Title: LIM Kinase and Slingshot Are Critical for Neurite Extension
Abstract: Cofilin and its closely related protein, actin-depolymerizing factor (ADF), are key regulators of actin cytoskeleton dynamics that have been implicated in growth cone motility and neurite extension. Cofilin/ADF are inactivated by LIM kinase (LIMK)-catalyzed phosphorylation and reactivated by Slingshot (SSH)-catalyzed dephosphorylation. Here we examined the roles of cofilin/ADF, LIMKs (LIMK1 and LIMK2), and SSHs (SSH1 and SSH2) in nerve growth factor (NGF)-induced neurite extension. Knockdown of cofilin/ADF by RNA interference almost completely inhibited NGF-induced neurite extension from PC12 cells, and double knockdown of SSH1/SSH2 significantly suppressed both NGF-induced cofilin/ADF dephosphorylation and neurite extension from PC12 cells, thus indicating that cofilin/ADF and their activating phosphatases SSH1/SSH2 are critical for neurite extension. Interestingly, NGF stimulated the activities of both LIMK1 and LIMK2 in PC12 cells, and suppression of LIMK1/LIMK2 expression or activity significantly reduced NGF-induced neurite extension from PC12 cells or chick dorsal root ganglion (DRG) neurons. Inhibition of LIMK1/LIMK2 activity reduced actin filament assembly in the peripheral region of the growth cone of chick DRG neurons. These results suggest that proper regulation of cofilin/ADF activities through control of phosphorylation by LIMKs and SSHs is critical for neurite extension and that LIMKs regulate actin filament assembly at the tip of the growth cone. Cofilin and its closely related protein, actin-depolymerizing factor (ADF), are key regulators of actin cytoskeleton dynamics that have been implicated in growth cone motility and neurite extension. Cofilin/ADF are inactivated by LIM kinase (LIMK)-catalyzed phosphorylation and reactivated by Slingshot (SSH)-catalyzed dephosphorylation. Here we examined the roles of cofilin/ADF, LIMKs (LIMK1 and LIMK2), and SSHs (SSH1 and SSH2) in nerve growth factor (NGF)-induced neurite extension. Knockdown of cofilin/ADF by RNA interference almost completely inhibited NGF-induced neurite extension from PC12 cells, and double knockdown of SSH1/SSH2 significantly suppressed both NGF-induced cofilin/ADF dephosphorylation and neurite extension from PC12 cells, thus indicating that cofilin/ADF and their activating phosphatases SSH1/SSH2 are critical for neurite extension. Interestingly, NGF stimulated the activities of both LIMK1 and LIMK2 in PC12 cells, and suppression of LIMK1/LIMK2 expression or activity significantly reduced NGF-induced neurite extension from PC12 cells or chick dorsal root ganglion (DRG) neurons. Inhibition of LIMK1/LIMK2 activity reduced actin filament assembly in the peripheral region of the growth cone of chick DRG neurons. These results suggest that proper regulation of cofilin/ADF activities through control of phosphorylation by LIMKs and SSHs is critical for neurite extension and that LIMKs regulate actin filament assembly at the tip of the growth cone. The regulation of actin cytoskeleton dynamics plays a fundamental role in cell shape change, motility, and migration in response to stimuli. In neurons, actin filaments accumulate at the distal tip of the growth cone in the growing neurite, and actin filament dynamics and reorganization are essential for controlling growth cone motility and morphology and determining the direction and speed of neurite extension (1Luo L. Nat. Rev. Neurosci. 2000; 1: 173-180Crossref PubMed Scopus (823) Google Scholar, 2Song H.J. Poo M.M. Nat. Cell Biol. 2001; 3: E81-E88Crossref PubMed Scopus (322) Google Scholar, 3Meyer G. Feldman E.L. J. Neurochem. 2002; 83: 490-503Crossref PubMed Scopus (151) Google Scholar, 4Dent E.W. Gertler F.B. Neuron. 2003; 40: 209-227Abstract Full Text Full Text PDF PubMed Scopus (696) Google Scholar). Cofilin and its closely related protein, actin depolymerizing factor (ADF), 2The abbreviations used are: ADF, actin depolymerizing factor; chLIMK1, chick LIM kinase 1; chLIMK2, chick LIM kinase 2; DRG, dorsal root ganglion; LIMK, LIM kinase; NGF, nerve growth factor; P-ADF, Ser-3-phosphorylated ADF; P-cofilin, Ser-3-phosphorylated cofilin; siRNA, small interfering RNA; SSH, Slingshot; YFP, yellow fluorescent protein; RV, reverse. 2The abbreviations used are: ADF, actin depolymerizing factor; chLIMK1, chick LIM kinase 1; chLIMK2, chick LIM kinase 2; DRG, dorsal root ganglion; LIMK, LIM kinase; NGF, nerve growth factor; P-ADF, Ser-3-phosphorylated ADF; P-cofilin, Ser-3-phosphorylated cofilin; siRNA, small interfering RNA; SSH, Slingshot; YFP, yellow fluorescent protein; RV, reverse. are key mediators of actin filament dynamics that act by stimulating the depolymerization and severing of actin filaments (5Bamburg J.R. Annu. Rev. Cell Dev. Biol. 1999; 15: 185-230Crossref PubMed Scopus (830) Google Scholar). The activities of cofilin and ADF are inhibited by phosphorylation at Ser-3 by LIM kinases (LIMKs, composed of LIMK1 and LIMK2) (6Arber S. Barbayannis F.A. Hanser H. Schneider C. Stanyon C.A. Bernard O. Caroni P. Nature. 1998; 393: 805-809Crossref PubMed Scopus (1140) Google Scholar, 7Yang N. Higuchi O. Ohashi K. Nagata K. Wada A. Kangawa K. Nishida E. Mizuno K. Nature. 1998; 393: 809-812Crossref PubMed Scopus (1038) Google Scholar) and TES kinases (TESKs, composed of TESK1 and TESK2) (8Toshima J. Toshima J.Y. Amano T. Yang N. Narumiya S. Mizuno K. Mol. Biol. Cell. 2001; 12: 1131-1145Crossref PubMed Scopus (218) Google Scholar); the inactive Ser-3-phosphorylated cofilin and ADF (P-cofilin/P-ADF) are reactivated by dephosphorylation by Slingshot (SSH) family protein phosphatases (SSH1, SSH2, and SSH3) (9Niwa R. Nagata-Ohashi K. Takeichi M. Mizuno K. Uemura T. Cell. 2002; 108: 233-246Abstract Full Text Full Text PDF PubMed Scopus (530) Google Scholar, 10Ohta Y. Kousaka K. Nagata-Ohashi K. Ohashi K. Muramoto A. Shima Y. Niwa R. Uemura T. Mizuno K. Genes Cells. 2003; 8: 811-824Crossref PubMed Scopus (95) Google Scholar) and chronophin (a haloacid dehalogenase) (11Gohla A. Birkenfeldm J. Bokoch G.M. Nat. Cell Biol. 2005; 7: 21-29Crossref PubMed Scopus (248) Google Scholar). Because cofilin/ADF and their upstream regulators, LIMKs and SSHs, are abundant in neuronal growth cones (12Bamburg J.R. Bray D. J. Cell Biol. 1987; 105: 2817-2825Crossref PubMed Scopus (155) Google Scholar, 13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar, 14Rosso S. Bollati F. Bisbal M. Peretti D. Sumi T. Nakamura T. Quiroga S. Ferreira A. Caceres A. Mol. Biol. Cell. 2004; 15: 3433-3449Crossref PubMed Scopus (111) Google Scholar, 15Gehler S. Shaw A.E. Sarmiere P.D. Bamburg J.R. Letourneau P.C. J. Neurosci. 2004; 24: 10741-10749Crossref PubMed Scopus (101) Google Scholar), these proteins have been implicated in the control of neurite extension and guidance through regulating actin filament dynamics. Neurotrophins are known to regulate neurite outgrowth and guidance (2Song H.J. Poo M.M. Nat. Cell Biol. 2001; 3: E81-E88Crossref PubMed Scopus (322) Google Scholar). Rat pheochromocytoma PC12 cells have been often used as a model system to investigate nerve growth factor (NGF)-induced neurite outgrowth. NGF induces cofilin/ADF dephosphorylation in PC12 cells (16Meberg P.J. Ono S. Minamide L.S. Takahashi M. Bamburg J.R. Cell Motil. Cytoskeleton. 1998; 39: 172-190Crossref PubMed Scopus (216) Google Scholar), and overexpression of cofilin/ADF or SSH1 enhances neurite extension from PC12 cells and primary cultured neurons, such as chick dorsal root ganglion (DRG) or rat cortical neurons (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar, 17Meberg P.J. Bamburg J.R. J. Neurosci. 2000; 20: 2459-2469Crossref PubMed Google Scholar). In contrast, overexpression of LIMK1 in neurons suppresses growth cone motility and extension (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar). In addition, the growth cone collapse induced by semaphorin-3A (a repulsive guidance molecule) or Nogo-66 (a myelin-associated inhibitor of axon regeneration) requires transient activation of LIMK1 and cofilin phosphorylation in chick DRG neurons (18Aizawa H. Wakatsuki S. Ishii A. Moriyama K. Sasaki Y. Ohashi K. Sekine-Aizawa Y. Sehara-Fujisawa A. Mizuno K. Goshima Y. Yahara I. Nat. Neurosci. 2001; 4: 367-373Crossref PubMed Scopus (292) Google Scholar, 19Hsieh S.H.K. Ferraro G.B. Fournier A.E. J. Neurosci. 2006; 26: 1006-1015Crossref PubMed Scopus (117) Google Scholar). These results suggest that LIMK1 acts as a negative regulator of neurite out-growth by inhibiting cofilin/ADF activity. However, recent studies have suggested that LIMK1 has a seemingly opposite function on neurite outgrowth: neurite extension from hippocampal neurons was enhanced by LIMK1 expression and suppressed by blockade of LIMK1 activation (14Rosso S. Bollati F. Bisbal M. Peretti D. Sumi T. Nakamura T. Quiroga S. Ferreira A. Caceres A. Mol. Biol. Cell. 2004; 15: 3433-3449Crossref PubMed Scopus (111) Google Scholar, 20Lee-Hoeflich S.T. Causing C.G. Podkowa M. Zhao X. Wrana J.L. Attisano L. EMBO J. 2004; 23: 4792-4801Crossref PubMed Scopus (175) Google Scholar, 21Yang E.J. Yoon J.H. Min D.S. Chung K.C. J. Biol. Chem. 2004; 279: 8903-8910Abstract Full Text Full Text PDF PubMed Scopus (47) Google Scholar, 22Tursun B. Schluter A. Peters M.A. Viehweger B. Ostendorff H.P. Soosairajah J. Drung A. Bossenz M. Johnsen S.A. Schweizer M. Bernard O. Bach I. Genes Dev. 2005; 19: 2307-2319Crossref PubMed Scopus (86) Google Scholar). Thus, further studies are required to understand the role of LIMK1 in neurite extension. In this study, we examined the roles of cofilin/ADF and its phosphoregulation in neurite extension by knocking down the expression of cofilin/ADF, LIMK1/LIMK2, and SSH1/SSH2 using small interfering RNAs (siRNAs). Knockdown of cofilin/ADF markedly blocked NGF-induced neurite extension of PC12 cells, and knockdown of SSH1/SSH2 suppressed NGF-induced cofilin/ADF dephosphorylation and neurite extension, indicating that SSH1/SSH2-mediated cofilin/ADF dephosphorylation is crucial for neurite extension. LIMK1 and LIMK2 were also activated after NGF stimulation of PC12 cells, and knockdown of LIMK1/LIMK2 significantly suppressed NGF-induced neurite extension from PC12 cells and chick DRG neurons. Our results indicate that both LIMK1/LIMK2-mediated phosphorylation and SSH1/SSH2-mediated dephosphorylation of cofilin/ADF are important for neurite extension. Materials—K252a, U73122, and wortmannin were purchased from Calbiochem (La Jolla, CA). Latrunculin A was from Molecular Probes (Eugene, OR). Rabbit polyclonal antibodies against rat/mouse ADF and SSH2 were generated against the C-terminal peptide of rat/mouse ADF and SSH2, respectively. Rabbit polyclonal antibodies against cofilin, P-cofilin, LIMK1 (C10), and SSH1 were prepared as described previously (8Toshima J. Toshima J.Y. Amano T. Yang N. Narumiya S. Mizuno K. Mol. Biol. Cell. 2001; 12: 1131-1145Crossref PubMed Scopus (218) Google Scholar, 23Okano I. Hiraoka J. Otera H. Nunoue K. Ohashi K. Iwashita S. Hirai M. Mizuno K. J. Biol. Chem. 1995; 270: 31321-31330Abstract Full Text Full Text PDF PubMed Scopus (169) Google Scholar, 24Kaji N. Ohashi K. Shuin M. Niwa R. Uemura T. Mizuno K. J. Biol. Chem. 2003; 273: 33450-33455Abstract Full Text Full Text PDF Scopus (84) Google Scholar). Rabbit polyclonal antibody against LIMK2 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). S3 and RV peptides were designed and synthesized as described previously (18Aizawa H. Wakatsuki S. Ishii A. Moriyama K. Sasaki Y. Ohashi K. Sekine-Aizawa Y. Sehara-Fujisawa A. Mizuno K. Goshima Y. Yahara I. Nat. Neurosci. 2001; 4: 367-373Crossref PubMed Scopus (292) Google Scholar, 25Nishita M. Aizawa H. Mizuno K. Mol. Cell. Biol. 2002; 22: 774-783Crossref PubMed Scopus (115) Google Scholar). Plasmids—The siRNA-targeting constructs were generated using pSUPER or pSUPER.retro.puro vectors (OligoEngine, Seattle, WA), as described previously (26Brummelkamp T.R. Bernards R. Agami R. Science. 2002; 296: 550-553Crossref PubMed Scopus (3928) Google Scholar). The 19-base targeting sequences were as follows: 5′-GACTTGCGTAGCCTTAAGA-3′ (rat LIMK1), 5′-GGACAAGAAGCTGAATCTG-3′ (rat LIMK2), 5′-GAGGAGCTGTCCCGATGAC-3′ (rat SSH1), 5′-TGCGTCAAACTTAGAGGAC-3′ (rat SSH2), 5′-GCACGAATTACAAGCTAAC-3′ (rat cofilin), 5′-GCACGAGTATCAAGCAAAT-3′ (rat ADF), 5′-GGAGCTGATCCGCTTTGAT-3′ (chick LIMK1), and 5′-CTGCCTAATCAAGTTGGAT-3′ (chick LIMK2). We also used the second siRNA sequences (termed siRNA2) targeting rat and chick LIMKs as follows: 5′-GAAGGACTACTGGGCCCGC-3′ (rat LIMK1), 5′-GTGAAAGAGGTCAACCGGA-3′ (rat LIMK2), 5′-GTTCATCGGAGTGCTTTAC-3′ (chick LIMK1), and 5′-GGACAAGAAGCTCAATCTC-3′ (chick LIMK2). As a control, we used a non-targeting sequence, 5′-TCTTCCCCCAAGAAAGATA-3′, which does not exist in the rat or chick genome. Expression plasmids coding for N-terminally Myc-tagged chick LIMK1 (chLIMK1) and chLIMK2 and C-terminally Myc-tagged chick cofilin and ADF were constructed by inserting their full-length cDNAs into pMyc-C1 or pcDNA3.1/Myc-His(+) mammalian expression vector containing the Myc epitope tag (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar, 27Ohashi K. Toshima J. Tajinda K. Nakamura T. Mizuno K. J. Biochem. (Tokyo). 1994; 111: 636-642Crossref Scopus (36) Google Scholar). Expression plasmid (pEYFP-C1) coding for yellow fluorescent protein (YFP) was purchased from Clontech. Cell Culture and Electroporation—PC12 cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 15% fetal bovine serum. Cells were plated in poly-l-lysine-coated dishes. For the RNA interference experiments, cells were mixed with pSUPER constructs in Opti-minimum essential medium and electroporated at 320 V and 975 μF using a Gene Pulser II (Bio-Rad). After recovery, the cells were selected by incubation with puromycin (3 μg/ml) for 2 days and then used for further analysis. Chick DRG explants were cultured as described previously (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar). For siRNA experiments for chick DRG neurons, we used the chick Neuron Nucleofector Kit (Amaxa, Gaithersburg, MD), according to the manufacture’s instructions. Briefly, chick DRGs were dissected from E7-8 chick embryos and digested with 0.25% trypsin for 10 min at 37 °C. Dissociated cells were centrifuged and resuspended in nucleofection solution, mixed with 3 μg of pSUPER plasmid and 1 μg of YFP plasmid and electroporated using the fixed program (G-13). Electroporated cells were transferred to dishes and incubated for 1 h at 37°C in 5% CO2 to remove non-neuronal cells. Medium containing non-adherent cells was collected, centrifuged, and suspended in Ham’s F-12 medium containing 10% fetal bovine serum, N2 supplement and 20 ng/ml NGF (Wako, Osaka, Japan), and then plated in poly-l-lysine-coated dishes. After 48 h of incubation, the neurons were resuspended in Ham’s F-12 medium containing 4% bovine serum albumin, N2 supplement, and 20 ng/ml NGF, replated in poly-l-lysine- and laminin-coated glass-bottom culture dishes, and further incubated for 12 h before time-lapse observation. CHO-K1 cells were maintained in minimum essential medium-α supplemented with 9% fetal bovine serum and transfected using Lipofectamine-2000 (Invitrogen), according to the manufacturer’s instructions. Generation of PC12 Cell Lines—Retrovirus stocks encoding human SSH1 or mouse SSH2 were obtained by transfection of pLNCX plasmids (Clontech) encoding human SSH1 or mouse SSH2 cDNAs with packaging vectors (retrovirus packaging kit Ampho, Takara Bio, Otsu, Japan) into 293T cells and by harvesting conditioned growth medium containing secreted retrovirus. Virus stocks were filtered through a 0.45-μm filter, supplemented with 8 μg/ml polybrene, and incubated with PC12 cells for 48 h. PC12 cells expressing retrovirus-encoding cDNAs were selected for growth in medium containing 800 μg/ml G418. Recombinant Herpes Simplex Virus Preparation and Infection—The recombinant herpes simplex virus stocks were prepared as described previously (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar). For infection, freshly dissociated DRG explants were allowed to adhere to dishes for 30 min and then incubated with recombinant viral stocks for 12 h before analysis. Immunoprecipitation and Immunoblot Analyses—Immunoprecipitation and immunoblot analyses were performed as described previously (28Ohashi K. Nagata K. Maekawa M. Ishizaki T. Narumiya S. Mizuno K. J. Biol. Chem. 2000; 275: 3577-3582Abstract Full Text Full Text PDF PubMed Scopus (408) Google Scholar). In Vitro Kinase Assay—Serum-starved PC12 cells were stimulated with 100 ng/ml NGF and lysed in kinase buffer (50 mm Hepes (pH 7.4), 150 mm NaCl, 1% Nonidet P-40, 5% glycerol, 1 mm MgCl2, 1 mm MnCl2, 10 mm NaF, 1 mm Na3VO4, 1 mm dithiothreitol, 1 mm phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin). LIMK1 or LIMK2 was immunoprecipitated with anti-LIMK1 or anti-LIMK2 antibodies and subjected to an in vitro kinase assay using His6-tagged cofilin as a substrate, as described (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar). Reaction mixtures were separated on SDS-PAGE and analyzed by autoradiography, Amido Black staining, and immunoblotting with anti-LIMK1 and anti-LIMK2 antibodies. Neurite Outgrowth Assay—PC12 cells were cotransfected with pSUPER and YFP plasmids (4:1) by electroporation, cultured for 32 h and further serum-starved for 16 h. Then, neurite outgrowth was stimulated with 50 ng/ml NGF and allowed to proceed for 48 h in Dulbecco’s modified Eagle’s medium containing 0.1% bovine serum albumin. The cells with neurites longer than two cell body lengths of YFP-positive cells were scored as the neurite-bearing cells. The mean lengths of the longest neurites were measured for YFP-positive cells with neurites longer than one cell body length. Time-lapse Video Fluorescence Image Analysis and Quantification—Live growth cones of chick DRG neurons were observed using an inverted fluorescence microscope (model DMIRBE, Leica) equipped with a 40× phase-contrast objective lens, as described (13Endo M. Ohashi K. Sasaki Y. Goshima Y. Niwa R. Uemura T. Mizuno K. J. Neurosci. 2003; 23: 2527-2537Crossref PubMed Google Scholar). Time-lapse fluorescence images were captured every 30 s for 15 min with 50-100-ms exposures, using a Coolsnap HQ-cooled CCD camera (Roper Scientific) driven by Q550FW Imaging Software (Leica). For quantification of areas of protrusion, a series of images was digitized. Growth cones were outlined at 30-s intervals, and the outlines were used to calculate areas of new protrusion of the growth cone perimeter by binary segmentation using IPLab image analysis software (Scanalytics, Fairfax, VA). The index of growth cone motility was calculated by dividing the average area of new protrusions of each growth cone measured at 30-s intervals over a total recording time of 3 min by the average growth cone perimeter during 3 min of recording. The rate of neurite extension was defined as the mean distance that the center of the growth cone migrated during 10 min of recording. Statistical analyses were performed using Student’s t test. Immunofluorescence and Quantification of Actin Assembly—DRG explants were fixed with 4% paraformaldehyde containing 10% sucrose for 20 min followed by -20 °C methanol for 5 min. The explants were blocked with 2% fetal bovine serum in phosphate-buffered saline for 1 h and then incubated with rabbit polyclonal anti-P-cofilin antibodies that recognize both P-cofilin and P-ADF (8Toshima J. Toshima J.Y. Amano T. Yang N. Narumiya S. Mizuno K. Mol. Biol. Cell. 2001; 12: 1131-1145Crossref PubMed Scopus (218) Google Scholar) and a monoclonal anti-β-actin antibody (AC-15, Sigma) overnight at 4 °C, washed three times with phosphate-buffered saline, and incubated with fluorescein- and rhodamine-conjugated secondary antibodies for 1 h at room temperature. To quantify the actin assembly at the leading edge of the growth cone, DRG neurons were stained with anti-β-actin antibody, and noncollapsed growth cones were randomly selected and fluorescence images were acquired using a Coolsnap HQ-cooled CCD camera driven by Q550FW Imaging Software. Growth cones were outlined on the fluorescence image by automatic outline tool and the average fluorescence intensity (mean pixel density) in a region of 5 μm width inside the front edge of the growth cone was measured using NIH image (Version 1.63). The fluorescence intensities were collected from 20-25 different growth cones per each sample group. Statistical analysis was performed using Student’s t test. Both SSH1 and SSH2 Are Involved in NGF-induced Cofilin/ADF Dephosphorylation in PC12 Cells—To investigate the effect of cofilin/ADF phosphoregulation on NGF-induced neurite extension, we first analyzed changes in P-cofilin/P-ADF levels in PC12 cells after NGF stimulation. Immunoblot analysis with an anti-P-cofilin antibody, which recognizes both P-cofilin and P-ADF (8Toshima J. Toshima J.Y. Amano T. Yang N. Narumiya S. Mizuno K. Mol. Biol. Cell. 2001; 12: 1131-1145Crossref PubMed Scopus (218) Google Scholar), revealed that NGF induced cofilin/ADF dephosphorylation in PC12 cells (Fig. 1A), as previously reported (16Meberg P.J. Ono S. Minamide L.S. Takahashi M. Bamburg J.R. Cell Motil. Cytoskeleton. 1998; 39: 172-190Crossref PubMed Scopus (216) Google Scholar). To examine whether SSHs are responsible for the NGF-induced cofilin/ADF dephosphorylation, we suppressed SSH1 and SSH2 expression with siRNAs, using a pSUPER.retro.puro vector containing the puromycin-resistance gene to allow selection of siRNA-transfected cells. Transfection of siRNA plasmids targeting rat SSH1 and SSH2 substantially reduced the expression of endogenous SSH1 and SSH2, respectively, in PC12 cells (Fig. 1B). Knockdown of SSH1 or SSH2 alone produced no apparent effect on NGF-induced cofilin/ADF dephosphorylation (Fig. 1C). In contrast, knockdown of both SSH1 and SSH2 at the same time notably inhibited NGF-induced dephosphorylation of cofilin/ADF (Fig. 1C), indicating that SSH1 and SSH2 function redundantly for NGF-induced cofilin/ADF dephosphorylation. In PC12 cell lines stably expressing human SSH1-Myc or mouse SSH2-Myc, whose expression was not affected by the siRNA plasmids targeting rat SSH1 or SSH2, double knockdown of rat SSH1/SSH2 had no apparent effect on NGF-induced cofilin/ADF dephosphorylation (Fig. 1D). Thus, the siRNA-resistant human SSH1 and mouse SSH2 restored NGF-induced cofilin/ADF dephosphorylation. NGF-induced cofilin/ADF dephosphorylation was blocked by K252a (an inhibitor of Trk tyrosine kinase receptors), U73122 (an inhibitor of phospholipase C), and latrunculin A (an inhibitor of actin filament assembly) (Fig. 1E), which indicates that NGF-induced cofilin/ADF dephosphorylation requires Trk, phospholipase C, and F-actin. Cofilin and ADF Are Essential for NGF-induced Neurite Extension from PC12 Cells—Cofilin and ADF are closely related proteins and PC12 cells express both (16Meberg P.J. Ono S. Minamide L.S. Takahashi M. Bamburg J.R. Cell Motil. Cytoskeleton. 1998; 39: 172-190Crossref PubMed Scopus (216) Google Scholar). To examine whether cofilin and ADF affect NGF-induced neurite out-growth from PC12 cells, we suppressed the expression of endogenous cofilin and ADF in PC12 cells by transfecting cofilin and ADF siRNA plasmids (Fig. 2A). The PC12 cells were cotransfected with YFP plasmids to permit the transfected cells to be visualized. The transfected cells were cultured for 32 h, serumstarved for 16 h, and then stimulated with NGF for 48 h (Fig. 2B). Transfection of cofilin/ADF siRNAs had no apparent effect on the viablity of PC12 cells (supplemental Fig. S1). To quantitate the effects of cofilin/ADF siRNA on neurite extension, we scored the percentage of neurite-bearing cells with neurites longer than two cell body lengths in YFP-positive cells (Fig. 2C). We also compared the mean lengths of the longest neurites of YFP-positive cells with neurites longer than one cell body (Fig. 2D). Individual knockdown of cofilin or ADF significantly reduced both the number of neurite-bearing cells and the mean neurite length, compared with cells treated with a control siRNA. ADF knockdown suppressed neurite outgrowth more prominently than cofilin knockdown. When both cofilin and ADF were knocked down simultaneously, the percentage of neurite-bearing cells drastically decreased (Fig. 2B, 2C). Cotransfection of chick cofilin or ADF, whose expression was not affected by rat cofilin or ADF siRNA, significantly blocked the inhibitory effects of cofilin/ADF knockdown on neurite extension (Fig. 2E), which indicates that the effects of cofilin/ADF siRNAs are due to the suppression of cofilin/ADF expression. These results suggest that cofilin/ADF are essential for NGF-induced neurite outgrowth from PC12 cells. SSH1 and SSH2 Are Critical for NGF-induced Neurite Extension from PC12 Cells—We next examined whether SSH1 and SSH2 are involved in NGF-induced neurite extension from PC12 cells by knocking down SSH1 and SSH2 with siRNAs (Fig. 3A, see also Fig. 1B). The percentage of neurite-bearing cells was significantly reduced by SSH1 or SSH2 single knockdown or SSH1/SSH2 double knockdown, compared with control siRNA (Fig. 3B). The mean neurite length was also reduced in single and double SSH knockdown cells (Fig. 3C). In PC12 cells stably expressing human SSH1 or mouse SSH2, the number of neurite-bearing cells (Fig. 3D) and the mean neurite length (not shown) were not affected by either single or double knockdown of endogenous rat SSH1/SSH2, which indicates that the effects of SSH1/SSH2 siRNA on neurite extension are attributable to the suppression of SSH1/SSH2 expression and both SSH1 and SSH2 act to promote NGF-induced neurite extension from PC12 cells. NGF Induces LIMK1/LIMK2 Activation in PC12 Cells—NGF induces cofilin/ADF dephosphorylation, raising the possibility that the kinase activity of LIMKs may be negatively regulated by NGF stimulation. However, when we examined the changes in the kinase activities of LIMK1 and LIMK2 in PC12 cells after NGF stimulation, both LIMK1 and LIMK2 were activated rather than repressed. LIMK1 activity increased 1.7-fold at 2 min after NGF treatment and then reverted to the basal level by 30 min (Fig. 4A). In contrast, LIMK2 activity gradually increased, reaching 1.5-fold by 30 min after NGF treatment (Fig. 4B). Pretreatment of PC12 cells with wortmannin, an inhibitor of phosphoinositide 3-kinase, blocked the NGF-induced activation of LIMK1 but not of LIMK2 (Fig. 4C, 4D). In contrast, Y-27632, a specific inhibitor of ROCK, had no apparent effect on NGF-induced LIMK1 or LIMK2 activation (Fig. 4, C and D). LIMK1 and LIMK2 Are Critical for NGF-induced Neurite Extension from PC12 Cells—To examine whether LIMK1 and LIMK2 contribute to NGF-induced neurite extension from PC12 cells, we introduced siRNA plasmids targeting rat LIMK1 and LIMK2 into the cells, which suppressed the expression of endogenous LIMK1 and LIMK2, respectively (Fig. 5A). The kinase activity of LIMK1 and LIMK2 in cells expressing the corresponding siRNA also decreased, according to the decrease in their expression levels (supplemental Fig. S2). Knockdown of LIMK1, LIMK2, or both significantly reduced the number of neurite-bearing cells and the mean neurite length, compared with control cells (Fig. 5, B-D). LIMK2 knockdown suppressed neurite extension more prominently than LIMK1 knockdown. Similar results were obtained by using another set of siRNAs for rat LIMK1 and LIMK2 (supplemental Fig. S3). These results suggest that both LIMK1 and LIMK2 are critical for NGF-induced neurite extension of PC12 cells. We also analyzed the P-cofilin/P-ADF levels in unstimulated PC12 cells expressing LIMK1/LIMK2 siRNAs. The P-cofilin/P-ADF levels reduced in cells expressing LIMK2 or LIMK1/LIMK2 double knockdown cells but not LIMK1 knockdown cells (supplemental Fig. S4), which suggests that LIMK2, but not LIMK1, is mainly involved in maintenance of the P-cofilin/P-ADF levels in unstimulated PC12 cells. LIMK1 and LIMK2 Are Critical for the Growth Cone Extension and Motility of Chick DRG Neurons—To assess the roles of LIMK1 and LIMK2 in the growth cone extension and motility of neurons in primary culture, we used time-lapse fluorescence microscopy to analyze the effects of LIMK1 and LIMK2 knock-down on the growth cone movement of chick DRG neurons. When cotransfected into CHO cells, the siRNA plasmids targeting chLIMK1 and chLIMK2 strongly silenced the expression of Myc-chLIMK1 and Myc-chLIMK2, respectively (Fig. 6A). To analyze the growth cone motility, chick DRG neurons were cotransfected with YFP plasmids and siRNA plasmids for chLIMK1 or chLIMK2 using Nucleofector II, cultured them for 48 h, replated the neurons in laminin-coated dishes, and further cultured them for 12 h before time-lapse observation. NGF was always included in the culture medium. Representative live images of growth cones are shown in Fig. 6B. Time-lapse observations revealed that single knockdown of LIMK1 or LIMK2, or double knockdown of LIMK1/LIMK2, markedly decreased the rate of growth cone extension, compared with the rate in DRG neurons transfected with a control siRNA (Fig. 6B). To quantitate the data, we measured the distance that the center of the growth cone migrated during 10 min of recording. The average rate of neurite extension of the neurons expressing LIMK1, LIMK2, and LIMK1/LIMK2 si