- Email: [email protected]
NXF1, the primary mRNA export receptor, specifically interacts with a neuronal RNA-binding protein HuD
BBRC Biochemical and Biophysical Research Communications 321 (2004) 291–297 www.elsevier.com/locate/ybbrc
TAP/NXF1, the primary mRNA export receptor, specifically interacts with a neuronal RNA-binding protein HuD Kuniaki Saito a, Toshinobu Fujiwara a, Jun Katahira b, Kunio Inoue a, Hiroshi Sakamoto a,* a b
Department of Biology, Graduate School of Science and Technology, Kobe University, 1-1 Rokkodaicho, Nadaku, Kobe 657-8501, Japan Biomolecular Dynamics Group, Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan Received 10 June 2004
Abstract Hu proteins are RNA-binding proteins that are implicated in the control of stabilization, nuclear export, and/or translation of specific mRNAs with AU-rich elements (AREs) in the 3 0 -untranslated region. Three neuron-specific Hu proteins (HuD, HuB, and HuC), but not a ubiquitously expressed Hu protein HuR, have an activity to induce neurite outgrowth when they are overexpressed in a rat neuronal cell line PC12. Here we show that TAP/NXF1, the primary export receptor for the bulk mRNA, is a specific binding partner for HuD. In vitro binding experiments using recombinant proteins revealed that the interaction between TAP and HuD is direct and that HuD can form a ternary complex together with both TAP and RNA. Interestingly, HuR does not interact with TAP. These results suggest that HuD acts as a novel adaptor protein to recruit TAP for efficient export of ARE-containing mRNAs in neuronal cells. Ó 2004 Elsevier Inc. All rights reserved. Keywords: HuD; AU-rich element; Neuron; TAP; mRNA export
Hu proteins are RNA-binding proteins that are homologous to Drosophila ELAV and are highly conserved among vertebrates [1,2]. In mammals, three members of Hu proteins (HuB, HuC, and HuD) are expressed essentially in a neuron-specific manner, while HuR is expressed in various tissues [3–6]. Hu proteins contain three RNA-binding domains (RBDs). The two aminoterminal RBDs and the carboxyl-terminal RBD bind AU-rich elements (AREs) and the poly(A) sequence, respectively [7,8]. Hu proteins are known to shuttle between the nucleus and the cytoplasm, although in the steady-state cells neuron-specific Hu proteins are predominantly localized in the cytoplasm whereas HuR is localized in the nucleus [9–11]. Based on these features, Hu proteins are implicated in the control of stabiliza*
Corresponding author. Fax: +81-78-803-5720. E-mail address: [email protected] (H. Sakamoto).
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.06.140
tion, nuclear export, and/or translation of ARE-containing mRNAs [12–17]. Indeed, there is cumulative evidence for the involvement of Hu proteins in the control of the stability and translation of ARE-containing mRNAs [16,18–21]. Moreover, nucleocytoplasmic shuttling of HuR was suggested to be required for the nuclear export of c-fos mRNA, which possibly depends upon a shuttling sequence of HuR, HNS [22]. Neuron-specific Hu proteins, however, seem to utilize a distinct pathway for nucleocytoplasmic shuttling since they do not contain any functional HNS-like sequence. In eukaryotic cells, export of the bulk mRNA from the nucleus depends upon the primary mRNA export receptor, TAP/NXF1 (here referred to as TAP) [23–25]. TAP was originally identified as a cellular factor to export type D retroviral RNAs to the cytoplasm by binding to the viral RNA element called the constitutive transport element (CTE) [26,27]. TAP interacts with
292
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
REF/Aly, a component of the multiprotein complex that is called exon–exon junction complex (EJC), which is formed on mRNA after splicing of individual introns [28,29]. TAP then exports the EJC-containing mRNAs to the cytoplasm by its nucleocytoplasmic shuttling function. However, it was recently reported that the components of EJC are not essential for the export of bulk mRNA, suggesting additional adaptor protein(s) for TAP-dependent mRNA export [30,31]. In this respect, it was shown that several shuttling SR proteins, which bind an intronless mRNA export element, interact with TAP and act as adaptor proteins for TAP-dependent mRNA export [32]. Since SR proteins are known to bind exonic splicing enhancer elements on mRNAs, the major RNA-binding adaptor for TAP-dependent mRNA export may be such shuttling SR proteins. Thus, direct or indirect interaction with TAP is one of the key mechanisms for RNA export from the nucleus [31,32], and it is possible that TAP adaptors unidentified so far may act to export specific mRNAs and contribute to specific functions in some types of cells, such as neurons. Based on the notion above, we came to think of a working model that neuron-specific Hu proteins utilize the interaction with TAP to exit the nucleus together with specific ARE-containing mRNAs. Here we show that TAP can directly interact with the RNA-bound state of HuD and that the TAP interaction occurs for HuD but not for HuR. Our results suggest that HuD acts as a novel adaptor protein to recruit TAP for efficient export of ARE-containing mRNAs in neurons.
Materials and methods Plasmid construction and preparation of recombinant proteins. The plasmids encoding T7-tagged fusion proteins were described previously [33]. The plasmids encoding GST-Hu proteins were made by introducing the same cDNA fragments as in T7-tagged Hu plasmids followed by the segment encoding six histidine residues into the expression vector pGEX3X (Amersham). The myc-tagged human TAP coding sequence (1–619) was introduced into between the BamHI and XhoI sites of pCDNA3.1 (Clontech). The same TAP cDNA fragment as in the TAP expression vector followed by the segment encoding six histidine residues was introduced into the expression vector pMAL-c2 (NEB). Plasmids of GST and MBP-fusion constructs were transformed into Escherichia coli XL1-blue. These fusion proteins were induced with 1 mM IPTG for 3 h and affinity-purified with His-Trap purification system (Amersham–Pharmacia). For the in vitro pulldown assay, myc-tagged TAP was synthesized in vitro using TNT T7 Coupled Reticulocyte Lysate System (Promega). Cell culture and transfection. PC12 and HeLa cells were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS) and 5% horse serum (for PC12 cells) or 10% FBS (for HeLa cells), respectively. Transient transfection into PC12 cells was performed by electroporation using a GenePulser (Bio-Rad). Transient transfection into HeLa cells was carried out using PolyFect transfection reagent (Qiagen). Immunoprecipitation and immunofluorescence analysis. For immunoprecipitation analysis, HeLa cells transfected with myc-TAP to-
gether with either T7-GFP, T7-HuD, or T7-HuDmt were lysed in TNE buffer (20 mM Tris–HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% NP40, 1 mM PMSF, and 10 lg/ml aprotinin, and 10 lg/ml leupeptin) and immunoprecipitated with anti-T7 monoclonal antibody (Novagen) or anti-myc monoclonal antibody 9E10 (Roche). The precipitated proteins were separated by electrophoresis and immunoblotted with anti-T7 or anti-myc monoclonal antibody. For indirect immunofluorescence analysis to detect endogenous Hu proteins and TAP, NGFdifferentiated PC12 cells were fixed with 4% formaldehyde and then incubated with anti-Hu monoclonal antibody 16A11 and anti-TAP polyclonal antibodies. Subsequently, the cells were incubated with FITC-labelled anti-mouse IgG secondary antibody (DAKO) and Alexa 546 anti-rabbit IgG and then analyzed using a confocal laserscanning microscope (Zeiss LSM5 Pascal). In vitro binding experiment. GST pull-down assay was performed as described previously [33] and the bound proteins were separated by SDS–PAGE and immunoblotted with anti-myc monoclonal antibody (Roche) or anti-MBP antibody (NEB). Poly(U) pull-down assay was performed as described previously [33].
Results Specific and direct interaction between TAP and HuD To test our working model that TAP interacts with HuD, myc-tagged TAP (myc-TAP) was transfected into HeLa cells together with T7-tagged mouse HuD (T7-HuD) or T7-tagged green fluorescent protein (T7-GFP), and then the cell extracts were subjected to immunoprecipitation with anti-T7 monoclonal antibody (Fig. 1A). myc-TAP was coprecipitated with T7-HuD but not with T7-GFP (T7-IP, a-myc). Conversely, when the same extracts were immunoprecipitated with antimyc monoclonal antibody, T7-HuD but not T7-GFP was coprecipitated with myc-TAP (myc-IP, a-T7). These results suggest that TAP specifically interacts with HuD. To examine whether the interaction between TAP and HuD is direct or indirect, we purified TAP fused with maltose-binding protein (MBP–TAP) and HuD fused with glutathione-S-transferase (GST-HuD), and then tested their physical interaction by GST pull-down assay (Fig. 1B). MBP–TAP was pulled down by GST-HuD but not GST alone (a-MBP). Since there was no interaction between GST-HuD and MBP alone, the interaction between TAP and HuD is direct. TAP is ubiquitously expressed in various cell types and is predominantly localized in the nucleus, whereas HuD is expressed specifically in neuronal cells including a rat neuronal cell line PC12 and is predominantly localized in the cytoplasm [11]. In addition, both TAP and HuD are suggested to shuttle between the nucleus and the cytoplasm. To verify the relevance of the TAP– HuD interaction we found, endogenous TAP expression and the interaction between endogenous TAP and HuD were examined in PC12 cells. We first prepared antiTAP rabbit polyclonal antibodies and confirmed that TAP is expressed both in HeLa cells and PC12 cells (Fig. 2A). We then immunostained nerve growth factor
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
Fig. 1. Interaction between TAP and HuD in vivo and in vitro. (A) myc-TAP was cotransfected with either T7-HuD or T7-GFP into HeLa cells. Extracts from the transfected cells were immunoprecipitated with anti-T7 (T7-IP) or anti-myc antibody (myc-IP) and the precipitates were immunoblotted with anti-myc (a-myc) and anti-T7 antibody (a-T7), respectively. For quantitative controls, each extract was immunoblotted with either anti-T7 antibody or anti-myc antibody (Cell Ext.). Asterisks indicate heavy and light chains of IgG used. (B) Affinity-purified MBP–TAP or MBP was incubated with GST-HuD or GST and pulled down with glutathione–Sepharose beads, followed by immunoblotting with anti-MBP antibody (a-MBP). Coomassie brilliant blue (CBB) staining of the same membrane is shown below. Also, CBB staining of the purified proteins used for the pull-down assay is shown on the left.
(NGF)-treated PC12 cells with the anti-TAP antibodies together with anti-Hu monoclonal antibody 16A11 which is specific to neuron-specific Hu proteins (Fig. 2B). A small fraction of endogenous TAP is present in the cytoplasm although it is predominantly localized in the nucleus, indicative of the shuttling nature of TAP. Importantly, TAP and neuron-specific Hu proteins are partly colocalized in the apical end of neurites as well as in the cytoplasm of the cell body. To see the interaction between endogenous TAP and HuD, we tried immunoprecipitation of the cell extracts
293
Fig. 2. Colocalization of endogenous TAP with neuron-specific Hu proteins and the interaction between TAP and HuD in PC12 cells. (A) Preparation of rabbit polyclonal antibodies specific to TAP. Extracts from HeLa cells and PC12 cells were immunoblotted with the antiTAP antibodies. Positions of molecular weight markers are shown on the left. (B) Colocalization of TAP with neuron-specific Hu proteins in PC12 cells. Confocal views of NGF-treated PC12 cells immunostained simultaneously with anti-TAP antibodies (a,d) and anti-Hu 16A11 antibody (b,e). Squared regions in panels a and b are enlarged (d,e). Nomarski view of the same cells is shown (c). A merged view of panels d and e is also shown (f). (C) Specific interaction of endogenous TAP with T7-HuD in PC12 cells. T7-HuD or T7-GFP was transfected into PC12 cells. Extracts from the transfected cells were immunoprecipitated with anti-T7 antibody and the precipitates were immunoblotted with anti-TAP polyclonal antibodies (a-TAP, top). The same membrane was reprobed with anti-T7 antibody (a-T7, bottom).
294
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
from PC12 cells with anti-Hu 16A11 antibody followed by immunoblotting with anti-TAP antibodies. However, coprecipitation of the endogenous TAP with HuD was barely detectable possibly because of some experimental problems such as competition of the anti-Hu monoclonal antibody with endogenous TAP for the HuD-binding (not shown). Therefore, we adopted transfection of T7-HuD into PC12 cells followed by immunoprecipitation with anti-T7 antibody (Fig. 2C). In this case, endogenous TAP was clearly coprecipitated with T7-HuD but not T7-GFP. Taken together, these results indicate that the specific TAP–HuD interaction really occurs in neuronal cells. TAP interacts with both RNA-bound and unbound states of HuD We show that TAP interacts directly with HuD (Fig. 1). HuD is known to bind mRNAs that contain both AREs and the poly(A) sequence [7,8]. Thus, we examine
Fig. 3. Interaction of TAP with HuD is independent of the RNAbinding state of HuD. (A) MBP–TAP was incubated with GST-HuD or GST and pulled down with poly(U)–Sepharose beads, followed by immunoblotting with anti-MBP antibody (a-MBP). One-fifth of the input MBP–TAP is also shown on the left. (B) myc-TAP was cotransfected with T7-HuD, T7-HuDmt or T7-GFP into HeLa cells. Extracts from the transfected cells were immunoprecipitated with antimyc antibody (myc-IP) or anti-T7 antibody (T7-IP) and the precipitates were immunoblotted with anti-T7 antibody (a-T7) or anti-myc antibody (a-myc), respectively. For quantitative controls, each extract was immunoblotted with either anti-T7 antibody or anti-myc antibody (Cell Ext.).
whether the RNA-bound state of HuD can interact with TAP. To do this, we took advantage of the poly(U)binding activity of HuD which represents its AREbinding activity [33]. MBP–TAP was incubated with GST-HuD or GST alone and then the reaction mixtures were subjected to pull-down assay with poly(U)–Sepharose beads. MBP–TAP was pulled down when it was incubated with GST-HuD but not with GST alone, clearly showing that the RNA-bound state of HuD interacts with TAP in vitro (Fig. 3A). Next using cotransfection of myc-TAP with T7-HuD or T7-HuDmt, we examined whether RNA-binding of HuD is required for the interaction with TAP in cells. T7-HuDmt is a mutant which lacks any RNA-binding activity and is localized predominantly in the cytoplasm like the wild type T7-HuD [33]. Both T7-HuD and T7HuDmt were coprecipitated with myc-TAP, and vice versa, myc-TAP was coprecipitated with T7-HuD or T7-HuDmt (Fig. 3B). Control T7-GFP did not show
Fig. 4. Mapping of the interaction domain of TAP with HuD. Various in vitro-translated TAP deletion mutants and EGFP were incubated with GST-HuD and pulled down with glutathione–Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). Structures of deletion mutants are schematically shown above. The regions known to be responsible for the interaction with various factors are also shown on the left. CBB staining of GST-HuD used for the reactions is also shown. One-fifth of the input in vitro-translated products are also shown below.
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
295
any interaction with myc-TAP. These results show that TAP interacts with HuD in cells irrespective of the RNA-binding state of HuD as in in vitro. Mapping of the interaction domains of TAP and HuD TAP is a large protein which contains multiple domains to exert its functions for nuclear export of the bulk cellular mRNA and some viral RNAs with CTE (for review, see [34]). We prepared various in vitrotranslated TAP deletion mutants and examined their interaction with GST-HuD (Fig. 4). Although small deletions from the amino- or carboxyl-termini of TAP allowed the interaction with HuD, the results show that a wide region of TAP-containing REF-, CTE-, and p15binding domains is required for the interaction with HuD. However, since deletions often disrupt the proper structure of their neighboring domains, the HuDinteraction domain of TAP might be located within the region smaller than that we determined here. We next prepared various GST-HuD deletion mutants and examined their interaction with in vitrotranslated myc-TAP (Fig. 5). The results show that the ARE-binding domain (RBD1 and RBD2) with a short carboxyl-flanking region is required and sufficient for the interaction with TAP, although the binding seems slightly weaker than the full-length HuD.
Fig. 6. In vitro-translated myc-TAP was incubated with GST-HuD, GST-HuR or GST and pulled down with glutathione–Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). CBB staining of GST-fusion proteins used for the reactions is shown below.
Correlation of the TAP interaction with the neuriteinducing activity We previously showed that three neuron-specific Hu proteins (HuB, HuC, and HuD) have an activity to induce neurite outgrowth when overexpressed in PC12 cells, whereas HuR has no such a neurite-inducing activity and that cytoplasmic localization of HuD is required for its neurite-inducing activity [11]. This indicates that neuron-specific Hu proteins have some specific function(s) in the cytoplasm in neuronal cells. If the interaction of HuD with TAP is involved in the ability of HuD to induce neurite outgrowth, it would be conceivable that TAP interacts with HuD but not HuR. To test this idea, using GST pull-down assay, we examined whether HuR is different from HuD with respect to the interaction with TAP (Fig. 6). As expected, the results show that TAP does not interact with HuR. Thus, between HuD and HuR, the ability to interact with TAP clearly correlates with the neurite-inducing activity in PC12 cells.
Discussion
Fig. 5. Mapping of the interaction domain of HuD with TAP. In vitro-translated myc-TAP was incubated with various affinity-purified GST-HuD deletion mutants and pulled down with glutathione– Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). Structures of deletion mutants are schematically shown above. CBB staining of GST-fusion proteins used for the reactions is shown below.
It is well accepted that TAP promotes cellular mRNA export through interaction with adaptor proteins that bind mRNAs rather than by direct binding to mRNAs. Such RNA-binding adaptor proteins known to date are only members of the REF family and the SR protein family. In this study, we show that TAP directly interacts with HuD that can bind ARE-containing mRNAs. The two criteria, direct interaction with TAP and RNAbinding ability, suggest that HuD is a novel adaptor protein for TAP-dependent export of ARE-containing mRNAs. Consistently, the RNA-bound state of HuD can interact with TAP in vitro.
296
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
We previously identified a short cis-acting sequence which is responsible for cytoplasmic localization of HuD [11], and recently renamed the cis-acting sequence to cytoplasmic localization signal (CLS). The CLS region (amino acids 277–315) is apparently different from the TAP-interacting region identified in this study (amino acids 52–249). In the previous study, however, we observed that HuD mutants containing the TAP-interacting region but not CLS still show cytoplasmic localization to some extent. Thus, the CLS-dependent cytoplasmic localization may be independent of the possible TAP adaptor function of HuD. Alternatively, an unidentified factor(s) that binds CLS may enhance the TAP adaptor function of HuD, leading to efficient TAP-dependent export of ARE-containing mRNAs. In this study, we also show that HuR does not interact with TAP. Therefore, HuR seems to shuttle in a TAP-independent manner. Indeed, nucleocytoplasmic shuttling of HuR is known to utilize two independent pathways mediated by different nuclear export receptors, TRN2 and CRM1 [22]. Interestingly, comparing HuD with HuR, the TAP interaction correlates with the neurite-inducing activity. This correlation suggests not only that HuB and HuC may also interact with TAP, but also that HuD may export a specific set of target mRNAs distinct from those exported by HuR in neuronal cells, although both sets of target mRNAs perhaps contain ARE in the 3 0 -untranslated region. HuD may also act together with other TAP adaptor proteins, REFs and SR proteins, and such collaboration of multiple export adaptors recruits multiple molecules of TAP on the single mRNAs to achieve efficient export of the target mRNAs for HuD. Moreover, we unexpectedly observed that TAP is colocalized with neuron-specific Hu proteins even in the neurites induced by NGF, suggesting that TAP might have an additional cytoplasmic role other than mRNA export, such as translational regulation [35], which is involved in the physiological function of neuron-specific Hu proteins. The identification of mRNAs that bind the TAP–HuD complex will provide important insights into the function of neuron-specific Hu proteins.
Acknowledgments We thank Dr. Hans J. Gross for critical comments and advice on the manuscript. This work was supported by the research Grant from MEXT (No. 14035235) to H.S.
References [1] A. Szabo, J. Dalmau, G. Manley, M. Rosenfeld, E. Wong, J. Henson, J.B. Posner, H.M. Furneaux, HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal, Cell 67 (1991) 325–333.
[2] T.D. Levine, F. Gao, P.H. King, L.G. Andrews, J.D. Keene, HelN1: an autoimmune RNA-binding protein with specificity for 3 0 uridylate-rich untranslated regions of growth factor mRNAs, Mol. Cell. Biol. 13 (1993) 3494–3504. [3] Y. Wakamatsu, J.A. Weston, Sequential expression and role of Hu RNA-binding proteins during neurogenesis, Development 124 (1997) 3449–3460. [4] H.J. Okano, R.B. Darnell, A hierarchy of Hu RNA binding proteins in developing and adult neurons, J. Neurosci. 17 (1997) 3024–3037. [5] P.H. King, Hel-N2: a novel isoform of Hel-N1 which is conserved in rat neural tissue and produced in early embryogenesis, Gene 151 (1994) 261–265. [6] R. Abe, Y. Uyeno, K. Yamamoto, H. Sakamoto, Tissue-specific expression of the gene encoding a mouse RNA binding protein homologous to human HuD antigen, DNA Res. 1 (1994) 175–180. [7] W.J. Ma, S. Chung, H. Furneaux, The Elav-like proteins bind to AU-rich elements and to the poly(A) tail of mRNA, Nucleic Acids Res. 25 (1997) 3564–3569. [8] R. Abe, E. Sakashita, K. Yamamoto, H. Sakamoto, Two different RNA binding activities for the AU-rich element and the poly (A) sequence of the mouse neuronal protein mHuC, Nucleic Acids Res. 24 (1996) 4895–4901. [9] X.C. Fan, J.A. Steitz, HNS, a nuclear-cytoplasmic shuttling sequence in HuR, Proc. Natl. Acad. Sci. USA 95 (1998) 15293– 15298. [10] X.C. Fan, J.A. Steitz, Overexpression of HuR, a nuclearcytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs, EMBO J. 17 (1998) 3448–3460. [11] K. Kasashima, K. Terashima, K. Yamamoto, E. Sakashita, H. Sakamoto, Cytoplasmic localization is required for the mammalian ELAV-like protein HuD to induce neuronal differentiation, Genes Cells 4 (1999) 667–683. [12] I. Yaman, J. Fernandez, B. Sarkar, R.J. Schneider, M.D. Snider, L.E. Nagy, M. Hatzoglou, Nutritional control of mRNA stability is mediated by a conserved AU-rich element that binds the cytoplasmic shuttling protein HuR, J. Biol. Chem. 277 (2002) 41539–41546. [13] W. Wang, H. Furneaux, H. Cheng, M.C. Caldwell, D. Hutter, Y. Liu, N. Holbrook, M. Gorospe, HuR regulates p21 mRNA stabilization by UV light, Mol. Cell. Biol. 20 (2000) 760–769. [14] S. Sengupta, B.C. Jang, M.T. Wu, J.H. Paik, H. Furneaux, T. Hla, The RNA-binding protein HuR regulates the expression of cyclooxygenase-2, J. Biol. Chem. 278 (2003) 25227–25233. [15] K. Sakai, Y. Kitagawa, G. Hirose, Binding of neuronal ELAVlike proteins to the uridine-rich sequence in the 3 0 -untranslated region of tumor necrosis factor-alpha messenger RNA, FEBS Lett. 446 (1999) 157–162. [16] K. Mazan-Mamczarz, S. Galban, I. Lopez de Silanes, J.L. Martindale, U. Atasoy, J.D. Keene, M. Gorospe, RNA-binding protein HuR enhances p53 translation in response to ultraviolet light irradiation, Proc. Natl. Acad. Sci. USA 100 (2003) 8354– 8359. [17] P.H. King, RNA-binding analyses of HuC and HuD with the VEGF and c-myc 3 0 -untranslated regions using a novel ELISAbased assay, Nucleic Acids Res. 28 (2000) E20. [18] W. Wang, M.C. Caldwell, S. Lin, H. Furneaux, M. Gorospe, HuR regulates cyclin A and cyclin B1 mRNA stability during cell proliferation, EMBO J. 19 (2000) 2340–2350. [19] A.C. Beckel-Mitchener, A. Miera, R. Keller, N.I. PerroneBizzozero, Poly (A) tail length-dependent stabilization of GAP43 mRNA by the RNA-binding protein HuD, J. Biol. Chem. 277 (2002) 27996–28002. [20] D. Antic, N. Lu, J.D. Keene, ELAV tumor antigen, Hel-N1, increases translation of neurofilament M mRNA and induces formation of neurites in human teratocarcinoma cells, Genes Dev. 13 (1999) 449–461.
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297 [21] C.F. Manohar, M.L. Short, A. Nguyen, N.N. Nguyen, D. Chagnovich, Q. Yang, S.L. Cohn, HuD, a neuronal-specific RNA-binding protein, increases the in vivo stability of MYCN RNA, J. Biol. Chem. 277 (2002) 1967–1973. [22] I.E. Gallouzi, J.A. Steitz, Delineation of mRNA export pathways by the use of cell-permeable peptides, Science 294 (2001) 1895– 1901. [23] J. Katahira, K. Strasser, A. Podtelejnikov, M. Mann, J.U. Jung, E. Hurt, The Mex67p-mediated nuclear mRNA export pathway is conserved from yeast to human, EMBO J. 18 (1999) 2593–2609. [24] G.S. Wilkie, V. Zimyanin, R. Kirby, C. Korey, H. Francis-Lang, D. Van Vactor, I. Davis, Small bristles, the Drosophila ortholog of NXF-1, is essential for mRNA export throughout development, RNA 7 (2001) 1781–1792. [25] W. Tan, A.S. Zolotukhin, J. Bear, D.J. Patenaude, B.K. Felber, The mRNA export in Caenorhabditis elegans is mediated by CeNXF-1, an ortholog of human TAP/NXF and Saccharomyces cerevisiae Mex67p, RNA 6 (2000) 1762–1772. [26] I.C. Braun, E. Rohrbach, C. Schmitt, E. Izaurralde, TAP binds to the constitutive transport element (CTE) through a novel RNAbinding motif that is sufficient to promote CTE-dependent RNA export from the nucleus, EMBO J. 18 (1999) 1953–1965. [27] Y. Kang, H.P. Bogerd, B.R. Cullen, Analysis of cellular factors that mediate nuclear export of RNAs bearing the Mason-Pfizer monkey virus constitutive transport element, J. Virol. 74 (2000) 5863–5871.
297
[28] V.N. Kim, J. Yong, N. Kataoka, L. Abel, M.D. Diem, G. Dreyfuss, The Y14 protein communicates to the cytoplasm the position of exon–exon junctions, EMBO J. 20 (2001) 2062–2068. [29] H. Le Hir, D. Gatfield, E. Izaurralde, M.J. Moore, The exon– exon junction complex provides a binding platform for factors involved in mRNA export and nonsense-mediated mRNA decay, EMBO J. 20 (2001) 4987–4997. [30] D. Longman, I.L. Johnstone, J.F. Caceres, The Ref/Aly proteins are dispensable for mRNA export and development in Caenorhabditis elegans, RNA 9 (2003) 881–891. [31] D. Gatfield, E. Izaurralde, REF1/Aly and the additional exon junction complex proteins are dispensable for nuclear mRNA export, J. Cell Biol. 159 (2002) 579–588. [32] Y. Huang, R. Gattoni, J. Stevenin, J.A. Steitz, SR splicing factors serve as adapter proteins for TAP-dependent mRNA export, Mol. Cell 11 (2003) 837–843. [33] K. Kasashima, E. Sakashita, K. Saito, H. Sakamoto, Complex formation of the neuron-specific ELAV-like Hu RNA-binding proteins, Nucleic Acids Res. 30 (2002) 4519–4526. [34] E. Izaurralde, Friedrich Miescher Prize awardee lecture review. A conserved family of nuclear export receptors mediates the exit of messenger RNA to the cytoplasm, Cell Mol. Life Sci. 58 (2001) 1105–1112. [35] L. Jin, B.W. Guzik, Y.C. Bor, D. Rekosh, M.L. Hammarskjold, Tap and NXT promote translation of unspliced mRNA, Genes Dev. 17 (2003) 3075–3086.
TAP/NXF1, the primary mRNA export receptor, specifically interacts with a neuronal RNA-binding protein HuD Kuniaki Saito a, Toshinobu Fujiwara a, Jun Katahira b, Kunio Inoue a, Hiroshi Sakamoto a,* a b
Department of Biology, Graduate School of Science and Technology, Kobe University, 1-1 Rokkodaicho, Nadaku, Kobe 657-8501, Japan Biomolecular Dynamics Group, Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan Received 10 June 2004
Abstract Hu proteins are RNA-binding proteins that are implicated in the control of stabilization, nuclear export, and/or translation of specific mRNAs with AU-rich elements (AREs) in the 3 0 -untranslated region. Three neuron-specific Hu proteins (HuD, HuB, and HuC), but not a ubiquitously expressed Hu protein HuR, have an activity to induce neurite outgrowth when they are overexpressed in a rat neuronal cell line PC12. Here we show that TAP/NXF1, the primary export receptor for the bulk mRNA, is a specific binding partner for HuD. In vitro binding experiments using recombinant proteins revealed that the interaction between TAP and HuD is direct and that HuD can form a ternary complex together with both TAP and RNA. Interestingly, HuR does not interact with TAP. These results suggest that HuD acts as a novel adaptor protein to recruit TAP for efficient export of ARE-containing mRNAs in neuronal cells. Ó 2004 Elsevier Inc. All rights reserved. Keywords: HuD; AU-rich element; Neuron; TAP; mRNA export
Hu proteins are RNA-binding proteins that are homologous to Drosophila ELAV and are highly conserved among vertebrates [1,2]. In mammals, three members of Hu proteins (HuB, HuC, and HuD) are expressed essentially in a neuron-specific manner, while HuR is expressed in various tissues [3–6]. Hu proteins contain three RNA-binding domains (RBDs). The two aminoterminal RBDs and the carboxyl-terminal RBD bind AU-rich elements (AREs) and the poly(A) sequence, respectively [7,8]. Hu proteins are known to shuttle between the nucleus and the cytoplasm, although in the steady-state cells neuron-specific Hu proteins are predominantly localized in the cytoplasm whereas HuR is localized in the nucleus [9–11]. Based on these features, Hu proteins are implicated in the control of stabiliza*
Corresponding author. Fax: +81-78-803-5720. E-mail address: [email protected] (H. Sakamoto).
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.06.140
tion, nuclear export, and/or translation of ARE-containing mRNAs [12–17]. Indeed, there is cumulative evidence for the involvement of Hu proteins in the control of the stability and translation of ARE-containing mRNAs [16,18–21]. Moreover, nucleocytoplasmic shuttling of HuR was suggested to be required for the nuclear export of c-fos mRNA, which possibly depends upon a shuttling sequence of HuR, HNS [22]. Neuron-specific Hu proteins, however, seem to utilize a distinct pathway for nucleocytoplasmic shuttling since they do not contain any functional HNS-like sequence. In eukaryotic cells, export of the bulk mRNA from the nucleus depends upon the primary mRNA export receptor, TAP/NXF1 (here referred to as TAP) [23–25]. TAP was originally identified as a cellular factor to export type D retroviral RNAs to the cytoplasm by binding to the viral RNA element called the constitutive transport element (CTE) [26,27]. TAP interacts with
292
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
REF/Aly, a component of the multiprotein complex that is called exon–exon junction complex (EJC), which is formed on mRNA after splicing of individual introns [28,29]. TAP then exports the EJC-containing mRNAs to the cytoplasm by its nucleocytoplasmic shuttling function. However, it was recently reported that the components of EJC are not essential for the export of bulk mRNA, suggesting additional adaptor protein(s) for TAP-dependent mRNA export [30,31]. In this respect, it was shown that several shuttling SR proteins, which bind an intronless mRNA export element, interact with TAP and act as adaptor proteins for TAP-dependent mRNA export [32]. Since SR proteins are known to bind exonic splicing enhancer elements on mRNAs, the major RNA-binding adaptor for TAP-dependent mRNA export may be such shuttling SR proteins. Thus, direct or indirect interaction with TAP is one of the key mechanisms for RNA export from the nucleus [31,32], and it is possible that TAP adaptors unidentified so far may act to export specific mRNAs and contribute to specific functions in some types of cells, such as neurons. Based on the notion above, we came to think of a working model that neuron-specific Hu proteins utilize the interaction with TAP to exit the nucleus together with specific ARE-containing mRNAs. Here we show that TAP can directly interact with the RNA-bound state of HuD and that the TAP interaction occurs for HuD but not for HuR. Our results suggest that HuD acts as a novel adaptor protein to recruit TAP for efficient export of ARE-containing mRNAs in neurons.
Materials and methods Plasmid construction and preparation of recombinant proteins. The plasmids encoding T7-tagged fusion proteins were described previously [33]. The plasmids encoding GST-Hu proteins were made by introducing the same cDNA fragments as in T7-tagged Hu plasmids followed by the segment encoding six histidine residues into the expression vector pGEX3X (Amersham). The myc-tagged human TAP coding sequence (1–619) was introduced into between the BamHI and XhoI sites of pCDNA3.1 (Clontech). The same TAP cDNA fragment as in the TAP expression vector followed by the segment encoding six histidine residues was introduced into the expression vector pMAL-c2 (NEB). Plasmids of GST and MBP-fusion constructs were transformed into Escherichia coli XL1-blue. These fusion proteins were induced with 1 mM IPTG for 3 h and affinity-purified with His-Trap purification system (Amersham–Pharmacia). For the in vitro pulldown assay, myc-tagged TAP was synthesized in vitro using TNT T7 Coupled Reticulocyte Lysate System (Promega). Cell culture and transfection. PC12 and HeLa cells were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS) and 5% horse serum (for PC12 cells) or 10% FBS (for HeLa cells), respectively. Transient transfection into PC12 cells was performed by electroporation using a GenePulser (Bio-Rad). Transient transfection into HeLa cells was carried out using PolyFect transfection reagent (Qiagen). Immunoprecipitation and immunofluorescence analysis. For immunoprecipitation analysis, HeLa cells transfected with myc-TAP to-
gether with either T7-GFP, T7-HuD, or T7-HuDmt were lysed in TNE buffer (20 mM Tris–HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% NP40, 1 mM PMSF, and 10 lg/ml aprotinin, and 10 lg/ml leupeptin) and immunoprecipitated with anti-T7 monoclonal antibody (Novagen) or anti-myc monoclonal antibody 9E10 (Roche). The precipitated proteins were separated by electrophoresis and immunoblotted with anti-T7 or anti-myc monoclonal antibody. For indirect immunofluorescence analysis to detect endogenous Hu proteins and TAP, NGFdifferentiated PC12 cells were fixed with 4% formaldehyde and then incubated with anti-Hu monoclonal antibody 16A11 and anti-TAP polyclonal antibodies. Subsequently, the cells were incubated with FITC-labelled anti-mouse IgG secondary antibody (DAKO) and Alexa 546 anti-rabbit IgG and then analyzed using a confocal laserscanning microscope (Zeiss LSM5 Pascal). In vitro binding experiment. GST pull-down assay was performed as described previously [33] and the bound proteins were separated by SDS–PAGE and immunoblotted with anti-myc monoclonal antibody (Roche) or anti-MBP antibody (NEB). Poly(U) pull-down assay was performed as described previously [33].
Results Specific and direct interaction between TAP and HuD To test our working model that TAP interacts with HuD, myc-tagged TAP (myc-TAP) was transfected into HeLa cells together with T7-tagged mouse HuD (T7-HuD) or T7-tagged green fluorescent protein (T7-GFP), and then the cell extracts were subjected to immunoprecipitation with anti-T7 monoclonal antibody (Fig. 1A). myc-TAP was coprecipitated with T7-HuD but not with T7-GFP (T7-IP, a-myc). Conversely, when the same extracts were immunoprecipitated with antimyc monoclonal antibody, T7-HuD but not T7-GFP was coprecipitated with myc-TAP (myc-IP, a-T7). These results suggest that TAP specifically interacts with HuD. To examine whether the interaction between TAP and HuD is direct or indirect, we purified TAP fused with maltose-binding protein (MBP–TAP) and HuD fused with glutathione-S-transferase (GST-HuD), and then tested their physical interaction by GST pull-down assay (Fig. 1B). MBP–TAP was pulled down by GST-HuD but not GST alone (a-MBP). Since there was no interaction between GST-HuD and MBP alone, the interaction between TAP and HuD is direct. TAP is ubiquitously expressed in various cell types and is predominantly localized in the nucleus, whereas HuD is expressed specifically in neuronal cells including a rat neuronal cell line PC12 and is predominantly localized in the cytoplasm [11]. In addition, both TAP and HuD are suggested to shuttle between the nucleus and the cytoplasm. To verify the relevance of the TAP– HuD interaction we found, endogenous TAP expression and the interaction between endogenous TAP and HuD were examined in PC12 cells. We first prepared antiTAP rabbit polyclonal antibodies and confirmed that TAP is expressed both in HeLa cells and PC12 cells (Fig. 2A). We then immunostained nerve growth factor
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
Fig. 1. Interaction between TAP and HuD in vivo and in vitro. (A) myc-TAP was cotransfected with either T7-HuD or T7-GFP into HeLa cells. Extracts from the transfected cells were immunoprecipitated with anti-T7 (T7-IP) or anti-myc antibody (myc-IP) and the precipitates were immunoblotted with anti-myc (a-myc) and anti-T7 antibody (a-T7), respectively. For quantitative controls, each extract was immunoblotted with either anti-T7 antibody or anti-myc antibody (Cell Ext.). Asterisks indicate heavy and light chains of IgG used. (B) Affinity-purified MBP–TAP or MBP was incubated with GST-HuD or GST and pulled down with glutathione–Sepharose beads, followed by immunoblotting with anti-MBP antibody (a-MBP). Coomassie brilliant blue (CBB) staining of the same membrane is shown below. Also, CBB staining of the purified proteins used for the pull-down assay is shown on the left.
(NGF)-treated PC12 cells with the anti-TAP antibodies together with anti-Hu monoclonal antibody 16A11 which is specific to neuron-specific Hu proteins (Fig. 2B). A small fraction of endogenous TAP is present in the cytoplasm although it is predominantly localized in the nucleus, indicative of the shuttling nature of TAP. Importantly, TAP and neuron-specific Hu proteins are partly colocalized in the apical end of neurites as well as in the cytoplasm of the cell body. To see the interaction between endogenous TAP and HuD, we tried immunoprecipitation of the cell extracts
293
Fig. 2. Colocalization of endogenous TAP with neuron-specific Hu proteins and the interaction between TAP and HuD in PC12 cells. (A) Preparation of rabbit polyclonal antibodies specific to TAP. Extracts from HeLa cells and PC12 cells were immunoblotted with the antiTAP antibodies. Positions of molecular weight markers are shown on the left. (B) Colocalization of TAP with neuron-specific Hu proteins in PC12 cells. Confocal views of NGF-treated PC12 cells immunostained simultaneously with anti-TAP antibodies (a,d) and anti-Hu 16A11 antibody (b,e). Squared regions in panels a and b are enlarged (d,e). Nomarski view of the same cells is shown (c). A merged view of panels d and e is also shown (f). (C) Specific interaction of endogenous TAP with T7-HuD in PC12 cells. T7-HuD or T7-GFP was transfected into PC12 cells. Extracts from the transfected cells were immunoprecipitated with anti-T7 antibody and the precipitates were immunoblotted with anti-TAP polyclonal antibodies (a-TAP, top). The same membrane was reprobed with anti-T7 antibody (a-T7, bottom).
294
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
from PC12 cells with anti-Hu 16A11 antibody followed by immunoblotting with anti-TAP antibodies. However, coprecipitation of the endogenous TAP with HuD was barely detectable possibly because of some experimental problems such as competition of the anti-Hu monoclonal antibody with endogenous TAP for the HuD-binding (not shown). Therefore, we adopted transfection of T7-HuD into PC12 cells followed by immunoprecipitation with anti-T7 antibody (Fig. 2C). In this case, endogenous TAP was clearly coprecipitated with T7-HuD but not T7-GFP. Taken together, these results indicate that the specific TAP–HuD interaction really occurs in neuronal cells. TAP interacts with both RNA-bound and unbound states of HuD We show that TAP interacts directly with HuD (Fig. 1). HuD is known to bind mRNAs that contain both AREs and the poly(A) sequence [7,8]. Thus, we examine
Fig. 3. Interaction of TAP with HuD is independent of the RNAbinding state of HuD. (A) MBP–TAP was incubated with GST-HuD or GST and pulled down with poly(U)–Sepharose beads, followed by immunoblotting with anti-MBP antibody (a-MBP). One-fifth of the input MBP–TAP is also shown on the left. (B) myc-TAP was cotransfected with T7-HuD, T7-HuDmt or T7-GFP into HeLa cells. Extracts from the transfected cells were immunoprecipitated with antimyc antibody (myc-IP) or anti-T7 antibody (T7-IP) and the precipitates were immunoblotted with anti-T7 antibody (a-T7) or anti-myc antibody (a-myc), respectively. For quantitative controls, each extract was immunoblotted with either anti-T7 antibody or anti-myc antibody (Cell Ext.).
whether the RNA-bound state of HuD can interact with TAP. To do this, we took advantage of the poly(U)binding activity of HuD which represents its AREbinding activity [33]. MBP–TAP was incubated with GST-HuD or GST alone and then the reaction mixtures were subjected to pull-down assay with poly(U)–Sepharose beads. MBP–TAP was pulled down when it was incubated with GST-HuD but not with GST alone, clearly showing that the RNA-bound state of HuD interacts with TAP in vitro (Fig. 3A). Next using cotransfection of myc-TAP with T7-HuD or T7-HuDmt, we examined whether RNA-binding of HuD is required for the interaction with TAP in cells. T7-HuDmt is a mutant which lacks any RNA-binding activity and is localized predominantly in the cytoplasm like the wild type T7-HuD [33]. Both T7-HuD and T7HuDmt were coprecipitated with myc-TAP, and vice versa, myc-TAP was coprecipitated with T7-HuD or T7-HuDmt (Fig. 3B). Control T7-GFP did not show
Fig. 4. Mapping of the interaction domain of TAP with HuD. Various in vitro-translated TAP deletion mutants and EGFP were incubated with GST-HuD and pulled down with glutathione–Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). Structures of deletion mutants are schematically shown above. The regions known to be responsible for the interaction with various factors are also shown on the left. CBB staining of GST-HuD used for the reactions is also shown. One-fifth of the input in vitro-translated products are also shown below.
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
295
any interaction with myc-TAP. These results show that TAP interacts with HuD in cells irrespective of the RNA-binding state of HuD as in in vitro. Mapping of the interaction domains of TAP and HuD TAP is a large protein which contains multiple domains to exert its functions for nuclear export of the bulk cellular mRNA and some viral RNAs with CTE (for review, see [34]). We prepared various in vitrotranslated TAP deletion mutants and examined their interaction with GST-HuD (Fig. 4). Although small deletions from the amino- or carboxyl-termini of TAP allowed the interaction with HuD, the results show that a wide region of TAP-containing REF-, CTE-, and p15binding domains is required for the interaction with HuD. However, since deletions often disrupt the proper structure of their neighboring domains, the HuDinteraction domain of TAP might be located within the region smaller than that we determined here. We next prepared various GST-HuD deletion mutants and examined their interaction with in vitrotranslated myc-TAP (Fig. 5). The results show that the ARE-binding domain (RBD1 and RBD2) with a short carboxyl-flanking region is required and sufficient for the interaction with TAP, although the binding seems slightly weaker than the full-length HuD.
Fig. 6. In vitro-translated myc-TAP was incubated with GST-HuD, GST-HuR or GST and pulled down with glutathione–Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). CBB staining of GST-fusion proteins used for the reactions is shown below.
Correlation of the TAP interaction with the neuriteinducing activity We previously showed that three neuron-specific Hu proteins (HuB, HuC, and HuD) have an activity to induce neurite outgrowth when overexpressed in PC12 cells, whereas HuR has no such a neurite-inducing activity and that cytoplasmic localization of HuD is required for its neurite-inducing activity [11]. This indicates that neuron-specific Hu proteins have some specific function(s) in the cytoplasm in neuronal cells. If the interaction of HuD with TAP is involved in the ability of HuD to induce neurite outgrowth, it would be conceivable that TAP interacts with HuD but not HuR. To test this idea, using GST pull-down assay, we examined whether HuR is different from HuD with respect to the interaction with TAP (Fig. 6). As expected, the results show that TAP does not interact with HuR. Thus, between HuD and HuR, the ability to interact with TAP clearly correlates with the neurite-inducing activity in PC12 cells.
Discussion
Fig. 5. Mapping of the interaction domain of HuD with TAP. In vitro-translated myc-TAP was incubated with various affinity-purified GST-HuD deletion mutants and pulled down with glutathione– Sepharose beads, followed by immunoblotting with the anti-myc antibody (a-myc). Structures of deletion mutants are schematically shown above. CBB staining of GST-fusion proteins used for the reactions is shown below.
It is well accepted that TAP promotes cellular mRNA export through interaction with adaptor proteins that bind mRNAs rather than by direct binding to mRNAs. Such RNA-binding adaptor proteins known to date are only members of the REF family and the SR protein family. In this study, we show that TAP directly interacts with HuD that can bind ARE-containing mRNAs. The two criteria, direct interaction with TAP and RNAbinding ability, suggest that HuD is a novel adaptor protein for TAP-dependent export of ARE-containing mRNAs. Consistently, the RNA-bound state of HuD can interact with TAP in vitro.
296
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297
We previously identified a short cis-acting sequence which is responsible for cytoplasmic localization of HuD [11], and recently renamed the cis-acting sequence to cytoplasmic localization signal (CLS). The CLS region (amino acids 277–315) is apparently different from the TAP-interacting region identified in this study (amino acids 52–249). In the previous study, however, we observed that HuD mutants containing the TAP-interacting region but not CLS still show cytoplasmic localization to some extent. Thus, the CLS-dependent cytoplasmic localization may be independent of the possible TAP adaptor function of HuD. Alternatively, an unidentified factor(s) that binds CLS may enhance the TAP adaptor function of HuD, leading to efficient TAP-dependent export of ARE-containing mRNAs. In this study, we also show that HuR does not interact with TAP. Therefore, HuR seems to shuttle in a TAP-independent manner. Indeed, nucleocytoplasmic shuttling of HuR is known to utilize two independent pathways mediated by different nuclear export receptors, TRN2 and CRM1 [22]. Interestingly, comparing HuD with HuR, the TAP interaction correlates with the neurite-inducing activity. This correlation suggests not only that HuB and HuC may also interact with TAP, but also that HuD may export a specific set of target mRNAs distinct from those exported by HuR in neuronal cells, although both sets of target mRNAs perhaps contain ARE in the 3 0 -untranslated region. HuD may also act together with other TAP adaptor proteins, REFs and SR proteins, and such collaboration of multiple export adaptors recruits multiple molecules of TAP on the single mRNAs to achieve efficient export of the target mRNAs for HuD. Moreover, we unexpectedly observed that TAP is colocalized with neuron-specific Hu proteins even in the neurites induced by NGF, suggesting that TAP might have an additional cytoplasmic role other than mRNA export, such as translational regulation [35], which is involved in the physiological function of neuron-specific Hu proteins. The identification of mRNAs that bind the TAP–HuD complex will provide important insights into the function of neuron-specific Hu proteins.
Acknowledgments We thank Dr. Hans J. Gross for critical comments and advice on the manuscript. This work was supported by the research Grant from MEXT (No. 14035235) to H.S.
References [1] A. Szabo, J. Dalmau, G. Manley, M. Rosenfeld, E. Wong, J. Henson, J.B. Posner, H.M. Furneaux, HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal, Cell 67 (1991) 325–333.
[2] T.D. Levine, F. Gao, P.H. King, L.G. Andrews, J.D. Keene, HelN1: an autoimmune RNA-binding protein with specificity for 3 0 uridylate-rich untranslated regions of growth factor mRNAs, Mol. Cell. Biol. 13 (1993) 3494–3504. [3] Y. Wakamatsu, J.A. Weston, Sequential expression and role of Hu RNA-binding proteins during neurogenesis, Development 124 (1997) 3449–3460. [4] H.J. Okano, R.B. Darnell, A hierarchy of Hu RNA binding proteins in developing and adult neurons, J. Neurosci. 17 (1997) 3024–3037. [5] P.H. King, Hel-N2: a novel isoform of Hel-N1 which is conserved in rat neural tissue and produced in early embryogenesis, Gene 151 (1994) 261–265. [6] R. Abe, Y. Uyeno, K. Yamamoto, H. Sakamoto, Tissue-specific expression of the gene encoding a mouse RNA binding protein homologous to human HuD antigen, DNA Res. 1 (1994) 175–180. [7] W.J. Ma, S. Chung, H. Furneaux, The Elav-like proteins bind to AU-rich elements and to the poly(A) tail of mRNA, Nucleic Acids Res. 25 (1997) 3564–3569. [8] R. Abe, E. Sakashita, K. Yamamoto, H. Sakamoto, Two different RNA binding activities for the AU-rich element and the poly (A) sequence of the mouse neuronal protein mHuC, Nucleic Acids Res. 24 (1996) 4895–4901. [9] X.C. Fan, J.A. Steitz, HNS, a nuclear-cytoplasmic shuttling sequence in HuR, Proc. Natl. Acad. Sci. USA 95 (1998) 15293– 15298. [10] X.C. Fan, J.A. Steitz, Overexpression of HuR, a nuclearcytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs, EMBO J. 17 (1998) 3448–3460. [11] K. Kasashima, K. Terashima, K. Yamamoto, E. Sakashita, H. Sakamoto, Cytoplasmic localization is required for the mammalian ELAV-like protein HuD to induce neuronal differentiation, Genes Cells 4 (1999) 667–683. [12] I. Yaman, J. Fernandez, B. Sarkar, R.J. Schneider, M.D. Snider, L.E. Nagy, M. Hatzoglou, Nutritional control of mRNA stability is mediated by a conserved AU-rich element that binds the cytoplasmic shuttling protein HuR, J. Biol. Chem. 277 (2002) 41539–41546. [13] W. Wang, H. Furneaux, H. Cheng, M.C. Caldwell, D. Hutter, Y. Liu, N. Holbrook, M. Gorospe, HuR regulates p21 mRNA stabilization by UV light, Mol. Cell. Biol. 20 (2000) 760–769. [14] S. Sengupta, B.C. Jang, M.T. Wu, J.H. Paik, H. Furneaux, T. Hla, The RNA-binding protein HuR regulates the expression of cyclooxygenase-2, J. Biol. Chem. 278 (2003) 25227–25233. [15] K. Sakai, Y. Kitagawa, G. Hirose, Binding of neuronal ELAVlike proteins to the uridine-rich sequence in the 3 0 -untranslated region of tumor necrosis factor-alpha messenger RNA, FEBS Lett. 446 (1999) 157–162. [16] K. Mazan-Mamczarz, S. Galban, I. Lopez de Silanes, J.L. Martindale, U. Atasoy, J.D. Keene, M. Gorospe, RNA-binding protein HuR enhances p53 translation in response to ultraviolet light irradiation, Proc. Natl. Acad. Sci. USA 100 (2003) 8354– 8359. [17] P.H. King, RNA-binding analyses of HuC and HuD with the VEGF and c-myc 3 0 -untranslated regions using a novel ELISAbased assay, Nucleic Acids Res. 28 (2000) E20. [18] W. Wang, M.C. Caldwell, S. Lin, H. Furneaux, M. Gorospe, HuR regulates cyclin A and cyclin B1 mRNA stability during cell proliferation, EMBO J. 19 (2000) 2340–2350. [19] A.C. Beckel-Mitchener, A. Miera, R. Keller, N.I. PerroneBizzozero, Poly (A) tail length-dependent stabilization of GAP43 mRNA by the RNA-binding protein HuD, J. Biol. Chem. 277 (2002) 27996–28002. [20] D. Antic, N. Lu, J.D. Keene, ELAV tumor antigen, Hel-N1, increases translation of neurofilament M mRNA and induces formation of neurites in human teratocarcinoma cells, Genes Dev. 13 (1999) 449–461.
K. Saito et al. / Biochemical and Biophysical Research Communications 321 (2004) 291–297 [21] C.F. Manohar, M.L. Short, A. Nguyen, N.N. Nguyen, D. Chagnovich, Q. Yang, S.L. Cohn, HuD, a neuronal-specific RNA-binding protein, increases the in vivo stability of MYCN RNA, J. Biol. Chem. 277 (2002) 1967–1973. [22] I.E. Gallouzi, J.A. Steitz, Delineation of mRNA export pathways by the use of cell-permeable peptides, Science 294 (2001) 1895– 1901. [23] J. Katahira, K. Strasser, A. Podtelejnikov, M. Mann, J.U. Jung, E. Hurt, The Mex67p-mediated nuclear mRNA export pathway is conserved from yeast to human, EMBO J. 18 (1999) 2593–2609. [24] G.S. Wilkie, V. Zimyanin, R. Kirby, C. Korey, H. Francis-Lang, D. Van Vactor, I. Davis, Small bristles, the Drosophila ortholog of NXF-1, is essential for mRNA export throughout development, RNA 7 (2001) 1781–1792. [25] W. Tan, A.S. Zolotukhin, J. Bear, D.J. Patenaude, B.K. Felber, The mRNA export in Caenorhabditis elegans is mediated by CeNXF-1, an ortholog of human TAP/NXF and Saccharomyces cerevisiae Mex67p, RNA 6 (2000) 1762–1772. [26] I.C. Braun, E. Rohrbach, C. Schmitt, E. Izaurralde, TAP binds to the constitutive transport element (CTE) through a novel RNAbinding motif that is sufficient to promote CTE-dependent RNA export from the nucleus, EMBO J. 18 (1999) 1953–1965. [27] Y. Kang, H.P. Bogerd, B.R. Cullen, Analysis of cellular factors that mediate nuclear export of RNAs bearing the Mason-Pfizer monkey virus constitutive transport element, J. Virol. 74 (2000) 5863–5871.
297
[28] V.N. Kim, J. Yong, N. Kataoka, L. Abel, M.D. Diem, G. Dreyfuss, The Y14 protein communicates to the cytoplasm the position of exon–exon junctions, EMBO J. 20 (2001) 2062–2068. [29] H. Le Hir, D. Gatfield, E. Izaurralde, M.J. Moore, The exon– exon junction complex provides a binding platform for factors involved in mRNA export and nonsense-mediated mRNA decay, EMBO J. 20 (2001) 4987–4997. [30] D. Longman, I.L. Johnstone, J.F. Caceres, The Ref/Aly proteins are dispensable for mRNA export and development in Caenorhabditis elegans, RNA 9 (2003) 881–891. [31] D. Gatfield, E. Izaurralde, REF1/Aly and the additional exon junction complex proteins are dispensable for nuclear mRNA export, J. Cell Biol. 159 (2002) 579–588. [32] Y. Huang, R. Gattoni, J. Stevenin, J.A. Steitz, SR splicing factors serve as adapter proteins for TAP-dependent mRNA export, Mol. Cell 11 (2003) 837–843. [33] K. Kasashima, E. Sakashita, K. Saito, H. Sakamoto, Complex formation of the neuron-specific ELAV-like Hu RNA-binding proteins, Nucleic Acids Res. 30 (2002) 4519–4526. [34] E. Izaurralde, Friedrich Miescher Prize awardee lecture review. A conserved family of nuclear export receptors mediates the exit of messenger RNA to the cytoplasm, Cell Mol. Life Sci. 58 (2001) 1105–1112. [35] L. Jin, B.W. Guzik, Y.C. Bor, D. Rekosh, M.L. Hammarskjold, Tap and NXT promote translation of unspliced mRNA, Genes Dev. 17 (2003) 3075–3086.