Nuclear Localization of Phospholipase D1 Mediates the Activation of Nuclear Protein Kinase Cα and Extracellular Signal-regulated Kinase Signaling Pathways

Abstract

Recent studies highlight the existence of a nuclear lipid metabolism related to cellular proliferation. However, the importance of nuclear phosphatidylcholine (PC) metabolism is poorly understood. Therefore, we were interested in nuclear PC as a source of second messengers and, particularly, nuclear localization of PC-specific phospholipase D (PLD). In the present study we have identified the nuclear localization sequence (NLS) of PLD1 whose mutation abolished its nuclear import. Recently, we reported that caspase-mediated cleavage of PLD1 generates the N-terminal fragment (NF-PLD1) and C-terminal fragment (CF-PLD1). Here we show that CF-PLD1 but not NF-PLD1, is exclusively imported into the nucleus via its functional NLS, whereas only some portions of intact PLD1 were localized into the nucleus. The NLS of intact PLD1 or CF-PLD1 is required for interaction with importin-β, which is known to mediate nuclear import. The amount of intact PLD1 or CF-PLD1 translocated into nucleus is correlated with its binding affinity with importin-β. Ultimately, nuclear localization of intact PLD1 but not CF-PLD1 mediates the activation of nuclear protein kinase Cα and extracellular signal-regulated kinase signaling pathways. Taken together, we propose that nuclear localization of PLD1 via the NLS and its interaction with importin-β may provide new insights on the functional role of nuclear PLD1 signaling.

Introduction

It is becoming increasingly evident that stimulation of nuclear lipid metabolism plays a central role in many signal transduction pathways that ultimately result in various cell responses including proliferation and differentiation. An increasing body of evidence has demonstrated that the nucleus is a site of active lipid metabolism for both synthesis and hydrolysis (1, 2). Nuclear lipid metabolism is regulated independently from that of the plasma membrane because there are extracellular stimuli that cause the generation of lipid signaling molecules only in the nucleus and not in the plasma membrane (3). Nuclear lipid second messengers are likely involved in the control of cell proliferation and gene expression (4, 5). Nuclear phosphoinositides have received a considerable degree of attention. Nevertheless, there are many reports that have dealt with metabolism and signaling activities of nuclear phosphatidylcholine (PC)² and sphingolipids (5, 6). Like its extranuclear counterpart, nuclear PC comprises the major phospholipid class within that compartment (6).

PLD is a signal-activated enzyme that catalyzes phosphatidylcholine, producing phosphatidic acid (PA) and choline (7). Two distinct isoforms of mammalian PLD, PLD1 and PLD2, have been identified (7). Although most research on PLD-mediated PC metabolism focused on cellular membrane, PC also has been associated with the nuclear envelope, chromatin, and the nuclear matrix (8, 9). Moreover, studies from several laboratories have demonstrated the presence of a PC-specific PLD activity associated with nuclei. G protein-dependent and oleate-dependent PLD activity has been identified in nuclei from Madin-Darby canine kidney cells (10), IIC9 fibroblasts (11), LA-N-1 neuroblastoma cells (12), and vascular smooth muscle cells (13). However, the isoform responsible and regulation of its nuclear localization have yet to be established. Changes in nucleocytoplasmic localization might represent a mode of PLD regulation in nuclear PC metabolism. However, the molecular mechanisms that regulate nuclear translocation of PLD remain unknown.

Depending on the size of the protein, nuclear import through the nuclear pore complex can occur through either passive diffusion (for small molecules of less than 40–50 kDa) or by an active process facilitated by the nuclear localization sequence (NLS) present in nuclear protein (14). NLSs are usually recognized by the heterodimeric receptor proteins importin-β and importin-α (15). Here, we have defined a functional NLS of PLD1. Recently, we have reported that caspase-mediated cleavage of PLD1 during apoptosis generates the N-terminal fragment (NF-PLD1) and the C-terminal fragment (CF-PLD1) (16). In the present study, we show that CF-PLD1 but not NF-PLD1 is exclusively imported into the nucleus via its functional NLS, whereas a portion of intact PLD1 was localized into nucleus. The NLS of intact PLD1 or CF-PLD1 associates with importin-β and mediates its nuclear import. The nuclear localization of intact PLD1 but not CF-PLD1 mediates the activation of protein kinase Cα (PKCα) and extracellular signal-regulated kinase (ERK) in the nucleus. Taken together, our current study sheds light on yet not understood molecular mechanisms regulating nuclear localization of PLD, a timely topic with important implications in nuclear signaling of PLD.

EXPERIMENTAL PROCEDURES

Cell Lines and Materials

HEK293, U87MG, U251MG, U373MG, and T98G cells were cultured at 37 °C in DMEM (Invitrogen) containing 10% fetal bovine serum and 1% antibiotic-antimycotic. Cells were grown to 60% confluence for transient transfection using Lipofectamine Plus (Invitrogen) according to the manufacturer's instructions. Leptomycin B (LMB), phorbol myristate acetate and platelet-derived growth factor (PDGF) were from Calbiochem. PLD1 inhibitor (VU0155069) was purchased from Cayman Chemical (Ann. Arbor, MI). The siRNA of 21-nucleotide sequences corresponding to human PLD1 sequences (nucleotides 1571–1591, AAGGUGGGACGACAAUGAGCA) was purchased from Dharmacon Research Inc (Lafayette, CO). PLD1 inhibitor (VU0155069) was from Cayman Chemical.

Construction of Plasmids and Site-directed Mutagenesis

Cloning of GFP-tagged full-length (FL)-, NF-, and CF-PLD1 has been previously described (16). GST-PLD1 fragments were kindly provided by Dr. Sung Ho Ryu (Pohang University of Science and Technology, Korea). GST-importin-β was kindly provided by Dr. Hoyun Lee (Department of Biochemistry, Microbiology and Immunology, University of Ottawa, Canada). Site-directed mutagenesis of the NLS (K553A/R555A/K556A/K559A/K564A) of PLD1 was generated by PCR using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA), according to the manufacturer's instructions. Mutation was then confirmed by DNA sequencing. Primers for generation of the NLS mutant PLD1 are as follows: NLM1 (K553A/R555A/K556A) (5′-forward (CTGAAAGGAATAGGAGCCCCAGCAGCCTTCTCCA-AATTTAG) and 3′-reverse (CTAAATTTGGAGAAGGCTGCTGGGGCTCCTATTCCTTTCAG); NLM (K553A/R555A/K556A/K559A/K564A) (5′-forward GCAGCCTTCTCCGCATTTAGTCTCTACGCGCAGCTCCAC) and 3′-reverse (GTGGAGCTGCGCGTAGAG ACTAAATGCGGAGAAGGCTGC). Mutagenesis for generation of NLM-CF- or NLM-FL-PLD1 was performed using NLM1-CF-PLD1 or NLM1-FL-PLD1 as a template.

PLD Activity Assay

For measurement of PLD activity, cells were labeled with [³H]myristic acid, and PLD activity was assessed by measurement of formation of [³H]phosphatidylbutanol, the product of PLD-mediated transphosphatidylation, in the presence of 1-butanol, as previously described (17).

In Vitro Binding Experiment

The lysates of HEK293 cells were incubated with 3 μg of GST fusion proteins immobilized on glutathione-Sepharose beads in a final volume of 500 μl of lysis buffer for 1.5 h at 4 °C. Protein complexes were collected by centrifugation and washed 4 times with washing buffer (1% Triton X-100, 150 mm NaCl, 20 mm Tris-HCl, pH 8.0, 20 mm NaF, 2 mm sodium orthovanadate, 1 mm PMSF, 10 μg/ml leupeptin, 10 μg/ml aprotinin). Associated protein complexes were resolved by SDS-PAGE and transferred to a nitrocellulose membrane. Immunoreactivity was detected using the indicated antibodies, horseradish peroxidase-conjugated secondary antibodies, and enhanced chemiluminescence (Amersham Biosciences) according to the manufacturer's instructions.

In Vitro Translation

In vitro transcription and translation of FL-PLD and CF-PLD1 were carried out using the TnT Quick Reticulocyte Lysate System (Promega, Madison, WI) according to the manufacturer's instructions. Freshly synthesized proteins were added to bead-bound GST fusion proteins and allowed to bind at 4 °C, and then associated protein complexes were resolved by SDS-PAGE and transferred to a nitrocellulose membrane. Immunoreactivity was detected using the same methods with the above in vitro binding experiment.

Western Blotting and Immunoprecipitation

Cell lysates were analyzed by immunoblot and/or immunoprecipitation as previously described (18). Enhanced chemiluminescence was used for signal detection. The following antibodies were used: anti-α-tubulin (Sigma), anti-GFP (Santa Cruz), anti-murine double minute 2 (Santa Cruz), anti-importin-β (Santa Cruz), anti-caspase3 (Cell Signaling), anti-lamin B (Santa Cruz), anti-phospho PKCα (Cell Signaling), anti-PKCα (Santa Cruz), anti-phospho ERK (Cell Signaling), anti-ERK (Cell Signaling). Rabbit polyclonal anti-PLD antibody that recognizes both PLD1 and PLD2 was generated as previously described (19).

Fluorescence Microscopy

Cells were transfected with various PLD1 constructs and incubated with media containing 1 mg/ml Hoechst (Molecular Probes, CA) for 20 min. Cells were visualized, and images were collected using a fluorescence microscope (Axiovert 200 m, Zwiss, Germany) and a confocal fluorescence microscopy (LXM510, Zeiss, Germany). For immunocytochemistry, cells were washed with PBS and fixed with 4% formaldehyde, PBS for 10 min followed by permeabilization with 0.2% Nonidet P-40, PBS for 5 min at room temperature. Cells were washed with PBS, blocked for 2 h in 20 mg/ml bovine serum albumin (BSA), and then incubated with primary antibody, BSA for 2 h. Cells were washed four times with PBS and then incubated with anti-rabbit Texas Red-conjugated secondary antibody or anti-mouse fluorescein isothiocyanate-conjugated secondary antibody (Jackson ImmunoResearch Laboratories) containing 1 mg/ml Hoechst for 1 h. Cells were washed again four times with PBS and mounted with antifading mounting medium (Vector Laboratories). Cells were visualized, and images were collected using a fluorescence microscope.

Separation of Nuclear and Cytosolic Fractionation

Using a commercially available kit, nuclear and cytosolic fractions were separated according to the manufacturer's protocol (Pierce). Purity of the nuclear and cytosolic fraction was confirmed by immunoblotting with antibody to lamin B and α-tubulin, respectively.

Statistics

Results are expressed as the mean ± S.D. of the number of determinations indicated. Statistical significance of differences was determined by analysis of variance. Significance was accepted when p < 0.01.

RESULTS

Endogenous PLD1 Is Detected in the Nucleus

Evidence is being accumulated on the importance of internal nuclear lipid metabolism. Nuclear lipid metabolism gives rise to several lipid second messengers that function within the nucleus. Thus, we examined the presence of nuclear PLD hydrolyzing nuclear PC. As shown in Fig. 1A, PLD1 was detected not only in the cytoplasm but also in the nucleus in HEK293 cells as analyzed by Western blot using antibody to PLD that recognizes both PLD1 and PLD2 (Fig. 1A). PLD1 is predominantly expressed in HEK293 cells. Purity of the nuclear and cytosolic fraction was confirmed by immunoblotting with antibody to nuclear marker, lamin B, or the cytosolic marker, α-tubulin. The relative percentage of cytosolic and nuclear PLD1 fraction from the total cellular PLD pool, was examined. Each sample represents the fraction from the same number of total cell equivalents. As shown in Fig. 1A, cytosolic PLD1 accounts for about 38% of the total cellular PLD pool, and nuclear PLD1 accounts for 13% of the total PLD pool. To further confirm the nuclear localization of PLD1, HEK293 cells were immunostained with anti-PLD antibody and analyzed by confocal scanning laser microscopy. The localization of PLD1 was comparable with that of the immunoblot, and PLD1 was detected in the nucleus as well as cytosol (Fig. 1B). No red labeling was observed when the primary antibody was omitted (data not shown). Moreover, we examined whether PLD is localized in the nucleus from various human glioma cells, which show highly invasive characteristics and elevated expression of PLD in comparison with that of normal glial cells (20). PLD1 and PLD2 were detected in the nucleus as well as cytosol in four kinds of glioma cells (Fig. 1C). These data suggest the presence of the PC-PLD protein in the nucleus.

FIGURE 1.

PLD1 is localized in both the cytoplasm and the nucleus. A, HEK293 cells were fractionated into cytosol (C) and nucleus (N) and analyzed by immunoblot using antibodies to PLD, lamin B, or α-tubulin. The relative percentage of cytosolic and nuclear PLD1 fraction from the total cellular PLD pool was examined. Total lysates (T). Each sample represents the fraction from the same number of total cell equivalents. These data are representative of results obtained from three experiments. B, localization of endogenous PLD in HEK293 cells was analyzed by immunostaining using antibody to PLD and Texas Red-labeled secondary antibody, and nuclei were stained with Hoechst and then observed using confocal microscopy. The white arrows represent nuclear localization of PLD. C, various human glioma cells were fractionated into cytosol (C) and nucleus (N) and analyzed by immunoblot using the indicated antibodies. The relative percentage of cytosolic and nuclear PLD1 fraction from the total cellular PLD pool was examined. Each sample represents the fraction from the same number of total cell equivalents. These data are representative of results obtained from three experiments.

Nuclear Localization of PLD1 Is Not Due to Inhibition of Nuclear Export

Nuclear localization of PLD1 could result from either import into the nucleus or the inhibition of nuclear export. To address this latter possibility, HEK293 cells expressing endogenous PLD1 were treated with an inhibitor of the CRM-1 exportin pathway, LMB, and the localization of PLD1 was examined (Fig. 2A). As a positive control, murine double minute 2 (MDM2), which is exported from the nucleus by CRM-1 exportin pathway, was predominantly localized in cytoplasm in untreated cells. Treatment with LMB resulted in the nuclear accumulation of MDM2, whereas localization of PLD1 was not affected by LMB treatment. To further confirm that nuclear PLD1 is not due to suppressed nuclear export, we performed Western blot analysis of the nuclear versus cytoplasmic fractions from control versus LMB-treated cells. LMB treatment induced nuclear accumulation of MDM2 but not of PLD1 (Fig. 2B). This result is comparable with that of Fig. 2A. The level of MDM2 and PLD1 was quantitated using densitometer analysis. The purity of the nuclear and cytosolic fraction was examined by immunoblotting with antibody to nuclear marker, lamin B, or the cytosolic marker, α-tubulin (Fig. 2B). These results suggest that the presence of PLD1 in the nucleus does not result from inhibition of nuclear export.

FIGURE 2.

Nuclear import of PLD1 is not due to inhibition of nuclear export. A, HEK293 cells were left untreated or treated with LMB (leptomycin B, 5 ng/ml) for 4 h, fixed, permeabilized, and stained with an FITC-conjugated antibody specific for MDM2 (green) and a Cy3-conjugated antibody specific for PLD (red) with Hoechst to identify the nuclei and then viewed by fluorescence microscopy. B, HEK293 cells were left untreated or treated with LMB, fractionated into cytosol (C) and nucleus (N), and analyzed by immunoblot using the indicated antibodies. The level of MDM2 or PLD1 was quantitated using densitometer analysis. These data are representative of results obtained from three experiments.

PLD1 Contains a Functional NLS

Next, we investigated the amino acid sequences involved in the nuclear import using the PSORTII program (21). We have identified a series of basic amino acids (⁵⁵³KPRK⁵⁵⁶) in the loop region, as seen in many classical NLSs (22). In addition, we noticed two additional basic amino acids (Lys-559 and Lys-564) located in the near of ⁵⁵³KRPK⁵⁵⁶ (Fig. 3A). These sequences are highly conserved in mammalian PLD1 from different species. We could not find the putative NLS of PLD2. To determine whether or not the putative NLS is functional for nuclear import of PLD1, an intact PLD1 construct was generated by mutation of lysine and arginine residues within the putative NLS. The NLS mutant (NLM)-PLD1 harbors mutations in five lysine residues, which were changed to alanine (K553A/R555A/K556A/K559A/K564A) (Fig. 3A). Wild type or the mutant form of PLD1 was transfected into HEK293 cells and fractionated into nuclear and cytosolic fraction, then analyzed by immunoblot. As shown in Fig. 3B, an intact GFP-PLD1 was detected in both cytosolic and nuclear fraction, whereas GFP-NLM-PLD1 was localized exclusively in the cytosol but not in the nucleus. These data demonstrate that the NLS of PLD1 is required for its nuclear import. To compare relative levels and localization of the endogenous versus heterologously expressed PLD1, Western blot was performed using anit-PLD1 antibody (Fig. 3B). Exogenous GFP-PLD1 and endogenous PLD1 showed similar localization pattern in the cytosolic and nuclear fractions. Exogenous GFP-PLD1 did not affect the localization of the endogenous PLD1. The endogenous PLD1 expression level was quantitated. Ectopic expression of PLD1 increased the protein level of the endogenous PLD1. This result is correlated with our recent report that PLD activity is coupled to selective induction of PLD1 expression (23). To examine whether nuclear localization of PLD1 affects its enzymatic activity, PLD activity was measured. We observed that basal and phorbol myristate acetate-stimulated PLD activity was independent of NLS (Fig. 3C). Taken together, these data demonstrate that the mutated residues are required for localization but not activity.

FIGURE 3.

PLD1 contains a functional NLS. A, alignment of the putative NLS in mammalian PLD1 from different species and schematic representation of mutations generated in the GFP-tagged PLD1 (NLM) are shown. B, after the indicated GFP-PLD1 constructs were transfected in HEK293 cell, the cells were fractionated into cytosol (C) and nucleus (N) and analyzed by immunoblot using antibodies to the indicated antibodies. The level of endogenous PLD1 was quantitated using densitometer analysis. These data are representative of results obtained from three experiments. Vec, vector. C, HEK293 cells were transiently transfected the indicated expression vector, treated with or without phorbol myristate acetate (50 nm) for 1 h, and labeled with [³H]myristic acid, and PLD activity was measured as described under “Experimental Procedures.” The data are the mean ± S.D. of the five independent experiments. NS, not significant; PtdBut, [³H]phosphatidylbutanol.

CF-PLD1, a Caspase Cleavage Product of PLD1, Is Localized into the Nucleus

Recently, we reported that caspase-mediated cleavage of PLD1 during apoptosis results in the production of NF-PLD1 and CF-PLD1 (16). Whereas NF-PLD1 lacks the NLS, CF-PLD1 contains the NLS. Thus, we examined the localization of PLD1 cleavage products using fluorescence microscopy and Western blot. To further define the functional NLS motif, two kinds of putative NLS mutants were generated using GFP-tagged CF-PLD1. NLM-CF-PLD1 contains both mutation of adjacent lysine residues (K559A, K564A) as well as one stretch of basic amino acids (⁵⁵³KPRK⁵⁵⁶), and NLM1-CF-PLD1 contains mutation of one stretch of basic amino acids (Fig. 4A). NF-PLD1 (1–545 amino acids) was exclusively localized in the cytoplasm as a diffuse distribution, whereas CF-PLD1 (546–1036 amino acids) was predominantly localized in the nucleus (Fig. 4B) as shown by fluorescence microscopy. To rule out any effect on localization of CF-PLD1 resulting from the GFP fusion, we examined the localization using FLAG-tagged CF-PLD1. FLAG-tagged CF-PLD1 was also localized exclusively in the nucleus (data not shown). As shown in fluorescence microscopy, NLM-CF-PLD1 exclusively shows cytosolic distribution but not nuclear localization, whereas NLM1-CF-PLD1 is evenly distributed in both the cytosol and nucleus (Fig. 4B). To further confirm the localization pattern of these proteins, HEK293 cells were transfected with these constructs, and the lysates were fractionated into cytosol and nucleus. As shown in Fig. 4C, NF-PLD1 and CF-PLD1 were exclusively localized in the cytosol and nucleus, respectively. NLM1-CF-PLD1 was detected evenly in cytosol and nucleus, whereas NLM-CF-PLD1 was predominantly detected in the cytosol but not in the nucleus. Thus, it is suggested that NLM but not NLM1 is sufficient for nuclear import of CF-PLD1. These results are consistent with those obtained from fluorescence microscopy.

FIGURE 4.

CF-PLD1, a caspase cleavage product of PLD1, is localized into nucleus. A, shown is schematic representation of caspase cleavage fragments (NF-PLD1, CF-PLD1), its NLS mutants (NLM1-CF-PLD1, NLM-CF-PLD1). B, HEK293 cells were transfected with the indicated GFP-PLD1 constructs, and its localization was observed using fluorescence microscopy. C, HEK293 cells were transfected with the indicated GFP-PLD1 constructs, and the lysates were fractionated into cytosol (C) and nucleus (N) and immunoblotted with the indicated antibodies. These data are representative of results obtained from three experiments.

Exogenous CF-PLD1 detected by anti-GFP and anti-PLD antibodies showed similar localization patterns (Fig. 4C). When GFP-tagged NF- or CF-PLD1 was transfected, exogenous PLD1 did not affect localization of endogenous PLD1. Taken together, these data demonstrate that nuclear import of CF-PLD1 involves the functional NLS (⁵⁵³KPRK⁵⁵⁶/K⁵⁵⁹/K⁵⁶⁴) motif.

NLS of PLD1 Is Required for Binding to Importin-β

To determine whether or not nuclear translocation of PLD1 involves the importin-β pathway, we examined the endogenous interaction between PLD1 and importin-β in HEK293 cells. As shown in Fig. 5A, PLD1 displayed specific binding to importin-β, as analyzed by co-immunoprecipitation and immunoblot. Their interaction was not observed by preimmune IgG. Then we investigated which region of PLD1 is associated with importin-β. Importin-β interacted with F3 fragment of PLD1 (499–604 amino acids), which contains NLS motif (553–564 amino acids) as analyzed by a glutathione S-transferase-PLD pull down assay (Fig. 5B). We then investigated the question of whether or not the association between PLD1 and importin-β is mediated through the NLS. As shown in Fig. 5C, wild type of FL-PLD1 was associated with importin-β, whereas NLM-PLD1 did not interact with importin-β. Moreover, NF-PLD1 not localized into nucleus did not interact with importin-β, whereas CF-PLD1 very strongly interacted with importin-β. Interestingly, importin-β showed more strong interaction with CF-PLD1 compared with that of FL-PLD1. Considering that only some portion of FL-PLD1 is localized in nucleus and CF-PLD1 is exclusively localized in nucleus, it is suggested that the amount of PLD1 imported into the nucleus is correlated with binding affinity with importin-β. Additionally, NLM-CF-PLD1 did not interact with importin-β, whereas NLM1-CF-PLD1 exhibited weak binding to importin-β compared with that of CF-PLD1 (Fig. 5D). Therefore, the binding affinity of NLM- or NLM1-CF-PLD1 with importin-β is also correlated with the amount of CF-PLD1 imported into nucleus.

FIGURE 5.

The NLS of PLD1 is required for binding to importin-β. A, lysates of HEK293 cells were co-immunoprecipitated (IP) and immunoblotted using the indicated antibodies. B, shown is a schematic representation of the structure of PLD1 (upper panel). PX, phox domain; PH, pleckstrin homology-like domain; HKD, catalytic motif. The equal amounts GST or GST fusion proteins (GST-PLD1 fragments, F1–F7) were incubated with HEK293 cell lysates as described under “Experimental Procedures.” The precipitated proteins were subjected to immunoblot analysis using antibody against importin-β, and the amount of the GST fusion protein was visualized by Western blotting using anti-GST antibody (lower panel). The results shown are representative of three separate experiments. C and D, HEK293 cells were transfected with the indicated GFP-PLD1 constructs, and the lysates were immunoprecipitated with antibody to importin-β and immunoblotted with antibody to GFP. The level of GFP-PLD1 and importin-β was immunoblotted using the indicated antibody. Relative binding capacity of CF- or FL-PLD1 with importin-β in the immune complex was normalized and quantified relative to the level of corresponding PLD1 or an endogenous importin-β protein in the lysate using densitometer analysis. Vec, vector. The equal amounts GST or GST-importin-β were incubated with in vitro translated FL-PLD1 or CF-PLD1 as described under “Experimental Procedures. The binding extent of GST-importin-β with FL-PLD1 or CF-PLD1 was quantitated as a relative level of input corresponding to the translated PLD1 using densitometer analysis. GST and GST-importin-β proteins used in E were stained with Coomassie Blue. These data are representative of results obtained from three experiments.

We have normalized and quantified the binding of CF-PLD1 or FL-PLD1 with importin-β relative to the level of ectopically expressed CF-PLD, FL-PLD1, or endogenous importin-β using densitometer analysis (Fig. 5, C and D). Furthermore, we also examined whether this interaction is direct using GST-importin-β and in vitro translated FL-PLD1 or CF-PLD1. Αs shown in Fig. 5E, importin-β directly interacted with FL-PLD1 or CF-PLD1. Especially, importin-β showed more strong binding with CF-PLD1 compared with FL-PLD1. These data are comparable to the result from immunoprecipitation. Taken together, these results suggest that the NLS motif of PLD1 associates with importin-β and, thus, mediates its nuclear translocation.

Nuclear PLD1 Mediates the Activation of Nuclear PKCα and ERK Signaling Pathways

There is general consensus over the fact that activation of nuclear PLD leads to increased levels of diacylglycerol (DG) in the nucleus, which stimulate classical protein kinase C (1). Thus, we examined whether nuclear localization of PLD1 may affect nuclear PKC or ERK signaling. As shown Fig. 6A, ectopic expression of wild type of FL-PLD1 in HEK293 cells increased the phosphorylation of both cytosolic and nuclear PKCα compared with that of vector-transfected cells. The phosphorylation of nuclear PKCα in ΝLΜ-PLD1-tranfected cells was suppressed compared with that of wild type PLD1-transfected cells PLD1, whereas NLM-PLD1 increased the phosphorylation of cytosolic PKCα, comparable to that of wild type PLD1. Moreover, wild type of PLD1 enhanced the phosphorylation of both cytosolic and nuclear ERK, whereas NLM-PLD1 induced the activation of only cytosolic ERK but not nuclear ERK. On the contrary, CF-PLD1 and NLM-CF-PLD1 did not affect phosphorylation of PKCα and ERK in both the cytosol and nucleus compared with those of vector-transfected cells (Fig. 6B). Because NLM-PLD1 has enzymatic activity similar to that of wild type PLD1 but is not present in the nucleus and CF-PLD1 has no PLD activity, it is suggested that enzymatic activity of PLD1 is responsible for the activation of PKCα and ERK in the nucleus and cytosol.

FIGURE 6.

Nuclear localization of PLD1 is correlated with the activation of nuclear PKCα and ERK. A and B, HEK293 cells were transfected with the indicated GFP-PLD1 constructs, and the lysates were fractionated into cytosol (C) and nucleus (N) and analyzed by immunoblot using the indicated antibodies. These data are representative of results obtained from three experiments. The level of p-PKCα, p-ERK relative to the level of PKCα, or ERK was quantitated using densitometer analysis. The values are the mean ± S.D. of the three independent experiments. p < 0.01 (*) and p < 0.01 (**) versus the vector cytosolic fraction, not significant (NS) versus the vector cytosolic fraction.

To further confirm whether nuclear PLD1 is required for the activation of nuclear PKCα and ERK, we examined the effect of knockdown of PLD1 on PDGF-induced phosphorylation of PKCα and ERK. As shown in Fig. 7A, PDGF increased the phosphorylation of PKCα and ERK in both the cytosol and nucleus in control siRNA-transfected HEK293 cells, whereas depletion of PLD1 suppressed PDGF-induced activation of PKCα and ERK in the nucleus as well as the cytosol. Furthermore, PLD1 inhibitor (VU0155069) abolished PDGF-induced phosphorylation of PKCα and ERK in both the cytosol and nucleus (Fig. 7B), suggesting that PLD activity is required for the activation of nuclear and cytosolic PKCα or ERK. We further tried to confirm that enzymatic activity of PLD1 is responsible for the activation of PKCα and ERK. As shown in Fig. 7C, the catalytically inactive form of PLD1 (KRM-PLD1; mutation of lysine to arginine in the catalytic region of PLD1) suppressed the activation of these kinases. Taken together, these data indicate that nuclear PLD1 mediates nuclear signaling of PKCα and ERK.

FIGURE 7.

Knockdown and inhibition of PLD1 disrupt the activation of nuclear PKCα and ERK. A, HEK293 cells were transfected with siRNA for control or PLD1, treated with or without PDGF (10 ng/ml) for 30 min, fractionated into cytosol (C) and nucleus (N), and analyzed by immunoblot using the indicated antibodies. B, HEK293 cells were pretreated with or without PLD1 inhibitor (VU0155069, 10 μm) for 30 min, stimulated with or without PDGF (10 ng/ml) for 30 min, fractionated into cytosol (C) and nucleus (N), and analyzed by immunoblot using the indicated antibodies. C, HEK293 cells were transfected with vector, WT-PLD1, or KRM-PLD1, fractionated into cytosol (C) and nucleus (N), and analyzed by immunoblot using the indicated antibodies. The level of p-PKCα, p-ERK relative to the level of PKCα, or ERK was quantitated using densitometer analysis. These data are representative of results obtained from three experiments. The values are the mean ± S.D. of the three independent experiments. *, p < 0.01 versus the vehicle cytosolic fraction for control siRNA.

DISCUSSION

To our knowledge this study is the first demonstration that nuclear localization of PLD1 mediates the activation of nuclear PKCα and ERK signaling pathways. In the literature there are reports of the presence and suggested functions of most of the phospholipid constituents of the cell nucleus. The PC cycle is an important signaling pathway for cell proliferation and differentiation. Despite the existence of nuclear PC metabolism and PC-PLD enzymatic activity in nucleus, mechanisms of the nuclear translocation of PC-PLD have not been clarified. Moreover, in most studies nuclear PLD has been identified only indirectly through overexpression of enzyme that could lead to a distribution of the PLD significantly different from endogenous enzyme. Using immunofluorescence and immunogold labeling, PLD1 has been localized to the nucleus of mammalian cells (GH3 cells), and upon Golgi apparatus collapse in response to brefeldin A, intranuclear staining of PLD1 increased significantly (24). The same group has reported that when the Golgi apparatus was disrupted by treatment with brefeldin A, PLD2 translocated to the nucleus, similarly to PLD1 (25). However, we could not observe any nuclear translocation of PLD1 by treatment with brefeldin A in HEK293 cells. These discrepancies might be attributable to differences in cell types and/or agonist-activated signaling pathway. The significance of the translocation of PLD isoform from the Golgi apparatus to the nucleus is at present unclear. Others have reported the intranuclear presence of PLD2 in renal carcinoma cells and have suggested this isoform might somehow be involved in tumorigenesis because its amount was very low in the nucleus of normal renal cells (26). Although we observed the presence of PLD2 in the nucleus from glioma cells, we could not find the putative NLS motif of PLD2 using the bioinformatic approach. It could be possible that PLD2 might contain nonclassical nuclear localization motif. We have been unable to identify conditions (e.g. growth factors, apoptosis, normal versus transformed phenotype) that might alter the relative distribution of endogenous PLD1 between the cytosolic and nuclear compartments. Although we have disclosed the importance of NLS and its interaction with β-importin in the nuclear localization of PLD1, further study for ligand or condition-induced nuclear translocation of endogenous PLD1 will strengthen the molecular mechanism of nucleocytoplasmic shuttling of PLD1 under physiological condition.

We have identified a functional and highly conserved NLS required for nuclear import of PLD1. We have recently reported that caspase-mediated cleavage of PLD1 during apoptosis results in production of NF-PLD1 and CF-PLD1. The NLS of PLD1 is located near the caspase cleavage site (Asp-545) (16). Interestingly, CF-PLD1 containing the functional NLS, but not NF-PLD1 lacking the NLS, is exclusively imported into the nucleus, whereas FL-PLD1 is partially imported into the nucleus. The extent of nuclear localization of CF-PLD1 and FL-PLD1 is different even though these two proteins have the same NLS motif. The amount of FL- or CF-PLD1 translocated into nucleus was correlated with its binding affinity with importin-β. Although we examined relative binding affinity of CF- or FL-PLD1 with importin-β using immunoprecipitation and an GST pulldown assay, future study using purified proteins is required for quantification or kinetics of these binding affinities. Thus, it is possible that CF-PLD1 may easily be exposed to NLS machinery such as importin-β compared with FL-PLD1 and, thus, exclusively localized in the nucleus. In addition, the NLS motif of PLD1 is required for its binding to importin-β.

It seems that CF-PLD1 showing no PLD activity may have a function in the nucleus regardless of the enzymatic activity of PLD. Although it is speculated that CF-PLD1 in the nucleus might play a role such as modulation of gene expression, at present we don't know the function for CF-PLD1 in the nucleus. In the future study we will investigate its function deeply using experiments such as microarray.

Recent findings showing that nuclear PLD1 in vascular smooth muscle cells is specifically stimulated by heterotrimeric G protein-coupled receptors but not by receptor tyrosine kinases favor a very specific role for nuclear phospholipid mediators generated by this enzyme, suggesting the emerging role of nuclear localization in PLD1 function, and more importantly, in pathological processes (13). Thus, it is possible that a nuclear PLD1 might regulate specific nuclear signaling pathways. An increase in PLD-generated nuclear DG mass has been reported in Madin-Darby canine kidney cells (10), HL60 human promyelocytic leukemia cells (28), and IIC9 fibroblasts (11). The profiles of the nuclear DG resembled the species profiles of PC but not of inositol lipid species (29). Interestingly, our finding that nuclear intact PLD1 mediates the activation of nuclear PKCα and ERK signaling pathways and suppression of its nuclear import abolished the activation of PKCα and ERK in the nucleus is in favor of specific role of nuclear PLD1 signaling pathway. Although CF-PLD1 is exclusively localized in the nucleus, it is not responsible for the activation of PKCα and ERK in the nucleus as well as cytosol, in which PLD activity is involved. Nuclear PKC isozymes are involved in the regulation of biological processes as important as cell proliferation and differentiation, gene expression, neoplastic transformation, and apoptosis (30). There is evidence for nuclear translocation of classical PKC isotypes that was mediated by DG produced along PLD pathway (31). Therefore, it has been proposed that nuclear DG is the attracting force that drives to the nucleus classical PKCs. Activated nuclear PKC has been shown to phosphorylate a number of proteins involved in cell division and appears to be critical for progression through the G₁/S (32) and G₂/M checkpoints of the cell cycle (33). The relocalization of ERK to the nucleus appears to be an important regulatory step for mitogen-induced gene expression and cell cycle reentry. Regulating the accessibility of ERK to the nucleus is a key signaling event by which cells may control the intensity and temporal activation of genes during cell growth and differentiation (34). Downstream signaling in the nucleus is another facet that requires in-depth investigation, because we know very little of the physiological targets of some nuclear lipid second messengers. At present there are little data regarding a possible role played by PA produced by PLD in the nucleus. In the cytoplasm PA has been implicated in the regulation of the actin cytoskeleton by inducing actin polymerization and formation of stress fibers. Moreover, PA can activate phosphatidylinositol 4-phosphate 5-kinase, increasing the synthesis of phosphatidylinositol 4,5-bisphosphate (35). Intriguingly, actin may be one of the most important constituents of the nuclear matrix (36), whereas phosphatidylinositol 4-phosphate 5-kinase is present in the speckle domains of the nucleus (11, 37), which have been implicated in pre-mRNA processing (27). Because type I phosphatidylinositol 4-phosphate 5-kinase activity is stimulated by PA (35), it is suggested that nuclear PLD1 might function in mRNA processing via regulation of phosphatidylinositol 4,5-bisphosphate synthesis. Additionally, because periodic S phase accumulation of DG from either PC-derived PA or phosphatidylinositol is essential for continued cell growth/division, nuclear PLD1-derived-PA might regulate the cell cycle (6).

Thus, in the future it will be of fundamental importance to investigate the role of PA generated within the nucleus. NLS of PLD1 and its interaction with importin-β provide new insights into nuclear localization of PLD1 and its nuclear signaling. Further clarification and extension of our proposal will come from the study of signal-dependent nuclear translocation of PLD1. In addition, future identification of the physiological roles of nuclear PLD1 or CF-PLD1 will shed light on nuclear PLD1-mediated signaling function.