PS (phosphatidylserine) in mammalian cells is synthesized by two distinct base-exchange enzymes, PSS1 (PS synthase 1) and PSS2, which are responsible for the conversion of PC (phosphatidylcholine) and PE (phosphatidylethanolamine) respectively into PS in intact cells. The PS synthesis in cultured mammalian cells is inhibited by exogenous PS, and this feedback control occurs through inhibition of PSSs by PS. In the present study, we purified epitope-tagged forms of human PSS1 and PSS2. The purified PSS2 was shown to catalyse the conversion of PE, but not PC, into PS, this being consistent with the substrate specificity observed in intact cells. On the other hand, the purified PSS1 was shown to catalyse the conversion of both PC and PE into PS, although PSS1 in intact cells had been shown not to contribute to the conversion of PE into PS to a significant extent. Furthermore, we found that the purified PSS2, but not the purified PSS1, was inhibited on the addition of PS to the enzyme assay mixture, raising the possibility that there was some difference between the mechanisms of the inhibitory actions of PS towards PSS1 and PSS2.

INTRODUCTION

PS (phosphatidylserine) is an essential phospholipid for the growth of mammalian cells [1,2], forming 3–10% of the total phospholipids in various mammalian tissues and cultured cells. PS synthesis in mammalian cells occurs through the exchange of L-serine with the choline moiety of PC (phosphatidylcholine) or the ethanolamine moiety of PE (phosphatidylethanolamine) [35], whereas in bacteria and yeast, PS is synthesized from CDP-diacylglycerol and L-serine [6,7]. The serine base exchange in mammalian cells is catalysed by PSS1 (PS synthase 1) and PSS2 [35], which, respectively, are responsible for the conversion of PC and PE into PS in intact cells [8].

The PS biosynthesis in CHO (Chinese-hamster ovary)-K1 cells is remarkably inhibited on the addition of PS to the culture medium [9], indicating that feedback control is involved in the regulation of PS biosynthesis. A CHO-K1 cell mutant, named 29, which is defective in the feedback control of PS biosynthesis, has been isolated [10]. The PS biosynthetic rate and PS content in the mutant 29 cells are ∼2-fold those in the parental CHO-K1 cells [10,11]. Unlike PS biosynthesis in CHO-K1 cells, that in the mutant 29 cells is not inhibited significantly on the addition of PS to the culture medium [10,11]. Furthermore, the serine base exchange for PS formation in the membrane fraction of CHO-K1 cells is inhibited by exogenous PS, whereas that of mutant 29 cells is resistant to inhibition by exogenous PS [10,11]. These abnormalities of mutant 29 cells with regard to PS synthesis have been shown to be caused by a point mutation in the PSS1 gene, resulting in the replacement of Arg-95 with lysine (R95K) [11]. In addition, like the R95K PSS1 mutant, the PSS2 mutant, in which Arg-97 is replaced with lysine, has been shown to be resistant to inhibition by PS in intact cells and an isolated membrane fraction [12]. These observations suggest that the feedback control of PS biosynthesis occurs not through regulation of expression of the PSS1 and PSS2 genes, but through inhibition of PSSs by PS.

Chinese-hamster PSS2 tagged with FLAG and HA (haemagglutinin) peptides has been purified to near homogeneity, and the purified enzyme, like the crude enzyme, was inhibited by PS [13], implying that the direct interaction of PS with PSS2 is critical for the inhibition of PSS2 by PS. However, PSS1 has not been purified so far, and therefore it remains unknown whether the inhibition of PSS1 by PS occurs through the direct interaction of PS with PSS1. Furthermore, the substrate specificity of PSS1 remains to be established by kinetic analysis of purified PSS1. In the present study, to address these issues, we purified an epitope-tagged form of human PSS1. In addition, we also purified an epitope-tagged form of human PSS2 to compare the enzymatic characteristics of human PSS1 and PSS2.

EXPERIMENTAL

Materials

Anti-FLAG M2 gel, FLAG peptide, 3× FLAG peptide, bovine brain PS, egg yolk PC and egg yolk PE were purchased from Sigma; asolectin was from Wako Pure Chemical Industries; sucrose monolaurate was from Dojindo Laboratories; and L-[U-14C]serine, [methyl-14C]choline and [2-14C]ethanolamine were from GE Healthcare.

Cloning of PSS1 and PSS2 cDNAs

Oligonucleotides corresponding to parts of the sequences of putative human PSS1 (GenBank® accession number D14694) and PSS2 (GenBank® accession number NM_030783) were used to amplify PSS1 and PSS2 cDNA fragments from a HeLa cell cDNA library (Invitrogen) by means of PCR. To amplify the PSS1 cDNA fragment, a sense primer containing a SalI site, ATAGTCGACAGGCGGGCCATGGCGTC, and an antisense primer containing a NotI site, ATAGCGGCCGCTCATTTCTTTCCAACGCCATTGGTG, were used for PCR. To amplify the PSS2 cDNA fragment, a sense primer containing a SalI site, ATAGTCGACCGAAACGCCATGCGGAGG, and an antisense primer containing a NotI site, ATAGCGGCCGCTCAGTTTGGAGTTGGTGCTCCC, were used for PCR. The PCR products were cloned into the pCR 2.1-TOPO TA cloning vector (Invitrogen). The resultant plasmids were designated pTOPO/hPSS1 and pTOPO/hPSS2 respectively.

Cell culture

HeLa cells were maintained in ES medium [13a], supplemented with 6% (w/v) fetal bovine serum, penicillin G (100 units/ml), streptomycin sulfate (100 mg/ml) and NaHCO3 (1.176 g/l) under a 5% CO2 atmosphere with saturated humidity at 37 °C.

Transient transfection of HeLa cells with the PSS1 and PSS2 cDNAs

The plasmids pTOPO/hPSS1 and pTOPO/hPSS2 were cleaved with SalI and NotI, and then the resultant hPSS1 and hPSS2 cDNA fragments were ligated with a plasmid, pCMV-Script2, cleaved with the same restriction enzymes. pCMV-Script2 was constructed by replacing the multicloning site of pCMV-Script (Stratagene) with a new multicloning site comprising SacI, SalI, EcoRI, NotI, SrfI, SmaI, XhoI, ApaI and KpnI sites in that order. The resultant plasmids, pCMV/hPSS1 and pCMV/hPSS2, were each introduced into HeLa cells using Lipofectamine™ reagent (Invitrogen) according to the manufacturer's instructions.

Construction of HeLa cell strains producing epitope-tagged forms of PSS1 and PSS2

A cDNA clone encoding a PSS1 protein with a FLAG peptide tag between the first methionine and second alanine and an HA peptide tag at the C-terminus was constructed by means of PCR. For construction of the PSS1 fusion cDNA clone, a sense primer containing a SalI site, 5′-ATTGTCGACGCCGCCATGGATTACAAGGATGACGACGATAAGGCGTCCTGCGTGGGGAG-3′, an antisense primer containing a NotI site, 5′-ATTGCGGCCGCTCAGGCGTAGTCCGGGACGTCATATGGGTATTTCTTTCCAACGCCATTGGTGACTTTTG-3′, and a template plasmid, pTOPO/hPSS1, were used for PCR. A cDNA clone encoding a PSS2 protein with a FLAG peptide tag between the first methionine and second arginine and a HA peptide tag at the C-terminus was also constructed by means of PCR. For construction of the PSS2 fusion cDNA clone, a sense primer containing a SalI site, 5′-ATTGTCGACGCCGCCATGGATTACAAGGATGACGACATAAGCGGAGGGGCGAGCACAG-3′, an antisense primer containing a NotI site, 5′-ATTGCGGCCGCTCAGGCGTAGTCCGGGACGTCATATGGGTAGTTTGGAGTTGGTGCTCCCTCGC-3′, and a template plasmid, pTOPO/hPSS2, were used for PCR. The PCR products were cleaved with SalI and NotI, and then ligated with pCMV-script2 cleaved with the same restriction enzymes. The resultant plasmids were designated pCMV/FH-hPSS1 and pCMV/FH-hPSS2 respectively.

HeLa cells were transfected with pCMV/FH-hPSS1 or pCMV/FH-hPSS2 using Lipofectamine™ reagent according to the manufacturer's instructions. After selection with G418 (1.2 mg/ml; Invitrogen), several colonies resistant to the drug were purified, propagated and assayed for PSS activity. Among pCMV/FH-hPSS1-transfected HeLa cell clones showing an increase in PSS activity, one clone designated HeLa/FH-hPSS1 was used for further experiments. Among pCMV/FH-hPSS2-transfected HeLa cell clones showing an increase in PSS activity, one clone designated HeLa/FH-hPSS2 was used for further experiments.

Preparation of membranes

HeLa cells (HeLa/FH-hPSS1 or -hPSS2) were cultivated in a culture bottle containing 2 litres of ES medium supplemented with 10% (w/v) fetal calf serum, penicillin G (100 units/ml), streptomycin sulfate (100 mg/ml) and NaHCO3 (1.176 g/l) under a 5% CO2 atmosphere with saturated humidity at 37 °C. Thereafter, all manipulations were carried out at 4 °C or on ice. The cells were harvested by centrifugation (700 g for 5 min), washed three times with 50 ml of PBS and then suspended in ∼12 ml of buffer A [0.25 M sucrose, 10 mM Hepes/NaOH, pH 7.5, 1 mM EDTA, 1 mg/ml pepstatin and one tablet per 10 ml of Complete™ Mini protease inhibitor cocktail (EDTA-free) (Roche Diagnostics)]. The cell suspension was homogenized with a Potter–Elvehjem-type Teflon/glass homogenizer. The homogenate was centrifuged at 700 g for 5 min, followed by centrifugation of the supernatant at 100000 g for 1 h. The precipitate was suspended in buffer A and then further centrifuged at 100000 g for 1 h. The resultant precipitate, as the intact membranes, was suspended in ∼3 ml of buffer B (5 mM Hepes/NaOH, pH 7.5, 0.1 mM L-serine, 20% (w/v) glycerol, 1 mg/ml pepstatin and one tablet per 10 ml of Complete™ Mini (EDTA-free)], and then stored at –70 °C until use.

Purification of epitope-tagged hPSS1 and hPSS2

A stock suspension of asolectin (50 mg/ml in water) was prepared by sonication at room temperature (20–25 °C). Thereafter, all manipulations were carried out at 4 °C or on ice. The intact membranes were incubated for 15 min at ∼5 mg of protein/ml in a buffer consisting of 3.5 mM Hepes/NaOH (pH 7.5), 0.07 mM L-serine, 14% glycerol, 10 mg/ml asolectin, 150 mM NaCl, 1 mM DTT (dithiothreitol), 0.1 mM EDTA, 1 mg/ml pepstatin, one tablet per 10 ml of Complete™ Mini (EDTA-free) and 1.0% sucrose monolaurate and then centrifuged at 100000 g for 30 min. The supernatant was recovered as the solubilized membrane fraction. The solubilized membrane fraction (∼11 ml) was incubated with 0.5 ml of a 50% slurry of anti-FLAG M2 affinity beads for 3 h with gentle shaking. The affinity beads had been equilibrated with buffer C consisting of 10 mM Hepes/NaOH (pH 7.5), 0.1 mM L-serine, 15% glycerol, 2 mg/ml asolectin, 150 mM NaCl, 0.1 mM EDTA and 0.2% sucrose monolaurate. The affinity beads incubated with the solubilized membrane fraction were washed six times with buffer C supplemented with 1 mM DTT. The washed beads were incubated for 1 h with 0.5 ml of buffer C supplemented with 1 mM DTT and 0.2 mg/ml of 3×FLAG peptide to elute the proteins bound to the beads. After a brief centrifugation, the supernatant was collected and then the beads was resuspended in 0.5 ml of buffer C supplemented with 1 mM DTT and 0.2 mg/ml 1×FLAG peptide. Then, the beads were precipitated by a brief centrifugation, and the supernatant was mixed with the fraction eluted with 3×FLAG peptide, and stored at −70 °C as the purified enzyme fraction.

Assaying of base-exchange activities

Method 1

The enzyme source was incubated in 0.1 ml of a standard assay buffer containing 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2, 5 mg/ml asolectin and either 0.2 mM L-[U-14C]serine (10 μCi/μmol), 0.2 mM [2-14C]ethanolamine (10 μCi/μmol) or 0.2 mM [methyl-14C]choline (20 μCi/μmol), at 37 °C for 20 or 30 min. When the dependence on phospholipid substrates, serine, Ca2+ and pH was examined, PC or PE suspension instead of asolectin was added to the assay buffer. The PC and PE suspension stocks (20 mg/ml in a buffer containing 0.25 M sucrose, 10 mM Hepes/NaOH, pH 7.5, and 1 mM EDTA) were prepared by sonication. To measure the base-exchange activities in cell homogenates of transient transfectants with PSS cDNA clones, cells were harvested at 48 h after transfection, resuspended in a sonication buffer (0.25 M sucrose, 10 mM Hepes/NaOH, pH 7.5, and 1 mM EDTA), and then sonicated on ice. The reaction was started by adding the enzyme source. After stopping the reaction by adding 50 μl of 100 mM EDTA containing 20 mM L-serine, 20 mM ethanolamine or 20 mM choline, lipids were extracted as described by Saito et al. [14], and radioactivity incorporated into lipids (the chloroform phase) was measured. When examined by TLC, the chloroform-soluble materials generated by the serine exchange catalysed by purified enzymes consisted of more than 95% PS.

Method 2

To examine inhibition of the activity by exogenous PS, PS liposomes or PS- and PE-mixed liposomes were prepared by sonication in a buffer containing 0.25 M sucrose, 10 mM Hepes/NaOH (pH 7.5), 1 mM EDTA and the enzyme source was pre-incubated with the liposomes in 89.5 μl of 56 mM Hepes (pH 7.5) in the presence or absence of the exogenous phospholipid substrate (2.2 mg/ml PC or 1.1 mg/ml PE) for ∼30 min on ice. After the pre-incubation, the reaction was initiated by adding 10.5 μl of 48 mM CaCl2 containing 1.9 mM L-[U-14C]serine (10 μCi/μmol). The mixture was incubated at 37 °C for 30 min. After stopping the reaction by adding 50 μl of 100 mM EDTA containing 20 mM L-serine, lipids were extracted as described by Saito et al. [14], and radioactivity incorporated into lipids (the chloroform phase) was measured.

SDS/PAGE, silver staining and Western blotting

SDS/PAGE was carried out by the method of Laemmli [15] with modifications. Samples for SDS/PAGE were incubated in an SDS sample buffer [0.05 M Tris/HCl, pH 6.7, containing 2% (w/v) SDS, 5% (v/v) 2-mercaptoethanol, 10% glycerol and 13 mg/ml Bromophenol Blue] at 37 °C for 1 h, and then proteins were separated by electrophoresis on 10% (w/v) polyacrylamide gel. For silver staining, proteins separated on gels were stained with a silver staining kit (Wako Pure Chemical Industries). For Western-blot analysis, proteins separated by SDS/PAGE were transferred to a PVDF membrane. The membrane was blocked with PBS containing 0.05% Tween 20 and 5% (w/v) non-fat dried skimmed milk powder and then incubated with the anti-HA epitope tag antibody (Sigma) in PBS containing 0.05% Tween 20 and 0.5% non-fat dried skimmed milk powder. The proteins that cross-reacted with the antibody were detected using horseradish peroxidase-conjugated anti-mouse IgG (Biosource) and an enhanced chemiluminescence reagent (PerkinElmer Life Sciences), and immunoreactive bands were detected using a cooled CCD camera (charge-coupled-device camera)-linked Cool Saver system (Atto Instruments).

Protein determination

The protein content of cell homogenates, membrane fractions and solubilized membrane fractions was determined by a Pierce BCA (bicinchoninic acid) protein assay kit using BSA as the standard. The protein content in the purified enzyme fractions was estimated by densitometric comparison of silver-stained proteins with stained calibration bands of BSA at known concentrations on SDS/PAGE.

RESULTS

Cloning of human PSS1 and PSS2 cDNAs

In the human genome, there are two genes that are inferred to encode PSS1 and PSS2 respectively from their nucleotide sequences. We amplified full-length cDNA clones of the putative human PSS1 and PSS2 from a HeLa cell cDNA library by means of PCR, and then examined whether these gene products were able to catalyse serine base exchange for PS formation by transient expression in HeLa cells. As shown in Figure 1, homogenates of the transient transfectants with the putative PSS1 and PSS2 cDNA clones respectively exhibited 2.9- and 3.5-fold higher specific serine base-exchange activity than a homogenate of HeLa cells transfected with the control vector, confirming that the putative PSS1 and PSS2 genes actually encoded PSS1 and PSS2. To enable affinity purification and detection by Western blotting, we constructed expression plasmids that, respectively, encoded human PSS1 and PSS2 having FLAG and HA peptide tags, named FH-hPSS1 and FH-hPSS2 respectively. When HeLa cells were transiently transfected with the plasmids, these fusion proteins were successfully produced as shown in Figure 1(C). The transfectants with the tagged PSS1 and PSS2 cDNAs exhibited serine base-exchange activities similar to those of the transfectants with the non-tagged native PSS1 and PSS2 cDNAs (Figure 1A), indicating that the tagging of PSS1 and PSS2 did not significantly affect the enzyme activity.

PSS activities (A, B) and expression of recombinant PSSs (C, D) in HeLa cells transfected with human PSS cDNA clones

Figure 1
PSS activities (A, B) and expression of recombinant PSSs (C, D) in HeLa cells transfected with human PSS cDNA clones

(A) Homogenates of HeLa cells transfected transiently with an empty vector (Vector), and the expression plasmids encoding human PSS1 (hPSS1), human PSS2 (hPSS2), FH-hPSS1 and FH-hPSS2 were incubated in the standard assay buffer containing L-[U-14C]serine without asolectin at 37 °C for 20 min by following ‘Method 1’ described in the Experimental section. Results are means±S.D. (n=3), with each experiment performed in duplicate. (B) Homogenates of HeLa cells, the stable transformant HeLa/FH-hPSS1 cells and the stable transformant HeLa/FH-hPSS cells were subjected to enzyme assay as described for (A). Results are means±S.D. (n=2), with each experiment performed in duplicate. (C) Equal amount of proteins of HeLa cells transfected transiently with an empty vector (Vector), and the expression plasmids encoding FH-hPSS1 and FH-hPSS2 were analysed by Western blotting with anti-FLAG, anti-HA and anti-actin antibodies. Molecular masses are indicated on the left-hand side (in kDa). (D) Equal amounts of proteins of the indicated cells were analysed by Western blotting with anti-FLAG, anti-HA and anti-actin antibodies. Molecular masses are indicated on the left-hand side (in kDa).

Figure 1
PSS activities (A, B) and expression of recombinant PSSs (C, D) in HeLa cells transfected with human PSS cDNA clones

(A) Homogenates of HeLa cells transfected transiently with an empty vector (Vector), and the expression plasmids encoding human PSS1 (hPSS1), human PSS2 (hPSS2), FH-hPSS1 and FH-hPSS2 were incubated in the standard assay buffer containing L-[U-14C]serine without asolectin at 37 °C for 20 min by following ‘Method 1’ described in the Experimental section. Results are means±S.D. (n=3), with each experiment performed in duplicate. (B) Homogenates of HeLa cells, the stable transformant HeLa/FH-hPSS1 cells and the stable transformant HeLa/FH-hPSS cells were subjected to enzyme assay as described for (A). Results are means±S.D. (n=2), with each experiment performed in duplicate. (C) Equal amount of proteins of HeLa cells transfected transiently with an empty vector (Vector), and the expression plasmids encoding FH-hPSS1 and FH-hPSS2 were analysed by Western blotting with anti-FLAG, anti-HA and anti-actin antibodies. Molecular masses are indicated on the left-hand side (in kDa). (D) Equal amounts of proteins of the indicated cells were analysed by Western blotting with anti-FLAG, anti-HA and anti-actin antibodies. Molecular masses are indicated on the left-hand side (in kDa).

Purification of FLAG- and HA peptide-tagged human PSS1 and PSS2

We isolated HeLa cell lines, named HeLa/FH-hPSS1 and HeLa/FH-hPSS2, which stably produced FH-hPSS1 and FH-hPSS2 respectively (Figures 1B and 1D) and used these cell lines as enzyme sources for purification. To solubilize FH-hPSS1 and FH-hPSS2, membranes prepared from HeLa/FH-hPSS1 and HeLa/FH-hPSS2 cells were incubated with a detergent, sucrose monolaurate. After centrifugation at 100000 g for 30 min, the supernatant was recovered as the solubilized membrane fraction. The solubilized membrane fractions prepared from HeLa/FH-hPSS1 and HeLa/FH-hPSS2 cells were subjected to affinity purification of FH-hPSS1 and FH-hPSS2 using anti-FLAG antibody-coupled agarose beads. The affinity purification of FH-hPSS1 resulted in ∼840-fold enrichment of PSS activity with a ∼6.7% recovery, whereas that of FH-hPSS2 resulted in ∼1200-fold enrichment of PSS activity with a ∼24% recovery (Table 1). When the protein pattern of the purified FH-hPSS1 fraction was analysed by SDS/PAGE followed by silver staining, a protein of ∼40 kDa was detected as the major protein (Figure 2A). The major ∼40 kDa protein cross-reacted with anti-HA and anti-FLAG antibodies (Figures 2B and 2E), and the immunostaining of this protein was competed by HA or FLAG peptides (Figure 2E). These results suggested that the major ∼40 kDa protein was the FH-hPSS1 protein. On analysis of the purified FH-hPSS2 fraction by SDS/PAGE followed by silver staining, a protein of ∼50 kDa was detected as the major protein (Figure 2C) which cross-reacted with anti-HA and anti-FLAG antibodies (Figures 2D and 2E). The immunostaining of the ∼50 kDa protein was competed by HA or FLAG peptides (Figure 2E). Thus the major ∼50 kDa protein appeared to be the FH-hPSS2 protein.

Table 1
Purification of FH-hPSS1 and FH-hPSS2

Results shown are from one of two sets of experiments, in which similar results were obtained.

PSS
FractionProtein (μg)Total activity (nmol/h)Specific activity (μmol/h per mg of protein)Enrichment (fold)
FH-hPSS1     
 Detergent-treated membranes 73000 1800 0.025 1.0 
 Solubilized membranes 60000 1100 0.018 0.72 
 Elution of anti-FLAG antibody beads 5.6* 120 21 840 
FH-hPSS2     
 Detergent-treated membranes 71000 3600 0.05 1.0 
 Solubilized membranes 58000 2900 0.05 1.0 
 Elution of anti-FLAG antibody beads 14* 850 61 1200 
PSS
FractionProtein (μg)Total activity (nmol/h)Specific activity (μmol/h per mg of protein)Enrichment (fold)
FH-hPSS1     
 Detergent-treated membranes 73000 1800 0.025 1.0 
 Solubilized membranes 60000 1100 0.018 0.72 
 Elution of anti-FLAG antibody beads 5.6* 120 21 840 
FH-hPSS2     
 Detergent-treated membranes 71000 3600 0.05 1.0 
 Solubilized membranes 58000 2900 0.05 1.0 
 Elution of anti-FLAG antibody beads 14* 850 61 1200 
*

Protein concentrations were estimated by densitometric comparison of silver-stained proteins in the fraction with stained calibration bands of BSA at known concentrations on SDS/PAGE.

SDS/PAGE and Western-blot analysis of proteins obtained at different steps of FH-hPSS1 (A, B, E) and FH-hPSS2 (C, D, E) purification

Figure 2
SDS/PAGE and Western-blot analysis of proteins obtained at different steps of FH-hPSS1 (A, B, E) and FH-hPSS2 (C, D, E) purification

(A, C) Proteins were separated on 10% acrylamide gels and then stained with a silver-staining kit. Lane 1, molecular-mass markers (in kDa); lane 2, membrane fraction; lane 3, solubilized membrane fraction; lane 4, unbound fraction on affinity purification with an anti-FLAG antibody; lane 5, eluted fraction on affinity purification with an anti-FLAG antibody. (B, D) Proteins separated by SDS/PAGE were analysed by Western blotting with an anti-HA antibody. Lane 1, membrane fraction; lane 2, solubilized membrane fraction; lane 3, unbound fraction on affinity purification with the anti-FLAG antibody; lane 4, eluted fraction on affinity purification with the anti-FLAG antibody. (E) The eluted fraction on affinity purification with the anti-FLAG antibody was subjected to Western-blot analysis with anti-FLAG or anti-HA antibodies; for the analysis, the blotting membranes were incubated with the antibodies in the presence (+) or absence (–) of 0.2 mg/ml of FLAG or HA peptides, as indicated. The immunostaining of the putative FH-hPSS1 protein (∼40 kDa) and FH-hPSS2 protein (∼50 kDa) bands is shown.

Figure 2
SDS/PAGE and Western-blot analysis of proteins obtained at different steps of FH-hPSS1 (A, B, E) and FH-hPSS2 (C, D, E) purification

(A, C) Proteins were separated on 10% acrylamide gels and then stained with a silver-staining kit. Lane 1, molecular-mass markers (in kDa); lane 2, membrane fraction; lane 3, solubilized membrane fraction; lane 4, unbound fraction on affinity purification with an anti-FLAG antibody; lane 5, eluted fraction on affinity purification with an anti-FLAG antibody. (B, D) Proteins separated by SDS/PAGE were analysed by Western blotting with an anti-HA antibody. Lane 1, membrane fraction; lane 2, solubilized membrane fraction; lane 3, unbound fraction on affinity purification with the anti-FLAG antibody; lane 4, eluted fraction on affinity purification with the anti-FLAG antibody. (E) The eluted fraction on affinity purification with the anti-FLAG antibody was subjected to Western-blot analysis with anti-FLAG or anti-HA antibodies; for the analysis, the blotting membranes were incubated with the antibodies in the presence (+) or absence (–) of 0.2 mg/ml of FLAG or HA peptides, as indicated. The immunostaining of the putative FH-hPSS1 protein (∼40 kDa) and FH-hPSS2 protein (∼50 kDa) bands is shown.

Enzymatic characterization of purified FH-hPSS1 and FH-hPSS2

The time courses of PS formation by the purified enzymes were both almost linear for 60 min (Figures 3A and 3B), and the PS formation activities depended proportionately on the amount of the purified enzyme, as shown in Figures 3(C) and 3(D).

Time and protein dependence of the purified enzyme activity

Figure 3
Time and protein dependence of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated in the standard assay buffer containing L-[U-14C]serine at 37 °C for various periods by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Various amounts of purified FH-hPSS1 (C) and FH-hPSS2 (D) were incubated in the standard assay buffer containing L-[U-14C]serine at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Results are means±S.D. (n=3), with each experiment performed in duplicate.

Figure 3
Time and protein dependence of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated in the standard assay buffer containing L-[U-14C]serine at 37 °C for various periods by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Various amounts of purified FH-hPSS1 (C) and FH-hPSS2 (D) were incubated in the standard assay buffer containing L-[U-14C]serine at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Results are means±S.D. (n=3), with each experiment performed in duplicate.

We also examined the choline and ethanolamine base-exchange activities of the purified FH-hPSS1, because previous experiments involving CHO-K1 cells overproducing Chinese-hamster PSS1 [16,17] had suggested that in a cell homogenate, PSS1 was able to catalyse choline base exchange for PC formation and ethanolamine base exchange for PE formation, as well as serine base exchange for PS formation. As shown in Figure 4(A), the purified FH-hPSS1 was able to catalyse all three base exchanges effectively, as was expected. On the other hand, the purified FH-hPSS2 was shown to catalyse ethanolamine and serine base exchange, but not choline base exchange (Figure 4B), this being consistent with the substrate specificity of the purified Chinese-hamster PSS2 with epitope tags [13].

Serine, ethanolamine and choline base-exchange activities of purified FH-hPSS1 (A) and FH-hPSS2 (B)

Figure 4
Serine, ethanolamine and choline base-exchange activities of purified FH-hPSS1 (A) and FH-hPSS2 (B)

The activities were measured by following ‘Method 1’ as described in the Experimental section. Results are means±S.D. (n=3), with each experiment performed in duplicate.

Figure 4
Serine, ethanolamine and choline base-exchange activities of purified FH-hPSS1 (A) and FH-hPSS2 (B)

The activities were measured by following ‘Method 1’ as described in the Experimental section. Results are means±S.D. (n=3), with each experiment performed in duplicate.

Under the standard assay conditions for serine base exchange, we used 5 mg/ml of asolectin, a heterogeneous phospholipid mixture comprising PC, PE and other phospholipids, as the phospholipid substrate source. To determine whether PC and/or PE were used as substrates, the PS formation by the purified enzymes was measured in the presence of PC or PE instead of asolectin. As shown in Figure 5(A), the rate of PS formation by the purified FH-hPSS1 increased on the addition of both PC and PE in a similar dose-dependent manner, indicating that FH-hPSS1 was able to catalyse the conversion of both PC and PE into PS. This result was unexpected, because PSS1 in intact cells had been shown to contribute to the conversion of PC, but not PE, into PS to a significant extent [8]. This finding led us to examine the effect of PE on FH-PSS1 in an intact membrane fraction. As shown in Figure 6, the rate of PS formation by a membrane fraction prepared from HeLa/FH-hPSS1 cells increased on the addition of PE in a dose-dependent manner. In addition, when assayed in the presence of 1 mg/ml PE, the membrane fraction of HeLa/FH-hPSS1 cells exhibited ∼2.5-fold higher PS formation activity than the membrane fraction of HeLa cells (Figure 6). These results also supported the idea that FH-hPSS1 was able to catalyse the conversion of PE into PS in a cell-free system. In contrast with the rate of PS formation by the purified FH-hPSS1, that by the purified FH-hPSS2 increased on the addition of PE, but not PC (Figure 5B), indicating that FH-hPSS2 was able to catalyse the conversion of PE, but not PC, into PS. This specificity of FH-hPSS2 coincided with the specificity of the purified Chinese-hamster PSS2 with epitope tags [13] and was consistent with the specificity of PSS2 in intact cells [8].

Effects of the potential phospholipid substrates PC and PE on PS synthesis by purified FH-hPSS1 and FH-hPSS2

Figure 5
Effects of the potential phospholipid substrates PC and PE on PS synthesis by purified FH-hPSS1 and FH-hPSS2

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section in a buffer consisting of PC (○) or PE (●) at the indicated concentrations, 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[U-14C]serine. The assay buffer included a small amount (0.04 mg/ml) of asolectin, which originated in the enzyme sources. After the incubation, the radioactivity incorporated into lipids was measured. Results are means±S.D. (n=3), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

Figure 5
Effects of the potential phospholipid substrates PC and PE on PS synthesis by purified FH-hPSS1 and FH-hPSS2

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section in a buffer consisting of PC (○) or PE (●) at the indicated concentrations, 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[U-14C]serine. The assay buffer included a small amount (0.04 mg/ml) of asolectin, which originated in the enzyme sources. After the incubation, the radioactivity incorporated into lipids was measured. Results are means±S.D. (n=3), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

Effect of PE on PS synthesis by intact membrane fractions prepared from HeLa and HeLa/FH-hPSS1 cells

Figure 6
Effect of PE on PS synthesis by intact membrane fractions prepared from HeLa and HeLa/FH-hPSS1 cells

Membrane fractions prepared from HeLa (○) and HeLa/FH-hPSS1 (●) cells were incubated at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section in a buffer consisting of 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[14C]serine, and PE at the indicated concentrations. After the incubation, the radioactivity incorporated into lipids was measured. Results are means±S.D. (n=2), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

Figure 6
Effect of PE on PS synthesis by intact membrane fractions prepared from HeLa and HeLa/FH-hPSS1 cells

Membrane fractions prepared from HeLa (○) and HeLa/FH-hPSS1 (●) cells were incubated at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section in a buffer consisting of 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[14C]serine, and PE at the indicated concentrations. After the incubation, the radioactivity incorporated into lipids was measured. Results are means±S.D. (n=2), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

To examine the substrate (L-serine) dependence of the purified enzymes, we measured PS formation in the presence of various concentrations of L-serine, and then PS formation versus L-serine concentration was plotted in a double reciprocal form (Figures 7A and 7B). The double reciprocal plots showed that the Km (app) for L-serine and the Vmax (app) of purified FH-hPSS1 in the presence of 2 mM PC were 67 μM and 0.051 nmol of PS/h per ng of protein respectively, that those in the presence of 1 mM PE were 24 μM and 0.061 nmol of PS/h per ng of protein respectively and that those of purified FH-hPSS2 in the presence of 1 mM PE were 120 μM and 0.57 nmol of PS/h per ng of protein respectively.

L-Serine dependence of the purified enzyme activity

Figure 7
L-Serine dependence of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated in a buffer containing various concentrations of L-[14C]serine, 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Results are means±S.D. (n=2), with each experiment performed in duplicate.

Figure 7
L-Serine dependence of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated in a buffer containing various concentrations of L-[14C]serine, 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min by following ‘Method 1’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. Results are means±S.D. (n=2), with each experiment performed in duplicate.

When Ca2+ dependence of PS formation by the purified enzymes was examined, all of the serine base-exchange activities of FH-hPSS1 in the presence of PC and PE and FH-hPSS2 in the presence of PE were dependent on Ca2+ (Figures 8A and 8B). The optimal pH for the purified FH-hPSS1 was between pH 7.0 and 7.5, irrespective of the phospholipid substrates PC and PE (Figure 8C) and that for the purified FH-hPSS2 was in the vicinity of pH 7.5 (Figure 8D).

Ca2+ (A, B) and pH dependence (C, D) of the purified enzyme activity

Figure 8
Ca2+ (A, B) and pH dependence (C, D) of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated by following ‘Method 1’ described in the Experimental section in an assay buffer containing various concentrations of CaCl2, 50 mM Hepes/NaOH (pH 7.5), 0.2 mM L-[U-14C]serine and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min. The radioactivity incorporated into lipids was then measured. Purified FH-hPSS1 (C) and FH-hPSS2 (D) were incubated by following ‘Method 1’ described in the Experimental section in a buffer containing 50 mM Hepes/NaOH (pH 6.0–9.0), 5 mM CaCl2, 0.2 mM L-[U-14C]serine and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min. The radioactivity incorporated into lipids was then measured. Two independent experiments gave similar results; values are expressed for one of two experiments and are means for duplicate determinations, with a variation of <10% between duplicates.

Figure 8
Ca2+ (A, B) and pH dependence (C, D) of the purified enzyme activity

Purified FH-hPSS1 (A) and FH-hPSS2 (B) were incubated by following ‘Method 1’ described in the Experimental section in an assay buffer containing various concentrations of CaCl2, 50 mM Hepes/NaOH (pH 7.5), 0.2 mM L-[U-14C]serine and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min. The radioactivity incorporated into lipids was then measured. Purified FH-hPSS1 (C) and FH-hPSS2 (D) were incubated by following ‘Method 1’ described in the Experimental section in a buffer containing 50 mM Hepes/NaOH (pH 6.0–9.0), 5 mM CaCl2, 0.2 mM L-[U-14C]serine and either 2 mg/ml of PC (○) or 1 mg/ml of PE (●) at 37 °C for 30 min. The radioactivity incorporated into lipids was then measured. Two independent experiments gave similar results; values are expressed for one of two experiments and are means for duplicate determinations, with a variation of <10% between duplicates.

Effects of exogenous PS on PS synthesis by the purified enzymes

When PS formation by a membrane fraction prepared from HeLa/FH-hPSS1 cells was measured without a detergent and with phospholipids in the membrane fraction as the phospholipid substrate, it was remarkably inhibited by exogenous PS in a dose-dependent manner, being reduced by ∼75% (Figure 9A). However, the purified FH-hPSS1 was activated, instead of inhibited, by exogenous PS, irrespective of the assay conditions, whether the phospholipid substrate PC or PE was added to the enzyme assay mixture (Figure 9C). To determine whether or not the detergent sucrose monolaurate, which was needed for the solubilization and purification of FH-hPSS1, affected the inhibition of PS formation by PS, we examined the effect of exogenous PS on PS formation by a solubilized membrane fraction prepared by treatment of HeLa/FH-hPSS1 cell membranes with sucrose monolaurate. As shown in Figure 9(B), when assayed in the presence of the phospholipid substrate PE, PS formation by the solubilized membrane fraction was enhanced by exogenous PS. In the case of the presence of the phospholipid substrate PC, PS formation by the solubilized membrane fraction was inhibited by exogenous PS, but was more resistant to inhibition than in the case of an intact membrane fraction (Figures 9A and 9B). In contrast with purified FH-hPSS1, purified FH-hPSS2 (Figure 9D), like membrane and solubilized membrane fractions prepared from HeLa/FH-hPSS2 cells (results not shown), was remarkably inhibited by exogenous PS, this being consistent with the previous results obtained on assaying of the activity of the purified Chinese-hamster PSS2 with epitope tags [13]. These results suggested that the treatment with sucrose monolaurate under the conditions used in the present study resulted in the loss of normal regulation of PSS1 activity by PS, whereas the regulation of PSS2 activity by PS was not affected by the same treatment, which therefore raised the possibility that there was some difference between the mechanisms of the inhibitory actions of PS towards PSS1 and PSS2.

Effect of exogenous PS on PSS activity

Figure 9
Effect of exogenous PS on PSS activity

(A) A membrane fraction prepared from HeLa/FH-hPSS1 cells was incubated in the standard assay buffer containing L-[U-14C]serine without asolectin at 37 °C for 30 min in the presence of PS at the concentrations indicated by following ‘Method 2’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. (BD) A solubilized membrane fraction prepared from HeLa/FH-hPSS1 cells (B), purified FH-hPSS1 (C) and purified FH-hPSS2 (D) was incubated by following ‘Method 2’ described in the Experimental section in a buffer containing 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[U-14C]serine at 37 °C for 30 min in the presence of PS at the concentrations indicated and either 2 mg/ml of PC (○) or 1 mg/ml of PE (● and ▲). The radioactivity incorporated into lipids was then measured. PS was added to the assay buffer as PS liposomes prepared separately from PC and PE liposomes (○ and ● respectively) or PS and PE mixed liposomes (▲). The results are expressed as percentages of the activity observed without exogenous PS. Results are means±S.D. (n=3), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

Figure 9
Effect of exogenous PS on PSS activity

(A) A membrane fraction prepared from HeLa/FH-hPSS1 cells was incubated in the standard assay buffer containing L-[U-14C]serine without asolectin at 37 °C for 30 min in the presence of PS at the concentrations indicated by following ‘Method 2’ described in the Experimental section. The radioactivity incorporated into lipids was then measured. (BD) A solubilized membrane fraction prepared from HeLa/FH-hPSS1 cells (B), purified FH-hPSS1 (C) and purified FH-hPSS2 (D) was incubated by following ‘Method 2’ described in the Experimental section in a buffer containing 50 mM Hepes/NaOH (pH 7.5), 5 mM CaCl2 and 0.2 mM L-[U-14C]serine at 37 °C for 30 min in the presence of PS at the concentrations indicated and either 2 mg/ml of PC (○) or 1 mg/ml of PE (● and ▲). The radioactivity incorporated into lipids was then measured. PS was added to the assay buffer as PS liposomes prepared separately from PC and PE liposomes (○ and ● respectively) or PS and PE mixed liposomes (▲). The results are expressed as percentages of the activity observed without exogenous PS. Results are means±S.D. (n=3), with each experiment performed in duplicate. In some instances, error bars are too small to be visible.

DISCUSSION

The purification of enzymes is a crucial step for elucidating their catalytic and regulatory mechanisms. In the present study, we purified human PSS1 and PSS2 tagged with FLAG and HA peptides, designated FH-hPSS1 and FH-hPSS2 respectively. The purified FH-hPSS1 was shown to catalyse serine, choline and ethanolamine base exchange for PS, PC and PE formation, whereas the purified FH-hPSS2, like the purified Chinese-hamster PSS2 with epitope tags [13], was shown to catalyse serine and ethanolamine base exchange, but not choline base exchange. These substrate specificities were consistent with those suggested by experiments involving crude enzymes in membrane fractions of CHO cell lines that overproduced Chinese-hamster PSS1 or PSS2 [16,17]. Furthermore, like the purified Chinese-hamster PSS2 with epitope tags [13], the purified FH-hPSS2 was shown to use PE, but not PC, as a phospholipid substrate. This was also consistent with the PSS2 substrate specificity suggested by experiments involving intact cells [8]. Unexpectedly, the rate of PS formation by the purified FH-hPSS1 was found to increase on the addition of both PC and PE in a similar dose-dependent manner, indicating that PSS1 was able to catalyse the conversion of both PC and PE into PS. This was a surprise to us, because the following observations had indicated that PSS1 in intact cells was able to catalyse the conversion of PC, but not PE, into PS to a significant extent: (i) A CHO cell line, CHO-K1, is able to convert both exogenous [32P]PC and [32P]PE into [32P]PS [8,18]. (ii) A CHO-K1 cell mutant, named PSA-3, which is deficient in PSS1, is defective in the conversion of exogenous [32P]PC into [32P]PS, but exhibits normal activity with regard to the conversion of exogenous [32P]PE into [32P]PS [8,18]. (iii) Another CHO-K1 cell mutant, named PSB-2, which is defective in both PSS1 and PSS2, is defective in the conversion of both exogenous [32P]PC and [32P]PE into [32P]PS [8]. (iv) The defect of PSB-2 cells in the conversion of exogenous [32P]PC into [32P]PS is complemented by the expression of PSS1 cDNA, but not PSS2 cDNA [8]. (v) The defect of PSB-2 cells in the conversion of exogenous [32P]PE into [32P]PS is complemented by the expression of PSS2 cDNA, but not by PSS1 cDNA [8].

Why does the purified PSS1, but not the PSS1 in intact cells, effectively catalyse the conversion of PE into PS? PSS1 and PSS2 are integral membrane proteins with several membrane-spanning segments [16,17], and are located primarily in MAMs (mitochondria-associated membranes) [19], which comprise a specialized domain of the ER (endoplasmic reticulum). Although the ER membranes appear to have an intrinsic capacity for transbilayer equilibration of phospholipids [20], the actual distribution of PC and PE in the two leaflets of the lipid bilayer of the ER membranes of CHO cells remains unknown. In addition, whether MAM, as well as the ER membranes, has the capacity for transbilayer equilibration of phospholipids remains to be elucidated. Therefore there is the possibility that PC and PE are asymmetrically distributed in the two leaflets of the lipid bilayer of the membranes, in which PSS1 of CHO cells is located. In that case, whether the catalytic site of PSS1 is located in the luminal or cytoplasmic leaflet of the membranes should influence the access of the phospholipid substrates PC and PE to the active site of this enzyme. Furthermore, there might be putative microdomains of membranes, which are deficient in PE. Thus one possible explanation for the discrepancy between the substrate specificities of the purified PSS1 and the PSS1 in intact cells is that in intact cells, access of PE to the active site of PSS1 was limited because of its localization in the membrane or membrane leaflet deficient in PE. Alternatively, there might be proteinous factor(s) preventing the access of PE to the active site of PSS1 in intact membranes, which could be eliminated during the purification procedures. In the present study, however, we found that PS formation by an intact membrane fraction prepared from HeLa/FH-hPSS1 cells increased strikingly on the addition of PE, implying the absence of such preventing factor(s) in the membrane. To examine these possibilities, determination of the membrane topology and localization of the active sites of PSS1 and PSS2 are imperative, and we are currently carrying out such investigations.

The PS-mediated inhibition of PSS1 and PSS2 has been shown to be critical for the maintenance of a normal cellular PS level in CHO-K1 cells [11,12]. However, the precise mechanisms underlying the PS-mediated inhibition of PSS are unknown. Whether the inhibition occurs through direct interaction of PS with the synthases or is mediated by unidentified factor(s) remains unknown. We previously purified Chinese-hamster PSS2 tagged with FLAG and HA peptides to near homogeneity [13]. In the present study, we purified FH-hPSS2, which resulted in ∼1200-fold enrichment of the activity. The purified FH-hPSS2, as well as the purified Chinese-hamster PSS2 with epitope tags [13], was inhibited by exogenous PS. These results implied that direct interaction of PS with PSS2 was important for the regulation of PSS2 activity. In contrast with the purified FH-hPSS2, the purified FH-hPSS1 was activated, instead of inhibited, by exogenous PS. In addition, it was suggested that the treatment of PSS1 with sucrose monolaurate resulted in the loss of normal regulation of PSS1 activity by PS. Therefore the detergent treatment might affect the intrinsic capacity of PSS1 to be regulated through direct binding with PS. Alternatively, the inhibition of PSS1 by PS in intact cells might be mediated by regulatory protein(s) that are dissociated from PSS1 by the detergent treatment and eliminated during the purification procedures.

Abbreviations

     
  • CHO cell

    Chinese-hamster ovary cell

  •  
  • DTT

    dithiothreitol

  •  
  • ER

    endoplasmic reticulum

  •  
  • HA

    haemagglutinin

  •  
  • MAM

    mitochondria-associated membrane

  •  
  • PC

    phosphatidylcholine

  •  
  • PE

    phosphatidylethanolamine

  •  
  • PS

    phosphatidylserine

  •  
  • PSS

    PS synthase

FUNDING

This work was supported by the Japan Science and Technology Agency; Core Research for Evolutional Science and Technology; and by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan [grant number 19590068].

References

References
1
Kuge
 
O.
Nishijima
 
M.
Akamatsu
 
Y.
 
Phosphatidylserine biosynthesis in cultured Chinese hamster ovary cells. II. Isolation and characterization of phosphatidylserine auxotrophs
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
5790
-
5794
)
2
Voelker
 
D. R.
Frazier
 
J. L.
 
Isolation and characterization of a Chinese hamster ovary cell line requiring ethanolamine or phosphatidylserine for growth and exhibiting defective phosphatidylserine synthase activity
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
1002
-
1008
)
3
Kuge
 
O.
Nishijima
 
M.
 
Phosphatidylserine synthase I and II of mammalian cells
Biochim. Biophys. Acta
1997
, vol. 
1348
 (pg. 
151
-
156
)
4
Kuge
 
O.
Nishijima
 
M.
 
Biosynthetic regulation and intracellular transport of phosphatidylserine in mammalian cells
J. Biochem.
2003
, vol. 
133
 (pg. 
397
-
403
)
5
Vance
 
J. E.
 
Phosphatidylserine and phosphatidylethanolamine in mammalian cells: two metabolically related aminophospholipids
J. Lipid Res.
2008
, vol. 
49
 (pg. 
1377
-
1387
)
6
Matsumoto
 
K.
 
Phosphatidylserine synthase from bacteria
Biochim. Biophys. Acta
1997
, vol. 
1348
 (pg. 
214
-
227
)
7
Yamashita
 
S.
Nikawa
 
J.
 
Phosphatidylserine synthase from yeast
Biochim. Biophys. Acta
1997
, vol. 
1348
 (pg. 
228
-
235
)
8
Saito
 
K.
Nishijima
 
M.
Kuge
 
O.
 
Genetic evidence that phosphatidylserine synthase II catalyzes the conversion of phosphatidylethanolamine to phosphatidylserine in Chinese hamster ovary cells
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
17199
-
17205
)
9
Nishijima
 
M.
Kuge
 
O.
Akamatsu
 
Y.
 
Phosphatidylserine biosynthesis in cultured Chinese hamster ovary cells. I. Inhibition of de novo phosphatidylserine biosynthesis by exogenous phosphatidylserine and its efficient incorporation
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
5784
-
5789
)
10
Hasegawa
 
K.
Kuge
 
O.
Nishijima
 
M.
Akamatsu
 
Y.
 
Isolation and characterization of a Chinese hamster ovary cell mutant with altered regulation of phosphatidylserine biosynthesis
J. Biol. Chem.
1989
, vol. 
264
 (pg. 
19887
-
19892
)
11
Kuge
 
O.
Hasegawa
 
K.
Saito
 
K.
Nishijima
 
M.
 
Control of phosphatidylserine biosynthesis through phosphatidylserine-mediated inhibition of phosphatidylserine synthase I in Chinese hamster ovary cells
Proc. Natl. Acad. Sci. U.S.A.
1998
, vol. 
95
 (pg. 
4199
-
4203
)
12
Kuge
 
O.
Saito
 
K.
Nishijima
 
M.
 
Control of phosphatidylserine synthase II activity in Chinese hamster ovary cells
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
23844
-
23849
)
13
Kuge
 
O.
Hasegawa
 
K.
Ohsawa
 
T.
Saito
 
K.
Nishijima
 
M.
 
Purification and characterization of Chinese hamster phosphatidylserine synthase 2
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
42692
-
42698
)
13a
Koyama
 
H.
Kodama
 
H.
 
Adenine phosphoribosyltransferase deficiency in cultured mouse mammary tumor FM3A cells resistant to 4-carbamoylimidazolium 5-olate
Cancer Res.
1982
, vol. 
42
 (pg. 
4210
-
4214
)
14
Saito
 
M.
Bourque
 
E.
Kanfer
 
J.
 
Studies on base-exchange reactions of phospholipids in rat brain particles and a ‘solubilized’ system
Arch. Biochem. Biophys.
1975
, vol. 
169
 (pg. 
304
-
317
)
15
Laemmli
 
U. K.
 
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
Nature
1970
, vol. 
227
 (pg. 
680
-
685
)
16
Kuge
 
O.
Nishijima
 
M.
Akamatsu
 
Y.
 
A Chinese hamster cDNA encoding a protein essential for phosphatidylserine synthase I activity
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
24184
-
24189
)
17
Kuge
 
O.
Saito
 
K.
Nishijima
 
M.
 
Cloning of a Chinese hamster ovary (CHO) cDNA encoding phosphatidylserine synthase (PSS) II, overexpression of which suppresses the phosphatidylserine biosynthetic defect of a PSS I-lacking mutant of CHO-K1 cells
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
19133
-
19139
)
18
Kuge
 
O.
Nishijima
 
M.
Akamatsu
 
Y.
 
Phosphatidylserine biosynthesis in cultured Chinese hamster ovary cells. III. Genetic evidence for utilization of phosphatidylcholine and phosphatidylethanolamine as precursors
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
5795
-
5798
)
19
Stone
 
S. J.
Vance
 
J. E.
 
Phosphatidylserine synthase-1 and -2 are localized to mitochondria-associated membranes
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
34534
-
34540
)
20
Kol
 
M. A.
de Kroon
 
A. I.
Killian
 
J. A.
de Kruijff
 
B.
 
Transbilayer movement of phospholipids in biogenic membranes
Biochemistry
2004
, vol. 
43
 (pg. 
2673
-
2681
)