Sphingosine kinase 2 is a nuclear protein and inhibits DNA synthesis.

Sphingosine kinase-1 (SPHK1) is a key enzyme catalyzing the formation of an important bioactive lipid messenger, sphingosine 1-phosphate, and is implicated in the regulation of cell proliferation and antiapoptotic processes. Biological features of another isozyme SPHK2, however, remain unclear. The present studies were undertaken to characterize SPHK2 by comparison with SPHK1. When SPHK2 was transiently expressed in various cell lines, it was localized in the nuclei as well as in the cytosol, whereas SPHK1 was distributed in the cytosol but not in the nucleus. We have mapped a functional nuclear localization signal (NLS) to the N-terminal region of SPHK2. We have observed that the expression of SPHK2 in various cell types causes inhibition of DNA synthesis, resulting in the cell cycle arrest at G1/S phase. We have also demonstrated that an NLS mutant of SPHK2, SPHK2R93E/R94E, failed to enter the nucleus and to inhibit DNA synthesis. Moreover, a fusion protein, NLS-SPHK1, where SPHK1 was fused to the NLS sequence of SPHK2 acquired the ability to enter nuclei and inhibited DNA synthesis. These results indicate that SPHK2 localizes in the nuclei and causes inhibition of DNA synthesis, and this may affect subsequent cellular events.

Sphingosine 1-phosphate (SPP) 1 is a bioactive lipid that regulates diverse biological processes such as calcium mobilization, cell growth, differentiation, survival, motility, and cytoskeletal reorganization, acting both inside and outside the cells (1,2). Recently, SPP was identified as the ligand for a family of G protein-coupled receptors known as the endothelial differentiation gene-1 family, now collectively renamed SPP receptors (3)(4)(5)(6), supporting a role for SPP as an extracellular ligand. However, the intracellular targets of SPP have not yet been identified.
Sphingosine kinase (SPHK), the enzyme that catalyzes the phosphorylation of sphingosine, regulates the intracellular levels of SPP. Two isoforms of mammalian SPHK (SPHK1 and SPHK2) have been cloned and characterized (7,8). SPHK1 predominantly localizes in the cytosol, and its overexpression induces cell proliferation by promoting the G 1 to S transition of the cell cycle as well as by inhibiting the apoptotic response to serum deprivation or ceramide treatment (9). Several cellular proteins have recently been identified as SPHK1-interacting molecules, namely TRAF2 (10), RPK118 (11), and AKAP-related protein (12), which should help facilitate the understanding of the regulation and intracellular site of action of SPHK1.
In contrast to SPHK1, little is known about the cellular actions of the other isozyme, SPHK2. In the present studies, we investigated the biological features of SPHK2. We have discovered that SPHK2 localizes in the nuclei of cells through its novel nuclear localization signal (NLS) sequence, depending on cell type and cell density. We have also demonstrated that nuclear localization of SPHK2 causes inhibition of DNA synthesis in various cell types.

EXPERIMENTAL PROCEDURES
Plasmid Construction and Site-directed Mutagenesis-The mSPHK1 and mSPHK2 cDNAs (DDBJ/EMBL/GenBank TM accession number AF068748 and AF245448, respectively) were amplified from a mouse brain and a mouse kidney cDNA library, respectively, by PCR using ExTaq polymerase (Takara, Otsu, Japan). The full-length mSPHK2 or mSPHK1 cDNA was subcloned into a mammalian expression vector pCMV5 with an influenza hemagglutinin (HA) epitope tag to express N-terminally HA-tagged fusion proteins. In some experiments, mSPHK1, mSPHK2, or their HA-fused constructs were also cloned into the expression vector pTB-701, in which the green fluorescent protein (GFP) sequence had been subcloned into an EcoRI site for C-terminally GFP-fused protein expression. The human SPHK2 (GenBank TM accession number NM-020126) cDNA was amplified from a human liver cDNA library by PCR using KOD-PLUS polymerase (Toyobo, Tokyo) with 5Ј-GGA TCC GTA CCC TGG TCA GGG CTA AG-3Ј and 5Ј-GGA TCC CTC AGG GCT CCC GCC C-3Ј sense and antisense primers, respectively. The PCR product was subcloned into pCMV5.
The oligonucleotide primers used to synthesize truncated SPHK2 constructs were designed using pTB-701-HA-mSPHK2-GFP as a template, and the amino acid numbering was based on the mSPHK2 sequence as reported by Liu et al. (8). To obtain middle portion truncation mutant (⌬353-478, mSPHK2⌬M), N-terminal fragment (residues 1-352) and C-terminal fragment (residues 479 -617) cDNAs were obtained first by PCR amplification. For the N-terminal fragment, the primers used were 5Ј-GAG GAA TTC GCC ACC ATG TAC CCA TAC GAT GTT CCA GAT TAC GCT ATG GCC CCA CCA CCA CTA C-3Ј (sense) and 5Ј-AGA GGT ACC CAA GGC TGG TTC TGT GGT AG-3Ј (antisense). For the C-terminal fragment, the primers used were 5Ј-AGA GGT ACC CCA GTG GAC CAC CTC CTC-3Ј (sense) and 5Ј-CTC GAA TTC GGC TTG TGG CTT TTG ACC TGC-3Ј (antisense). These products were then ligated at a KpnI site and subcloned into an EcoRI site in pTB-701.
Site-directed mutagenesis for mSPHK2R93E/R94E was performed using mutagenic synthetic oligonucleotides (5Ј-GC CGT CGA GGG GGC GAG GAG AGA GCT ACG CGG ACC-3Ј and its reverse complement) and pTB-701-HA-mSPHK2-GFP as a template with a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). All of the constructs were verified by DNA sequencing.
The HA-mSPHK2-GFP cDNA was also subcloned into pUHG10-3 plasmid for the secondary stable transfection. The HeLa Tet-On cells (Clontech) were co-transfected with the plasmids pUHG10-3-HA-mSPHK2-GFP and pSV2bsr (Kaken Pharmaceutical Co., Tokyo). Drugresistant clones were isolated in the presence of 2.5 g/ml blastocidin. From the blastocidin-resistant clones, the cells harboring the pUHG10 -3-HA-mSPHK2-GFP were selected on the basis of the highest inducibility of HA-mSPHK2-GFP expression in the presence of 100 ng/ml doxycycline (Dox).
Immunocytochemistry-Cells were grown on four-chambered slides (Nalge/Nunc) and transfected using Fugene-6 reagent according to the manufacturer's instructions (Roche Applied Science). Subcellular localization studies using confocal microscopy were performed as described previously (11). A rabbit polyclonal anti-hSPHK2 antibody was raised against the synthetic peptide SQALHIQRLRPKPEARPR (amino acid residues 35-52) conjugated to keyhole limpet hemocyanin. The antibody was further purified by the immunogen peptide-immobilized Sepharose 4B.
Isolation of Nuclei-The separation of nuclei from cells was carried out essentially as described (13), with some modifications (14). For the preparation of cytosolic fraction, the supernatant fractions above the , or NIH 3T3 cells (C) with different epitope tags as indicated. Two days after transfection, cells were fixed and analyzed by confocal microscopy. In D-G, after fixation, cells were permeabilized and stained for confocal microscopy analyses with anti-hSPHK2 antibody (green). Nuclei were also stained with 2 g/ml DAPI (blue). Bars, 10 m.
sucrose cushion obtained after 15,000 ϫ g step were further centrifuged at 100,000 ϫ g for 30 min. The resultant nuclear and cytosolic fractions were assayed for enzymatic activity. For immunoprecipitation studies, nuclear fractions were obtained as described (15).
Measurement of in Vitro SPHK Activity-SPHK activity was determined in the presence of sphingosine, prepared as a complex with 4 mg/ml bovine serum albumin and [␥-32 P]ATP in kinase buffer containing 200 mM KCl as described previously (8) using either cell extracts or purified nuclear fractions as an enzyme source. In Fig. 2, the nuclear fractions were purified from HeLa cells grown in 10-cm dishes as described in Ref. 15 and were lysed by sonication in cold lysis buffer (20 mM Tris-HCl, pH 7.4, 130 mM NaCl, 1% (w/v) Triton X-100, and protease inhibitors (Roche Applied Science)). The nuclear lysates were clarified by centrifugation for 15 min at 10,000 ϫ g and incubated for 1 h with anti-hSPHK2 antibody (9 g). The immunoprecipitates were collected by protein A/G Plus-agarose (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and then washed three times with the lysis buffer and used as an enzyme source. [ 32 P]SPP was separated by thin layer chromatography and quantitated using a Fujix Bio-Imaging Analyzer BAS 2000 (Fuji Photo Film).
Measurement of DNA Synthesis-Bromodeoxyuridine (BrdUrd) incorporation was measured using BrdUrd labeling and detection kit I (Roche Applied Science). Briefly, NIH 3T3 cells grown on chambered slides were transiently transfected with either the GFP control plasmid or various plasmid constructs as indicated in Fig. 5 legend. Two days after transfection, cells were incubated for 3 h with 10 M BrdUrd and then fixed in 70% ethanol containing 50 mM glycine (pH 2.0) for 20 min at Ϫ20°C. After washing with PBS, cells were incubated for 1 h at 37°C with mouse monoclonal anti-BrdUrd and rat monoclonal anti-HA antibodies. After washing with the washing buffer, cells were incubated for 1 h with Alexa 594-conjugated goat anti-mouse IgG and Alexa 488conjugated goat anti-rat IgG in PBS containing 0.1% bovine serum albumin. Five random fields from each sample were analyzed with a confocal microscope. The percentage of transfected cells (green HA fluorescence) incorporating BrdUrd (red nuclei) was calculated as an average of four sample slides with a minimum of 100 cells scored per field.
For [ 3 H]thymidine incorporation studies, HeLa cells stably expressing SPHK2 were seeded in six-well plates at a density of 10 4 cells/well in Dulbecco's modified Eagle's medium supplemented with 10% FCS with various concentrations of Dox as specified in the figure legends. After 24 h, the serum in the medium was reduced from 10 to 0.5%, while Dox concentrations were maintained. Medium containing 0.5% FCS plus various concentrations of Dox was changed every day. After 8 days, cells were pulsed with 1 Ci/well of [methyl-1,2-3 H]thymidine (Amersham Biosciences) for 16 h, and the radioactivity incorporated into trichloroacetic acid-insoluble material was measured as previously described (16). Values are the means Ϯ S.D. of triplicate determinations.
Flow Cytometric Evaluations-NIH 3T3 cells were synchronized by a double thymidine block at the beginning of the S phase and released into thymidine-free medium (17). Briefly, cells in 60-mm dishes were first transfected with SPHK2-GFP or control GFP plasmid vectors (3 g each) and maintained in Dulbecco's modified Eagle's medium with 10% FCS and 2.5 mM thymidine for 19 h. Cells were washed twice with PBS and maintained further in Dulbecco's modified Eagle's medium with 10% FCS for 10 h. Cells were treated again with 2.5 mM thymidine for 17 h. Cells were washed twice with PBS and used for cell cycle analysis. At various time intervals after incubation of cells in Dulbecco's modified Eagle's medium with 10% FCS, cells were collected by trypsinization, washed once with PBS, and resuspended in PBS containing 0.75% formaldehyde for 10 min at room temperature. The cells were washed with PBS, resuspended in ice-cold 70% ethanol in PBS, and incubated for 10 min. The cells were washed once with PBS and resuspended in PBS containing 50 g/ml propidium iodide (PI) and 125 units/ml ribonuclease A and incubated for 1 h at room temperature in the dark.
Flow cytometric analysis for GFP and PI fluorescence was performed using a four-color FACSCalibur (BD Biosciences). Electronic compensation was used among the fluorescence channels to remove residual spectral overlap. GFP and PI fluorescence data were collected on a logarithmic and a linear scale, respectively. For each sample, 40,000 events were collected. Analyses of the multivariate data and DNA histograms were performed with CELLQuest and ModFitLT softwares, respectively (Verity Software House, Topsham, ME) that are part of the FACSCalibur operating system.
Other Procedures-Lactate dehydrogenase activity was measured according to the supplier's protocol (Kyokuto Chemicals, Tokyo). Protein was determined by the method of Bradford (18).

SPHK2 Is Predominantly Localized in the Nucleus-During
immunocytochemical analyses of SPHK2, we observed that SPHK2 was predominantly localized in the nuclei of various cell types. When mouse SPHK2 was fused with GFP and transiently expressed in COS7 cells, mSPHK2-GFP was mainly localized in the nuclei and to a lesser extent in the cytoplasm (Fig. 1A). Nuclear localization of SPHK2 was also observed when it was transiently expressed in HeLa cells (Fig. 1B) and NIH 3T3 cells (Fig. 1C). To exclude the possibility that GFP fusion itself had caused SPHK2 to mislocalize into the nuclei and to show that the nuclear localization of SPHK2 is a general characteristic independent of species differences, human SPHK2 (without any added tag sequence) was transiently expressed in HeLa cells, and its intracellular distribution was examined using a specific antibody raised against a peptide sequence common to both human and mouse SPHK2. Similar to mSPHK2, hSPHK2 was mainly localized in the nuclei with some cytoplasmic staining in these cells (Fig. 1D). The nuclear localization of hSPHK2 was confirmed by the nucleus-specific 4,6-diamidino-2-phenylindole (DAPI) staining of HeLa cells (Fig. 1E) and the colocalization of DAPI with hSPHK2 staining (Fig. 1F). Next, the other SPHK isozyme, SPHK1, was fused with GFP, and its intracellular distribution was compared with SPHK2. In COS7 cells, mSPHK1-GFP was localized mainly in the cytoplasm but not in the nucleus (Fig. 1H), an observation that is consistent with a previous report (9).
It is important to test whether the nucleo-cytoplasmic distribution of SPHK2 may change depending on cell type and cell confluence. When hSPHK2 was transiently expressed in HeLa cells, it was predominantly localized in the nuclei under all tested conditions (Table I). In the case of COS7 cells, the intracellular distribution of hSPHK2 was dramatically influenced by cell density; nuclear localization was prominent (61.4%) particularly in high cell density cultures, whereas it was only 10.3% in low cell density cultures. In contrast, hSPHK2 was predominantly localized in the cytosol irrespective of cell density in HEK293 (Table I).
Next, we tried to determine whether endogenous SPHK2 was also localized in the nucleus similar to exogenously ex-TABLE I Cell density-dependent and -independent changes in nucleocytoplasmic distribution of SPHK2 in various cell types hSPHK2 was transiently expressed in various cell lines either with high (3 ϫ 10 4 /mm 2 ) or with low cell densities (6 ϫ 10 3 /mm 2 ). Two days after transfection cells were fixed, permeabilized, immunostained using anti-hSPHK2 antibody, and analyzed by confocal microscopy as described under "Experimental Procedures." Cells expressing hSPHK2 were subdivided into three populations: predominantly nuclear localization (N Ͼ C), equal distribution between nuclus and cytosol (N ϭ C), and predominantly cytosolic localization (N Ͻ C), depending on SPHK2 staining pattern. Data are means Ϯ S. E. of duplicate cultures from a representative experiment. Five different fields were analyzed with a minimum of 100 cells.
Cell types and density Cell populations expressing SPHK2

SPHK2 Is a Nuclear Protein
pressed recombinant SPHK2. To address this issue, endogenous SPHK2 from purified nuclear extracts of HeLa cells was immunoprecipitated using anti-hSPHK2 antibody and analyzed for its activity. The immunoprecipitates showed clear SPP production, whereas the immunoprecipitates prepared in the presence of the immunogen peptide contained no detectable radioactivity corresponding to SPP, suggesting that the purified nuclear fractions contained SPHK activity that bound specifically to the antibody ( Fig. 2A). The aliquots of the immunoprecipitates prepared in the absence of the immunogen peptide were further analyzed by immunoblot using anti-hSPHK2 antibody (Fig. 2B). The immunoprecipitates from nuclear fractions revealed a clear band around 70 kDa, which corresponded to the position of affinity-purified recombinant HA-hSPHK2. The immunoreactive band around 70 kDa sometimes appeared as a doublet, suggesting the existence of post-translational modification of the protein. The anti-hSPHK2 antibody did not cross-react with HA-SPHK1. When the immunoprecipitates were immunoblotted with anti-SPHK1 antibody, there was no immunoreactive band either around authentic HA-SPHK1 or HA-SPHK2. These results strongly suggest that the anti-hSPHK2 antibody specifically recognizes endogenous SPHK2 but not SPHK1 and that the purified nuclear fractions contain endogenous SPHK2. This conclusion was further confirmed by immunocytochemical analyses. After fixation and staining of HeLa cells with anti-hSPHK2 antibody, endogenous SPHK2 was found to be distributed mainly in the nucleus and to a lesser extent in the cytosol (Fig. 2C), which is consistent with FIG. 2. Detection of endogenous SPHK2. A, detection of endogenous SPHK2 activity. Endogenous SPHK2 from the purified nuclear lysates from HeLa cells was immunoprecipitated (IP) using anti-hSPHK2 antibody in the absence or presence of 15 g/ml immunogen peptide. The immunoprecipitates were washed and assayed for SPHK activity. The endogenous SPHK2 produced 1.4 pmol of SPP/min/tube. Contamination of the purified nuclear fractions from cytosolic, endosomal, or endoplasmic reticulum fractions was estimated to be less than 3, 3, and 9%, respectively, as judged by the content of several specific protein markers: lactate dehydrogenase for cytosol, early endosomal antigen 1 for endosomes, and BiP/GRP78 for endoplasmic reticulum (see Supplemental Material). B, the aliquots of immunoprecipitates without immunogen peptide, affinity-purified HA-SPHK1, and HA-SPHK2 were subjected to 12.5% SDS-PAGE followed by immunoblot analyses using anti-hSPHK2 or anti-SPHK1 antibody (Abcam, Cambridge, UK). The molecular masses of standard proteins and the position of heavy and light chains of immunoglobulin are indicated. The arrows mark the positions of SPHK1 and SPHK2. C, HeLa cells were fixed, permeabilized, and stained for confocal microscopy analyses using anti-hSPHK2 antibody (red). Nuclei were also stained with 2 g/ml DAPI (blue). Bar, 10 m. the findings observed using a recombinant overexpressed system ( Fig. 1 and Table I).
NLS Is Localized in the N-terminal Portion of SPHK2-mSPHK2 encodes a protein, which is larger than mSPHK1 and contains an additional 236 amino acids (8). Compared with SPHK1, SPHK2 contains two largely unrelated sequences, one at the amino terminus and the other in the middle of the protein (Fig. 3A). It may not be unreasonable to assume that the differences in the intracellular distribution between mSPHK1 and mSPHK2 are attributable to these additional unrelated sequences, which exist only in mSPHK2. To test this possibility, we constructed two deletion mutants: mSPHK2⌬M, which lacks the middle portion of the unrelated sequence, and mSPHK2⌬N, which lacks the amino-terminal portion of the protein (Fig. 3A). When mSPHK2⌬M-GFP was transiently expressed in COS7 cells, the protein was localized mainly in the nuclei showing that the middle portion of SPHK2 had no influence on the nuclear localization of this protein (Fig. 1J). On the other hand, mSPHK2⌬N-GFP could not enter the nuclei when it was expressed in COS7 cells (Fig. 1K). These results strongly suggest that an NLS sequence resides within the amino terminus of mSPHK2.
Identification of the NLS Sequence of SPHK2-Although the sequence of mSPHK2 has been analyzed for its domain structure, so far no one has reported the presence of an NLS identical to the previously known NLS sequences. However, our analyses showed that SPHK2 possesses a putative NLS (RGRRGGRRR) with an arginine cluster, which is highly similar to a type of monopartite NLS enriched in arginine residues identified in the Tat and Rev proteins of human immunodeficiency virus type 1 (19), the Rex protein of human T-cell leukemia virus type 1 (20), and the atypical protein kinase C (21) (Fig. 3B). To demonstrate that this putative NLS functions as a true NLS in the cells, another mutant mSPHK2R93E/R94E-GFP, where both Arg-93 and Arg-94 of mSPHK2-GFP were mutated to glutamic acid, was constructed and analyzed. A similar mutation in the NLS sequence of protein kinase C has been reported to abolish the nuclear accumulation of protein kinase C (21). When mSPHK2R93E/R94E-GFP was transiently expressed in COS7 cells, the mutant protein could not enter the nuclei (Fig. 1L), indicating that these arginine residues were critical for the functioning of this NLS.
To show that this tentative NLS sequence of SPHK2 functions as a true NLS, this sequence was fused with a cytoplasmic protein, SPHK1, and its nuclear targeting efficiency was analyzed. When NLS-mSPHK1-GFP was transiently expressed in COS7 cells, the protein accumulated mainly in the nuclei (Fig.  1M). These observations indicate that the NLS sequence of mSPHK2 was necessary and sufficient to direct a complete nuclear import of SPHK2.
SPHK2 Causes Inhibition of DNA Synthesis-Nuclear localization of SPHK2 as observed in Figs. 1 and 2 raised the possibility that this enzyme may be involved in some nuclear function. We next analyzed DNA synthesis as measured by [ 3 H]thymidine incorporation into cells stably expressing SPHK2. For these experiments, HA-mSPHK2-GFP was stably expressed in HeLa Tet-On cells in a Dox-inducible manner. The cells expressing SPHK2 (induced by 0.1 g/ml Dox) showed a 50% inhibition in [ 3 H]thymidine incorporation compared with the cells expressing no exogenous SPHK2 under Dox-deficient conditions (Fig. 4A). Dox-inducible expression of SPHK2 was confirmed by immunoblot analyses using anti-HA antibody as shown in Fig. 4B. In the cells carrying empty vector, Dox at the concentrations used here had no significant effect on thymidine incorporation (data not shown).
Nuclear Localization Is Necessary for SPHK2 to Inhibit DNA Synthesis-The mechanism of the inhibition of DNA synthesis by SPHK2 was further assessed by measuring BrdUrd incorporation into nascent DNA in NIH 3T3 cells transiently expressing various constructs of SPHK and its mutants. When HA-mSPHK2-GFP was expressed in NIH 3T3 cells, BrdUrd incorporation was strongly inhibited as compared with the cells transfected with a control vector (Fig. 5). In contrast, cells expressing HA-mSPHK1 showed a 40% increase in BrdUrd incorporation, in agreement with a previous report (9). It is important to determine whether the nuclear localization of SPHK2 is required for the inhibition of DNA synthesis. The next experiments were designed to show the importance of nuclear localization of SPHK2 for inhibition of DNA synthesis by manipulating the ability of nuclear localization of SPHK2. Upon transient transfection in NIH 3T3 cells, the NLS mutant, HA-mSPHK2R93E/R94E, which failed to enter the nucleus (Fig. 1L), lost its ability to inhibit DNA synthesis (Fig. 5). More importantly, when mSPHK1 acquired nuclear localization ability upon fusion with the NLS sequence from mSPHK2, it showed inhibitory effects on DNA synthesis, suggesting that nuclear localization, and not the other features of SPHK2, is important for the inhibition of DNA synthesis. When hSPHK2 without any added tag sequences was transiently expressed in HeLa cells, essentially the same inhibitory effect on BrdUrd uptake was observed (data not shown).
Next, we have checked the protein expression and enzymatic activity of various SPHK constructs. As shown in Fig. 6A, vector-transfected COS7 cells have low levels of SPHK activity. In cells transfected with HA-mSPHK1, NLS-HA-mSPHK1, or hSPHK2, in vitro SPHK activity was increased by 63-, 53-, and 52-fold, respectively. We have further demonstrated the in vitro accumulation of SPP in the nucleus using purified intact nuclear fractions. HeLa cells transiently expressing various SPHK proteins were disrupted, and the nuclei were isolated through high sucrose solution (13). In cells expressing hSPHK2, SPP accumulated mostly in the nuclear fractions (75%) (Fig. 6B), which supports the results obtained from the morphological determinations (Table I). On the other hand, SPP accumulated mainly in the cytosol of cells expressing HA-mSPHK1, consistent with a previous report (9). NLS-HA-mSPHK1 showed a clear shift in nuclear accumulation of SPP, confirming the results from the confocal microscopic analyses (Fig. 1M). In all nuclear preparations, lactate dehydrogenase, a typical cytosolic marker protein, was almost undetectable, whereas the typical nuclear membrane marker nucleoporin was detected exclusively in the nuclear fractions, validating the reliability of the cell fractionation technique.
SPHK2 Expression Causes Cell Cycle Arrest at G 1 /S Phase-Next, the effect of SPHK2 on cell cycle was studied. NIH 3T3 cells transiently transfected with mSPHK2-GFP were synchronized to the beginning of the S phase by a double thymidine block, and the DNA content of cells expressing the SPHK2 protein was analyzed by flow cytometry. About 100% of the nonexpressing and SPHK2-expressing cells had not entered G 2 /M phase at 0 h as a result of double thymidine blockinduced synchronization of the cell cycle (Fig. 7B). Upon release from the thymidine block, the percentage of cells in various phases of the cell cycle remains very different in the two populations. In the SPHK2-nonexpressing cells, a majority of cells entered the S phase after 3 h, and the bulk of the cells left the S phase and entered the G 2 /M phase by 9 h (Fig. 7B). In contrast, the SPHK2-expressing cells lagged in their progression through the cell cycle and persisted in the G 1 phase much longer, showing a typical G 1 /S arrest, such that at the 9-h time point when nearly 40% of the nonexpressing cells were in the G 2 /M, only a very modest 7% of the SPHK2-expressing cells were in the G 2 /M phase of the cell cycle. NIH 3T3 cells expressing GFP alone showed a pattern essentially similar to the nonexpressing cells as in Fig. 7B (data not shown). This G 1 /S arrest eventually did not lead to apoptosis as judged by the occurrence of cells in the sub-G 1 region (hypodiploidy) in cell cycle analysis. The results from flow cytometric analyses strongly suggest that SPHK2 causes inhibition of DNA synthesis through cell cycle arrest at G 1 /S phase. DISCUSSION This is the first report to show that SPHK2 is predominantly localized in the nuclei (Figs. 1 and 2), depending on the cell type FIG. 6. SPHK2 activity in purified nuclear fractions. A, COS7 cells were transiently transfected with hSPHK2, HA-mSPHK1, NLS-HA-mSPHK1, or plasmid vector. Two days after transfection, cell lysates were subjected to SDS-PAGE followed by immunoblot analyses either with anti-HA or -SPHK2 antibody (insets). Aliquots of cell lysates were also assayed for enzymatic activity. B, hSPHK2, HA-mSPHK1, NLS-HA-mSPHK1, or plasmid vector was transiently transfected in HeLa cells. Two days after transfection, cells were disrupted and fractionated. Cytosolic (C) and nuclear (N) fractions were assayed for SPP accumulation or lactate dehydrogenase (LDH) activity. The SPHK activity of control cells transfected with plasmid vector alone was subtracted from those expressing various SPHK constructs. Data are expressed as percentage of total activity and are the mean Ϯ S.E. of triplicate determinations. and cell confluence ( Table I). The present results as summarized in Table I are also consistent with a previous report that SPHK2 transiently transfected in HEK293 cells mainly localizes in the cytosol (8). However, in HeLa cells, hSPHK2 is almost exclusively in the nucleus, and in COS7 cells it is dominantly nuclear in high cell density cultures. Whether SPHK2 is involved in contact inhibition of proliferation in COS7 cells is an attractive possibility that needs to be explored carefully. Molecular mechanisms underlying the regulation of nucleo-cytoplasmic shuttling of SPHK2 are at present unknown and need to be studied using various model cell systems. Recently, Kleuser et al. (14) have reported that platelet-derived growth factor stimulates nuclear SPHK activity in Swiss3T3 cells. They have suggested that SPHK1 translocates from the cytosol to the nuclear envelope upon stimulation of cells by platelet-derived growth factor. However, the present data dem- onstrate that it is SPHK2, which is localized mainly in the nucleoplasm and not in the nuclear envelope (Fig. 1).
We have also demonstrated that nuclear localization of SPHK2 causes inhibition of the DNA synthesis (Figs. 4 and 5). That this observation is not an artifactual effect caused by overexpressing exogenously introduced genes but represents an intrinsic feature of SPHK2 is supported by several arguments. First, the inhibition of DNA synthesis is specific to SPHK2 expression but not to a control, empty vector or SPHK1 overexpression (Fig. 5). Second, the inhibitory effect of SPHK2 expression on DNA synthesis was cancelled by introducing the NLS mutant (Fig. 5), although the levels of expression of the wild type SPHK2 and the mutant SPHK2R93E/R94E were similar (Fig. 1, compare A and L). Third, SPHK1 acquired the ability to inhibit DNA synthesis when the NLS sequence of SPHK2 was fused to this isozyme (Fig. 5). Finally, expression of even a small amount of SPHK2, as manipulated by varying doxycycline concentrations in HeLa Tet-on cells stably transfected with SPHK2, showed similar nuclear localization and inhibitory effects on DNA synthesis (Fig. 4). The inhibitory effect of SPHK2 on DNA synthesis is not due to the inactivation of the enzyme in the nuclei, because SPHK2 and NLS-SPHK1 in purified nuclei were just as active as cytosolic enzymes as judged by their specific activity ( Fig. 6 and data not shown).
So far, SPHK1 has been implicated both in cell proliferation and antiapoptotic processes (1,2,9). In the present studies, we have confirmed that SPHK1 mainly localizes in the cytosol and not in the nucleus (Fig. 1H) and that the enzyme stimulates DNA synthesis (Fig. 5). On the other hand, we have also demonstrated that SPHK2 mainly localizes in the nuclei of several cells tested (Fig. 1) and that its nuclear localization corresponds to the inhibition of DNA synthesis (Figs. 4 and 5). It is interesting to speculate that cell proliferation or cessation of the proliferation process may be controlled by the balance of SPP content between cytosol and nucleus as well as the cellular balance between ceramide plus sphingosine versus SPP level as explained by the so-called "rheostat" model (1,2). Thus, the binding proteins for SPP in the cytosol and the nucleus may determine the fate of cells. Identification of the binding protein(s) for SPP is an urgent subject that needs to be studied for a proper understanding of the molecular mechanism of proliferation, differentiation, and apoptosis.