- Email: [email protected]
CLP36 interacts with palladin in dorsal root ganglion neurons
Neuroscience Letters 476 (2010) 53–57
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
CLP36 interacts with palladin in dorsal root ganglion neurons Tomohiko Hasegawa a,b , Koji Ohno a,∗ , Shinji Funahashi a,b , Kazufumi Miyazaki a , Akira Nagano b , Kohji Sato a a b
Department of Anatomy, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan Department of Orthopedics, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan
a r t i c l e
i n f o
Article history: Received 16 January 2010 Received in revised form 31 March 2010 Accepted 31 March 2010 Keywords: CLP-36 PDLIM1 PDZ–LIM protein Nerve injury Axotomy
a b s t r a c t CLP36, a member of the ␣-actinin-associated LIM protein (ALP)/enigma protein family, plays a role in neurite outgrowth in the peripheral nervous system. However, the underlying molecular mechanisms are not known. In this study, we performed yeast two-hybrid screening of an E18 mouse whole-body cDNA library with CLP36 as the bait and isolated palladin as a CLP36-binding protein. Palladin is an actin-binding protein and it was shown to have a role in the extension of cortical neurons. A coimmunoprecipitation study showed that CLP36 and palladin formed a complex in the dorsal root ganglion (DRG). In addition, CLP36 and palladin were colocalized in the neurites and cell bodies of primary DRG neurons. Furthermore, sciatic nerve transection increased the expression of both CLP36 and palladin mRNAs in DRG neurons, with the increase in CLP36 mRNA being more prominent. This implies that CLP36 has a more specific role in nerve regeneration than palladin. Our results suggest that CLP36 may interact with palladin to influence neurite outgrowth during sciatic nerve regeneration. © 2010 Elsevier Ireland Ltd. All rights reserved.
Peripheral nerves have the ability to regenerate after injury. After a peripheral nerve is injured, the proximal stump undergoes transformation into a growth cone and the nerve begins to regenerate. Growth cones are motile structures and their formation and movement are based on rearrangement of the actin cytoskeleton. To date, much research has been done in attempts to clarify the mechanism by which the actin cytoskeleton is regulated in growth cones. While these studies have suggested that a large number of actin-binding proteins (ABPs) may be involved, their actual roles in the growth cone are not well understood [4]. CLP36 is a member of the ALP/enigma family of proteins that have a PDZ domain at the N-terminal and 1–3 LIM domains at the C-terminal. Because ALP/enigma proteins interact with ABPs such as ␣-actinin [1,11,17,19,21,22] and -tropomyosin [7], these proteins are thought to modulate the structure and/or dynamics of the actin cytoskeleton. Recently, we reported that CLP36 forms a complex with ␣-actinin in DRG and PC12 cells [13]. CLP36 is localized to growth cones and its expression is markedly increased in peripheral sensory and motor neurons after nerve injury. In addition, knockdown of CLP36 expression was shown to promote neuritogenesis with enhancement of growth cone movement in nerve growth factor (NGF)-treated PC12 cells. These findings suggest that CLP36 may play an important role in rearrangement of the actin cytoskeleton during peripheral nerve regeneration. However, there is no
∗ Corresponding author. Tel.: +81 53 435 2290; fax: +81 53 435 2290. E-mail address: [email protected] (K. Ohno). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.03.081
information available about the molecular mechanisms underlying this function. To further investigate the role of CLP36 in neurite outgrowth, including the functional significance of the CLP36-␣actinin interaction, we considered that it was necessary to identify other proteins that bind with CLP36 in neurons. Yeast two-hybrid screening was carried out with the DupLexA hybrid system kit (Origene Technologies). The cDNA fragment encoding CLP36 (a kind gift from H. Betz, Frankfurt, Germany) was subcloned in frame within the inducible pGilda bait vector. Then this bait construct was used to screen 108 independent recombinant clones of a mouse E18 whole-body cDNA library in pJG 4-5 transformed into yeast strain EGY48. We obtained 100 blue colonies that grew on the original galactose-inducible selection plates and analyzed the cDNA fragments contained within these colonies. For mapping of the sites involved in binding between CLP36 and palladin, yeast two-hybrid assays were performed. A deletion mutant of palladin lacking the last three amino acids (palladin EDL) was amplified by the polymerase chain reaction (PCR) and was cloned into the pJG 4-5 prey vector. Then five CLP36 deletion mutant fragments (PDZ + MID amino acids 2–262, MID + LIM 8–327, PDZ alone 2–89, MID alone 80–262, and LIM alone 253–327) were amplified by PCR and cloned into the pGilda vector. To test the interaction of palladin with RIL and mystique, as well as that of CLP36 with Pick1, syntenin, and PSD-95, cDNA fragments for RIL and mystique were amplified by reverse transcriptase (RT)-PCR from rat brain total RNA and cloned into the pGilda vector. Similarly, pJG 4-5 Pick1, syntenin, and PSD-95 cDNA fragments were obtained
54
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
by RT-PCR and cloned into the pJG 4-5 vectors. Yeast two-hybrid assays were employed to test the above-mentioned interactions. F2408 fibroblast cells were provided by the Health Science Research Resources Bank (Osaka, Japan). Primary DRG neurons were prepared from adult male Wistar rats (200 g), as described previously [15]. For immunofluorescence, cells were fixed with 4% formaldehyde for 20 min, permeabilized with 0.2% Triton X-100 and 4% normal goat serum in PBS for 20 min, and then were reacted with anti-CLP36 mAb (BD Transduction Laboratories, Biosciences, 1:200) and anti-palladin polyclonal Ab (Proteintech, 1:200) for 3 h at RT. Primary antibodies were detected by incubation with antimouse or anti-rabbit Arexa 488- or 594-conjugated secondary IgG antibodies (Molecular Probes, 1:500) for 1 h at RT. For visualization of actin filaments, permeabilized cells were reacted with Alexa 594-conjugated phalloidin (Molecular Probes, 1:200). Immunofluorescence images were captured with a BZ-8000 (Keyence) and were processed using the BZ-analyzer software (Keyence). For coimmunoprecipitation experiments, the Triton X-100 soluble fraction extracted from F2408 cells or rat dorsal root ganglia was incubated for 3 h at 4 ◦ C with anti-CLP36 mAb (0.75 g) or control mouse IgG conjugated with protein G sepharose (Amersham Biosciences). After washing, the reaction products were subjected to western blot analysis. The primary and secondary antibodies were as follows: anti-CLP36 mAb (1:500); anti-palladin polyclonal Ab (1:1000); anti-␣-actinin mAb (Sigma, 1:400); anti-actin mAb (Sigma, 1:2000); horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgG (Nakalai Tesque, 1: 5000). For the absorption test, GST-palladin (Proteintech) and GST were bound to glutathione-sepharose beads (Amersham Biosciences) and then incubated overnight with an anti-palladin polyclonal antibody. After centrifugation, the supernatants were used for western blot analysis. GST protein was purified according the manufacturer’s protocol (Amersham Biosciences). Sciatic nerve transection was performed in adult male Wistar rats (200 g). The animals were deeply anesthetized with sodium pentobarbitol (50 mg/kg i.p.), and sciatic nerve transection and control surgery were performed as described previously [13]. At 3 days after surgery, the rats were anesthetized again and exsanguinated by cutting the carotid vessels, after which the L3–L5 DRGs were immediately removed from the control and operated sides. Frozen sections (18 m thick) were cut on a cryostat. The sequences of the probes for in situ hybridization were checked in a database search to exclude significant homology with other genes. For CLP36 mRNA, we used probes with previously published sequences [13]. For palladin mRNA, a specific oligonucleotide probe (BC127081: nucleotides 2267–2306) was employed. These probes were labeled with [35 S]dATP using terminal deoxynucleotidyl transferase. Sections incubated with control sense probes for CLP36 and palladin did not display any signals (data not shown). For quantitative analysis, DRG neurons of a similar size (n = 30) were selected on both the operated and control sides, and the number of silver grains localized on these neurons was counted. Statistical evaluation was performed using Student’s t-test. For immunohistochemistry, frozen sections were fixed with 4% paraformaldehyde and pretreated with 0.35% hydrogen peroxidase. They were reacted with anti-CLP36 polyclonal Ab (1:200, Proteintech) or anti-palladin polyclonal Ab (1:200), then with polymer-horseradish peroxidaselabeled anti-rabbit secondary Ab (Dako), and were finally visualized with diaminobenzidine and hydrogen peroxidase. The specificity of staining was confirmed by absorption of the antibodies with GST-CLP36 [13] or GST-palladin. Real-time PCR was performed with SYBR Premix ExTaq (Takara) and a Thermal Cycler Dice Real Time System (Takara). Total RNAs from L3 to L5 DRGs were isolated with an RNeasy mini kit (Quiagen) and converted to cDNA by reverse transcription. Primer sequences used for PCR analysis were listed in Supplemental Table 1. Levels of
CLP36, palladin, mystique, and RIL mRNAs were normalized to that of GAPDH mRNA and were shown as ratios to the value of control CLP36 mRNA. PCR-amplified products were also electrophoresed on 1% agarose gels to confirm that single bands were amplified. By screening the E18 mouse whole-body cDNA library, we isolated three clones encoding the in-frame C-terminal region of palladin cDNA, two clones encoding the in-frame full-length ␥filamin, ␣-actinin, telethonin binding protein (FATZ) cDNA, and one clone each encoding in-frame actinin-2, actinin-3, and actinin-4 cDNA fragments. The yeast two-hybrid assay confirmed the specificity of the interaction of these clones with full-length CLP36 (data not shown). Because FATZ is exclusively expressed in muscle tissue [5], further investigation was focused on the relationship between CLP36 and palladin. CLP36 has a PDZ domain at its N-terminal region and palladin carries a putative PDZ-binding motif (EDL) at the end of its C-terminal region. Thus, we speculated that the interaction between CLP36 and palladin involved the PDZ domain of CLP36 and the C-terminus of palladin. To test this hypothesis, we generated deletion mutants of CLP36 and examined their binding to the C-terminal region of palladin by the yeast two-hybrid assay (Fig. 1A). The results showed that all of CLP36 mutants that involve its PDZ domain could bind to the C-terminal region of palladin. However, a palladin mutant lacking the last three amino acids (EDL) was unable to bind with CLP36 (Fig. 1B). These findings supported our hypothesis and indicated that the C-terminus of palladin binds with the PDZ domain of CLP36. To examine the specificity of this interaction, we tested whether the C-terminus of palladin could bind with the PDZ domains of other proteins. We found that the C-terminus of palladin did not interact with the PDZ domain of PSD-95, Pick1, or syntenin (Fig. 1C). However, it interacted with mystique and RIL, which are from the ALP protein family along with CLP36 (Fig. 1D). These results suggested that the C-terminus of palladin preferentially binds to the PDZ domain of proteins from the ALP family. We next investigated whether CLP36 and palladin could form an intracellular protein complex. Initially, we searched for a CLP36–palladin complex in fibroblasts (F2408 cells) because both CLP36 and palladin have been shown to play an important role in stress fiber formation. In F2408 cell lysates, anti-palladin antibody reacted with a major band of ∼90 kDa and also recognized bands of 50 and 70 kDa. These bands coincided with palladin isoforms that have been reported previously [6]. Furthermore, two additional bands were observed at 90–100 kDa. All of these bands disappeared after preadsorption with the antigenic peptide (Fig. 2A, left). A coimmunoprecipitation study demonstrated that palladin was precipitated with anti-CLP36 antibody but not with control IgG (Fig. 2A, right). Interestingly, among several palladin isoforms, the high molecular weight 90 kDa isoform was only precipitated with anti-CLP36 antibody (asterisk). In addition, ␣-actinin was found in the precipitates. To compare the subcellular localization of palladin and CLP36, we stained F2408 cells with an anti-palladin antibody (Fig. 2B, upper) and an anti-CLP36 antibody (Fig. 2B, lower). We found that palladin and CLP36 were localized at stress fibers, with both molecules being detected at the same sites (Fig. 2B, right panel, arrowheads). These findings indicate that CLP36 and palladin are associated with each other in F2408 cells. Similarly, endogenous palladin and ␣-actinin were precipitated from DRG tissue specimens with anti-CLP36 antibody. As seen in F2408 cells, several palladin isoforms were also expressed in DRG tissues and a high molecular weight 90 kDa isoform was only precipitated with antiCLP36 antibody (Fig. 2C). Primary DRG neurons showed intense staining for CLP36 and palladin in the cell body, and staining was also observed along neurites including the growth cones (Fig. 2D). It has been shown that both CLP36 and palladin play a role in neurite outgrowth [2,13]. It is known that neurons increase the synthesis of proteins required for regeneration after nerve injury.
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
55
Fig. 1. Analysis of the interaction between CLP36 and palladin in a yeast two-hybrid system. (A) Interaction of full-length CLP36 and its deletion mutants with the C-terminal region of palladin. All of CLP36 mutants that involve the PDZ domain interacted with palladin. (B) The last three amino acids at the C-terminal of palladin are necessary for binding with CLP36. (C) Palladin does not interact with the PDZ domain of synaptic proteins such as PSD-95, Pick1, and syntenin. (D) Palladin interacts with proteins of the ALP/enigma family, including RIL and mystique as well as CLP36.
Indeed, we recently reported that CLP36 expression is upregulated in peripheral sensory and motor neurons after nerve injury. To investigate whether palladin expression is also regulated by nerve injury, we examined changes of palladin mRNA expression in DRG neurons after sciatic nerve transection and compared the results
with those for CLP36 mRNA. Under normal conditions, CLP36 mRNA was weakly expressed in DRG neurons, but its expression was dramatically upregulated and reached the peak at 3 days after sciatic nerve transection (Fig. 3A). Palladin mRNA expression was very weak in DRG neurons under normal conditions. Sciatic nerve tran-
Fig. 2. Association of endogenous CLP36 with palladin in fibroblasts and DRG neurons. (A) Equal amounts of protein (20 g/lane) from F2408 cells were subjected to SDS-PAGE, transferred to a nitrocellulose membrane, and probed with anti-palladin antibody. For the absorption test, anti-palladin antibody was incubated overnight with GST-palladin or GST (control) (left). F2408 cell lysates were precipitated with anti-CL36 antibody and the immunoprecipitants were blotted with anti-palladin, ␣-actinin, and CLP36 antibodies. Asterisk indicates the band precipitated by anti-CLP36 antibody (right). (B) F2408 cells were double-stained with anti-CLP36 and palladin antibodies. Bar = 10 m. Right photos are regions of interest boxed in the left ones. Arrowheads indicate some examples of sites on stress fibers where CLP36 and palladin are colocalized. (C) DRG tissue lysates were precipitated with anti-CLP36 antibody and the immnoprecipitants were blotted with anti-palladin, ␣-actinin, and CLP36 antibodies. (D) Primary DRG neurons were double-stained with anti-CLP36 and palladin antibodies. Asterisk indicates the band precipitated by anti-CLP36 antibody. Panels on the right show high-magnification views of a growth cone. Arrows indicate colocalization of CLP36 and palladin in the same growth cone. Bars = 20 m.
56
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
Fig. 3. Changes of CLP36 and palladin expressions in DRG neurons after sciatic nerve transection. (A) Real-time PCR analysis of CLP36 and palladin mRNAs in DRG tissues after sciatic nerve transection. The bars represent the standard error of the mean (n = 3). (B) At 3 days after sciatic nerve transection, CLP36 and palladin mRNA expressions were examined by in situ hybridization histochemistry. Dark-field photomicrographs show that CLP36 mRNA is markedly increased in the DRG after sciatic nerve transection (a, b). Bright-field photomicrographs show accumulation of hybridization signals for CLP36 mRNA on DRG neurons (c, d). Sciatic nerve transection also increased palladin mRNA expression in DRG neurons (e–h). (C) Quantitative analysis of the changes of CLP36 and palladin mRNA expressions in DRG neurons after sciatic nerve transection. Statistical evaluation was performed with Student’s t-test (*p < 0.01). (D) Western blot analysis shows an increase in CLP36 protein on day 3 after transection, while there is no significant change of palladin protein. (E) On day 3 after transection, DRG neurons were labeled with anti-CLP36 and palladin antibodies. Right photos show control sections incubated with the respective antibodies preabsorbed by GST-CLP36 (upper) or GST-palladin (lower). Arrows indicate some examples of DRG neurons. Scale bar = 100 m. (F) RT-PCR analysis of RIL and mystique mRNA expressions in DRG tissues on day 3 after sciatic nerve transection. Mystique mRNA was not detected in either of control or operated DRG tissues. (G) Real-time PCR analysis confirmed that sciatic nerve transection did not affect the expression of RIL mRNA in DRG tissues. The error bars represent the standard error of the mean (n = 3). Statistical evaluation was performed with Student’s t-test (*p < 0.01).
section also induced palladin mRNA expression in DRG neurons, but the increase was not so dramatic as that of CLP36 mRNA (Fig. 3A–C). Consistent with these findings, CLP36 and palladin proteins were localized to DRG neurons and there was no significant change in the level of palladin protein in DRG tissues after sciatic nerve transection, in contrast to the increase of CLP36 protein (Fig. 3D and E). These findings indicate that the CLP36 and palladin genes respond differently to nerve injury. While CLP36 mainly partici-
pates in nerve regeneration, palladin is thought to have a role in both normal and regenerating sciatic nerves. As shown in Fig. 1D, RIL and mystique also have the ability to interact with palladin. We wanted to know whether these molecules are expressed in DRG neurons and whether their expression is upregulated similarly to that of CLP36 by sciatic nerve transection. RT-PCR showed that RIL mRNA was expressed in the DRG, whereas mystique mRNA was almost undetectable (Fig. 3F).
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
Real-time PCR then demonstrated that RIL and mystique mRNA expressions were not affected by sciatic nerve transection (Fig. 3G). These findings suggest that RIL and palladin form a protein complex in DRG neurons. However, it seems possible that CLP36 and RIL participate in different cellular functions after interacting with palladin. In this study, we performed yeast two-hybrid screening and isolated palladin as a binding partner of CLP36. Palladin is an actin cross-linking protein that plays an important role in the assembly and maintenance of various actin-based structures [6]. Both CLP36 and palladin are widely distributed in epithelial and mesenchymal cells. These two proteins are localized at stress fibers and are essential for stress fiber formation in BeWo cells [16] or Rcho-1 cells [14], respectively. In addition, we showed here that endogenous CLP36 and palladin form a complex in F2408 fibroblasts. Very recently, an interaction between palladin and CLP36 has been reported by two different groups [10,12]. Corresponding to our findings, these groups showed that the C-terminus of palladin specifically interacts with the PDZ domain of CLP36. We also demonstrated that a CLP36–palladin complex exists in DRG neurons and in F2408 cells. CLP36 and palladin were colocalized in the cell bodies and neurites of primary DRG neurons. Maeda et al. [10] suggested that a CLP36–palladin complex might interact with ␣-actinin to regulate the actin cytoskeleton. Since CLP36 forms a complex with ␣-actinin in DRG neurons [13], it is therefore possible that CLP36 also forms a functional complex with palladin and ␣-actinin in these neurons. In the nervous system, CLP36 is predominantly expressed by peripheral sensory and motor neurons [13]. We showed that CLP36 mRNA expression is dramatically increased in DRG neurons after sciatic nerve injury, suggesting that CLP36 plays a critical role in the regeneration of primary sensory nerves rather than maintaining their structure or function. Similarly, because palladin has been shown to play a role in neurite extension by cortical neurons [2], we expected that sciatic nerve transection might also increase palladin mRNA expression in DRG neurons. However, there was little change of palladin after nerve injury. These findings led us to the hypothesis that palladin has a role in maintenance of the sensory system under normal conditions and is also involved in regeneration after nerve injury. Palladin interacts with a number of actin-associated proteins, including ezrin, VASP, profiling, LASP-1, and Eps8, and it is a cytoskeletal scaffold protein [6]. Since strong palladin staining has been observed in the substantia gelatinosa of the spinal cord [8], it is probable that palladin is localized to sciatic nerve terminals and participates in organization of the actin cytoskeleton together with palladin-binding proteins. After sciatic nerve injury occurs, palladin may switch its binding partners and form a different complex with CLP36 to regulate rearrangement of the actin cytoskeleton for the promotion of nerve regeneration. Although a variety of ABPs are involved in the mechanism of nerve regeneration, it remains unclear how the functions of these ABPs are coordinated. Considering that CLP36 knockdown led to excessive neuritogenesis in PC12 cells [13], CLP36 is not only a promoter of nerve regeneration but seems to play an important role in this coordination mechanism. Our results indicated that CLP36 might act as an adaptor that links ␣-actinin with palladin in the regenerating sciatic nerve. It has been reported that members of the ALP/enigma protein family bind signaling molecules via the LIM domain [3,9,20,22], and the LIM domain of CLP36 was reported to bind serine/threonine kinase Clik1 [18]. It is therefore possible that signaling proteins such as Clik1 kinase also participate in the CLP36 complex and modulate the functions of ␣-actinin and/or palladin in the regenerating sciatic nerve. However, further studies will be required to examine this hypothesis.
57
Acknowledgements We thank Prof. Heinrich Betz for providing rat CLP36 cDNA. This work was partly supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.neulet.2010.03.081. References [1] K. Bauer, M. Kratzer, M. Otte, K.L. De Quintana, J. Hagmann, G.J. Arnold, C. Eckerskorn, F. Lottspeich, W. Siess, Human CLP36, a PDZ-domain and LIMdomain protein, binds to alpha-actinin-1 and associates with actin filaments and stress fibers in activated platelets and endothelial cells, Blood 96 (2000) 4236–4245. [2] M. Boukhelifa, M.M. Parast, J.G. Valtschanoff, A.S. LaMantia, R.B. Meeker, C.A. Otey, A role for the cytoskeleton-associated protein palladin in neurite outgrowth, Mol. Biol. Cell 12 (2001) 2721–2729. [3] E. Cuppen, H. Gerritis, B. Pepers, B. Wieringa, W. Hendriks, PDZ motifs in PTP-BL and RIL bind to internal protein segments in the LIM domain protein RIL, Mol. Biol. Cell 9 (1998) 671–683. [4] E.W. Dent, F.B. Gertler, Cytoskeletal dynamics and transport in growth cone motility and axon guidance, Neuron 40 (2003) 209–227. [5] G. Faulkner, A. Pallavicini, A. Comelli, M. Salamon, G. Bortoletto, C. Ievolella, S. Trevisan, S. Kojic, F. Dalla Vecchia, P. Laveder, G. Valle, G. Lanfranchi, FATZ, a filamin-, actinin, and telethonin-binding protein of the Z-disc of skeletal muscle, J. Biol. Chem. 275 (2000) 41234–41242. [6] S.M. Goicoechea, D. Arneman, C.A. Otey, The role of palladin in actin organization and cell motility, Eur. J. Cell Biol. 87 (2008) 517–525. [7] P.M. Guy, D.A. Kenny, G.N. Gill, The PDZ domain of the LIM protein enigma binds to beta-tropomyosin, Mol. Biol. Cell 10 (1999) 1973–1983. [8] S.J. Hwang, S. Pagliardini, M. Boukhelifa, M.M. Parast, C.A. Otey, A. Rustioni, J.G. Valtschanoff, Palladin is expressed preferentially in excitatory terminals in the rat central nervous system, J. Comp. Neurol. 436 (2001) 211–224. [9] S. Kuroda, C. Tokunaga, Y. Kiyohara, O. Higuchi, H. Konishi, K. Mizuno, G.N. Gill, U. Kikkawa, Protein–protein interaction of zinc finger LIM domains with protein kinase C, J. Biol. Chem. 271 (1996) 31029–31032. [10] M. Maeda, E. Asano, D. Ito, S. Ito, Y. Hasegawa, M. Hamaguchi, T. Senga, Characterization of interaction between CLP36 and palladin, FEBS J. 276 (2009) 2775–2785. [11] N. Nakagawa, M. Hoshijima, M. Oyasu, N. Saito, K. Tanizawa, S. Kuroda, ENH, containing PDZ and LIM domains, heart/skeletal muscle-specific protein, associates with cytoskeletal proteins through the PDZ domain, Biochem. Biophys. Res. Commun. 272 (2000) 505–512. [12] P. von Nandelstadh, M. Ismail, C. Gardin, H. Suila, I. Zara, A. Belgrano, G. Valle, O. Carpen, G. Faulkner, A class III PDZ binding motif in the myotilin and FATZ families binds enigma family proteins: a common link for Z-disc myopathies, Mol. Cell. Biol. 29 (2009) 822–834. [13] K. Ohno, H. Kato, S. Funahashi, T. Hasegawa, K. Sato, Characterization of CLP36/Elfin/PDLIM1 in the nervous system, J. Neurochem. 111 (2009) 790–800. [14] M.M. Parast, C.A. Otey, Characterization of palladin, a novel protein localized to stress fibers and cell adhesions, J. Cell Biol. 150 (2000) 643–656. [15] D.S. Smith, J.H.P. Skene, A transcriptional-dependent switch controls competence of adult neurons for distinct models of axon growth, J. Neurosci. 17 (1997) 646–658. [16] N. Tamura, K. Ohno, T. Katayama, N. Kanayama, K. Sato, The PDZ–LIM protein CLP36 is required for actin stress fiber formation and focal adhesion assembly in BeWo cells, Biochem. Biophys. Res. Commun. 364 (2007) 589–594. [17] M. Torrado, V.V. Senatorov, R. Trivedi, R.N. Fariss, S.I. Tomarev, Pdlim2, a novel PDZ–LIM domain protein, interacts with alpha-actinins and filamin A, Invest. Ophth. Vis. Sci. 45 (2004) 3955–3963. [18] T. Vallenius, T.P. Mäkelä, Clik1: a novel kinase targeted to actin stress fibers by the CLP-36 PDZ–LIM protein, J. Cell Sci. 15 (2002) 2067–2073. [19] T. Vallenius, B. Scharm, A. Vesikansa, K. Luukko, R. Schäfer, T.P. Mäkelä, The PDZ–LIM protein RIL modulates actin stress fiber turnover and enhances the association of alpha-actinin with F-actin, Exp. Cell Res. 293 (2004) 117–128. [20] R. Wu, K. Durick, Z. Songyang, L.C. Cantley, S.S. Taylor, G.N. Gill, Specificity of LIM domain interactions with receptor tyrosine kinases, J. Biol. Chem. 271 (1996) 15934–15941. [21] H. Xia, S.T. Winokur, W.L. Kuo, M.R. Altherr, D.S. Bredt, Actinin-associated LIM protein: identification of a domain interaction between PDZ and spectrin-like repeated motifs, J. Cell Biol. 139 (1997) 507–515. [22] Q. Zhou, P. Ruiz-Lozano, M.E. Martone, J. Chen, Cypher, a striated musclerestricted PDZ and LIM domain containing protein, binds to alpha-actinin-2 and protein kinase C, J. Biol. Chem. 274 (1999) 19807–19813.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
CLP36 interacts with palladin in dorsal root ganglion neurons Tomohiko Hasegawa a,b , Koji Ohno a,∗ , Shinji Funahashi a,b , Kazufumi Miyazaki a , Akira Nagano b , Kohji Sato a a b
Department of Anatomy, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan Department of Orthopedics, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431-3192, Japan
a r t i c l e
i n f o
Article history: Received 16 January 2010 Received in revised form 31 March 2010 Accepted 31 March 2010 Keywords: CLP-36 PDLIM1 PDZ–LIM protein Nerve injury Axotomy
a b s t r a c t CLP36, a member of the ␣-actinin-associated LIM protein (ALP)/enigma protein family, plays a role in neurite outgrowth in the peripheral nervous system. However, the underlying molecular mechanisms are not known. In this study, we performed yeast two-hybrid screening of an E18 mouse whole-body cDNA library with CLP36 as the bait and isolated palladin as a CLP36-binding protein. Palladin is an actin-binding protein and it was shown to have a role in the extension of cortical neurons. A coimmunoprecipitation study showed that CLP36 and palladin formed a complex in the dorsal root ganglion (DRG). In addition, CLP36 and palladin were colocalized in the neurites and cell bodies of primary DRG neurons. Furthermore, sciatic nerve transection increased the expression of both CLP36 and palladin mRNAs in DRG neurons, with the increase in CLP36 mRNA being more prominent. This implies that CLP36 has a more specific role in nerve regeneration than palladin. Our results suggest that CLP36 may interact with palladin to influence neurite outgrowth during sciatic nerve regeneration. © 2010 Elsevier Ireland Ltd. All rights reserved.
Peripheral nerves have the ability to regenerate after injury. After a peripheral nerve is injured, the proximal stump undergoes transformation into a growth cone and the nerve begins to regenerate. Growth cones are motile structures and their formation and movement are based on rearrangement of the actin cytoskeleton. To date, much research has been done in attempts to clarify the mechanism by which the actin cytoskeleton is regulated in growth cones. While these studies have suggested that a large number of actin-binding proteins (ABPs) may be involved, their actual roles in the growth cone are not well understood [4]. CLP36 is a member of the ALP/enigma family of proteins that have a PDZ domain at the N-terminal and 1–3 LIM domains at the C-terminal. Because ALP/enigma proteins interact with ABPs such as ␣-actinin [1,11,17,19,21,22] and -tropomyosin [7], these proteins are thought to modulate the structure and/or dynamics of the actin cytoskeleton. Recently, we reported that CLP36 forms a complex with ␣-actinin in DRG and PC12 cells [13]. CLP36 is localized to growth cones and its expression is markedly increased in peripheral sensory and motor neurons after nerve injury. In addition, knockdown of CLP36 expression was shown to promote neuritogenesis with enhancement of growth cone movement in nerve growth factor (NGF)-treated PC12 cells. These findings suggest that CLP36 may play an important role in rearrangement of the actin cytoskeleton during peripheral nerve regeneration. However, there is no
∗ Corresponding author. Tel.: +81 53 435 2290; fax: +81 53 435 2290. E-mail address: [email protected] (K. Ohno). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.03.081
information available about the molecular mechanisms underlying this function. To further investigate the role of CLP36 in neurite outgrowth, including the functional significance of the CLP36-␣actinin interaction, we considered that it was necessary to identify other proteins that bind with CLP36 in neurons. Yeast two-hybrid screening was carried out with the DupLexA hybrid system kit (Origene Technologies). The cDNA fragment encoding CLP36 (a kind gift from H. Betz, Frankfurt, Germany) was subcloned in frame within the inducible pGilda bait vector. Then this bait construct was used to screen 108 independent recombinant clones of a mouse E18 whole-body cDNA library in pJG 4-5 transformed into yeast strain EGY48. We obtained 100 blue colonies that grew on the original galactose-inducible selection plates and analyzed the cDNA fragments contained within these colonies. For mapping of the sites involved in binding between CLP36 and palladin, yeast two-hybrid assays were performed. A deletion mutant of palladin lacking the last three amino acids (palladin EDL) was amplified by the polymerase chain reaction (PCR) and was cloned into the pJG 4-5 prey vector. Then five CLP36 deletion mutant fragments (PDZ + MID amino acids 2–262, MID + LIM 8–327, PDZ alone 2–89, MID alone 80–262, and LIM alone 253–327) were amplified by PCR and cloned into the pGilda vector. To test the interaction of palladin with RIL and mystique, as well as that of CLP36 with Pick1, syntenin, and PSD-95, cDNA fragments for RIL and mystique were amplified by reverse transcriptase (RT)-PCR from rat brain total RNA and cloned into the pGilda vector. Similarly, pJG 4-5 Pick1, syntenin, and PSD-95 cDNA fragments were obtained
54
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
by RT-PCR and cloned into the pJG 4-5 vectors. Yeast two-hybrid assays were employed to test the above-mentioned interactions. F2408 fibroblast cells were provided by the Health Science Research Resources Bank (Osaka, Japan). Primary DRG neurons were prepared from adult male Wistar rats (200 g), as described previously [15]. For immunofluorescence, cells were fixed with 4% formaldehyde for 20 min, permeabilized with 0.2% Triton X-100 and 4% normal goat serum in PBS for 20 min, and then were reacted with anti-CLP36 mAb (BD Transduction Laboratories, Biosciences, 1:200) and anti-palladin polyclonal Ab (Proteintech, 1:200) for 3 h at RT. Primary antibodies were detected by incubation with antimouse or anti-rabbit Arexa 488- or 594-conjugated secondary IgG antibodies (Molecular Probes, 1:500) for 1 h at RT. For visualization of actin filaments, permeabilized cells were reacted with Alexa 594-conjugated phalloidin (Molecular Probes, 1:200). Immunofluorescence images were captured with a BZ-8000 (Keyence) and were processed using the BZ-analyzer software (Keyence). For coimmunoprecipitation experiments, the Triton X-100 soluble fraction extracted from F2408 cells or rat dorsal root ganglia was incubated for 3 h at 4 ◦ C with anti-CLP36 mAb (0.75 g) or control mouse IgG conjugated with protein G sepharose (Amersham Biosciences). After washing, the reaction products were subjected to western blot analysis. The primary and secondary antibodies were as follows: anti-CLP36 mAb (1:500); anti-palladin polyclonal Ab (1:1000); anti-␣-actinin mAb (Sigma, 1:400); anti-actin mAb (Sigma, 1:2000); horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgG (Nakalai Tesque, 1: 5000). For the absorption test, GST-palladin (Proteintech) and GST were bound to glutathione-sepharose beads (Amersham Biosciences) and then incubated overnight with an anti-palladin polyclonal antibody. After centrifugation, the supernatants were used for western blot analysis. GST protein was purified according the manufacturer’s protocol (Amersham Biosciences). Sciatic nerve transection was performed in adult male Wistar rats (200 g). The animals were deeply anesthetized with sodium pentobarbitol (50 mg/kg i.p.), and sciatic nerve transection and control surgery were performed as described previously [13]. At 3 days after surgery, the rats were anesthetized again and exsanguinated by cutting the carotid vessels, after which the L3–L5 DRGs were immediately removed from the control and operated sides. Frozen sections (18 m thick) were cut on a cryostat. The sequences of the probes for in situ hybridization were checked in a database search to exclude significant homology with other genes. For CLP36 mRNA, we used probes with previously published sequences [13]. For palladin mRNA, a specific oligonucleotide probe (BC127081: nucleotides 2267–2306) was employed. These probes were labeled with [35 S]dATP using terminal deoxynucleotidyl transferase. Sections incubated with control sense probes for CLP36 and palladin did not display any signals (data not shown). For quantitative analysis, DRG neurons of a similar size (n = 30) were selected on both the operated and control sides, and the number of silver grains localized on these neurons was counted. Statistical evaluation was performed using Student’s t-test. For immunohistochemistry, frozen sections were fixed with 4% paraformaldehyde and pretreated with 0.35% hydrogen peroxidase. They were reacted with anti-CLP36 polyclonal Ab (1:200, Proteintech) or anti-palladin polyclonal Ab (1:200), then with polymer-horseradish peroxidaselabeled anti-rabbit secondary Ab (Dako), and were finally visualized with diaminobenzidine and hydrogen peroxidase. The specificity of staining was confirmed by absorption of the antibodies with GST-CLP36 [13] or GST-palladin. Real-time PCR was performed with SYBR Premix ExTaq (Takara) and a Thermal Cycler Dice Real Time System (Takara). Total RNAs from L3 to L5 DRGs were isolated with an RNeasy mini kit (Quiagen) and converted to cDNA by reverse transcription. Primer sequences used for PCR analysis were listed in Supplemental Table 1. Levels of
CLP36, palladin, mystique, and RIL mRNAs were normalized to that of GAPDH mRNA and were shown as ratios to the value of control CLP36 mRNA. PCR-amplified products were also electrophoresed on 1% agarose gels to confirm that single bands were amplified. By screening the E18 mouse whole-body cDNA library, we isolated three clones encoding the in-frame C-terminal region of palladin cDNA, two clones encoding the in-frame full-length ␥filamin, ␣-actinin, telethonin binding protein (FATZ) cDNA, and one clone each encoding in-frame actinin-2, actinin-3, and actinin-4 cDNA fragments. The yeast two-hybrid assay confirmed the specificity of the interaction of these clones with full-length CLP36 (data not shown). Because FATZ is exclusively expressed in muscle tissue [5], further investigation was focused on the relationship between CLP36 and palladin. CLP36 has a PDZ domain at its N-terminal region and palladin carries a putative PDZ-binding motif (EDL) at the end of its C-terminal region. Thus, we speculated that the interaction between CLP36 and palladin involved the PDZ domain of CLP36 and the C-terminus of palladin. To test this hypothesis, we generated deletion mutants of CLP36 and examined their binding to the C-terminal region of palladin by the yeast two-hybrid assay (Fig. 1A). The results showed that all of CLP36 mutants that involve its PDZ domain could bind to the C-terminal region of palladin. However, a palladin mutant lacking the last three amino acids (EDL) was unable to bind with CLP36 (Fig. 1B). These findings supported our hypothesis and indicated that the C-terminus of palladin binds with the PDZ domain of CLP36. To examine the specificity of this interaction, we tested whether the C-terminus of palladin could bind with the PDZ domains of other proteins. We found that the C-terminus of palladin did not interact with the PDZ domain of PSD-95, Pick1, or syntenin (Fig. 1C). However, it interacted with mystique and RIL, which are from the ALP protein family along with CLP36 (Fig. 1D). These results suggested that the C-terminus of palladin preferentially binds to the PDZ domain of proteins from the ALP family. We next investigated whether CLP36 and palladin could form an intracellular protein complex. Initially, we searched for a CLP36–palladin complex in fibroblasts (F2408 cells) because both CLP36 and palladin have been shown to play an important role in stress fiber formation. In F2408 cell lysates, anti-palladin antibody reacted with a major band of ∼90 kDa and also recognized bands of 50 and 70 kDa. These bands coincided with palladin isoforms that have been reported previously [6]. Furthermore, two additional bands were observed at 90–100 kDa. All of these bands disappeared after preadsorption with the antigenic peptide (Fig. 2A, left). A coimmunoprecipitation study demonstrated that palladin was precipitated with anti-CLP36 antibody but not with control IgG (Fig. 2A, right). Interestingly, among several palladin isoforms, the high molecular weight 90 kDa isoform was only precipitated with anti-CLP36 antibody (asterisk). In addition, ␣-actinin was found in the precipitates. To compare the subcellular localization of palladin and CLP36, we stained F2408 cells with an anti-palladin antibody (Fig. 2B, upper) and an anti-CLP36 antibody (Fig. 2B, lower). We found that palladin and CLP36 were localized at stress fibers, with both molecules being detected at the same sites (Fig. 2B, right panel, arrowheads). These findings indicate that CLP36 and palladin are associated with each other in F2408 cells. Similarly, endogenous palladin and ␣-actinin were precipitated from DRG tissue specimens with anti-CLP36 antibody. As seen in F2408 cells, several palladin isoforms were also expressed in DRG tissues and a high molecular weight 90 kDa isoform was only precipitated with antiCLP36 antibody (Fig. 2C). Primary DRG neurons showed intense staining for CLP36 and palladin in the cell body, and staining was also observed along neurites including the growth cones (Fig. 2D). It has been shown that both CLP36 and palladin play a role in neurite outgrowth [2,13]. It is known that neurons increase the synthesis of proteins required for regeneration after nerve injury.
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
55
Fig. 1. Analysis of the interaction between CLP36 and palladin in a yeast two-hybrid system. (A) Interaction of full-length CLP36 and its deletion mutants with the C-terminal region of palladin. All of CLP36 mutants that involve the PDZ domain interacted with palladin. (B) The last three amino acids at the C-terminal of palladin are necessary for binding with CLP36. (C) Palladin does not interact with the PDZ domain of synaptic proteins such as PSD-95, Pick1, and syntenin. (D) Palladin interacts with proteins of the ALP/enigma family, including RIL and mystique as well as CLP36.
Indeed, we recently reported that CLP36 expression is upregulated in peripheral sensory and motor neurons after nerve injury. To investigate whether palladin expression is also regulated by nerve injury, we examined changes of palladin mRNA expression in DRG neurons after sciatic nerve transection and compared the results
with those for CLP36 mRNA. Under normal conditions, CLP36 mRNA was weakly expressed in DRG neurons, but its expression was dramatically upregulated and reached the peak at 3 days after sciatic nerve transection (Fig. 3A). Palladin mRNA expression was very weak in DRG neurons under normal conditions. Sciatic nerve tran-
Fig. 2. Association of endogenous CLP36 with palladin in fibroblasts and DRG neurons. (A) Equal amounts of protein (20 g/lane) from F2408 cells were subjected to SDS-PAGE, transferred to a nitrocellulose membrane, and probed with anti-palladin antibody. For the absorption test, anti-palladin antibody was incubated overnight with GST-palladin or GST (control) (left). F2408 cell lysates were precipitated with anti-CL36 antibody and the immunoprecipitants were blotted with anti-palladin, ␣-actinin, and CLP36 antibodies. Asterisk indicates the band precipitated by anti-CLP36 antibody (right). (B) F2408 cells were double-stained with anti-CLP36 and palladin antibodies. Bar = 10 m. Right photos are regions of interest boxed in the left ones. Arrowheads indicate some examples of sites on stress fibers where CLP36 and palladin are colocalized. (C) DRG tissue lysates were precipitated with anti-CLP36 antibody and the immnoprecipitants were blotted with anti-palladin, ␣-actinin, and CLP36 antibodies. (D) Primary DRG neurons were double-stained with anti-CLP36 and palladin antibodies. Asterisk indicates the band precipitated by anti-CLP36 antibody. Panels on the right show high-magnification views of a growth cone. Arrows indicate colocalization of CLP36 and palladin in the same growth cone. Bars = 20 m.
56
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
Fig. 3. Changes of CLP36 and palladin expressions in DRG neurons after sciatic nerve transection. (A) Real-time PCR analysis of CLP36 and palladin mRNAs in DRG tissues after sciatic nerve transection. The bars represent the standard error of the mean (n = 3). (B) At 3 days after sciatic nerve transection, CLP36 and palladin mRNA expressions were examined by in situ hybridization histochemistry. Dark-field photomicrographs show that CLP36 mRNA is markedly increased in the DRG after sciatic nerve transection (a, b). Bright-field photomicrographs show accumulation of hybridization signals for CLP36 mRNA on DRG neurons (c, d). Sciatic nerve transection also increased palladin mRNA expression in DRG neurons (e–h). (C) Quantitative analysis of the changes of CLP36 and palladin mRNA expressions in DRG neurons after sciatic nerve transection. Statistical evaluation was performed with Student’s t-test (*p < 0.01). (D) Western blot analysis shows an increase in CLP36 protein on day 3 after transection, while there is no significant change of palladin protein. (E) On day 3 after transection, DRG neurons were labeled with anti-CLP36 and palladin antibodies. Right photos show control sections incubated with the respective antibodies preabsorbed by GST-CLP36 (upper) or GST-palladin (lower). Arrows indicate some examples of DRG neurons. Scale bar = 100 m. (F) RT-PCR analysis of RIL and mystique mRNA expressions in DRG tissues on day 3 after sciatic nerve transection. Mystique mRNA was not detected in either of control or operated DRG tissues. (G) Real-time PCR analysis confirmed that sciatic nerve transection did not affect the expression of RIL mRNA in DRG tissues. The error bars represent the standard error of the mean (n = 3). Statistical evaluation was performed with Student’s t-test (*p < 0.01).
section also induced palladin mRNA expression in DRG neurons, but the increase was not so dramatic as that of CLP36 mRNA (Fig. 3A–C). Consistent with these findings, CLP36 and palladin proteins were localized to DRG neurons and there was no significant change in the level of palladin protein in DRG tissues after sciatic nerve transection, in contrast to the increase of CLP36 protein (Fig. 3D and E). These findings indicate that the CLP36 and palladin genes respond differently to nerve injury. While CLP36 mainly partici-
pates in nerve regeneration, palladin is thought to have a role in both normal and regenerating sciatic nerves. As shown in Fig. 1D, RIL and mystique also have the ability to interact with palladin. We wanted to know whether these molecules are expressed in DRG neurons and whether their expression is upregulated similarly to that of CLP36 by sciatic nerve transection. RT-PCR showed that RIL mRNA was expressed in the DRG, whereas mystique mRNA was almost undetectable (Fig. 3F).
T. Hasegawa et al. / Neuroscience Letters 476 (2010) 53–57
Real-time PCR then demonstrated that RIL and mystique mRNA expressions were not affected by sciatic nerve transection (Fig. 3G). These findings suggest that RIL and palladin form a protein complex in DRG neurons. However, it seems possible that CLP36 and RIL participate in different cellular functions after interacting with palladin. In this study, we performed yeast two-hybrid screening and isolated palladin as a binding partner of CLP36. Palladin is an actin cross-linking protein that plays an important role in the assembly and maintenance of various actin-based structures [6]. Both CLP36 and palladin are widely distributed in epithelial and mesenchymal cells. These two proteins are localized at stress fibers and are essential for stress fiber formation in BeWo cells [16] or Rcho-1 cells [14], respectively. In addition, we showed here that endogenous CLP36 and palladin form a complex in F2408 fibroblasts. Very recently, an interaction between palladin and CLP36 has been reported by two different groups [10,12]. Corresponding to our findings, these groups showed that the C-terminus of palladin specifically interacts with the PDZ domain of CLP36. We also demonstrated that a CLP36–palladin complex exists in DRG neurons and in F2408 cells. CLP36 and palladin were colocalized in the cell bodies and neurites of primary DRG neurons. Maeda et al. [10] suggested that a CLP36–palladin complex might interact with ␣-actinin to regulate the actin cytoskeleton. Since CLP36 forms a complex with ␣-actinin in DRG neurons [13], it is therefore possible that CLP36 also forms a functional complex with palladin and ␣-actinin in these neurons. In the nervous system, CLP36 is predominantly expressed by peripheral sensory and motor neurons [13]. We showed that CLP36 mRNA expression is dramatically increased in DRG neurons after sciatic nerve injury, suggesting that CLP36 plays a critical role in the regeneration of primary sensory nerves rather than maintaining their structure or function. Similarly, because palladin has been shown to play a role in neurite extension by cortical neurons [2], we expected that sciatic nerve transection might also increase palladin mRNA expression in DRG neurons. However, there was little change of palladin after nerve injury. These findings led us to the hypothesis that palladin has a role in maintenance of the sensory system under normal conditions and is also involved in regeneration after nerve injury. Palladin interacts with a number of actin-associated proteins, including ezrin, VASP, profiling, LASP-1, and Eps8, and it is a cytoskeletal scaffold protein [6]. Since strong palladin staining has been observed in the substantia gelatinosa of the spinal cord [8], it is probable that palladin is localized to sciatic nerve terminals and participates in organization of the actin cytoskeleton together with palladin-binding proteins. After sciatic nerve injury occurs, palladin may switch its binding partners and form a different complex with CLP36 to regulate rearrangement of the actin cytoskeleton for the promotion of nerve regeneration. Although a variety of ABPs are involved in the mechanism of nerve regeneration, it remains unclear how the functions of these ABPs are coordinated. Considering that CLP36 knockdown led to excessive neuritogenesis in PC12 cells [13], CLP36 is not only a promoter of nerve regeneration but seems to play an important role in this coordination mechanism. Our results indicated that CLP36 might act as an adaptor that links ␣-actinin with palladin in the regenerating sciatic nerve. It has been reported that members of the ALP/enigma protein family bind signaling molecules via the LIM domain [3,9,20,22], and the LIM domain of CLP36 was reported to bind serine/threonine kinase Clik1 [18]. It is therefore possible that signaling proteins such as Clik1 kinase also participate in the CLP36 complex and modulate the functions of ␣-actinin and/or palladin in the regenerating sciatic nerve. However, further studies will be required to examine this hypothesis.
57
Acknowledgements We thank Prof. Heinrich Betz for providing rat CLP36 cDNA. This work was partly supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.neulet.2010.03.081. References [1] K. Bauer, M. Kratzer, M. Otte, K.L. De Quintana, J. Hagmann, G.J. Arnold, C. Eckerskorn, F. Lottspeich, W. Siess, Human CLP36, a PDZ-domain and LIMdomain protein, binds to alpha-actinin-1 and associates with actin filaments and stress fibers in activated platelets and endothelial cells, Blood 96 (2000) 4236–4245. [2] M. Boukhelifa, M.M. Parast, J.G. Valtschanoff, A.S. LaMantia, R.B. Meeker, C.A. Otey, A role for the cytoskeleton-associated protein palladin in neurite outgrowth, Mol. Biol. Cell 12 (2001) 2721–2729. [3] E. Cuppen, H. Gerritis, B. Pepers, B. Wieringa, W. Hendriks, PDZ motifs in PTP-BL and RIL bind to internal protein segments in the LIM domain protein RIL, Mol. Biol. Cell 9 (1998) 671–683. [4] E.W. Dent, F.B. Gertler, Cytoskeletal dynamics and transport in growth cone motility and axon guidance, Neuron 40 (2003) 209–227. [5] G. Faulkner, A. Pallavicini, A. Comelli, M. Salamon, G. Bortoletto, C. Ievolella, S. Trevisan, S. Kojic, F. Dalla Vecchia, P. Laveder, G. Valle, G. Lanfranchi, FATZ, a filamin-, actinin, and telethonin-binding protein of the Z-disc of skeletal muscle, J. Biol. Chem. 275 (2000) 41234–41242. [6] S.M. Goicoechea, D. Arneman, C.A. Otey, The role of palladin in actin organization and cell motility, Eur. J. Cell Biol. 87 (2008) 517–525. [7] P.M. Guy, D.A. Kenny, G.N. Gill, The PDZ domain of the LIM protein enigma binds to beta-tropomyosin, Mol. Biol. Cell 10 (1999) 1973–1983. [8] S.J. Hwang, S. Pagliardini, M. Boukhelifa, M.M. Parast, C.A. Otey, A. Rustioni, J.G. Valtschanoff, Palladin is expressed preferentially in excitatory terminals in the rat central nervous system, J. Comp. Neurol. 436 (2001) 211–224. [9] S. Kuroda, C. Tokunaga, Y. Kiyohara, O. Higuchi, H. Konishi, K. Mizuno, G.N. Gill, U. Kikkawa, Protein–protein interaction of zinc finger LIM domains with protein kinase C, J. Biol. Chem. 271 (1996) 31029–31032. [10] M. Maeda, E. Asano, D. Ito, S. Ito, Y. Hasegawa, M. Hamaguchi, T. Senga, Characterization of interaction between CLP36 and palladin, FEBS J. 276 (2009) 2775–2785. [11] N. Nakagawa, M. Hoshijima, M. Oyasu, N. Saito, K. Tanizawa, S. Kuroda, ENH, containing PDZ and LIM domains, heart/skeletal muscle-specific protein, associates with cytoskeletal proteins through the PDZ domain, Biochem. Biophys. Res. Commun. 272 (2000) 505–512. [12] P. von Nandelstadh, M. Ismail, C. Gardin, H. Suila, I. Zara, A. Belgrano, G. Valle, O. Carpen, G. Faulkner, A class III PDZ binding motif in the myotilin and FATZ families binds enigma family proteins: a common link for Z-disc myopathies, Mol. Cell. Biol. 29 (2009) 822–834. [13] K. Ohno, H. Kato, S. Funahashi, T. Hasegawa, K. Sato, Characterization of CLP36/Elfin/PDLIM1 in the nervous system, J. Neurochem. 111 (2009) 790–800. [14] M.M. Parast, C.A. Otey, Characterization of palladin, a novel protein localized to stress fibers and cell adhesions, J. Cell Biol. 150 (2000) 643–656. [15] D.S. Smith, J.H.P. Skene, A transcriptional-dependent switch controls competence of adult neurons for distinct models of axon growth, J. Neurosci. 17 (1997) 646–658. [16] N. Tamura, K. Ohno, T. Katayama, N. Kanayama, K. Sato, The PDZ–LIM protein CLP36 is required for actin stress fiber formation and focal adhesion assembly in BeWo cells, Biochem. Biophys. Res. Commun. 364 (2007) 589–594. [17] M. Torrado, V.V. Senatorov, R. Trivedi, R.N. Fariss, S.I. Tomarev, Pdlim2, a novel PDZ–LIM domain protein, interacts with alpha-actinins and filamin A, Invest. Ophth. Vis. Sci. 45 (2004) 3955–3963. [18] T. Vallenius, T.P. Mäkelä, Clik1: a novel kinase targeted to actin stress fibers by the CLP-36 PDZ–LIM protein, J. Cell Sci. 15 (2002) 2067–2073. [19] T. Vallenius, B. Scharm, A. Vesikansa, K. Luukko, R. Schäfer, T.P. Mäkelä, The PDZ–LIM protein RIL modulates actin stress fiber turnover and enhances the association of alpha-actinin with F-actin, Exp. Cell Res. 293 (2004) 117–128. [20] R. Wu, K. Durick, Z. Songyang, L.C. Cantley, S.S. Taylor, G.N. Gill, Specificity of LIM domain interactions with receptor tyrosine kinases, J. Biol. Chem. 271 (1996) 15934–15941. [21] H. Xia, S.T. Winokur, W.L. Kuo, M.R. Altherr, D.S. Bredt, Actinin-associated LIM protein: identification of a domain interaction between PDZ and spectrin-like repeated motifs, J. Cell Biol. 139 (1997) 507–515. [22] Q. Zhou, P. Ruiz-Lozano, M.E. Martone, J. Chen, Cypher, a striated musclerestricted PDZ and LIM domain containing protein, binds to alpha-actinin-2 and protein kinase C, J. Biol. Chem. 274 (1999) 19807–19813.