolamine to an acceptor lipid. These may include, but are in no way limited to, diacylglycerols, alkylacylglycerols and alkylalkylglycerols. Thus, the above-described enzymes may be used to synthesize lipids in vivo and in vitro.

Furthermore, hCEPT1 and hCEPT2 can be used in assay systems for assessing the effect of different compounds on membrane composition which would in turn allow for the rapid identification of compounds capable of modulating lipid content.

Given that the products of hCEPT1 and hCEPT2 include platelet activating factor, which is a mediator of inflammation and phosphatidylcholine, which is a component of lung surfactant, lipoproteins and bile, it is clear that compounds effective in treating a number of disorders could be identified in this way. Similarly, hCEPT1 and hCEPT2 could be used as gene therapy targets or antisense targets or as enzyme therapy agents for treating these disorders. Furthermore, hCEPT1 and hCEPT2 can be used to identify related enzymes. That is, nucleotide probes may be used to identify other isoforms of hCEPT1 and/or hCEPT2 and antibodies directed against the peptides or fragments thereof may be used to identify related enzymes in other organisms.

The invention will now be described by way of examples, although the invention is in no way limited to the examples.

EXAMPLE I-MATERIALS [a-32P] dATP (3000Ci/mmol) and [Y-32P] ATP (3000Ci/mmol) were purchased from DuPont/NEN. [Methyl-14C] CDP-choline and [methyl-14C] choline were purchased from American Radiolabeled Chemicals. [Ethanolamine-1,2- '4C] CDP-ethanolamine and [1,2-14C] ethanolamine were products of ICN. XGT11 forward sequencing primer, -GT1 1 human Quick-CloneT cDNA libraries, and AdvantageT cDNA polymerase were purchased from Clontech. The pCMV-Sport human brain cDNA library, custom oligonucleotides, and T4 DNA ligase were products of Life Technologies. Manual dideoxy sequencing was performed utilizing the T7 sequencing kit (Pharmacia). Lipids were purchased from Avanti Polar Lipids. All other reagents were of the highest quality commercially available.

EXAMPLE II-ISOLATION AND EXPRESSION OF A FULL LENGTH CHOLINEPHOSPHOTRANSFERASE cDNA.

A tblastn search (McMaster et al, 1996, Biochem J 313: 729-735) was performed versus the Expressed Sequence Tag (EST) data base using default parameters versus the predicted amino acid sequences of both the entire S. cerevisiae Cpt1p coding region and the Cpt1p CDP-alcohol phosphotransferase motif (Hjelmstad et al, 1994). An EST (human Jurkat T cell, Genbank accession number AA312638) was identified and sequenced in its entirety on both strands. A proposed coding region within the EST was amplifie by PCR and subcloned into the Bam HI/Sal I sites of the E. coli expression vector pET23a (Novagen) resulting in the addition of an 11 amino acid T7 epitope tag (Williams and McMaster, 1998, J Biol Chem 273: 13482-13487) to the N-terminus of the protein (pAH5). The T7 tagged version of hCEPT1 was excised from pAH5 with Bgl II and Sal I and subcloned into the constitutive S. cerevisiae expression vector p416 GPD (Gish and States, 1993, Nature Genet 3: 266-272) creating pAH9. All PCR derived products were sequenced in their entirety.

EXAMPLE III-ISOLATION OF A FULL LENGTH hCEPT2 cDNA The 365 bases of sequence deposited for the 1.1 kb EST clone 67440 (isolated from a Stratagene human placenta cDNA library) was initially identified based on significant homology to the active site of S. cerevisiae cholinephosphotransferase. EST clone 67440 was obtained from the IMAGE consortium and sequenced in its entirety on both strands utilizing a combination of manual dideoxy (Sambrook, J, Fritsch, EF, and Maniatis, T., 1989 in Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, New York) and automated techniques (Li-cor apparatus, National Research Council of Canada/Dalhousie University joint laboratory). Comparison of the completed sequence of the EST 67440 cDNA to the known yeast (Hjelmstad and Bell, 1990, J. Biol. Chem 265: 1755-1764; Hjelmstad and Bell, 1991, J. Biol. Chem. 266: 5094-5103), plant (Dewey et al, 1994, Plant Cell 6: 1495-1507), and our recently cloned human choline/ethanolaminephosphotransferase 1 sequence described above revealed that the EST was incomplete at the 5'end. The 5'end was extended through two successive rapid amplification of cDNA ends (RACE) protocols (Innis, M. A. et al, 1990 in PCR Protocols: A Guide to Methods and Applications, Academic Press, Inc., San Diego) from human kGT11 cDNA libraries. First, a human placenta cDNA library (Clontech) was amplifie utilizing XGT11 forward primer and EST 67440 specific primers using a nested PCR approach. The first PCR reaction contained 20 mM Tris-HCI (pH 8.4), 50 mM KCI, 3 mM MgCI2,0.2 mM dNTPs, 0.8 pM GT11 forward primer, 0.8 uM EST 67440 specific primer GTCTGCCAATGAGCGCAATAA, 4 pi XGT11 human placenta cDNA library, 2.5 units Taq DNA polymerase (Life Technologies). PCR reaction conditions were performed for 30 cycles of 94°C for 1min, 55°C for 1 min, and 72°C for 2 min. A 5 pl aliquot was removed from this PCR reaction and used as template for a second PCR reaction performed under identical conditions as the above reaction except a nested EST 67440 specific primer CCAGTCAGGATAAGTTCCTAAGCGA was utilized. A 3 pi aliquot of this second PCR reaction was TA cloned into pCR2.1 (Invitrogen) and transformed into E. coli.

Transformants were selected on LB plates containing 100 ug/ml ampicillin. To identify inserts containing EST 67440 5'sequences, a 150 base pair probe was generated by PCR amplification of EST 67440 utilizing the primers CCAGTCAGGATAAGTTCCTAAGCGA and GGCACGAGTACCAGTCAC. This EST 67440 specific probe was subjected to random prime labeling (Amersham MultiprimeT DNA labelling system). E. coli colonies containing TA cloned inserts were transferred to Hybond-N nyFon membranes (Amersham) and screened by colony hybridization (Sambrook, J, Fritsch, EF, and Maniatis, T., 1989 in Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, New York) utilizing the labeled EST 67440 specific probe. Plasmids were prepared from positive colonies and inserts were sequenced. This strategy extended the sequence 164 bases towards the 5'end of the cDNA, however, after several attempts a complete cDNA was never recovered using the above primer set. To further extend the sequence, a universal human XGT11 cDNA library (Clontech) was subjected to the identical RACE strategy using primers derived from the extended sequence. The sequence of the exterior primer was CAGTAGGAGATGAGCACGAGCGTG and the nested primer was GTGACCACGTTGACGGCGAGCCCC. Colonies were screened using the oligo CTGCTCCAGTGGATCCCGCTCTGG end labelle with T4 kinase (Sambrook, J, Fritsch, EF, and Maniatis, T., 1989 in Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, New York). The cDNA sequence was extended a further 158 bases using this strategy, however, the cDNA was still incomplete.

A full length cDNA was isolated using the Genetrapper system technology (Life Technologies). A primer corresponding to the most 5'RACE extneded sequence in hand, TATCGGCGGGTGGAGCACCGCTACAGC was biotinylated using terminal deoxynucleotide transferase and hybridized in solution to a single stranded human brain cDNA library that had been generated by digestion of the library with Gene II protein and exonuclease 111. Avidin bound to magnetic beads was used to selectively precipitate the biotinylated probe/hybridized cDNA complex, thus enriching the library for cDNAs complementary to the oligonucleotide used. The single stranded cDNAs were extended using Klenow and E. coli were transformed and screened for positive clones with the end labeled antisense oligo, CCAGTCAGGATAAGTTCCTAAGCGA, found within the original EST 67440.

Plasmid DNA from positive colonies was amplified and hCEPT2 specific inserts confirmed and sized by PCR. The library was constructed in a vector for over- expression of the encoded cDNA in cell culture, hence, one of the positive hCEPT2 plasmids was selected for transfection of HEK293s cells (human placenta kidney cell line) for subsequent enzyme assay analysis to confirm function.

EXAMPLE IV-NORTHERN BLOT.

Random primed [32p] labeled probes (Roberts and Green, 1994, Nature 371: 717-720) were synthesized versus either the entire 1.2 kb coding region of hCEPT1, the entire 1.2 kb coding region of hCEPT2 or a 2.0 kb region of human a-action cDNA. Multiple human tissue Northern blots (Clontech) were hybridized at 68°C in ExpressHybw solution (Munberg et al, 1995, Gene 156: 119- 122) for 1 hr and washed as per manufacturer's instructions. Blots were exposed to X-ray film for 1-3 days, as shown in Figure 4.

EXAMPLE V-WESTERN BLOT Microsomal membranes were prepared from BL21 (DE3) pLysS E. coli and HJ091 S. cerevisiae cells (Hjelmstad et al, 1994) grown to mid-log phase in appropriate media to ensure plasmid maintenance (Feinberg and Vogelstein, 1984, Anal Biochem 137: 266-267). Specifically, membrane preparations from S. cerevisiae strain HJ091 (cpt1:: LEU2 ept1-) containing pAH9 or p416 GPD, or from E. coli BL21 (DE3) pLysS containing pAH5 ( induction of protein with 0.4 mM IPTG for 2 hours at 25°C) were separated by SDS-PAGE and blotted with a monoclonal antibody to the T7 epitope. Blots were probed with a T7 epitope tag specific monoclonal Ab (1: 5,000, Novagen) coupled to horse radish peroxidase for detection using the ECL (Amersham) system, as shown in Figure 3.

Alternatively, membranes were prepared from HJ091 S. cerevisiae cells (cpt1:: LEU2 ept1-) (Williams and McMaster, 1998, J. Biol. Chem 273: 13482- 13487) grown to mid-log phase in appropriate media to ensure plasmid maintenance (Kaiser, C., Michaelis, S., and Mitchell, A., 1994 in Methods in Yeast Genetics, Cold Spring Harbor Press, New York). The HJ091 strain of S. cerevisiae is devoid of endogenous choline-or ethanolamine-phosphotransferase activity due to inactivated alleles at the loci encoding for these activities (cpt1:: LEU2 ept1-).

Proteins were transferred to PVDF membranes (Harlow, E. and Lane, D., 1988 in Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York) and blots were probed with a T7 epitope tag specific monoclonal Ab (1: 5,000, Novagen) coupled to horse radish peroxidase for subsequent detection using the ECL (Amersham) system.

EXAMPLE VI-SITE-DIRECTED MUTAGENESIS Plasmid pAH9 (Henneberry, AL, and McMaster, CR, 1999, Biochem.

J. 339: 291-298) contains a T7 epitope tagged version of the hCEPT1 cDNA in the constitutive yeast expression vector pGPD416 (Munberg, D., Muller, R., and Funk, M., 1995, Gene 156: 119-122) and was used as the template for all mutagenesis reactions. The T4 DNA polymerase directed MORPH plasmid DNA mutagenesis protocolTM (5 prime-3 prime) was used with the appropriate mutagenic oligonucleotides as directed by the manufacturer. All mutations were confirmed by DNA sequencing. It is of note that as can be seen in Figures 7 and 8, mutation of glycine 156 to alanine, serine or cysteine abolished ethanolaminephosphotransferase activity. Thus, the residue that alters enzyme specificity has been identified, as mutants are no longer able to use CDP- ethanolamine as a substrate.

EXAMPLE VIl-ENZYME ASSAYS Cholinephosphotransferase and ethanolaminephosphotransferase activities were determined using the method of O et al. (O., K-M., Siow, Y. L., and Choy, P. C., 1989, Biochem. Cell Biol. 67: 680-686) from microsomal membrane preparations of either HEK293s cells (human kidney cell line) transformed with pcDNA3.1 (mock), pMM6 (hCEPT1) or pCMVsport-hCEPT2 (hCEPT2), or S. cerevisiae HJ091 cells (cpt1:: LEU2 ept1-, yeast cells devoid of endogenous choline-and ethanolamine-phosphotransferases) transformed with p416GPD (mock), or pAH9 (hCEPT1) and its various mutant derivatives. Diradylglycerols or ceramides were dried under nitrogen gas and resuspended in 0.015% (w/v) Tween 20tam by sonication. Unless otherwise indicated, assay buffer contained 100 mM Tris-HCI (pH 8.0), 20 mM MgCI2,1 mM EDTA, 1 mM diradylglycerol or ceramide (final Tween 20 concentration of 0.00375%, w/v), and 10 ug of microsomal protein.

Components were incubated at room temperature for 5 min followed by the addition of CDP-choline or CDP-ethanolamine (0.2 mM, 2000 dpm/nmol). Assays were incubated at 37°C for 15 min and activity terminated by the addition of 3 mi CHCI3/CH30H (2/1, v/v) followed by 1.5 ml 0.9% (w/v) KCI. Tubes were vortexed and phase separation facilitated by centrifugation at 2000 g for 10 min. The aqueous phase was aspirated and the organic phase was washed twice with 1.5 ml 40% (v/v) CH30H. An aliquot of the organic phase was dried in a scintillation vial and radioactivity was determined. Samples were routinely analyzed by thin layer chromatography on silica gel plates in a solvent system of CHCI3/CH30H/NH40H/H2) (70/30/4/2, v/v) to confirm the synthesis of PtdCho, PtdEtn, and PAF and their analogues.

EXAMPLE VIII-PROTEIN AND LIPID DETERMINATION Protein was determined by the method of Lowry et al, 1951, J Biol Chem 193: 265-275 using bovine serum albumin as standard. Diradylglycerols not available commercially were prepared from PtdCho by Bacillus cereus phospholipase C digestion (Boehringer Mannheim) as directed by the manufacturer and yield was estimated using the method of Stern and Shapiro (Stern, I., and Shapiro, B., 1953, J. Clin. Pathol. 6: 158-160). Phospholipid phosphorus was determined by the method of Ames and Dubin (Ames, B. N. and Dubin, D. T., 1960, J. Biol. Chem. 235: 769-775).

EXAMPLE IX-ANALYSIS OF THE hCEPT1 cDNA PRODUCT The hCEPT1 coding region predicts a protein of molecular weight 46,550 comprised of 416 amino acid residues of which 48.6% are hydrophobic, as shown in Figure 2 (SEQ ID NO: 2). No signal or sorting sequences were apparent.

An expression system devoid of endogenous choline-and ethanolamine- phosphotransferase activities was sought for subsequent enzymological analysis of hCEPT1 p. Two systems were explored: (i) prokaryotes are devoid of the Kennedy pathways so hCEPT1 was expressed in E. coli; and (ii) hCEPT1 was produced in S. cerevisiae strain HJ091 (cpt1:: LEU2 ept1-) which contains null mutations at the loci coding for its endogenous choline- (CPT1) and choline/ethanolamine- (EPT1) phosphotransferases (Hjelmstad and Bell, 1990; Hjelmstad et al, 1994). Western blot analysis of T7 epitope tagged hCEPT1p revealed one band of the expected molecular weight of 46,550 was produced in both organisms, as shown in Figure 3. All of the hCEPT1 p detected in S. cerevisiae was full length, however, the majority of hCEPT1p produced in E. coli had undergone proteolytic degradation. Heterologous expression of hCEPT1 in S. cerevisiae HJ091 was utilized for subsequent analyses.

Obvious differences in the efficacy of hCEPT1 p to catalyze phosphobase transfer from either CDP-choline or CDP-ethanolamine to various diradylglycerols were apparent (Table 1). Of the diacylglycerols tested, CDP- choline preferred di10: 0 » di16: 1>di8: 0>di18: 1 while CDP ethanolamine preferred 16: 0/18: 1=di18: 1=di16: 1. High activities were also obtained using alkylacyl 16: 0/2: 0 or and its diacyl analogue 18: 1/2: 0, in concert with CDP-choline, resulting in the synthesis of PAF and acyl-PAF.

Kinetic analysis of hCEPT1p (using di18: 1 as the diradylglycerol substrate) revealed a Km (app) value of 55 pM for CDP-choline with a Vmax (app) of 40.2 nmol min-1 mg-1, and a Km (app) of 109 uM and Vmax (app) of 12.8 nmol min-1 mg-1 for CDP-ethanolamine. The addition of DTT to the assay mix (0-5 mM) did not affect cholinephosphotransferase activity regardless of the diradylglycerol substrate supplied (data not shown).

EXAMPLE X-TRANSCRIPT LEVELS A multiple human tissue Northern blot was hybridized with (i) a random primed probe synthesized versus the entire 1.2 kb hCEPT1 coding region, and (ii) a 2.0 kb region of human R-actin. One hCEPT1 transcript of 2.3 kb was observed in all tissue examined and there was no obvious enrichment in any one cell type when normalized to R-actin, as shown in Figure 4 and as discussed below. Similarly, hCEPT2 transcripts were detected in all cell types tested and were not particularly enriched in any tissue.

EXAMPLE XI-METABOLIC LABELLING S. cerevisiae HJ091 cells (cpt1:: LEU2 ept1-) transformed with p416GPD (mock) or pAH9 (hCEPT1) and its various mutant derivatives, were grown to mid-log phase in synthetic dextrose media containing appropriate nutritional supplements to ensure plasmid maintenance (Kaiser, C., Michaelis, S., and Mitchell, A., 1994 in Methods in Yeast Genetics, Cold Spring Harbor Press, New York). [14 C] Choline (10 pM, 1 x 105 dpm/nmol) was added to the cultures for 1 hr. For [14C] ethanolamine experiments, the cells were washed twice in synthetic dextrose media plus required nutritional supplements but without ammonium sulfate, and resuspended in minus ammonium sulfate media before the addition of [14C] ethanolamine (6.7 uM, 2.2 x 105 dpm/nmol) for 1 hr. The reduced nitrogen containing media was required for the efficient uptake of ethanolamine.

Subsequent to incubation with radiolabel, cells were concentrated by centrifugation, washed twice with water, and resuspended in 1 ml CHCI3/CH30H (1/1, v/v). Cells were disrupted for 1 min at 4°C using a BioSpec Multi Bead Beater containing 0.5 g of 0.5 mm acid washed glass beads. The beads were washed with 1.5 mi CHCI3/CH30H (2/1, v/v). To facilitate phase separation 1.5 ml water and 0.5 ml CHCI3 were added. Phospholipids in the organic phase were analyzed by thin layer chromatography on Whatman silica gel 60A plates using the solvent system CHCI3/CH30H//H20/CH3COOH (70/30/4/2, v/v). Aqueous metabolites were concentrated under vacuum, resuspended in H20 and separated by thin layer chromatography on Whatman silica gel 60A plates. Choline containing metabolites were separated in a solvent system consisting of CH30H/0.6% NaCI/NH40H (50/50/5, v/v). Ethanolamine containing metabolites were separated using CH3CH20H/2% NH40H (1/2, v/v). Radiolabel was detected using a BIOSCAN System 200 imaging scanner to identify and integrate the radioactive bands. This demonstrated the production of the lipid products PtdCho, and PtdEtn in cells expressing active enzyme, and the build up of substrate, CDP- choline or CDP-ethanolamine (and concomitant lack of PtdCho or PtdEtn production) in cells in which inactive enzyme was expressed.

EXAMPLE XII-STRUCTURAL PREDICTIONS The CDP-aminoalcohol phosphotransferases Cpt1 p and Ept1 p from yeast (Hjelmstad and Bell, 1990; Hjelmstad and Bell, 1987), and AAPT1 p from soybean (Dewey et al, 1994, Plant Cell 6: 1495-1507), were aligned with hCEPT1 p, as shown in Figure 5. Specifically, The CLUSTAL-W alignment algorithm set at default parameters was utilized. Underlined residues indicate the positioning of the catalytic CDP-alcohol phosphotransferase motif. Overall identity and similarity levels when compared to hCEPT1p were: 23.1% and 38.2% for yeast Cpt1p; 21.9% and 34.6% for yeast Ept1p, and; 26.9% and 41.8% for plant AAPTlp. Inspection revealed that the CDP-alcohol phosphotransferase motif, DG (x) 2AR (x) 8G (x) 3D (x) 3D, spans amino acid residues 136-158 of hCEPT1p and is located in a similar position within the primary sequence of each enzyme, as is evident in Figure 5.

Membrane spanning domains for each enzyme were estimated using the TmpredT" algorithm (Hofmann and Stoffel, 1993, Biol Chem Hoppe-Seyler 374: 166-172) and positioned within each sequence. Specifically, two separate secondary structure algorithms, nnpredict and PHD (Rost and Sander, 1993, Proc Natl Acad Sci USA 90: 7558-7562; Kneller et al, 1990, J Mol Biol 214: 171-182), were used to estimate the location of a-helix and P-sheet secondary structures in hCEPT1p. Furthermore, the positioning of amino acid residues of hCEPT1p in a predicted amphipathic helix within the catalytic domain. A strong membrane spanning helix prediction within three out of four of the aligned CDP-aminoalcohol phosphotransferases was the criteria used to position each bilayer spanning region, as shown in Figure 6.

EXAMPLE XIII-DISCUSSION The cloned hCEPT1 and hCEPT2 cDNAs code for a choline/ethanolaminephosphotransferase capable of utilizing both CDP-choline and CDP-ethanolamine as phosphobase donors to a broad range of diradylglycerols resulting in the synthesis of PtdCho, PtdEtn, and PAF, as shown in Table 2. Kinetic analysis revealed that CDP-choline was the preferred phosphobase donor with a Km (app) of 55 lux, compared to 109 uM for CDP- ethanolamine. The highest specific activities were obtained using CDP-choline for phosphoryl transfer to 1-0-16: 0/2: 0 alkylacylglycerol, its structural analogue 18: 1/2: 0, or di10: 0 diacylglycerol, resulting in the synthesis of PAF, acyl-PAF, and di10: 0 PtdCho. Previous reports observed several cholinephosphotransferase activities present in eukaryotic cell membranes and DTT sensitivity was used to differentiate an activity capable of synthesizing PAF (DTT-insensitive) from PtdCho (DTT-sensitive) (Snyder, 1997); hCEPT1p cholinephosphotransferase activity was insensitive to DTT regardless of the diradylglycerol substrate supplied.

The hCEPT1 transcript was detected in all cell types tested and was not enriched in any particular tissue. The ubiquitous nature of the hCEPT1 mRNA implies that the enzyme has the capacity to synthesize PtdCho and PtdEtn in vivo, however, its ability to synthesize PAF is difficult to reconcile with its tissue distribution. This is because PAF is normally not present in appreciable amounts but is synthesized in response to extracellular agonists via a specific deacylation/acetylation cycle active in cells of the immune system (Snyder, 1997; Leslie, 1997; Venable et al, 1993). The role of a de novo route for PAF synthesis, and indeed whether the supply of specific diradylglycerols and CDP aminoalcohols affects the ability of hCEPT1 p to partition the PtdCho, PtdEtn, and PAF biosynthetic pathways, deserves further characterization. Diacylglycerols have previously been demonstrated to limit the synthesis of both PtdCho (Jamil et al, 1992, J Biol Chem 267: 1752-1760) and PtdEtn (Tijburg et al, 1989, Biochem J 257: 645-659), consistent with the hypothesis that diradylglycerol levels may affect hCEPT1 p activity and specificity in vivo.

The availability of predicted amino acid sequence data for several CDP-aminoalcohol phosphotransferases facilitated a reexamination of the theorized membrane spanning helices of this class of enzymes from those originally predicted upon the initial cloning of the first cholinephosphotransferase gene, CPT1, from S. cerevisiae (Hjelmstad and Bell, 1990; Hoffman and Stoffel, 1993). A rearrangement of the postulated membrane spanning helices is proposed, specifically: amino acid residues corresponding to positions 180-201 within hCEPT1 p were original predicted to be exist within the solvent but are now envisioned to span the bilayer, and; residues corresponding to 348-368 of hCEPT1p are no longer predicted to reside within the membrane but are now present in the solvent. Of note is an amphipathic helix spanning residues 151-168 of hCEPT1p that is also present within the corresponding region of each CDP- alcohol phosphotransferase (McMaster and Bell, 1997; Williams and McMaster, 1998); the final two aspartate residues of the CDP-alcohol phosphotransferase catalytic motif for each enzyme lie within this amphipathic helix, as shown in Figure 6. These two aspartate residues are responsible for the nucleophilic attack of the hydroxyl of the hydrophobic diradylglycerol on the phosphoester bond of the water soluble phosphobase substrate (Williams and McMaster, 1998). It is hypothesized that the amphilicity of this helix is required to allow for interface of the two substrates.

A search for regulatory domains within hCEPT1p did not reveal any obvious motifs, however, several interesting regions of similarity were noted.

Amino acid residues 18-34 of hCEPT1 p align with residues 193-209 of arginosuccinate lyase. A previous study observed cholinephosphotransferase became rate-limiting for PtdCho synthesis in livers of fasted hamsters (Jamil et al, 1992). The inhibitor was purified and identified as arginosuccinate. In addition, residues 252-277 of hCEPT1p align with 89-114 of the PAF receptor (Snyder, 1997; Leslie, 1997; Venable et al, 1993). The S. cerevisiae Cpt1p and Ept1p enzymes require activation by their products, PtdCho and PtdCho/PtdEtn, respectively, in amounts indicative of precise phospholipid binding site (s) within the enzyme.

Expression vectors containing host-specific control elements operably linked to the cDNA sequence of either hCEPT1, hCEPT2 or the mutant forms thereof described above can be constructed. As discussed above, these enzymes can be used for transferring phosphocholine or phosphoethanolamine to an acceptor lipid. These may include, but are in no way limited to, diacylglycerols, alkylacylglycerols and alkylalkylglycerols. Thus, the above-described expression vectors may be used to synthesize lipids in vivo, thereby altering the lipid composition of the host organism. This in turn could be used to produce host organisms having a nutritionally more desirable lipid content or having a preferred growth phenotype, as alterations in fatty acid content can also alter membrane fluidity and growth temperatures of organisms.

In other embodiments, purified hCEPT1 and hCEPT2 can be used in vitro to synthesize lipids having a precise fatty acid content for use as a food additive or for other industrial or commercial purposes.

Furthermore, hCEPT1 and hCEPT2 as described above can be used in assay systems for assessing the effect of different compounds on lipid content <BR> which can be carried out either in vivo or in vitro. Specifically, hCEPT1 and/or hCEPT2 are provided access to suitable substrates in the presence of at least one compound proposed to have an effect on lipid metabolism. The lipids synthesized by the enzymes can then be analyzed which will in turn indicate what effect the compound is having, that is, whether enzyme activity is being enhanced or repressed or if some substrates or products are preferred. This would allow for the rapid identification of compounds capable of modulating lipid metabolism. Given that the products of hCEPT1 and hCEPT2 include platelet activating factor, which is a mediator of inflammation and phosphatidylcholine, which is a component of lung surfactant, lipoproteins and bile, it is clear that compounds effective in treating a number of disorders, such as chronic and acute pain and inflammation, asthma, allergies, and induction of labour, to name a few, could be identified in this way.

Similarly, these compounds could be used to affect other cellular processes, for example, lipoprotein secretion, lung surfactant generation and bile secretion.

Furthermore, hCEPT1 and hCEPT2 also produce lipid second messengers, which in turn may be metabolized into leukeotriens and prostaglandins which are in turn mediators of inflammation, it is clear that these compounds could also modulate inflammation. In addition, these compounds could also be used as a means to alter nutritional content and growth phenotypes of other organisms, as discussed above.

Similarly, hCEPT1 and hCEPT2 could be used as gene therapy targets or antisense targets or as enzyme therapy agents for treating any of the disorders listed above. Specifically, antisense RNA directed against either hCEPT1 or hCEPT2 or fragments thereof may be prepared and encapsulated, thereby producing a therapeutic compound for regulating lipid metabolism. Alternatively, purified hCEPT1 and hCEPT2 may be encapsulated and used as a therapeutic compound. In yet other embodiments, DNA sequences encoding hCEPT1 or hCEPT2 may be operably linked to host-specific control sequences for use as a reagent in gene replacement therapy. In these embodiments, the vector would include sequences arranged to direct retention of the vector in the host cell either by positive or negative selection or by directing integration of the DNA into the host genome by means known in the art. In these latter embodiments, the vector may include for example baculovirus sequences, Ti DNA or Autographa Californica nuclear polyhedral virus sequences, or other viral vectors.

Furthermore, reagents such as probes derived from the hCEPT1 and hCEPT2 gene sequences and antibodies directed against the hCEPT1 and hCEPT2 proteins can also be generated. These reagents can in turn be used as diagnostic reagents for diagnosing diseases or genotyping. Alternatively, these reagents can be used to identify related enzymes. That is, nucleotide probes may be used to identify other isoforms of hCEPT1 and/or hCEPT2 for example by library screening or PCR amplification and antibodies directed against the peptides or fragments thereof may be used to identify related enzymes in other organisms for example by library screening.

As discussed above, mutation of glycine 156 of hCEPT1 abolished ethanolaminephosphotransferase activity while cholinephosphotransferase activity remained intact. This residue is also conserved in hCEPT2, indicating that a similar mutant in hCEPT2 would have a similar phenotype. Thus, the residue that alters the enzyme specificity from using both CDP-choline and CDP-ethanolamine to just CDP-choline. This means that the substrate binding/specificity site has been identified. Furthermore, this means that instead of using activity as a measure of enzyme function, this region of either hCEPT1 or hCEPT2 could be expressed and a search for binding/no-binding of substrates, that is, by binding competition experiments to search for inhibitors/drugs could be carried out without actually assaying activity. Specifically, our mutagenesis study has identified the region required for substrate binding. Hence, routine prodedures would allow this region of amino acids to be synthesized, or the DNA coding for this region to be expressed as a glutathione-S-transferase fusion protein, His-tagged protein, or other available one step affinity purification systems, for purification of this region.

Once purified, radiolabelled substrate (CDP-choline, CDP-ethanolamine, or other CDP-aminoacohol deriviatives) would be predicted to bind this region. ELISA or RIA competition binding experiments (Harlow, E. and Lane, D., 1988 in Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York) could be used versus a panel of unknown compounds that would prevent this binding (inhibitors) and thus identify compounds that would alter the activity of hCEPT1 or hCEPT2 enzymes.

Since various modifications can be made in our invention as herein above described, and many apparent widely different embodiments of same made within the spirit and scope of the claims without department from such spirit and scope, it is intended that all matter contained in the accompanying specification shall be interpreted as illustrative only and not in a limiting sense.

Enzyme activity mg-1)min-1 Diradylglycerol CDP-choline CDP-ethanolamine di8: 0 54.9 # 6.4 3.1 # 0. 4 <BR> <BR> <BR> <BR> <BR> #9.83.7#0.4di10:0172.5 <BR> <BR> <BR> <BR> <BR> <BR> <BR> dil2:0 27. 2 + 6. 9 2. 7 + 0.3 dil4: 0 6.7 0. 2 3. 7+0.1 di16 : 0 4.3 0. 2 3. 5+0.2 dil6: l 71. 28. 3 13. 21.1 dil8: 1 40.6 # 2.7 12.2 # 1. 2 16: 0/18: 1 37.3 # 2.0 14.0 # 0.8 18: 1/2: 0 261.5 # 18.7 n.d.1 1-0-16: 0/2: 0 106.0 # 5.6 n. d.

TABLE 1 Microsome dpmnmolnmol/minmeandpm Inmol Img hCEPT2 choline 3172 3668 0. 865 10.638 10.673 choline 3193 3668 0. 871 10.708 ethanolamine 551 3126 0.176 2. 169 2.427 ethanolamine 682 3126 0. 218 2.685 293-S choline 465 3668 0. 127 1. 711 1.674 36680.1211.637choline445 31260.0470.6350.615ethanolamine147 ethanolamine 0.0440.5953126 1030636682.81013.59313.333hCEPT1choline choline 9910 3668 2. 702 13.072 31260.8043.8923.910ethanolamine2514 ethanolamine 2538 3126 0. 812 3.928 TABLE 2