- Email: [email protected]
Binding of Cdc42 to phospholipase D1 is important in neurite outgrowth of neural stem cells
BBRC Biochemical and Biophysical Research Communications 347 (2006) 594–600 www.elsevier.com/locate/ybbrc
Binding of Cdc42 to phospholipase D1 is important in neurite outgrowth of neural stem cells Mee-Sup Yoon a, Chan Ho Cho a, Ki Sung Lee b, Joong-Soo Han a
a,*
Department of Biochemistry and Molecular Biology, College of Medicine, Hanyang University, Seoul 133-791, Republic of Korea b The Research Center for Biomedicinal Resources, Paichai University, Taejon 302-735, Korea, Republic of Korea Received 23 May 2006 Available online 27 June 2006
Abstract We previously demonstrated that phospholipase D (PLD) expression and PLD activity are upregulated during neuronal differentiation. In the present study, employing neural stem cells from the brain cortex of E14 rat embryos, we investigated the role of Rho family GTPases in PLD activation and in neurite outgrowth of neural stem cells during differentiation. As neuronal differentiation progressed, the expression levels of Cdc42 and RhoA increased. Furthermore, Cdc42 and PLD1 were mainly localized in neurite, whereas RhoA was localized in cytosol. Co-immunoprecipitation revealed that Cdc42 was bound to PLD1 during differentiation, whereas RhoA was associated with PLD1 during both proliferation and differentiation. These results indicate that the association between Cdc42 and PLD1 is related to neuronal differentiation. To examine the effect of Cdc42 on PLD activation and neurite outgrowth, we transfected dominant negative Cdc42 (Cdc42N17) and constitutively active Cdc42 (Cdc42V12) into neural stem cells, respectively. Overexpression of Cdc42N17 decreased both PLD activity and neurite outgrowth, whereas co-transfection with Cdc42N17 and PLD1 restored them. On the other hand, Cdc42V12 increased both PLD activity and neurite outgrowth, suggesting that active state of Cdc42 is important in upregulation of PLD activity which is responsible for the increase of neurite outgrowth. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Phospholipase D1; Cdc42; RhoA; Rho family GTPases; Neural stem cells; Neurite outgrowth; Neuronal differentiation; Cortex
Neurite outgrowth is required to communicate with each other and with effector cells [1]. Leading the movement at the tips of the developing neurites, the growth cones extend by fusing membrane vesicles with their plasma membrane and also by constant growth and reorganization of the cytoskeleton below this plasma membrane, thereby pushing the leading edge forward [2]. However, much less is known about the mechanism of cytoskeletal dynamics in neurite outgrowth. Rho family GTPases have been implicated in regulating the actin cytoskeleton rearrangement [3]. Like other small GTPases, the Rho family GTPases serve as a molecular switch by converting from an inactive GDP-bound state to an active GTP-bound state and, once activated, they can interact with their specific effectors. Notably, Hall *
Corresponding author. Fax: +82 2 2294 6270. E-mail address: [email protected] (J.-S. Han).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.06.111
and colleagues have used microinjection technique to demonstrate that the Rho GTPases, including RhoA, Rac-1, and Cdc42, function coordinately to regulate cytoskeletal reorganization associated with cell motility [4]. In addition, recent reports suggest that Rho, Rac, and Cdc42 play a central role in dendritic development, and that differential activation of Rho-related GTPases contributes to the generation of morphological diversity in the developing cortex [5]. Rho family GTPases are also important modulators that control the activity of phospholipase D (PLD) [6]. PLD catalyzes the hydrolysis of phosphatidylcholine at their terminal phosphodiester bond to yield phosphatidic acid and choline. Phosphatidic acid is a second messenger involved in membrane remodeling events such as vesicle trafficking and cytoskeletal reorganization that are critical to cell growth [7]. To date, two distinct isoforms of mammalian PLD have been cloned and characterized: PLD1
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
is found throughout the cell, but particularly in perinuclear, Golgi, and heavy membrane fractions, whereas PLD2 is localized almost exclusively on the plasma membrane in the light membrane ‘‘lipid raft’’ fractions [8]. Both the isoforms require phosphatidylinositol-4,5-bisphosphate (PIP2) as a cofactor for their activities. PLD1 is activated by the ADP ribosylation factor (ARF)-, Ral-, and Rho-family GTPases, as well as protein kinase C-a [9]. In contrast, PLD2 activity is known to be constitutively active in vitro, and its activity is elevated by fatty acid, but unaffected by GTPases or PKC-a [7]. Among PLD1 activators, Rho family GTPases are direct activators, and transfection experiments have demonstrated that activated Rho, Rac, and Cdc42 can each activate PLD1 [6]. However, the distinct physiological functions of GTPases raise some questions about the specificity of PLD activation. We previously reported that PLD1 regulates neurite outgrowth by promoting neuronal differentiation via increase of synapsin I in rat cortex neural stem cell [10]. In the present study, we examined the role of Rho family GTPases on PLD activation during neuronal differentiation, especially focusing on neurite outgrowth, and found that Cdc42 increased PLD activation by binding to PLD1, and GTP–GDP state of Cdc42 regulated PLD1 activation in neural stem cells. We also demonstrated that PLD activation by Cdc42 increased neurite outgrowth, suggesting that PLD activity is closely related to neurite outgrowth. Materials and methods Materials. Coon’s modified Ham’s F-12 medium, human insulin, human transferrin, and penicillin/streptomycin solution were purchased from Sigma Chemical (St. Louis, MO, USA), [3H]-palmitate was from Du Pont—New England Nuclear (USA), and bFGF was from R&D systems (Minneapolis, MN, USA), and protein G–agarose was from Roche (Mannheim, Germany). Antibodies used were as follows: RhoA, Cdc42 monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and b-actin monoclonal antibodies (ABcam, Cambridge, MA). A polyclonal antibody that recognizes both PLD1 and PLD2 was provided by Dr. D.S. Min (Pusan National University, Korea). The constitutively active and dominant negative mutants of Cdc42 (Cdc42V12 and Cdc42N17) were kindly provided by Dr. Y.S. Juhnn (Seoul National University, Korea). L-a-PBt was from Avanti Polar Lipid Inc. (Alabaster, AL, USA), and the Silica Gel 60A plates for thin-layer chromatography (TLC) were purchased from Whatman (Clifton, NJ, USA). All other chemical agents were of analytical grade. Primary CNS precursor cultures. Sprague–Dawley time-pregnant rats (SD-rats) were purchased from Dae Han Biolink (Seoul, Korea). Embryonic brain cortices were dissected from embryonic day 14 rat. Cells were isolated by mechanical dissociation in Ca2+-/Mg2+-free Hanks’ balanced salt solution (CMF-HBSS; Invitrogen, Carlsbad, CA, USA), and transferred to 6 cm cell culture dishes precoated with polyornithine/fibronectin [poly-L-ornithine (15 lg/ml, Sigma, St. Louis, MO, USA) overnight at 37 °C incubator, followed by fibronectin (1 lg/ml, Sigma or Invitrogen, USA) for at least 2 h]. The cells were allowed to proliferate in the presence of 20 ng/ml basic fibroblast growth factor (bFGF) in serumfree defined medium (N2) for 5 days. The cells were allowed to continue proliferating in the presence of bFGF before induction of differentiation by withdrawal of the mitogen, bFGF. Immunocytochemistry. For immunocytochemical test, the neural stem cells were plated on a glass coverslip coated with fibronectin. Cells were
595
fixed in 4% paraformaldehyde/0.15% picric acid in phosphate-buffered saline (PBS) and were incubated overnight with primary antibodies at 4 °C. The following primary antibodies were used at the concentration given: Cdc42 monoclonal (1:50), RhoA monoclonal (1:50), b-tubulin type III (Tuj1) monoclonal (1:500) or polyclonal (1:2000) (both Babco, Richmond, CA, USA), and green fluorescent protein (GFP) monoclonal antibody 1:400 (Roche Molecular, Biochemical, Basel, Switzerland). For detection of primary antibodies, fluorescent-labeled (Cy3) secondary antibody (Jackson Immunoresearch Laboratories, West Grove, PA, USA) was used according to the manual of the manufacturer. Cells were mounted in VECTASHIELDÒ with DAPI (Vector laboratories, Burlingame, CA, USA) mounting medium for fluorescence and photographed using a fluorescent microscope (Nikon, Japan). Determination of PLD activity. PLD activity was determined as previously described by measuring [3H]-PEt produced via PLD catalyzed transphosphatidylation in [3H]-palmitic acid labeled cells [11]. Briefly, neural stem cells cultured on six-well plates were metabolically labeled with 1 lCi/ml of [3H]-palmitate for 24 h. The cells were then pretreated with 1% (v/v) ethanol for 15 min, and were quickly washed with ice-cold PBS and suspended in ice-cold methanol. Lipids were extracted according to the method of Bligh and Dyer [12], and [3H]-PEt was separated from other phospholipids by TLC on Silica G-60 plates, using a solvent system of ethyl acetate/iso-octane/acetic acid/water (130:20:30:100, v/v). The regions corresponding to the authentic PEt bands were identified with 0.002% (w/v) primulin in 80% (v/v) acetone, scraped and counted by a scintillation counter. Preparation of cytosol and membrane fractions. Neural stem cells were differentiated by withdrawal of bFGF for 4 days, scraped in ice-cold PBS, and harvested by microcentrifugation. The cells were then resuspended in a buffer solution [50 mM NaCl, 1 mM MgCl2, 2 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 lg/ml aprotinin, 10 lg/ ml leupeptin, and 50 mM Hepes, pH 7.5], disrupted by sonication, and centrifuged for 100,000g for 1 h. The supernatant was used as the cytosolic fraction, and the pellet (membrane fraction) was resuspended in lysis buffer containing 1% (v/v) Triton X-100. Immunoprecipitation. Cells were washed twice with ice-cold PBS and then lysed in washing buffer 1 (50 mM Tris–Cl, pH 7.5, 150 mM NaCl, 0.7 lg/ml pepstatin, 0.5% sodium deoxycholate, and 1% Nonidet P-40, and protease inhibitor mixture). The resulting cell lysates were spun at 15,000g for 10 min at 4 °C to pellet the unbroken cells. The supernatant was then precleared for 4 h with protein G–agarose at 4 °C with rocking. Equal protein aliquots of precleared cell lysates (1 mg) were incubated with the indicated antibodies (1 lg) for 1 h at 4 °C and captured with 30 ll of protein G–agarose beads for 4 h at 4 °C. The beads were collected by centrifugation, washed two times with washing buffer 1, two times with washing buffer 2 (50 mM Tris–HCl, pH 7.5, 500 mM NaCl, 0.05% sodium deoxycholate, and 1% Nonidet P-40), and one time with washing buffer 3 (50 mM Tris–HCl, pH 7.5, 0.05% sodium deoxycholate, and 1% Nonidet P-40), and resuspended in sample buffer. The immunoprecipitated proteins were separated by SDS–polyacrylamide gel electrophoresis and analyzed by Western blotting with anti-RhoA, Cdc42, or PLD1 antibodies. Western blot analysis. Cells were lysed in a solution containing 20 mM Tris–HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1% Triton X-100, 1 mM PMSF, and 1 mM Na3VO4. Proteins (20–30 lg) were resolved on SDS–polyacrylamide gels (10%) and then transferred to nitrocellulose membrane (Amersham Pharmacia Biotech, England). After blocking with 5% non-fat dried milk for 1 h, the membrane was incubated with primary antibodies. The blots were then incubated with horseradish peroxidase-conjugated secondary antibody (1:2000, New England Biolabs Inc., Beverly, MA, USA). Immunoreactive bands were detected by enhanced chemiluminescence (Amersham Pharmacia Biotech, England). Transfection of neural stem cells. Neural stem cells expanded in vitro with bFGF for 4–6 days were transfected by using Nucleofectorä Kit (Amaxa, Germany) as described by the manufacturer’s protocol. Transfection controls were carried out with the same amount of empty vector. One day after transfection, the cells were differentiated by bFGF withdrawal from the media and differentiated for an additional 4 days in vitro.
596
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
Reverse transcription-PCR (RT-PCR) for expression of plasmid DNA was used as an internal control to determine the transfection efficiency. RNA extraction and reverse transcription-PCR (RT-PCR) analysis. Total cellular RNA was prepared using TRI reagent (Molecular Reagent Center, Inc., Cincinnati, OH, USA), according to the recommendations of the manufacturer. Superscript kit (Invitrogen, USA) was used for the cDNA synthesis. The PCRs were carried out according to the standard protocols. Before the actual process, optimal MgCl2 concentrations and cycle numbers for linear amplification range were determined. Primer sequences (forward and backward) were as follows: GAPDH (5 0 -GGC ATTGCTCTCAATGACAA-3 0 ,5 0 -AGGGCCTCTCTCTTGCTCTC-3 0 ), Cdc42 (5 0 -CCgGGATCCATGCAGACAATTAAGTGTGTTGTT-3 0 , 5 0 -GCCGAATTCTTAGAATATACAGCACTTCCTTTT-3 0 ), and rPLD1 (5 0 -CGGTGGCGTTTGTGGGTGGGA-3 0 , 5 0 -GTCCTTGAAAAAGT TGCA-3 0 ). RT-PCR products were analyzed in an ethidium bromide containing agarose gel. Measurement of neurite outgrowth. The cells from randomly selected areas of at least five cultures from three independent experiments were photographed. Morphological characteristics were quantitated using Sigmascan Pro. (SPSS Inc., Chicago, IL, USA). Clusters of cells were excluded from the morphometric analysis. The length of primary neurite was defined as the distance from the soma to the tip of the longest branch. Statistical analysis. All experiments were performed at least three to five times, and data were analyzed using one-way ANOVA and considered to be significantly different at p < 0.05.
Results Changes of expressions and cellular distributions of Rho family GTPases during neuronal differentiation Neural precursors were expanded in serum-free medium in the presence of bFGF. After bFGF-proliferation for additional 4–6 days, over 95% of the cells were found to be immunoreactive for the intermediate filament nestin, a marker of neural precursor cells (data not shown), and less than 3% of the cells expressed neuronal (Tuj1) or astroglial (GFAP) markers. None of the nestin positive cells was positive for Tuj1 and less than 0.5% coexpressed GFAP. The similar proportion of the cells immunoreactive for neural precursors was observed in the passaged cultures, demonstrating selectiveness of the culture conditions for neural precursors. First, in order to examine whether Rho-related GTPases were expressed in the cortex during neuronal differentiation, we determined the expressions of RhoA, Cdc42, and Rac1 by Western blot analysis. We recently showed that PLD1 protein expression increased within 1 day and continued to increase by 660% until 4 days, and PLD activity was also increased during neuronal differentiation [10]. As shown in Fig. 1A, the expression of RhoA increased in a time-dependent manner, whereas the expression of Cdc42 increased rapidly within 1 day up to 224% and the expression level sustained thereafter. These data indicate that the increase of PLD activator proteins, Cdc42 and RhoA, is correlated with the increase of PLD activity. However, only a very small quantity of Rac1 was detected by Western blot analysis (data not shown). Next, in order to investigate the relationship between RhoA, Cdc42, and PLD1 during neuronal differentiation, we examined the location of PLD1,
Fig. 1. Differential expression (A) and cellular distribution (B) of Cdc42, RhoA, and PLD1 during neuronal differentiation. (A) Neural stem cells were differentiated for the indicated periods. Cells were lysed and then subjected to Western blotting by using anti-Cdc42, -RhoA or b-actin antibodies. The amount of protein samples used for PAGE was 30 lg per lane. (B) The cells differentiated for indicated periods were fractionated into cytosol and membrane fractions as described in Materials and methods. An equal amount of protein was analyzed with Western blotting by using anti-PLD1, Cdc42, and RhoA antibodies.
Cdc42, and RhoA during neuronal differentiation by Western blot analysis of cytosolic and membrane fractions (Fig. 1B). Although PLD1 protein expression increased in the both fractions, it was predominantly located in the membrane fraction, and increased. Concurrently, the expression level of Cdc42 in the membrane fraction increased mainly as the differentiation progressed. On the other hand, RhoA in the cytosolic fraction increased and only a small quantity was detected in the membrane fraction during neuronal differentiation. These results together indicate that PLD1 and Cdc42 co-locate in the neurite during neuronal differentiation. We employed immunocytochemical analysis to validate the location of Cdc42 and RhoA. As seen in Fig. 2A, during proliferation period of the neural stem cells with bFGF, Cdc42 was mainly located in the cytosol, whereas RhoA was confirmed to be in the membrane. After the cells were differentiated for 2 days by withdrawal of bFGF, Cdc42 was localized to the cell
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
597
Effects of Cdc42 on PLD activity and neurite outgrowth
Fig. 2. Immunocytochemical determination of localization of Cdc42 and RhoA in neural stem cells during proliferation (A) and neuronal differentiation (B). Neural stem cells were incubated with bFGF for proliferation or without bFGF for differentiation for 2 days.
membrane and neurite. Conversely, RhoA was predominantly located in the cytosol (Fig. 2B). These results confirm that PLD1 and Cdc42 are co-localized and may play a role in the neurite outgrowth during neuronal differentiation. Interaction of PLD with Cdc42 during neuronal differentiation To investigate whether PLD interacted with Cdc42 in the course of neuronal differentiation, co-immunoprecipitation experiments were performed. After treatment with or without bFGF for 4 days, the differentiated and undifferentiated cell lysates were immunoprecipitated with either anti-RhoA, -Cdc42 or -PLD antibodies, respectively, and the immunoprecipitates were probed with either anti-PLD, -Cdc42 or -RhoA antibody by Western blot analysis. As shown in Fig. 3A, Cdc42 was found to be associated with PLD1 during neuronal differentiation, but not during proliferative condition (Fig. 3C). However, RhoA expression increased and was associated with PLD1 in both differentiated and proliferative cells (Fig. 3B and C). Therefore, we ruled out RhoA as the regulator of PLD1 during neuronal differentiation. Moreover, to validate the association of PLD1 with Cdc42 in vivo, we analyzed PLD1 and Cdc42 immunocytochemically, and found that Cdc42 and PLD1 were co-localized in the neurite during neuronal differentiation after withdrawal of bFGF for 2 days (data not shown). These findings suggest that the association of PLD1 and Cdc42 is coupled with neurite outgrowth during neuronal differentiation and may play a role in the maintenance of the differentiated state.
Cdc42 is known to stimulate PLD activity, and PLD activation by Cdc42 is GTP-dependent [13]. In order to investigate the role of Cdc42 in PLD activation, we transfected dominant negative Cdc42 (Cdc42N17) and constitutively active Cdc42 (Cdc42V12) into neural stem cells, respectively, by electroporation. The transfection efficiency was verified by RT-PCR. Fig. 4A shows that Cdc42N17, Cdc42V12, and rPLD1 were successfully transfected by electroporation. One day after transfection, neural stem cells were differentiated by bFGF withdrawal and kept differentiating for another 2 days. When the neural stem cells were differentiated for 2 days by bFGF removal, PLD activity increased up to 157%, compared to non-differentiated cell (Fig. 4B). This increased PLD activity was remarkably attenuated by transfection of the cells with Cdc42N17, and PLD activity was restored by co-transfection of rPLD1 with Cdc42N17. In contrast, however, transfection of the cells with Cdc42V12 increased PLD activity by 29%, compared to vector transfected cells. Next, in order to examine the effect of Cdc42 on neurite outgrowth, either Cdc42N17, Cdc42N17/rPLD1, or Cdc42V12 was cotransfected into the neural stem cells along with EGFP-C1, and neurite length in GFP positive cells was subsequently measured during differentiation condition. Here, EGFPC1 was used as a marker for the transfected cells, because the transfection efficiency of co-transfected target plasmids with EGFP-C1 in hippocampal neuronal cells has been reported to be about 80% [14]. We could detect the transfected cells under a fluorescence microscope, because only the transfected cells turned green under fluorescence, and the neural stem cells were visible as red by immunostaining with Tuj1. Consistent with PLD activity, the active state of Cdc42 affected the neurite outgrowth (Fig. 4C). After 2 days of differentiation, transfection with Cdc42N17 decreased the neurite outgrowth by 30%, compared to vector-transfected cells, whereas transfection with Cdc42V12 increased it by 33%, compared to vector transfected cells. Moreover, the inhibition of neurite length by Cdc42N17 was recovered by the transfection of rPLD1. Taken together, these results suggest that active state of Cdc42 is important for PLD activation and neurite outgrowth, indicating that activated Cdc42 induces neurite outgrowth by increasing PLD activity. Discussion In this study, we investigated the role of interaction between PLD1 and Cdc42 in regulating neurite outgrowth. The expressions of Cdc42 and PLD1 were found to increase in neural stem cell as the neurite outgrowth progressed. Concurrently, Cdc42 and PLD were localized mainly in the neurite during neuronal differentiation. Three independent experiments confirmed that the binding of Cdc42 to PLD1 is related to increase of neurite outgrowth. First, PLD1 was associated with Cdc42 during neuronal
598
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
Fig. 3. Association between PLD1 and Cdc42 or RhoA during neuronal differentiation (A,B) or proliferation (C). (A,B) Neural stem cells were differentiated for the indicated periods. Cells were lysed and immunoprecipitates were prepared using anti-PLD, anti-Cdc42 or -RhoA antibodies, respectively. The immunoprecipitates were then subjected to Western blotting with anti-PLD, -Cdc42 or -RhoA antibodies. (C) After bFGF-expansion, cells were lysed and immunoprecipitates were prepared using anti-Cdc42 or RhoA antibodies. The immunoprecipitates were then subjected to Western blotting with anti-PLD and Cdc42 or RhoA antibody.
differentiation, but not during proliferation. Second, transfection with Cdc42N17 decreased both PLD activity and neurite outgrowth which was restored by co-transfection of rPLD1 with Cdc42N17. Third, transfection with Cdc42V12 increased both PLD activity and neurite outgrowth. These results led us to conclude that the association between PLD1 and Cdc42 regulates neurite outgrowth through modulation of PLD activity. A recent study shows that the elaboration of neurite in cortical neuron can be regulated by neurotrophins [15]. Since neurotrophins signal through receptor tyrosine kinases and Rho GTPases have been implicated in receptor tyrosine kinase [16], it is tempting to speculate that the phenotypic consequences of neurotrophins on neurite development may be mediated via the activation of Rho GTPases. Furthermore, Threagill and Bobb reported that neurotrophins affect neurite outgrowth in cultured cortical neurons which is regulated by Rho GTPase [5]. Our experiments showed that withdrawal of bFGF induced the expression of BDNF, NT3, and NT4/5 (data not shown). Therefore, it is quite likely that withdrawal of bFGF induced the
release of neurotrophins and Rho GTPases activation in the present neuronal differentiation system. It has been reported that activation of Cdc42 induces actin-rich surface protrusions, called filopodia [4,17]. Kozma et al. reported that activation of Cdc42 in neuroblastoma cell line, N1E-115, promotes the formation of filopodia along neurite extension and this process is blocked by introducing dominant negative Cdc42 into these cells [18]. In PC12 cells, dominant negative constructs of Cdc42 completely block NGF-induced neurite outgrowth, supporting the role of Cdc42 in neurite outgrowth [19]. However, neurite outgrowth requires not only Cdc42 activation, but also their appropriate localization to sites where neuritis is formed and extended [20]. In the present study, Cdc42 was localized in the membrane, especially in the neurite, during neuronal differentiation, suggesting that the localization of Cdc42 affects neurite outgrowth in neural stem cells. Cdc42 has been recently reported to be involved in actin assembly pathways, known as WASP and Arp2/3 [21], and Rohatgi et al. suggested that GTPbound Cdc42 associates with WASP regulatory domain,
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
599
Fig. 4. Effects of transfection with Cdc42N17 or Cdc42V12 on PLD activity (B) and neurite outgrowth (C). Neural stem cells were transfected with 5 lg of either vector, Cdc42N17, Cdc42V12, or rPLD1 using electroporation (Amaxa, Germany). (A) RT-PCR analyses for transfection efficiency. RNA was isolated from neural stem cells and RT-PCR analyses were performed for rPLD1 and Cdc42. (B) Neural stem cells differentiated for the indicated periods were labeled with [3H]-palmitic acid. PLD activity was measured by formation of [13H]-PEt by the method described in Materials and methods. (C) Comparison of neurite outgrowth between Cdc42N17- and Cdc42V12-transfected cells. Cells were co-transfected with 5 lg of empty parental control vector, Cdc42N17, or Cdc42V12 plus 1 lg of EGFP-C1 vectors. One day after transfection, transfected neural stem cells were differentiated by bFGF withdrawal. After differentiation for 2 days, the GFP-positive cells were observed by fluorescence microscopy, and the cells were stained with Tuj1 to visualize neurite extensions. The neurite growth of cells from randomly selected areas of three independent experiments was photographed.
thus freeing the C-terminal domain to bind to Arp2/3, and resulting in efficient polymerization of actin [22]. Another Cdc42-binding adaptor protein, IRS-58, has also been implicated as facilitiating neurite outgrowth through actin cytoskeleton reorganization [23]. These results, therefore, suggest that interaction between Cdc42 and regulators allows the former in an appropriate place in cell and stimulates its ability to induce actin polymerization. Rho family GTPases were originally identified as activators of PLD1 and have been shown by direct physical interaction to bind PLD1 [6]. Rho family proteins in human and rat appear to bind to PLD1 sequence between amino acids 984 and 1000 [24], and Walker et al. demonstrated that binding of Cdc42 to PLD1 and subsequent activation are GTP-dependent [13] which is affected by serine 124 residue in Cdc42 insert helix [25]. In the present study, we demonstrated that overexpression of Cdc42N17 decreased PLD activity and neurite outgrowth, whereas overexpression of
Cdc42V12 increased both PLD activity and neurite outgrowth, implying that the active state of Cdc42 is important in the binding to PLD1 and then PLD activation is responsible for neurite outgrowth. Furthermore, RhoA was found to be associated with PLD1 not only during neuronal differentiation, but also during proliferation, and RhoA was localized in the cytosol during differentiation, suggesting that association between RhoA with PLD1 was not involved in neuronal differentiation. PLD activation has been suggested to regulate neurite outgrowth during neuronal differentiation [10]. In PC12 cells, PLD is activated and associated with PKCa/bII which are the potent activators of PLD during NGF-induced neuronal differentiation, and the association between PLD and PKCa/bII has been suggested to be involved in PLD activation and neuronal differentiation [26]. Moreover, Cai et al. recently reported that the impaired capacity of neurite outgrowth in familial Alzheimer’s disease mutant neurons
600
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
was ameliorated by introducing PLD1 into these cells, indicating that PLD1 may provide a therapeutic target for rescuing compromised neuronal function in Alzheimer’s disease [27]. In conclusion, we have shown that association between Cdc42 and PLD1 is important for PLD activation and neurite outgrowth during neuronal differentiation. These characteristics suggest that PLD1 is one of the important regulators in neuronal differentiation. Further studies to elucidate PLD downstream signaling events leading to neuronal differentiation would broaden our understanding of molecular mechanism underlying neuronal differentiation. Acknowledgments This work was supported mainly by a Grant from the Korea Research Foundation (015-E00070, 2004) and partially by the KOSEF and MOST through the Research Center for Biomedicinal Resources (Bio-Med RRC), PaiChai University. References [1] K. Aoki, T. Nakamura, M. Matsuda, Spatio-temporal regulation of Rac1 and Cdc42 activity during nerve growth factor-induced neurite outgrowth in PC12 cells, J. Biol. Chem. 279 (2004) 713–719. [2] P. Alberts, R. Rudge, T. Irinopoulou, L. Danglot, C. GauthierRouviere, T. Galli, Cdc42 and actin control polarized expression of TI-VAMP vesicles to neuronal growth cones and their fusion with the plasma membrane, Mol. Biol. Cell 17 (2006) 1194–1203. [3] A. Hall, Rho GTPases and actin cytoskeleton, Science 279 (1998) 509–514. [4] C.D. Nobes, A. Hall, Rho, Rac, and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia, Cell 81 (1995) 53–62. [5] R. Threadgill, K. Bobb, Regulation of dendritic growth and remodeling by Rho, Rac, and Cdc42, Neuron 19 (1997) 625–634. [6] D. Powner, M.J. Wakelam, The regulation of phospholipase D by inositol phospholipids and small GTPases, FEBS Lett. 531 (2002) 62–64. [7] J.H. Exton, Regulation of phospholipase D, FEBS Lett. 531 (2002) 58–61. [8] Z. Freyberg, D. Sweeney, A. Siddhanta, S. Bourgoin, M. Frohman, D. Shields, Intracellular localization of phospholipase D1 in mammalian cells, Mol. Biol. Cell 12 (2001) 943–955. [9] M. Liscovitch, M. Czarny, G. Fiucci, X. Tang, Phospholipase D: molecular and cell biology of a novel gene family, Biochem. J. 345 (2000) 401–415. [10] M.S. Yoon, C. Yon, S.Y. Park, D.Y. Oh, A.H. Han, D.S. Kim, J.S. Han, Role of phospholipase D1 in neurite outgrowth of neural stem cells, Biochem. Biophys. Res. Commun. 329 (2005) 804–811.
[11] J.W. Oh, E.Y. Kim, B.S. Koo, H.B. Lee, K.S. Lee, Y.S. Kim, J.S. Han, Der f 2 activates phospholipase D in human T lymphocytes from Dermatophagoides farinae specific allergic individuals: involvement of protein kinase C-alpha, Exp. Mol. Med. 36 (2004) 486–492. [12] K.M. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959) 911–917. [13] S.J. Walker, W.J. Wu, R.A. Cerione, H.A. Brown, Activation of phospholipase D1 by Cdc42 requires the Rho insert region, J. Biol. Chem. 275 (2000) 15665–15668. [14] W.L. Kuo, M. Abe, J. Rhee, E.M. Eves, S.A. McCarthy, M. Yan, D.J. Templeton, M. McMahon, M.R. Rosner, Raf, but not MEK or ERK, is sufficient for differentiation of hippocampal neuronal cells, Mol. Cell. Biol. 16 (1996) 1458–1470. [15] A.K. McAllister, D.C. Lo, L.C. Katz, Neurotrophins regulate dendritic growth in developing visual cortex, Neuron 15 (1995) 791–803. [16] T. Hunter, Oncoprotein networks, Cell 88 (1997) 333–346. [17] E. Giniger, How do Rho family GTPases direct axon growth and guidance? A proposal relating signaling pathways to growth cone mechanics, Differentiation 70 (2002) 385–396. [18] R. Kozma, S. Sarner, S. Ahmed, L. Lim, Rho family GTPases and neuronal growth cone remodelling: relationship between increased complexity induced by Cdc42Hs, Rac1, and acetylcholine and collapse induced by RhoA and lysophosphatidic acid, Mol. Cell Biol. 17 (1997) 1201–1211. [19] R.H. Daniels, P.S. Hall, G.M. Bokoch, Membrane targeting of p21activated kinase1 (PAK1) induces neurite outgrowth from PC12 cells, EMBO J. 17 (1998) 754–764. [20] M. Neigishi, H. Katoh, Rho family GTPase as key regulators for neuronal network formation, J. Biochem. 132 (2002) 157–166. [21] R.D. Mullins, T.D. Pollard, Rho-family GTPases require the Arp2/3 complex to stimulate actin polymerization in Acanthamoeba extracts, Curr. Biol. 9 (1999) 405–415. [22] R. Rohatgi, L. Ma, H. Miki, M. Lopez, T. Kirchhausen, T. Takenawa, M.W. Kirschner, The interaction between N-WASP and the Arp2/3 complex links Cdc42-dependent signals to actin assembly, Cell (1999) 221–231. [23] S. Govind, R. Kozma, C. Monfries, L. Lim, S. Ahmed, Cdc42Hs facilitates cytoskeletal reorganization and neurite outgrowth by localizing the 58-kD insulin receptor substrate to filamentous actin, J. Cell Biol. 152 (2001) 579–594. [24] S. Cai, J.H. Exton, Determination of interaction sites of phospholipase D1 and RhoA, Biochem. J. 355 (2001) 779–785. [25] S.J. Walker, H.A. Brown, Specificity of Rho insert-mediated activation of phospholipase D1, J. Biol. Chem. 277 (2002) 26260– 26267. [26] D.S. Min, B.H. Ahn, D.J. Rhie, S.H. Yoon, S.J. Hahn, M.S. Kim, Y.H. Jo, Expression and regulation of phospholipase D during neuronal differentiation of PC12 cells, Neuropharmacology 41 (2001) 384–391. [27] D. Cai, M. Zhong, R. Wang, W.J. Netzer, D. Shields, H. Zheng, S.S. Sisodia, D.A. Foster, F.S. Gorelick, H. Xu, P. Greengard, Phospholipase D1 corrects impaired b-APP trafficking and neurite outgrowth in familial Alzheimer’s disease-linked presenilin-1 mutant neurons, Proc. Natl. Acad. Sci. USA 103 (2006) 1936–1940.
Binding of Cdc42 to phospholipase D1 is important in neurite outgrowth of neural stem cells Mee-Sup Yoon a, Chan Ho Cho a, Ki Sung Lee b, Joong-Soo Han a
a,*
Department of Biochemistry and Molecular Biology, College of Medicine, Hanyang University, Seoul 133-791, Republic of Korea b The Research Center for Biomedicinal Resources, Paichai University, Taejon 302-735, Korea, Republic of Korea Received 23 May 2006 Available online 27 June 2006
Abstract We previously demonstrated that phospholipase D (PLD) expression and PLD activity are upregulated during neuronal differentiation. In the present study, employing neural stem cells from the brain cortex of E14 rat embryos, we investigated the role of Rho family GTPases in PLD activation and in neurite outgrowth of neural stem cells during differentiation. As neuronal differentiation progressed, the expression levels of Cdc42 and RhoA increased. Furthermore, Cdc42 and PLD1 were mainly localized in neurite, whereas RhoA was localized in cytosol. Co-immunoprecipitation revealed that Cdc42 was bound to PLD1 during differentiation, whereas RhoA was associated with PLD1 during both proliferation and differentiation. These results indicate that the association between Cdc42 and PLD1 is related to neuronal differentiation. To examine the effect of Cdc42 on PLD activation and neurite outgrowth, we transfected dominant negative Cdc42 (Cdc42N17) and constitutively active Cdc42 (Cdc42V12) into neural stem cells, respectively. Overexpression of Cdc42N17 decreased both PLD activity and neurite outgrowth, whereas co-transfection with Cdc42N17 and PLD1 restored them. On the other hand, Cdc42V12 increased both PLD activity and neurite outgrowth, suggesting that active state of Cdc42 is important in upregulation of PLD activity which is responsible for the increase of neurite outgrowth. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Phospholipase D1; Cdc42; RhoA; Rho family GTPases; Neural stem cells; Neurite outgrowth; Neuronal differentiation; Cortex
Neurite outgrowth is required to communicate with each other and with effector cells [1]. Leading the movement at the tips of the developing neurites, the growth cones extend by fusing membrane vesicles with their plasma membrane and also by constant growth and reorganization of the cytoskeleton below this plasma membrane, thereby pushing the leading edge forward [2]. However, much less is known about the mechanism of cytoskeletal dynamics in neurite outgrowth. Rho family GTPases have been implicated in regulating the actin cytoskeleton rearrangement [3]. Like other small GTPases, the Rho family GTPases serve as a molecular switch by converting from an inactive GDP-bound state to an active GTP-bound state and, once activated, they can interact with their specific effectors. Notably, Hall *
Corresponding author. Fax: +82 2 2294 6270. E-mail address: [email protected] (J.-S. Han).
0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.06.111
and colleagues have used microinjection technique to demonstrate that the Rho GTPases, including RhoA, Rac-1, and Cdc42, function coordinately to regulate cytoskeletal reorganization associated with cell motility [4]. In addition, recent reports suggest that Rho, Rac, and Cdc42 play a central role in dendritic development, and that differential activation of Rho-related GTPases contributes to the generation of morphological diversity in the developing cortex [5]. Rho family GTPases are also important modulators that control the activity of phospholipase D (PLD) [6]. PLD catalyzes the hydrolysis of phosphatidylcholine at their terminal phosphodiester bond to yield phosphatidic acid and choline. Phosphatidic acid is a second messenger involved in membrane remodeling events such as vesicle trafficking and cytoskeletal reorganization that are critical to cell growth [7]. To date, two distinct isoforms of mammalian PLD have been cloned and characterized: PLD1
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
is found throughout the cell, but particularly in perinuclear, Golgi, and heavy membrane fractions, whereas PLD2 is localized almost exclusively on the plasma membrane in the light membrane ‘‘lipid raft’’ fractions [8]. Both the isoforms require phosphatidylinositol-4,5-bisphosphate (PIP2) as a cofactor for their activities. PLD1 is activated by the ADP ribosylation factor (ARF)-, Ral-, and Rho-family GTPases, as well as protein kinase C-a [9]. In contrast, PLD2 activity is known to be constitutively active in vitro, and its activity is elevated by fatty acid, but unaffected by GTPases or PKC-a [7]. Among PLD1 activators, Rho family GTPases are direct activators, and transfection experiments have demonstrated that activated Rho, Rac, and Cdc42 can each activate PLD1 [6]. However, the distinct physiological functions of GTPases raise some questions about the specificity of PLD activation. We previously reported that PLD1 regulates neurite outgrowth by promoting neuronal differentiation via increase of synapsin I in rat cortex neural stem cell [10]. In the present study, we examined the role of Rho family GTPases on PLD activation during neuronal differentiation, especially focusing on neurite outgrowth, and found that Cdc42 increased PLD activation by binding to PLD1, and GTP–GDP state of Cdc42 regulated PLD1 activation in neural stem cells. We also demonstrated that PLD activation by Cdc42 increased neurite outgrowth, suggesting that PLD activity is closely related to neurite outgrowth. Materials and methods Materials. Coon’s modified Ham’s F-12 medium, human insulin, human transferrin, and penicillin/streptomycin solution were purchased from Sigma Chemical (St. Louis, MO, USA), [3H]-palmitate was from Du Pont—New England Nuclear (USA), and bFGF was from R&D systems (Minneapolis, MN, USA), and protein G–agarose was from Roche (Mannheim, Germany). Antibodies used were as follows: RhoA, Cdc42 monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and b-actin monoclonal antibodies (ABcam, Cambridge, MA). A polyclonal antibody that recognizes both PLD1 and PLD2 was provided by Dr. D.S. Min (Pusan National University, Korea). The constitutively active and dominant negative mutants of Cdc42 (Cdc42V12 and Cdc42N17) were kindly provided by Dr. Y.S. Juhnn (Seoul National University, Korea). L-a-PBt was from Avanti Polar Lipid Inc. (Alabaster, AL, USA), and the Silica Gel 60A plates for thin-layer chromatography (TLC) were purchased from Whatman (Clifton, NJ, USA). All other chemical agents were of analytical grade. Primary CNS precursor cultures. Sprague–Dawley time-pregnant rats (SD-rats) were purchased from Dae Han Biolink (Seoul, Korea). Embryonic brain cortices were dissected from embryonic day 14 rat. Cells were isolated by mechanical dissociation in Ca2+-/Mg2+-free Hanks’ balanced salt solution (CMF-HBSS; Invitrogen, Carlsbad, CA, USA), and transferred to 6 cm cell culture dishes precoated with polyornithine/fibronectin [poly-L-ornithine (15 lg/ml, Sigma, St. Louis, MO, USA) overnight at 37 °C incubator, followed by fibronectin (1 lg/ml, Sigma or Invitrogen, USA) for at least 2 h]. The cells were allowed to proliferate in the presence of 20 ng/ml basic fibroblast growth factor (bFGF) in serumfree defined medium (N2) for 5 days. The cells were allowed to continue proliferating in the presence of bFGF before induction of differentiation by withdrawal of the mitogen, bFGF. Immunocytochemistry. For immunocytochemical test, the neural stem cells were plated on a glass coverslip coated with fibronectin. Cells were
595
fixed in 4% paraformaldehyde/0.15% picric acid in phosphate-buffered saline (PBS) and were incubated overnight with primary antibodies at 4 °C. The following primary antibodies were used at the concentration given: Cdc42 monoclonal (1:50), RhoA monoclonal (1:50), b-tubulin type III (Tuj1) monoclonal (1:500) or polyclonal (1:2000) (both Babco, Richmond, CA, USA), and green fluorescent protein (GFP) monoclonal antibody 1:400 (Roche Molecular, Biochemical, Basel, Switzerland). For detection of primary antibodies, fluorescent-labeled (Cy3) secondary antibody (Jackson Immunoresearch Laboratories, West Grove, PA, USA) was used according to the manual of the manufacturer. Cells were mounted in VECTASHIELDÒ with DAPI (Vector laboratories, Burlingame, CA, USA) mounting medium for fluorescence and photographed using a fluorescent microscope (Nikon, Japan). Determination of PLD activity. PLD activity was determined as previously described by measuring [3H]-PEt produced via PLD catalyzed transphosphatidylation in [3H]-palmitic acid labeled cells [11]. Briefly, neural stem cells cultured on six-well plates were metabolically labeled with 1 lCi/ml of [3H]-palmitate for 24 h. The cells were then pretreated with 1% (v/v) ethanol for 15 min, and were quickly washed with ice-cold PBS and suspended in ice-cold methanol. Lipids were extracted according to the method of Bligh and Dyer [12], and [3H]-PEt was separated from other phospholipids by TLC on Silica G-60 plates, using a solvent system of ethyl acetate/iso-octane/acetic acid/water (130:20:30:100, v/v). The regions corresponding to the authentic PEt bands were identified with 0.002% (w/v) primulin in 80% (v/v) acetone, scraped and counted by a scintillation counter. Preparation of cytosol and membrane fractions. Neural stem cells were differentiated by withdrawal of bFGF for 4 days, scraped in ice-cold PBS, and harvested by microcentrifugation. The cells were then resuspended in a buffer solution [50 mM NaCl, 1 mM MgCl2, 2 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 lg/ml aprotinin, 10 lg/ ml leupeptin, and 50 mM Hepes, pH 7.5], disrupted by sonication, and centrifuged for 100,000g for 1 h. The supernatant was used as the cytosolic fraction, and the pellet (membrane fraction) was resuspended in lysis buffer containing 1% (v/v) Triton X-100. Immunoprecipitation. Cells were washed twice with ice-cold PBS and then lysed in washing buffer 1 (50 mM Tris–Cl, pH 7.5, 150 mM NaCl, 0.7 lg/ml pepstatin, 0.5% sodium deoxycholate, and 1% Nonidet P-40, and protease inhibitor mixture). The resulting cell lysates were spun at 15,000g for 10 min at 4 °C to pellet the unbroken cells. The supernatant was then precleared for 4 h with protein G–agarose at 4 °C with rocking. Equal protein aliquots of precleared cell lysates (1 mg) were incubated with the indicated antibodies (1 lg) for 1 h at 4 °C and captured with 30 ll of protein G–agarose beads for 4 h at 4 °C. The beads were collected by centrifugation, washed two times with washing buffer 1, two times with washing buffer 2 (50 mM Tris–HCl, pH 7.5, 500 mM NaCl, 0.05% sodium deoxycholate, and 1% Nonidet P-40), and one time with washing buffer 3 (50 mM Tris–HCl, pH 7.5, 0.05% sodium deoxycholate, and 1% Nonidet P-40), and resuspended in sample buffer. The immunoprecipitated proteins were separated by SDS–polyacrylamide gel electrophoresis and analyzed by Western blotting with anti-RhoA, Cdc42, or PLD1 antibodies. Western blot analysis. Cells were lysed in a solution containing 20 mM Tris–HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1% Triton X-100, 1 mM PMSF, and 1 mM Na3VO4. Proteins (20–30 lg) were resolved on SDS–polyacrylamide gels (10%) and then transferred to nitrocellulose membrane (Amersham Pharmacia Biotech, England). After blocking with 5% non-fat dried milk for 1 h, the membrane was incubated with primary antibodies. The blots were then incubated with horseradish peroxidase-conjugated secondary antibody (1:2000, New England Biolabs Inc., Beverly, MA, USA). Immunoreactive bands were detected by enhanced chemiluminescence (Amersham Pharmacia Biotech, England). Transfection of neural stem cells. Neural stem cells expanded in vitro with bFGF for 4–6 days were transfected by using Nucleofectorä Kit (Amaxa, Germany) as described by the manufacturer’s protocol. Transfection controls were carried out with the same amount of empty vector. One day after transfection, the cells were differentiated by bFGF withdrawal from the media and differentiated for an additional 4 days in vitro.
596
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
Reverse transcription-PCR (RT-PCR) for expression of plasmid DNA was used as an internal control to determine the transfection efficiency. RNA extraction and reverse transcription-PCR (RT-PCR) analysis. Total cellular RNA was prepared using TRI reagent (Molecular Reagent Center, Inc., Cincinnati, OH, USA), according to the recommendations of the manufacturer. Superscript kit (Invitrogen, USA) was used for the cDNA synthesis. The PCRs were carried out according to the standard protocols. Before the actual process, optimal MgCl2 concentrations and cycle numbers for linear amplification range were determined. Primer sequences (forward and backward) were as follows: GAPDH (5 0 -GGC ATTGCTCTCAATGACAA-3 0 ,5 0 -AGGGCCTCTCTCTTGCTCTC-3 0 ), Cdc42 (5 0 -CCgGGATCCATGCAGACAATTAAGTGTGTTGTT-3 0 , 5 0 -GCCGAATTCTTAGAATATACAGCACTTCCTTTT-3 0 ), and rPLD1 (5 0 -CGGTGGCGTTTGTGGGTGGGA-3 0 , 5 0 -GTCCTTGAAAAAGT TGCA-3 0 ). RT-PCR products were analyzed in an ethidium bromide containing agarose gel. Measurement of neurite outgrowth. The cells from randomly selected areas of at least five cultures from three independent experiments were photographed. Morphological characteristics were quantitated using Sigmascan Pro. (SPSS Inc., Chicago, IL, USA). Clusters of cells were excluded from the morphometric analysis. The length of primary neurite was defined as the distance from the soma to the tip of the longest branch. Statistical analysis. All experiments were performed at least three to five times, and data were analyzed using one-way ANOVA and considered to be significantly different at p < 0.05.
Results Changes of expressions and cellular distributions of Rho family GTPases during neuronal differentiation Neural precursors were expanded in serum-free medium in the presence of bFGF. After bFGF-proliferation for additional 4–6 days, over 95% of the cells were found to be immunoreactive for the intermediate filament nestin, a marker of neural precursor cells (data not shown), and less than 3% of the cells expressed neuronal (Tuj1) or astroglial (GFAP) markers. None of the nestin positive cells was positive for Tuj1 and less than 0.5% coexpressed GFAP. The similar proportion of the cells immunoreactive for neural precursors was observed in the passaged cultures, demonstrating selectiveness of the culture conditions for neural precursors. First, in order to examine whether Rho-related GTPases were expressed in the cortex during neuronal differentiation, we determined the expressions of RhoA, Cdc42, and Rac1 by Western blot analysis. We recently showed that PLD1 protein expression increased within 1 day and continued to increase by 660% until 4 days, and PLD activity was also increased during neuronal differentiation [10]. As shown in Fig. 1A, the expression of RhoA increased in a time-dependent manner, whereas the expression of Cdc42 increased rapidly within 1 day up to 224% and the expression level sustained thereafter. These data indicate that the increase of PLD activator proteins, Cdc42 and RhoA, is correlated with the increase of PLD activity. However, only a very small quantity of Rac1 was detected by Western blot analysis (data not shown). Next, in order to investigate the relationship between RhoA, Cdc42, and PLD1 during neuronal differentiation, we examined the location of PLD1,
Fig. 1. Differential expression (A) and cellular distribution (B) of Cdc42, RhoA, and PLD1 during neuronal differentiation. (A) Neural stem cells were differentiated for the indicated periods. Cells were lysed and then subjected to Western blotting by using anti-Cdc42, -RhoA or b-actin antibodies. The amount of protein samples used for PAGE was 30 lg per lane. (B) The cells differentiated for indicated periods were fractionated into cytosol and membrane fractions as described in Materials and methods. An equal amount of protein was analyzed with Western blotting by using anti-PLD1, Cdc42, and RhoA antibodies.
Cdc42, and RhoA during neuronal differentiation by Western blot analysis of cytosolic and membrane fractions (Fig. 1B). Although PLD1 protein expression increased in the both fractions, it was predominantly located in the membrane fraction, and increased. Concurrently, the expression level of Cdc42 in the membrane fraction increased mainly as the differentiation progressed. On the other hand, RhoA in the cytosolic fraction increased and only a small quantity was detected in the membrane fraction during neuronal differentiation. These results together indicate that PLD1 and Cdc42 co-locate in the neurite during neuronal differentiation. We employed immunocytochemical analysis to validate the location of Cdc42 and RhoA. As seen in Fig. 2A, during proliferation period of the neural stem cells with bFGF, Cdc42 was mainly located in the cytosol, whereas RhoA was confirmed to be in the membrane. After the cells were differentiated for 2 days by withdrawal of bFGF, Cdc42 was localized to the cell
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
597
Effects of Cdc42 on PLD activity and neurite outgrowth
Fig. 2. Immunocytochemical determination of localization of Cdc42 and RhoA in neural stem cells during proliferation (A) and neuronal differentiation (B). Neural stem cells were incubated with bFGF for proliferation or without bFGF for differentiation for 2 days.
membrane and neurite. Conversely, RhoA was predominantly located in the cytosol (Fig. 2B). These results confirm that PLD1 and Cdc42 are co-localized and may play a role in the neurite outgrowth during neuronal differentiation. Interaction of PLD with Cdc42 during neuronal differentiation To investigate whether PLD interacted with Cdc42 in the course of neuronal differentiation, co-immunoprecipitation experiments were performed. After treatment with or without bFGF for 4 days, the differentiated and undifferentiated cell lysates were immunoprecipitated with either anti-RhoA, -Cdc42 or -PLD antibodies, respectively, and the immunoprecipitates were probed with either anti-PLD, -Cdc42 or -RhoA antibody by Western blot analysis. As shown in Fig. 3A, Cdc42 was found to be associated with PLD1 during neuronal differentiation, but not during proliferative condition (Fig. 3C). However, RhoA expression increased and was associated with PLD1 in both differentiated and proliferative cells (Fig. 3B and C). Therefore, we ruled out RhoA as the regulator of PLD1 during neuronal differentiation. Moreover, to validate the association of PLD1 with Cdc42 in vivo, we analyzed PLD1 and Cdc42 immunocytochemically, and found that Cdc42 and PLD1 were co-localized in the neurite during neuronal differentiation after withdrawal of bFGF for 2 days (data not shown). These findings suggest that the association of PLD1 and Cdc42 is coupled with neurite outgrowth during neuronal differentiation and may play a role in the maintenance of the differentiated state.
Cdc42 is known to stimulate PLD activity, and PLD activation by Cdc42 is GTP-dependent [13]. In order to investigate the role of Cdc42 in PLD activation, we transfected dominant negative Cdc42 (Cdc42N17) and constitutively active Cdc42 (Cdc42V12) into neural stem cells, respectively, by electroporation. The transfection efficiency was verified by RT-PCR. Fig. 4A shows that Cdc42N17, Cdc42V12, and rPLD1 were successfully transfected by electroporation. One day after transfection, neural stem cells were differentiated by bFGF withdrawal and kept differentiating for another 2 days. When the neural stem cells were differentiated for 2 days by bFGF removal, PLD activity increased up to 157%, compared to non-differentiated cell (Fig. 4B). This increased PLD activity was remarkably attenuated by transfection of the cells with Cdc42N17, and PLD activity was restored by co-transfection of rPLD1 with Cdc42N17. In contrast, however, transfection of the cells with Cdc42V12 increased PLD activity by 29%, compared to vector transfected cells. Next, in order to examine the effect of Cdc42 on neurite outgrowth, either Cdc42N17, Cdc42N17/rPLD1, or Cdc42V12 was cotransfected into the neural stem cells along with EGFP-C1, and neurite length in GFP positive cells was subsequently measured during differentiation condition. Here, EGFPC1 was used as a marker for the transfected cells, because the transfection efficiency of co-transfected target plasmids with EGFP-C1 in hippocampal neuronal cells has been reported to be about 80% [14]. We could detect the transfected cells under a fluorescence microscope, because only the transfected cells turned green under fluorescence, and the neural stem cells were visible as red by immunostaining with Tuj1. Consistent with PLD activity, the active state of Cdc42 affected the neurite outgrowth (Fig. 4C). After 2 days of differentiation, transfection with Cdc42N17 decreased the neurite outgrowth by 30%, compared to vector-transfected cells, whereas transfection with Cdc42V12 increased it by 33%, compared to vector transfected cells. Moreover, the inhibition of neurite length by Cdc42N17 was recovered by the transfection of rPLD1. Taken together, these results suggest that active state of Cdc42 is important for PLD activation and neurite outgrowth, indicating that activated Cdc42 induces neurite outgrowth by increasing PLD activity. Discussion In this study, we investigated the role of interaction between PLD1 and Cdc42 in regulating neurite outgrowth. The expressions of Cdc42 and PLD1 were found to increase in neural stem cell as the neurite outgrowth progressed. Concurrently, Cdc42 and PLD were localized mainly in the neurite during neuronal differentiation. Three independent experiments confirmed that the binding of Cdc42 to PLD1 is related to increase of neurite outgrowth. First, PLD1 was associated with Cdc42 during neuronal
598
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
Fig. 3. Association between PLD1 and Cdc42 or RhoA during neuronal differentiation (A,B) or proliferation (C). (A,B) Neural stem cells were differentiated for the indicated periods. Cells were lysed and immunoprecipitates were prepared using anti-PLD, anti-Cdc42 or -RhoA antibodies, respectively. The immunoprecipitates were then subjected to Western blotting with anti-PLD, -Cdc42 or -RhoA antibodies. (C) After bFGF-expansion, cells were lysed and immunoprecipitates were prepared using anti-Cdc42 or RhoA antibodies. The immunoprecipitates were then subjected to Western blotting with anti-PLD and Cdc42 or RhoA antibody.
differentiation, but not during proliferation. Second, transfection with Cdc42N17 decreased both PLD activity and neurite outgrowth which was restored by co-transfection of rPLD1 with Cdc42N17. Third, transfection with Cdc42V12 increased both PLD activity and neurite outgrowth. These results led us to conclude that the association between PLD1 and Cdc42 regulates neurite outgrowth through modulation of PLD activity. A recent study shows that the elaboration of neurite in cortical neuron can be regulated by neurotrophins [15]. Since neurotrophins signal through receptor tyrosine kinases and Rho GTPases have been implicated in receptor tyrosine kinase [16], it is tempting to speculate that the phenotypic consequences of neurotrophins on neurite development may be mediated via the activation of Rho GTPases. Furthermore, Threagill and Bobb reported that neurotrophins affect neurite outgrowth in cultured cortical neurons which is regulated by Rho GTPase [5]. Our experiments showed that withdrawal of bFGF induced the expression of BDNF, NT3, and NT4/5 (data not shown). Therefore, it is quite likely that withdrawal of bFGF induced the
release of neurotrophins and Rho GTPases activation in the present neuronal differentiation system. It has been reported that activation of Cdc42 induces actin-rich surface protrusions, called filopodia [4,17]. Kozma et al. reported that activation of Cdc42 in neuroblastoma cell line, N1E-115, promotes the formation of filopodia along neurite extension and this process is blocked by introducing dominant negative Cdc42 into these cells [18]. In PC12 cells, dominant negative constructs of Cdc42 completely block NGF-induced neurite outgrowth, supporting the role of Cdc42 in neurite outgrowth [19]. However, neurite outgrowth requires not only Cdc42 activation, but also their appropriate localization to sites where neuritis is formed and extended [20]. In the present study, Cdc42 was localized in the membrane, especially in the neurite, during neuronal differentiation, suggesting that the localization of Cdc42 affects neurite outgrowth in neural stem cells. Cdc42 has been recently reported to be involved in actin assembly pathways, known as WASP and Arp2/3 [21], and Rohatgi et al. suggested that GTPbound Cdc42 associates with WASP regulatory domain,
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
599
Fig. 4. Effects of transfection with Cdc42N17 or Cdc42V12 on PLD activity (B) and neurite outgrowth (C). Neural stem cells were transfected with 5 lg of either vector, Cdc42N17, Cdc42V12, or rPLD1 using electroporation (Amaxa, Germany). (A) RT-PCR analyses for transfection efficiency. RNA was isolated from neural stem cells and RT-PCR analyses were performed for rPLD1 and Cdc42. (B) Neural stem cells differentiated for the indicated periods were labeled with [3H]-palmitic acid. PLD activity was measured by formation of [13H]-PEt by the method described in Materials and methods. (C) Comparison of neurite outgrowth between Cdc42N17- and Cdc42V12-transfected cells. Cells were co-transfected with 5 lg of empty parental control vector, Cdc42N17, or Cdc42V12 plus 1 lg of EGFP-C1 vectors. One day after transfection, transfected neural stem cells were differentiated by bFGF withdrawal. After differentiation for 2 days, the GFP-positive cells were observed by fluorescence microscopy, and the cells were stained with Tuj1 to visualize neurite extensions. The neurite growth of cells from randomly selected areas of three independent experiments was photographed.
thus freeing the C-terminal domain to bind to Arp2/3, and resulting in efficient polymerization of actin [22]. Another Cdc42-binding adaptor protein, IRS-58, has also been implicated as facilitiating neurite outgrowth through actin cytoskeleton reorganization [23]. These results, therefore, suggest that interaction between Cdc42 and regulators allows the former in an appropriate place in cell and stimulates its ability to induce actin polymerization. Rho family GTPases were originally identified as activators of PLD1 and have been shown by direct physical interaction to bind PLD1 [6]. Rho family proteins in human and rat appear to bind to PLD1 sequence between amino acids 984 and 1000 [24], and Walker et al. demonstrated that binding of Cdc42 to PLD1 and subsequent activation are GTP-dependent [13] which is affected by serine 124 residue in Cdc42 insert helix [25]. In the present study, we demonstrated that overexpression of Cdc42N17 decreased PLD activity and neurite outgrowth, whereas overexpression of
Cdc42V12 increased both PLD activity and neurite outgrowth, implying that the active state of Cdc42 is important in the binding to PLD1 and then PLD activation is responsible for neurite outgrowth. Furthermore, RhoA was found to be associated with PLD1 not only during neuronal differentiation, but also during proliferation, and RhoA was localized in the cytosol during differentiation, suggesting that association between RhoA with PLD1 was not involved in neuronal differentiation. PLD activation has been suggested to regulate neurite outgrowth during neuronal differentiation [10]. In PC12 cells, PLD is activated and associated with PKCa/bII which are the potent activators of PLD during NGF-induced neuronal differentiation, and the association between PLD and PKCa/bII has been suggested to be involved in PLD activation and neuronal differentiation [26]. Moreover, Cai et al. recently reported that the impaired capacity of neurite outgrowth in familial Alzheimer’s disease mutant neurons
600
M.-S. Yoon et al. / Biochemical and Biophysical Research Communications 347 (2006) 594–600
was ameliorated by introducing PLD1 into these cells, indicating that PLD1 may provide a therapeutic target for rescuing compromised neuronal function in Alzheimer’s disease [27]. In conclusion, we have shown that association between Cdc42 and PLD1 is important for PLD activation and neurite outgrowth during neuronal differentiation. These characteristics suggest that PLD1 is one of the important regulators in neuronal differentiation. Further studies to elucidate PLD downstream signaling events leading to neuronal differentiation would broaden our understanding of molecular mechanism underlying neuronal differentiation. Acknowledgments This work was supported mainly by a Grant from the Korea Research Foundation (015-E00070, 2004) and partially by the KOSEF and MOST through the Research Center for Biomedicinal Resources (Bio-Med RRC), PaiChai University. References [1] K. Aoki, T. Nakamura, M. Matsuda, Spatio-temporal regulation of Rac1 and Cdc42 activity during nerve growth factor-induced neurite outgrowth in PC12 cells, J. Biol. Chem. 279 (2004) 713–719. [2] P. Alberts, R. Rudge, T. Irinopoulou, L. Danglot, C. GauthierRouviere, T. Galli, Cdc42 and actin control polarized expression of TI-VAMP vesicles to neuronal growth cones and their fusion with the plasma membrane, Mol. Biol. Cell 17 (2006) 1194–1203. [3] A. Hall, Rho GTPases and actin cytoskeleton, Science 279 (1998) 509–514. [4] C.D. Nobes, A. Hall, Rho, Rac, and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia, Cell 81 (1995) 53–62. [5] R. Threadgill, K. Bobb, Regulation of dendritic growth and remodeling by Rho, Rac, and Cdc42, Neuron 19 (1997) 625–634. [6] D. Powner, M.J. Wakelam, The regulation of phospholipase D by inositol phospholipids and small GTPases, FEBS Lett. 531 (2002) 62–64. [7] J.H. Exton, Regulation of phospholipase D, FEBS Lett. 531 (2002) 58–61. [8] Z. Freyberg, D. Sweeney, A. Siddhanta, S. Bourgoin, M. Frohman, D. Shields, Intracellular localization of phospholipase D1 in mammalian cells, Mol. Biol. Cell 12 (2001) 943–955. [9] M. Liscovitch, M. Czarny, G. Fiucci, X. Tang, Phospholipase D: molecular and cell biology of a novel gene family, Biochem. J. 345 (2000) 401–415. [10] M.S. Yoon, C. Yon, S.Y. Park, D.Y. Oh, A.H. Han, D.S. Kim, J.S. Han, Role of phospholipase D1 in neurite outgrowth of neural stem cells, Biochem. Biophys. Res. Commun. 329 (2005) 804–811.
[11] J.W. Oh, E.Y. Kim, B.S. Koo, H.B. Lee, K.S. Lee, Y.S. Kim, J.S. Han, Der f 2 activates phospholipase D in human T lymphocytes from Dermatophagoides farinae specific allergic individuals: involvement of protein kinase C-alpha, Exp. Mol. Med. 36 (2004) 486–492. [12] K.M. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959) 911–917. [13] S.J. Walker, W.J. Wu, R.A. Cerione, H.A. Brown, Activation of phospholipase D1 by Cdc42 requires the Rho insert region, J. Biol. Chem. 275 (2000) 15665–15668. [14] W.L. Kuo, M. Abe, J. Rhee, E.M. Eves, S.A. McCarthy, M. Yan, D.J. Templeton, M. McMahon, M.R. Rosner, Raf, but not MEK or ERK, is sufficient for differentiation of hippocampal neuronal cells, Mol. Cell. Biol. 16 (1996) 1458–1470. [15] A.K. McAllister, D.C. Lo, L.C. Katz, Neurotrophins regulate dendritic growth in developing visual cortex, Neuron 15 (1995) 791–803. [16] T. Hunter, Oncoprotein networks, Cell 88 (1997) 333–346. [17] E. Giniger, How do Rho family GTPases direct axon growth and guidance? A proposal relating signaling pathways to growth cone mechanics, Differentiation 70 (2002) 385–396. [18] R. Kozma, S. Sarner, S. Ahmed, L. Lim, Rho family GTPases and neuronal growth cone remodelling: relationship between increased complexity induced by Cdc42Hs, Rac1, and acetylcholine and collapse induced by RhoA and lysophosphatidic acid, Mol. Cell Biol. 17 (1997) 1201–1211. [19] R.H. Daniels, P.S. Hall, G.M. Bokoch, Membrane targeting of p21activated kinase1 (PAK1) induces neurite outgrowth from PC12 cells, EMBO J. 17 (1998) 754–764. [20] M. Neigishi, H. Katoh, Rho family GTPase as key regulators for neuronal network formation, J. Biochem. 132 (2002) 157–166. [21] R.D. Mullins, T.D. Pollard, Rho-family GTPases require the Arp2/3 complex to stimulate actin polymerization in Acanthamoeba extracts, Curr. Biol. 9 (1999) 405–415. [22] R. Rohatgi, L. Ma, H. Miki, M. Lopez, T. Kirchhausen, T. Takenawa, M.W. Kirschner, The interaction between N-WASP and the Arp2/3 complex links Cdc42-dependent signals to actin assembly, Cell (1999) 221–231. [23] S. Govind, R. Kozma, C. Monfries, L. Lim, S. Ahmed, Cdc42Hs facilitates cytoskeletal reorganization and neurite outgrowth by localizing the 58-kD insulin receptor substrate to filamentous actin, J. Cell Biol. 152 (2001) 579–594. [24] S. Cai, J.H. Exton, Determination of interaction sites of phospholipase D1 and RhoA, Biochem. J. 355 (2001) 779–785. [25] S.J. Walker, H.A. Brown, Specificity of Rho insert-mediated activation of phospholipase D1, J. Biol. Chem. 277 (2002) 26260– 26267. [26] D.S. Min, B.H. Ahn, D.J. Rhie, S.H. Yoon, S.J. Hahn, M.S. Kim, Y.H. Jo, Expression and regulation of phospholipase D during neuronal differentiation of PC12 cells, Neuropharmacology 41 (2001) 384–391. [27] D. Cai, M. Zhong, R. Wang, W.J. Netzer, D. Shields, H. Zheng, S.S. Sisodia, D.A. Foster, F.S. Gorelick, H. Xu, P. Greengard, Phospholipase D1 corrects impaired b-APP trafficking and neurite outgrowth in familial Alzheimer’s disease-linked presenilin-1 mutant neurons, Proc. Natl. Acad. Sci. USA 103 (2006) 1936–1940.