JBC Avanti Polar Lipids

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M400123200 on July 9, 2004

J. Biol. Chem., Vol. 279, Issue 38, 40153-40160, September 17, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
279/38/40153    most recent
M400123200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sheng, Z.
Right arrow Articles by Chirico, W. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sheng, Z.
Right arrow Articles by Chirico, W. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Nuclear and Nucleolar Localization of 18-kDa Fibroblast Growth Factor-2 Is Controlled by C-terminal Signals*

Zhi Sheng{ddagger}, John A. Lewis{ddagger}§, and William J. Chirico{ddagger}§

From the {ddagger}Molecular and Cellular Biology Program, School of Graduate Studies, §Department of Anatomy and Cell Biology, State University of New York Downstate Medical Center, Brooklyn, New York 11203

Received for publication, January 7, 2004 , and in revised form, June 10, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Members of high (22-, 22.5-, 24-, and 34-kDa) and low (18-kDa) molecular mass forms of fibroblast growth factor-2 (FGF-2) regulate cell proliferation, differentiation, and migration. FGF-2s have been previously shown to accumulate in the nucleus and nucleolus. Although high molecular weight forms of FGF-2 contain at least one nuclear localization signal (NLS) in their N-terminal extension, the 18-kDa FGF-2 does not contain a standard NLS. To determine signals controlling the nuclear and subnuclear localization of the 18-kDa FGF-2, its full-length cDNA was fused to that of green fluorescent protein (GFP). The fusion protein was primarily localized to the nucleus of COS-7 and HeLa cells and accumulated in the nucleolus. The subcellular distribution was confirmed using wild type FGF-2 and FGF-2 tagged with a FLAG epitope. A 17-amino acid sequence containing two groups of basic amino acid residues separated by eight amino acid residues directed GFP and a GFP dimer into the nucleus. We systematically mutated the basic amino acid residues in this nonclassical NLS and determined the effect on nuclear and nucleolar accumulation of 18-kDa FGF-2. Lys119 and Arg129 are the key amino acid residues in both nuclear and nucleolar localization, whereas Lys128 regulates only nucleolar localization of 18-kDa FGF-2. Together, these results demonstrate that the 18-kDa FGF-2 harbors a C-terminal nonclassical bipartite NLS, a portion of which also regulates its nucleolar localization.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
FGF-21 belongs to a large family of small polypeptides (17-34 kDa) that mediates cell proliferation, differentiation, and migration (reviewed in Ref. 1). A single FGF-2 mRNA encodes five protein isoforms. Differential initiation of translation from upstream CUG codons yields four high molecular mass isoforms of FGF-2 (22-34 kDa). The high molecular weight FGF-2s contain an N-terminal extension in which two nuclear localization signals are located (2, 3). The 18-kDa FGF-2 isoform is translated from the AUG codon and can be secreted through a pathway independent of the endoplasm reticulum and Golgi apparatus (reviewed in Ref. 4). The 18-kDa FGF-2 binds and activates cell surface FGF receptors. Activation of FGF receptors triggers downstream signal transduction and regulates gene expression (reviewed in Ref. 5). In addition to this signaling pathway, the presence of 18-kDa FGF-2, high molecular weight FGF-2s, and FGF-2 receptors in the nucleus suggests that FGF-2 participates in an intracrine signaling pathway (reviewed in Ref. 6).

Like other polypeptides, such as insulin and interleukin-1, 18-kDa FGF-2 can translocate into the nucleus after internalization (7). 18-kDa FGF-2, when added to adult bovine aortic endothelial cells, enters the nucleus and up-regulates the synthesis of ribosomal RNA (7, 8). Lysosomal inhibitors and microtubule-disrupting agents did not prevent accumulation of 18-kDa FGF-2 in the nucleus of human umbilical vein endothelial cells, indicating that nuclear translocation of 18-kDa FGF-2 is independent of lysosomes and microtubules (9). Recently, Hsia et al. (10) found that nuclear translocation of 18-kDa FGF-2 was dramatically inhibited by heparinase treatment or in heparin-deficient Chinese hamster ovary cells. These results suggest that heparan sulfate proteoglycans on the cell surface are involved in the nuclear translocation of exogenous 18-kDa FGF-2.

Endogenous 18-kDa FGF-2 also accumulates in the nucleus (11-17). In transfected baby hamster kidney (BHK-21) cells, 18-kDa FGF-2 concentrated in the nucleus, whereas in untransfected BHK-21 cells endogenous FGF-2 was detected in both cytoplasm and nucleus (11). Suramine, a compound that specifically inhibits the binding of FGF-2 to its receptors, blocked the endocytosis of FGF-2 in GM7372 bovine endothelial cells. However, the amount of 18-kDa FGF-2 in the nucleus of these cells did not change, suggesting that 18-kDa FGF-2 may translocate into the nucleus directly from the cytoplasm (12). Human and rat 18-kDa FGF-2 overexpressed in COS cells was present in both nuclei and cytoplasm (13, 14). 18-kDa FGF-2 tagged with fluorescence proteins (GFP or dsRed) localized in the nuclei of rabbit corneal endothelial cells (CEC) and rat Schwann cells (15, 16). Recently, Foletti et al. (17) reported that mouse and human 18-kDa FGF-2 accumulated in the nucleus when overexpressed in mouse fibroblast NIH 3T3 and human embryonic kidney 293T cells. However, other reports suggested that 18-kDa FGF-2 was predominantly cytoplasmic (18-20).

Presta et al. (21) evaluated the nuclear localization capabilities of the sequence (27KDPKR31) in the N terminus of 18-kDa FGF-2 based on its similarity to the NLS of FGF-1 (23YKKPK27) (22). However, altering basic amino acid residues to glutamine in the sequence did not prevent nuclear translocation of 18-kDa FGF-2 in COS-1 cells, suggesting that the NLS is located in another region (21). Claus et al. (16) found that mutating two arginine residues to glycine residues (R116G/R118G) in the C terminus of FGF-2 abolished nuclear localization of both 18- and 23-kDa FGF-2 in rat Schwann cells. A serial deletion of the C terminus of mouse 18-kDa FGF-2 revealed that the nuclear localization of 18-kDa FGF-2 is regulated by C-terminal sequences, but the specific amino acid residues responsible for the localization were not identified (17). Those data suggest that the C terminus of FGF-2 contains an NLS.

The nucleolus is the center of ribosomal biogenesis (reviewed in Ref. 23). Growth factors or growth-regulatory proteins such as FGF-1, FGF-2, FGF-3, angiogenin, and parathyroid hormone-related peptide have been detected in the nucleolus (reviewed in Ref. 24). Exogenously added 18-kDa FGF-2 translocated into the nucleus and accumulated in the nucleolus (7, 8). Similarly, endogenous 18-kDa FGF-2 overexpressed in the CEC and Schwann cells also localized to the nucleolus (15, 16). Studies of nucleolar proteins revealed that signals are required for nucleolar localization, but no consensus sequence has been determined (25-28). Several nucleolar localization sequences, resembling NLSs (clusters of basic amino acid residues) have been identified in a nucleocapsid protein of porcine reproductive and respiratory syndrome virus and ribosomal proteins S22, S25, and L22 (25-27). An interesting feature of these nucleolar localization sequences is that they overlap with NLSs in these proteins (25-27). However, a nucleolar localization sequence found in the N terminus of US11, a herpes simplex virus type 1 protein, lacks clusters of basic amino acid residues (28). Signals that are essential for nucleolar localization of 18-kDa FGF-2 have not been reported.

Here we demonstrate that 18-kDa FGF-2 accumulates in the nucleus and nucleolus of COS-7 and HeLa cells. Examination of the sequence of FGF-2 revealed a nonclassical bipartite NLS in the C terminus. Fusing the NLS to GFP drove the resulting fusion protein into the nucleus. Mutating basic amino acid residues in this region markedly decreased the nuclear and nucleolar localization of 18-kDa FGF-2. The identification of this motif clarifies the mechanism underlying the nuclear translocation and nucleolar accumulation of 18-kDa FGF-2.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Plasmid Construction—To construct pEGFP-N1-FGF-2 encoding the fusion protein GFP-FGF-2, the full-length cDNA of human 18-kDa FGF-2 (gift of R. Florkiewicz) was amplified in a PCR using forward (5'-ACGAGCTGTACAAGATGGCAGCCGGGAGCATCACC-3') and reverse (5'-AGAGTCGCGGCCGCTTAAAATCAGCTCTTAGCAGA-3') primers containing BsrGI and NotI sites, respectively. The amplified DNA was digested with BsrGI and NotI and then ligated to the corresponding sites at the 3'-end of GFP cDNA in pEGFP-N1 (Clontech). The full-length cDNA of FGF-2 was also placed at the 5'-end of GFP cDNA by first amplifying it using forward (5'-TTTAAGCTTATGGCAGCCGGGAGCATCACC-3') and reverse (5'-TTTCTGCAGGCTCTTAGCAGACATTGGAAG-3') primers containing HindIII and PstI sites, respectively. To construct plasmid pFGF-2-3xFLAG-CMVTM-14 encoding FGF-2-FLAG, the cDNA of 18-kDa FGF-2 was amplified using a forward primer (5'-ATATATAAGCTTATGGCAGCCGGGAGCATCACC-3') containing a HindIII site and a reverse primer (5'-ATATGGTACCGAGCTCTTAGCAGACATTGGAAGAAA-3') containing a KpnI site. After digesting the amplified DNA with HindIII and KpnI, the resulting product was inserted into the corresponding sites in p3xFLAG-CMVTM-14 (Sigma). The pN1 control plasmid and plasmids coding for untagged versions of FGF-2 (pN1-FGF-2) and FGF-2 mutants (pN1-FGF-2-K128G, pN1-FGF-2-K128G/R129G, pN1-FGF-2-R116G/R118G/K119G, and pN1-FGF-2-R116G/R118G/K119G/K128G/R129G) were constructed by using HindIII and BsrGI to remove the cDNA of GFP from the corresponding pEGFP-N1 plasmids described above and below (see "Site-directed Mutagenesis"). The overhangs were filled in with Klenow, and the resulting blunt ends were ligated.

A 17-amino acid sequence containing the NLS of the 18-kDa form of FGF-2 was fused to GFP by synthesizing two complementary DNAs corresponding to the sequence 114TYRSRKYTSWYVALKRT130 and containing a 5'-overhang of an XhoI site and a 3'-overhang of a BamHI site. After annealing, the double-stranded DNA was inserted into the corresponding sites at the 5'-end of GFP cDNA in plasmid pEGFP-N1. The two complementary DNAs were 5'-TCGAGGAATTCATGACTTACCGGTCAAGGAAATACACCAGTTGGTATGTGGCACTGAAACGAACTCTG-3' and 5'-GATCCAGTTCGTTTCAGTGCCACATACCAACTGGTGTATTTCCTTGACCGGTAAGTCATGAATTCC-3'. To generate pNLS-dEGFP-N1, in which the NLS of 18-kDa FGF-2 is fused to a GFP dimer, GFP cDNA was amplified with primers containing BsrGI sites. The forward and reverse primers were 5'-ATATATTGTACAAGATGGTGAGCAAGGGCGAGGAGCTGTTC-3' and 5'-CTTTACTTGTACAGCTCGTCCATGCCGAG-3', respectively. The amplified GFP cDNA was then digested with BsrGI and inserted into the corresponding site at the 3'-end of GFP cDNA in pNLS-EGFP-N1. The plasmid pdEGFP-N1 encoding a GFP dimer was constructed by first digesting pNLS-dEGFP-N1 with XhoI and BamHI to remove the NLS, blunting the product with DNA polymerase, and finally ligating the ends together.

Site-directed Mutagenesis—To obtain a plasmid encoding wild type 18-kDa FGF-2, two stop codons were introduced between FGF-2 and GFP cDNA in pFGF-2-EGFP-N1 using the QuikChange® site-directed mutagenesis kit (Stratagene) according to the manufacturer's instruction. The forward and reverse primers were 5'-CCAATGTCTGCTAAGAGCTGATGACTGCAGTCGACG-3' and 5'-CGGTACCGTCGACTGCAGTCATCAGCTCTTAGCAGA-3', respectively. The resulting plasmid (pFGF-2STOP-EGFP-N1) encodes 18-kDa FGF-2. To make mutants R116G, R118G, K119G, K128G, R129G, R116G/R118G, R118G/K119G, and K128G/R129G, mutations were introduced into GFP-FGF-2 using site-directed mutagenesis. The forward and reverse primers for mutant R116G were 5'-AACTACAATACTTACGGGTCAAGGAAATACACCAGTTGG-3' and 5'-CCAACTGGTGTATTTCCTTGACCCGTAAGTATTGTAGTT-3', respectively. Forward and reverse primers for R118G were 5'-TACAATACTTACCGGTCAGGGAAATACACCAGTT-3' and 5'-AACTGGTGTATTTCCCTGACCGGTAAGTATTGTA-3', respectively. Forward and reverse primers for K119G were 5'-ACAATACTTACCGGTCAAGGGGGTACACCAGTTGGTAT-3' and 5'-ATACCAACTGGTGTACCCCCTTGACCGGTAAGTATTGT-3', respectively. Forward and reverse primers for mutant K128G were 5'-TGGTATGTGGCACTGGGACGAACTGGGCAGTAT-3' and 5'-ATACTGCCCAGTTCGTCCCAGTGCCACATACCA-3', respectively. Forward and reverse primers for mutant R129G were 5'-TATGTGGCACTGAAAGGAACTGGGCAGTATAAA-3' and 5'-TTTATACTGCCCAGTTCCTTTCAGTGCCACATA-3', respectively. Forward and reverse primers for mutant R116G/R118G were 5'-AACTACAATACTTACGGGTCAGGGAAATACACCAGTTGG-3' and 5'-CCAACTGGTGTATTTCCCTGACCCGTAAGTATTGTAGTT-3', respectively. Forward and reverse primers for mutant R118G/K119G were 5'-AATACTTACCGGTCAGGGGGGTACACCAGTTGGTAT-3' and 5'-ATACCAACTGGTGTACCCCCCTGACCGGTAAGTATT-3', respectively. Forward and reverse primers for mutant K128G/R-129G were 5'-TGGTATGTGGCACTGGGAGGAACTGGGCAGTATAAA-3' and 5'-TTTATACTGCCCAGTTCCTCCCAGTGCCACATACCA-3', respectively. To make mutants R116G/K119G, R116G/R118G/K119G, and R116G/K128G/R129G, K119G, R118G/K119G, and K128GR129G mutations were introduced, respectively, into the mutant R116G. Forward and reverse primers for R116G/K119G were 5'-ACTTACGGGTCAAGGGGGTACACCAGTTGGTAT-3' and 5'-ATACCAACTGGTGTACCCCCTTGACCCGTAAGT-3', respectively. The forward and reverse primers for R116G/R118G/K119G were 5'-AACTACAATACTTACGGGTCAGGGGGGTACACCAGTTGGTAT-3' and 5'-ATACCAACTGGTGTACCCCCCTGACCCGTAAGTATTGTAGTT-3', respectively. Primers for R116G/K128G/R129G were the same as those used in generating K128G/R129G. Mutants R118G/K128G/R129G, R119G/K128G/R129G, R116G/R118G/K128G/R129G, and R118G/K119G/K128G/R129G were generated, respectively, by introducing R118G, R119G, R116G/R118G, or R118G/K119G into the mutant K128G/R129G using primers described above. K119G was introduced into the mutant R116G/K128G/R129G to obtain R116G/K119G/K128G/R129G using the same primers for K119G. Mutations K128G and R129G were introduced into the mutant R116G/R118G to generate R116G/R118G/K128G and R116G/R118G/R129G, respectively, using the corresponding primers described above. R116G/R118G/K119G/K128G/R129G was obtained by introducing K128G/R129G into the mutant R116G/R118G/K119G using the same primers for K128G/R129G described above. The sequences of all inserts and mutations were confirmed by automated DNA sequencing.

Cell Culture and Transient Transfection—Monkey kidney fibroblast COS-7 cells (American Type Culture Collection) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% calf bovine serum, 100 µg/ml penicillin, 100 units/ml streptomycin. Human cervix adenocarcinoma HeLa cells (American Type Culture Collection) were grown in Eagle's minimum essential medium supplemented with 10% calf bovine serum, 100 µg/ml penicillin, 100 units/ml streptomycin. For transient transfection, COS-7 or HeLa cells were trypsinized and counted. Cells (8 x 104) were plated onto a glass coverslip in a 24-well plate. After overnight incubation, 1 µg of plasmid DNA was transfected into cells using 1 µl of LipofectAMINETM 2000 (Invitrogen) according to the manufacturer's instructions.

Immunofluorescence Microscopy—Immunofluorescence microscopy was performed as described previously with minor modifications (29). In brief, transiently transfected cells on glass coverslips were fixed in Buffer A (4% paraformaldehyde in full phosphate-buffered saline containing 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 0.5 mM MgCl2, and 1 mM CaCl2) for 30 min at room temperature. Then cells were permeabilized in Buffer B (0.2% Triton X-100 in full phosphate-buffered saline) for 10 min at room temperature. Permeabilized cells were blocked in Buffer C (1% goat serum and 3% bovine serum albumin in full phosphate-buffered saline) for at least 1 h at room temperature. Mouse monoclonal FGF-2 antibody (1:20-100 diluted in Buffer C) (Oncogene) or mouse monoclonal nucleolin antibody (1:10 diluted in Buffer C) (gift from Dr. Piñol-Roma, New York) was incubated with cells for 16-24 h at 4 °C. Afterward, cells were incubated with Texas Red-conjugated goat anti-mouse IgG (1:100 diluted in Buffer C) (Jackson ImmunoResearch Laboratories) for 2 h at room temperature. For some experiments, FGF-2 and its fusion protein FGF-2-FLAG were visualized using fluorescein isothiocyanate-conjugated goat anti-mouse IgG (1:50 diluted in Buffer C) (Jackson ImmunoResearch Laboratories), and the nuclei were stained with propidium iodide (4 µg/ml in Buffer C) (Sigma). Samples were mounted in SlowFadeTM antifade solution (Molecular Probes, Inc., Eugene, OR) and observed using a Radiance 2000 confocal microscope (Bio-Rad). Confocal images (1024 x 768 pixels) were obtained using a 40x objective lens.

Nuclear Localization Analysis—NIH Image J version 1.30 software was used to quantify the nuclear localization of GFP fusion proteins. All confocal images were quantified using settings where the intensity of GFP fluorescence was linear and ranged from 10 to 160 pixel values (30). Usually, 8-10 sections encompassing an entire cell were taken at 2.0-µm intervals. To examine the nuclear localization of GFP and GFP fusion proteins, a square with an area of 225 pixels was used to measure the mean intensities of three different regions in the nucleus and the cytoplasm of a representative cell from each of three transfections. Thus, three representative cells yielded a total of nine areas from the nucleus and nine areas from the cytoplasm. Relative nuclear localization is reported as the ratio of the mean intensity of GFP fluorescence in the nucleus and cytoplasm (nuclear/cytoplasmic ratio).

Protein Extraction and Western Blotting—Transfected cells were collected by scraping and then centrifuged for 5 min in a microcentrifuge. SDS gel-loading buffer (50 mM Tris-HCl, pH 6.8, 2% (w/v) SDS, 0.1% (w/v) bromphenol blue, 10% (v/v) glycerol) was added directly to the cell pellets. Cell lysates were clarified using sonication (10 s) and then centrifuged for 2 min in a microcentrifuge. Because heating converts FGF-2 into small peptides (31) and FGF-2 has a propensity to oligomerize (32), cell lysates were treated with 20 mM dithiothreitol at room temperature for 10 min and then incubated with 100 mM iodoacetamide (Sigma) for 10 min at 37 °C before SDS-PAGE (33). Western blotting was performed as described previously (33). Briefly, separated proteins were transferred onto the polyvinylidene difluoride sequencing membrane (ImmobilonTM-PSQ; Millipore Corp.), and the blot was blocked overnight in 5% nonfat powdered milk in TBST (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 0.1% (v/v) Tween 20). GFP and GFP fusion proteins were detected using the mouse monoclonal GFP antibody (1:2000 diluted in TBST) (Clontech) and the mouse monoclonal FGF-2 antibody (1:1000 diluted in TBST) (Oncogene), respectively. Peroxidase-conjugated goat anti-mouse IgG (1:10000 diluted in TBST) (Jackson Immunoresearch Laboratories) was used as the secondary antibody. The secondary antibody was detected using SuperSignal chemiluminescent substrate (Pierce) according to the manufacturer's instructions.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
18-kDa FGF-2 Accumulated in the Nucleus—Although the nuclear translocation of exogenous and endogenous 18-kDa FGF-2 has been reported (7-16), the targeting signals have not been deciphered completely. In order to investigate signals controlling the subcellular localization of 18-kDa FGF-2, we fused its cDNA to the 3'-end of GFP cDNA. Plasmids encoding either GFP or GFP-FGF-2 were transfected into COS-7 cells, and expression of the proteins was verified using Western blotting. A 45-kDa protein was recognized by a GFP and an FGF-2 antibody, indicating that it was a fusion of GFP (27 kDa) and FGF-2 (18 kDa) (Fig. 1A, lanes 2 and 4). GFP, but not FGF-2, was detected in cells transfected with the control plasmid (Fig. 1A, lanes 1 and 3). Endogenous FGF-2 was not detected, because it is expressed at very low levels in COS-7 cells (34).



View larger version (36K):
[in this window]
[in a new window]
  		FIG. 1.
GFP-FGF-2 localizes to the nucleus. A, extracts were prepared from COS-7 cells 20 h after transfection with either pEGFP-N1 (lanes 1 and 3) or pEGFP-N1-FGF-2 (lanes 2 and 4) and separated on SDS-polyacrylamide gels. The proteins were transferred to polyvinylidene difluoride membranes and probed with anti-GFP antibody (lanes 1 and 2) or anti-FGF-2 antibody (lanes 3 and 4). The numbers on the left represent molecular weight markers. B, confocal microscope images of COS-7 cells 20 h after transfection with pEGFP-N1 (a-c) or pEGFP-N1-FGF-2 (d-f). Green fluorescence indicates the position of GFP (a) and GFP-FGF-2 (d), whereas the red fluorescence (propidium iodide (PI)) indicates the position of the nucleus (b and e). The extent of nuclear localization of GFP and GFP-FGF-2 is indicated by yellow in the merged images in c and f, respectively. C, the nuclear/cytoplasmic (N/C) ratio was obtained from confocal images of GFP and GFP-FGF-2 expressed in COS-7 (gray bars) and HeLa (white bars) cells as described under "Experimental Procedures." The error bars represent the S.D. of three determinations.

 
The subcellular localization of GFP and the GFP-FGF-2 fusion protein in COS-7 and HeLa cells was examined using fluorescence microscopy. GFP was distributed throughout the cell (Fig. 1B, a), whereas GFP-FGF-2 was predominantly localized to the nucleus (Fig. 1B, d). The position of the nucleus was determined using propidium iodide (Fig. 1B, b and e). Merged images confirmed that GFP-FGF-2 accumulated in the nucleus (Fig. 1B, c and f). Quantifying confocal images revealed that the nuclear/cytoplasmic ratio of GFP and GFP-FGF-2 in COS-7 cells was 2.4 and 10, respectively (Fig. 1C, gray bars). The same subcellular distribution of GFP and GFP-FGF-2 was observed in HeLa cells (Fig. 1C, white bars). We also placed FGF-2 cDNA at the 5'-end of GFP cDNA to produce an FGF-2-GFP fusion protein (Supplemental Data, Fig. S4, lane 1). The fluorescence patterns indicated that the subcellular distributions of FGF-2-GFP (Supplemental Data, Fig. S1, a) and GFP-FGF-2 (Fig. 1B, d) were identical.

The subcellular distribution of GFP-FGF-2 was also probed with an FGF-2 antibody (Fig. 2). Autofluorescence (Fig. 2A, d) and indirect immunofluorescence (Fig. 2A, e) indicated that GFP-FGF-2 fusion protein remained intact and localized to the nucleus. The merged images supported this conclusion (Fig. 2A, f). Endogenous FGF-2 was not detected in cells transfected with pEGFP-N1 (Fig. 2A, b). Together, these results indicate that FGF-2 drives GFP into the nucleus.



View larger version (53K):
[in this window]
[in a new window]
  		FIG. 2.
Integrity of GFP-FGF-2 and nuclear localization of FGF-2-FLAG. A, confocal images were obtained 20 h after COS-7 cells were transfected with pEGFP-N1 (a-c) or pEGFP-N1-FGF-2 (d-f). The position of GFP was determined by autofluorescence (a), whereas the position of GFP-FGF-2 was determined by both autofluorescence (d) and indirect immunofluorescence using an FGF-2 antibody and Texas Red-conjugated secondary antibody (e). Merged images are shown in c and f. B, confocal images of COS-7 cells 20 h after transfection with either pFGF-2STOP-EGFP-N1 (wild type FGF-2) (a-c) or pFGF-2-3xFLAG-CMVTM-14 (FGF-2-FLAG) (d-f). The position of the nucleus was determined using propidium iodide (PI). Nuclear localization FGF-2 (c) and FGF-2-3xFLAG (f) is indicated by yellow in the merged images.

 
We considered the possibility that the nuclear accumulation of the GFP-FGF-2 was an artifact of high expression levels. However, a time course of expression showed that GFP-FGF-2 accumulated in the nucleus as early as 6 h after transfection when expression levels were low (Supplemental Data, Fig. S2).

Our finding that GFP-FGF-2 accumulates in the nucleus does not agree with some earlier reports indicating that 18-kDa FGF-2 is predominantly cytosolic (18-20). Perhaps fusing GFP to FGF-2 exposed a cryptic NLS in FGF-2. We tested this possibility by fusing an alternative tag, a FLAG epitope, to FGF-2 and expressing the resulting protein in COS-7 cells. The 18-kDa FGF-2 and FGF-2-FLAG proteins were located predominantly in the nucleus (Fig. 2B, a and d). The pattern of FGF-2-FLAG was almost identical to that of GFP-FGF-2 (Fig. 1B, d, and Fig. 2A, d). The nuclei of cells were stained with propidium iodide (Fig. 2B, b and e). The merged images confirmed the nuclear localization of wild type FGF-2 and FGF-2-FLAG fusion proteins (Fig. 2B, c and f). The expression of wild type 18-kDa FGF-2 and FGF-2-FLAG in COS-7 cells was corroborated by Western blotting (Supplemental Data, Fig. S3). Together, these results demonstrate that 18-kDa FGF-2 expressed alone or as a fusion protein localizes predominantly to the nucleus.

Identification of a Nuclear Localization Signal in 18-kDa FGF-2—The above results led us to hypothesize that 18-kDa FGF-2 contains an NLS. A putative NLS in the N terminus of 18-kDa FGF-2 had been proposed (22). However, mutations in this sequence did not prevent the nuclear accumulation of 18-kDa FGF-2 (22). To determine whether the N terminus (amino acids 1-120) of FGF-2 harbors a different putative NLS, we fused it to GFP. The resulting fusion protein did not accumulate in the nucleus, indicating that the N terminus lacks an NLS (Supplemental Data, Fig. S1, b, and Fig. S4, lane 4). We also examined the protein sequence of 18-kDa FGF-2 using the programs PROSITE (35) and PredictNLS (36). Although PROSITE and PredictNLS did not detect an NLS, we found a sequence (116RSRKYTSWYVALKR129) in the C terminus of 18-kDa FGF-2 that resembles a nonclassical bipartite NLS. This sequence contains two clusters of two or three basic amino acid residues and an eight-residue spacer. Usually, a bipartite NLS contains two clusters and 9-12 residues in the spacer (36).

The function of a putative NLS can be tested by fusing it to a cytoplasmic reporter protein or by mutating it. If the sequence is an NLS, it will drive the cytoplasmic reporter protein into the nucleus. In the latter case, a protein containing a defective NLS will fail to enter the nucleus. We synthesized a DNA fragment corresponding to the sequence (114TYRSRKYTSWYVALKRT130) containing the putative NLS and then ligated it to the 5' terminus of GFP cDNA, yielding NLS-GFP. This sequence directed GFP into the nucleus (Fig. 3A, compare a and c). The nuclear/cytoplasmic ratio of NLS-GFP in COS-7 and HeLa cells was about 4.4 and 3.4, respectively, whereas that of GFP was about 2.4 and 1.8, respectively (Fig. 3B). We also made a larger reporter protein, NLS-dGFP (~60 kDa), a fusion protein containing the NLS fused to a GFP dimer. The 17-mer containing the NLS drove dGFP into the nucleus, whereas dGFP control distributed throughout the cell (compare Fig. 3A, b and d). Quantifying confocal images revealed that the nuclear/cytoplasmic ratio of NLS-dGFP in COS-7 and HeLa cells was about 2.8 and 3, respectively, whereas that of dGFP was about 1.4 (Fig. 3B). The expression of NLS-GFP, dGFP, and NLS-dGFP was confirmed by Western blotting (Supplemental Data, Fig. S4). Together, these results indicate that the 17-mer containing the NLS of 18-kDa FGF-2 can direct GFP and the GFP dimer into the nucleus but not as efficiently as full-length FGF-2 (compare Fig. 1B, d, and Fig. 3A, c and d). Since the putative NLS is located in the C terminus of 18-kDa FGF-2, we tested whether sequence position affects localization by placing this 17-amino acid sequence at the C terminus of GFP (Supplemental Data, Fig. S4). The subcellular distributions of NLS-GFP and GFP-NLS were almost identical, demonstrating that the position of the NLS in the sequence does not affect its function (compare Fig. 3A, c, and Supplemental Data, Fig. S1, c).



View larger version (48K):
[in this window]
[in a new window]
  		FIG. 3.
The NLS of FGF-2 drives GFP and GFP dimer into the nucleus. A, confocal images of COS-7 cells transfected with pEGFP-N1 (GFP; a), pdEGFP-N1 (dGFP; b), pNLS-EGFP-N1 (NLS-GFP; c), and pNLS-dEGFP-N1 (NLS-dGFP; d). B, the nuclear/cytoplasmic (N/C) ratio was obtained from confocal images of GFP, NLS-GFP, dGFP, and NLS-dGFP expressed in COS-7 (gray bars) and HeLa (white bars) cells as described under "Experimental Procedures." The error bars represent S.D. of three determinations.

 
Role of Basic Amino Acid Residues in Nuclear Localization of 18-kDa FGF-2—To determine the key elements of the NLS, we altered its basic amino acid residues using site-directed mutagenesis. Five basic amino acid residues form two clusters (Table I). We systematically altered each basic amino acid residue alone or in combination with others in the first cluster (116RSRK119) and determined the effect on nuclear localization of 18-kDa FGF-2. The subcellular distribution of mutants R116G and R118G were similar to that of GFP-FGF-2 (compare Fig. 1B, d, and Fig. 4A, a and b). Quantifying confocal images revealed that the nuclear/cytoplasmic ratio of R116G in COS-7 and HeLa cells was about 9 and 10, respectively, and that of R118G was about 13 in both COS-7 and HeLa cells (Fig. 5). In contrast, changing Lys119 into glycine moderately decreased nuclear localization of GFP-FGF-2 (compare Fig. 1B, d, and Fig. 4A, c). The nuclear/cytoplasmic ratio of K119G in COS-7 and HeLa cells was about 6 and 5, respectively (Fig. 5). The expression of all mutants was confirmed by Western blotting (Supplemental Data, Fig. S5). Our results indicate that Lys119 plays a more important role than Arg116 and Arg118 in nuclear localization of 18-kDa FGF-2.


View this table:
[in this window]
[in a new window]
  		TABLE I
Wild type and mutant versions of nuclear localization signal of FGF-2

The mutated amino acids are in boldface type and underlined.

 



View larger version (52K):
[in this window]
[in a new window]
  		FIG. 4.
Perturbation of GFP-FGF-2 nuclear localization by site-directed mutagenesis. Confocal fluorescence images were obtained from COS-7 cells transfected with plasmids encoding the following GFP-FGF-2 mutants. A, mutations in the first cluster of basic amino residues: R116G (a), R118G (b), K119G (c), R116G/R118G (d), R116G/K119G (e), R118G/K119G (f), and R116G/R118G/K119G (g). B, mutations in the second cluster of basic amino acid residues: K128G (a), R129G (b), and K128G/R129G (c). C, mutations in both clusters of basic amino acid residues: R116G/R118G/K128G (a), R116G/R118G/R129G (b), and R116G/R118G/K119G/K128G/R129G (c).

 



View larger version (32K):
[in this window]
[in a new window]
  		FIG. 5.
Quantitation of nuclear localization of GFP, GFPFGF-2, and GFP-FGF-2 mutants. The nuclear/cytoplasmic (N/C) ratios were obtained from confocal images of GFP, GFP-FGF-2, and GFP-FGF-2 mutants expressed in COS-7 (gray bars) and HeLa (white bars) cells as described under "Experimental Procedures." The specific amino acid changes in the GFP-FGF-2 mutants are indicated by the plus signs below the bars. The error bars represent S.D. of three determinations.

 
Changing both Arg116 and Arg118 into glycine residues moderately inhibited GFP-FGF-2 nuclear localization (Fig. 4A, d). The nuclear/cytoplasmic ratio of R116G/R118G in COS-7 and HeLa cells was about 4 and 5, respectively (Fig. 5). Our results do not agree with those of Claus et al. (16), who reported that R116G and R118G mutations abolished nuclear localization of 18- and 23-kDa FGF-2s in Schwann cells. Changing both Arg116 and Lys119 into glycine (R116G/K119G) or mutating both Arg118 and Lys119 in this cluster (R118G/K119G) markedly reduced nuclear localization (Fig. 4A, e and f). The nuclear/cytoplasmic ratio of R116G/K119G in COS-7 and HeLa cells was about 1.7 and 1.6, respectively, whereas that of R118G/K119G in HeLa and COS-7 cells was about 1.6 and 1.5, respectively (Fig. 5). Mutating all three basic amino acids in the first cluster (R116G/R118G/K119G) inhibited the nuclear accumulation of GFP-FGF-2 (Fig. 4A, g). Similarly, nuclear accumulation of untagged FGF-2-R116G/R118G/K119G was greatly reduced (Fig. 6, compare b and d). The nuclear/cytoplasmic ratio of this GFP-FGF-2 mutant in COS-7 and HeLa cells was about 1.4 and 1.3, respectively (Fig. 5), and that of the untagged FGF-2 mutant in COS-7 cells was 1.8 ± 0.6. Expression of this and other untagged FGF-2 mutants was confirmed by Western blotting (Supplemental Data, Fig. S6B). These results confirmed that Lys119 is the most critical residue of the first cluster in the regulation of nuclear localization of 18-kDa FGF-2.



View larger version (78K):
[in this window]
[in a new window]
  		FIG. 6.
Site-directed mutagenesis of untagged FGF-2. COS-7 cells were transfected with a control plasmid (pN1; a) or pN1-FGF-2 (b), pN1-FGF-2-K128G (c), pN1-FGF-2-R116G/R118G/K119G (d), pN1-FGF-2-K128G/R129G (e), or pN1-FGF-2-R116G/R118G/K119G/K128G/R129G (f). FGF-2 and its mutants were visualized by confocal immunofluorescence microscopy using an FGF-2 antibody and Texas Redconjugated secondary antibody.

 
To understand the role of the second cluster in nuclear localization of 18-kDa FGF-2, we mutated each basic amino acid residue alone or together. The K128G mutant accumulated in the nucleus, whereas the nuclear localization of R129G and K128G/R129G mutants was moderately reduced (Fig. 4B, a-c). The nuclear/cytoplasmic ratio of K128G, R129G, and K128G/R129G was about 10, 5, and 5, respectively, in COS-7 cells and 11, 5, and 3, respectively, in HeLa cells (Fig. 5). Similarly, untagged FGF-2-K128G and FGF-2-K128G/R129G mutants were localized to the nucleus less efficiently than wild-type FGF-2 (Fig. 6, compare b, c, and e). The nuclear/cytoplasmic ratios of FGF-2, FGF-2-K128G, and FGF-2-K128G/R129G were 7 ± 2, 5.6 ± 0.5, and 2.9 ± 0.2, respectively. Together, these results indicated that Lys128 is dispensable in nuclear localization of 18-kDa FGF-2. In contrast, Arg129 plays an important role in nuclear localization.

Since complete inhibition of nuclear localization had not been observed in the above mutants, we hypothesized that basic amino acid residues in both clusters may cooperate with each other in regulating nuclear localization. Combining the R116G/R118G and K128G mutations (Fig. 4A, d, and Fig. 4B, a) into a single mutant did not further inhibit nuclear localization (Fig. 4C, a). The nuclear/cytoplasmic ratio of R116G/R118G/K128G in COS-7 and HeLa cells was about 4 and 5, respectively (Fig. 5). These results confirm that Lys128 does not play a major role in the nuclear localization of FGF-2. However, combining the R116G/R118G and R129G mutations dramatically reduced nuclear localization (Fig. 4C, b). The nuclear/cytoplasmic ratio in COS-7 and HeLa cells was about 1.2 (Fig. 5). These results confirm that Arg129 plays an important role in nuclear localization and indicate that both clusters work together in regulating nuclear localization of 18-kDa FGF-2.

Compared with R116G, R118G, K119G, and K128G/R129G (Fig. 4, A (a-c) and B (c)), mutants R116G/K128G/R129G, R118G/K128G/R129G, and K119G/K128G/R129G remained in the cytoplasm (Fig. 5). Combining K128G/R129G mutations with any two basic amino acid residues in the first cluster (R116G/R118G/K128G/R129G, R116G/K119G/K128G/R129G, R118G/K119G/K128G/R129G) blocked nuclear localization (data not shown). The nuclear/cytoplasmic ratio of these mutants in HeLa and COS-7 cells was less than or equal to 1 (Fig. 5). Changing all five basic amino acid residues into glycine residues resulted in the exclusion of the GFP-FGF-2 (Fig. 4C, c) or FGF-2 (Fig. 6f) mutants from the nucleus. The nuclear/cytoplasmic ratio of this GFP mutant in HeLa and COS-7 cells was about 0.8 (Fig. 5), and that of the FGF-2 mutant in COS-7 cells was about 0.8 ± 0.1. These results confirm that both clusters are required for efficient nuclear localization of 18-kDa FGF-2. However, the first cluster is critical for nuclear localization. Our data also indicate that Lys119 and Arg129 are the key amino acid residues of the first and second clusters, respectively, in the nuclear localization of 18-kDa FGF-2.

Nucleolar Localization of 18-kDa FGF-2—Exogenously added 18-kDa FGF-2 has been shown to translocate into the nucleus and accumulate in the nucleolus (7, 8). The 18-kDa form of FGF-2 fused with fluorescent proteins also accumulated in the nucleolus of CEC and Schwann cells (16, 17). However, the motif that controls the nucleolar localization has not been determined. To visualize the nucleolus, we used an antibody that recognizes nucleolin (37, 38), one of the most abundant nucleolar proteins (Fig. 7, nucleolin panels). GFPFGF-2, but not GFP, accumulated in the nucleolus of COS-7 cells (Fig. 7A, a and d) and HeLa cells (data not shown). Nucleolin distributed evenly in the nucleolus (Fig. 7A, b and e). The merged images confirmed the nucleolar accumulation of GFP-FGF-2 (Fig. 7A, c and f).



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 7.
Amino acid residues of GFP-FGF-2 necessary for nucleolar localization. A, nucleolar localization of GFP-FGF-2. Confocal fluorescence images were obtained from COS-7 cells transfected with pEGFP-N1 (a-c) or pEGFP-N1-FGF-2 (d-f). The positions of the nucleoli were determined using anti-nucleolin antibody and a Texas Redconjugated secondary antibody (b and e). The extent of localization of GFP (c) and GFP-FGF-2 (f) to the nucleolus is indicated by yellow in the merged images. B, perturbation of nucleolar localization of GFP-FGF-2 by site-directed mutagenesis. Confocal fluorescence images were obtained from COS-7 cells transfected with the following pEGFP-N1-FGF-2 mutants: R116G (a-c), R118G (d-f), K119G (g-i), R129G (j-l), K128G (m-o), and R116G/R118G (p-r). Autofluorescence from GFP is shown in a, d, g, j, m, and p. The position of the nucleoli (nucleolin) is shown by red fluorescence in b, e, h, k, n, and q. The extent of nucleolar localization of GFP-FGF-2 mutants is indicated by yellow in the merged images (c, f, i, l, o, and r). C, perturbation of nucleolar localization of untagged FGF-2 and FGF-2-K128G. Confocal microscope images of COS-7 cells were taken 20 h after transfection with pEGFPC1-nucleolin and either pN1 (a-c), pN1-FGF-2 (d-f), or pN1-FGF-2-K128G (g-i). FGF-2 and its mutant (red fluorescence) were detected using an FGF-2 antibody and a Texas Red-conjugated secondary antibody (b, e, and h). The nucleolus was located using autofluorescence of GFP-nucleolin (a, d, and g). The merged images (c, f, and i) reveal co-localization of GFP-nucleolin and FGF-2 but not FGF-2-K128G.

 
The nucleolar targeting signals of a nucleocapsid protein of porcine reproductive and respiratory syndrome virus and several ribosomal proteins (S22, S25, and L22) overlap with their nuclear localization signals (25-27). Thus, we hypothesized that the sequence controlling the nucleolar localization of 18-kDa FGF-2 may also overlap with its NLS. To test this hypothesis, we examined the nucleolar localization of NLS-GFP and NLS-dGFP. Although NLS drove GFP into the nucleus, NLSGFP and NLS-dGFP did not accumulate in the nucleolus (Fig. 3A, c and d). The position of the nucleolus was determined using a nucleolin antibody (data not shown).

The failure of the NLS to drive GFP into the nucleolus led us to hypothesize that the nucleolar targeting sequence may function only in the context of full-length FGF-2. A common feature of nucleolar targeting sequences is the presence of basic amino acid residues (24-26). To investigate the role of basic amino acid residues of the NLS in nucleolar localization, we reexamined our set of GFP-FGF-2 mutants. Changing Arg116 or Arg118 into glycine did not affect nucleolar localization of FGF-2 (Fig. 7, compare A (d) and B (a and d)). The merged images confirmed their nucleolar localization (Fig. 7B, c and f). However, changing both Arg116 and Arg118 into glycine residues essentially blocked their nucleolar localization (Fig. 7B, p-r) but only moderately inhibited the nuclear localization of FGF-2 (Figs. 4A (d) and 5). These results indicate that Arg116 and Arg118 together, but neither alone, play an important role in nucleolar localization of FGF-2. The K119G and R129G mutants did not localize to the nucleolus in COS-7 (Fig. 7B, g-l) and HeLa cells (data not shown), demonstrating that these amino acid residues are indispensable for nucleolar localization. Furthermore, among the GFP-FGF-2 mutants studied (Table I), those containing K119G, R129G, or both mutations were blocked from the nucleolus (data not shown). The K128G mutant produced a unique pattern. This mutation did not change the nuclear localization of GFP-FGF-2 (Figs. 4B (a) and 5) but blocked nucleolar targeting (Fig. 7B, m-o). Similarly, the untagged FGF-2-K128G mutant localized to the nucleus efficiently but did not accumulate in the nucleolus (Fig. 7C, g-i). Furthermore, all GFP-FGF-2 mutants containing a K128G alteration (Table I) failed to accumulate in the nucleolus (data not shown). These results indicate that Lys128 controls nucleolar, but not nuclear, localization of FGF-2.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
We demonstrated above that the C terminus of 18-kDa FGF-2 harbors an unconventional bipartite NLS (116RSRKYTSWYVALKR129) that controls its nuclear and nucleolar localization. The NLS contains two small clusters of basic amino acid residues separated by a hydrophobic region containing eight amino acid residues. The first cluster consists of 116RSRK119, and the second cluster consists of 128KR129. Adding a 17-amino acid sequence containing this NLS to GFP drove the resulting fusion protein into the nucleus. Mutating basic amino acid residues in this sequence differentially modulated the nuclear localization of 18-kDa FGF-2. We demonstrated that basic amino acid residues Lys119 and Arg129 in this NLS are critical in regulating both nuclear and nucleolar localization of 18-kDa FGF-2, whereas Lys128 modulates only nucleolar targeting. These findings reveal the key determinants of nuclear and nucleolar localization of 18-kDa FGF-2.

The subcellular distribution of FGF-2 has been controversial (7-20). Our results showing that 18-kDa FGF-2 is predominantly localized to the nucleus are consistent with most but not all (7-16) studies of subcellular distribution of 18-kDa FGF-2. Some reports show that 18-kDa FGF-2 is predominantly cytosolic (17-20). The reason for the discrepancy is not clear.

We showed above that NLS directed fusion proteins into the nucleus but not as efficiently as full-length FGF-2. In the context of the full-length FGF-2, the NLS may adopt a conformation recognized by the nuclear translocation machinery. Most nuclear-cytoplasmic transport of proteins requires an NLS in the cargo and the nuclear transport receptors (karyopherins) that recognize NLSs (reviewed in Refs. 39 and 40). Recently, Fontes et al. (41) proposed a consensus sequence of a conventional bipartite NLS (KRX10-12KRRK) for proteins targeted to the nucleus. Basic amino acid residues in this sequence are responsible for binding to the importin {alpha} receptor. Examination of the three-dimensional model of wild type FGF-2 (Protein Data Bank code 1BLA [PDB] ) (42) indicates that the two basic clusters in the NLS are located on the outer surface of 18-kDa FGF-2. However, in the NLS-GFP fusion protein, the basic clusters of NLS may not be as accessible or in the same conformation as they are in the full-length FGF-2 fused to GFP (GFP-FGF-2), thereby resulting in less efficient targeting.

The motif harboring the NLS of 18-kDa FGF-2 is multifunctional. It overlaps with binding sites for its receptors and for casein kinase II (43-45). A model for the nuclear activity of FGF-2 has been proposed (44-47). Binding of FGF-2 to casein kinase II triggers nuclear translocation of casein kinase II and casein kinase II-dependent phosphorylation of nucleolin. Nucleolin, but not phosphorylated nucleolin, inhibits rDNA transcription. However, our findings raise the possibility that the NLS reported here affects rDNA transcription by regulating the interaction between FGF-2 and its nuclear and nucleolar targets. The role of these residues in nuclear and nucleolar activities of 18-kDa FGF-2 is under investigation.

Claus et al. (16) reported that mutating two arginine residues to glycine residues in the C terminus of FGF-2 (Arg148 and Arg151 in 23-kDa FGF-2, Arg116 and Arg118 in 18-kDa FGF-2) abolished nuclear localization of both 18- and 23-kDa FGF-2s in rat Schwann cells. However, our results demonstrate that the NLS extends from Arg116 to Arg129 and is bipartite. It is possible that the mechanism of nuclear translocation of 18-kDa FGF-2 in various cell types is different.

Several growth factors have been detected in the nucleolus, but the functional relevance of this translocation remains unclear (28). Exogenously added FGF-2 was found to accumulate in the nucleolus and stimulate the synthesis of ribosomal RNA (7, 8). We also found that GFP-FGF-2 and FGF-2 expressed in COS-7 and HeLa cells accumulated in the nucleolus. This finding is consistent with that of Claus et al. (16) using rat Schwann cells. Nuclear and nucleolar localization of FGF-3, another FGF family member with less than 50% homology in amino acid sequence to FGF-2, is governed by different motifs within the protein (48). However, we found that the accumulation of 18-kDa FGF-2 in the nucleus and nucleolus is controlled by one motif. Furthermore, our analysis of basic amino acid residues in the NLS revealed that Lys119 and Arg129 are key elements for nuclear and nucleolar localization, whereas Lys128 modulates only nucleolar localization.


   FOOTNOTES
 
* This work was supported by an award (0151261T) from the American Heart Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

The on-line version of this article (available at http://www.jbc.org) contains six additional figures. Back

To whom correspondence should be addressed: Dept. of Anatomy and Cell Biology, Box 5, State University of New York Downstate Medical Center, 450 Clarkson Ave., Brooklyn, NY 11203. Tel.: 718-270-1308; Fax: 718-270-3732; E-mail: william.chirico{at}downstate.edu.

1 The abbreviations used are: FGF-2, fibroblast growth factor-2; GFP, green fluorescence protein; NLS-GFP, a fusion protein in which the NLS of 18-kDa FGF-2 was fused to GFP; NLS-dGFP, a fusion protein in which the NLS of 18-kDa FGF-2 was fused to a GFP dimer; NLS, nuclear localization signal; CEC, corneal endothelial cell(s). Back


   ACKNOWLEDGMENTS
 
We thank Qingchun Tong and Bill Oxberry for assistance with the Confocal Facility at SUNY Downstate, Dr. George Ojakian for helpful discussions, Dolly Huq for technical assistance, Dr. Serafín Piñol-Roma for providing the nucleolin antibody, and Dr. Sui Huang for providing pEGFP-C1-nucleolin.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ornitz, D. M., and Itoh, N. (2001) Genome Biol. 2, 3005.1-3005.9
  2. Dono, R., James, D., and Zeller, R. (1998) Oncogene 16, 2151-2158[CrossRef][Medline] [Order article via Infotrieve]
  3. Arnaud, E., Touriol, C., Boutonnet, C., Gensac, M. C., Vagner, S., Prats, H., and Prats, A. C. (1999) Mol. Cell. Biol. 19, 505-514[Abstract/Free Full Text]
  4. Bikfalvi, A., Klein, S., Pintucci, G., and Rifkin, D. B. (1997) Endocr. Rev. 18, 26-45[Abstract/Free Full Text]
  5. Powers, C. J., McLeskey, S. W., and Wellstein, A. (2000) Endocr. Relat. Cancer 7, 165-197[Abstract]
  6. Keresztes, M., and Boonstra, J. (1999) J. Cell Biol. 145, 421-424[Free Full Text]
  7. Bouche, G., Gas, N., Prats, H., Baldin, V., Tauber, J.P., Teissie, J., and Amalric, F. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 6770-6774[Abstract/Free Full Text]
  8. Baldin, V., Roman, AM., Bosc-Bierne, I., Amalric, F., and Bouche, G. (1990) EMBO J. 9, 1511-1517[Medline] [Order article via Infotrieve]
  9. Hu, G., Kin, H., Xu, C., and Riordan, J.F. (2000) Biochem. Biophys. Res. Commun. 273, 551-556[CrossRef][Medline] [Order article via Infotrieve]
  10. Hsia, E., Richardson, T. P., and Nugent M. A. (2003) J. Cell. Biochem. 88, 1214-1225[CrossRef][Medline] [Order article via Infotrieve]
  11. Tessler, S., and Neufeld, G. (1990) J. Cell. Physiol. 145, 310-317[CrossRef][Medline] [Order article via Infotrieve]
  12. Dell'Era, P., Presta, M., and Ragnotti, G. (1991) Exp. Cell Res. 192, 505-510[CrossRef][Medline] [Order article via Infotrieve]
  13. Powell, P.P., and Klagsbrun, M. (1991) J. Cell. Physiol. 148, 202-210[CrossRef][Medline] [Order article via Infotrieve]
  14. Florkiewicz, R., Baird, A., and Gonzalez, A. M. (1991) Growth Factors 4, 265-275[Medline] [Order article via Infotrieve]
  15. Choi, J., Ko, M. K., and Kay, E. P. (2000) Mol. Vis. 6, 222-231[Medline] [Order article via Infotrieve]
  16. Claus, P., Döring, F., Gringel, S., Müller-Ostermeyer, F., Fuhlrott, J., Kraft, T., and Grothe, C. (2003) J. Biol. Chem. 278, 479-485[Abstract/Free Full Text]
  17. Foletti, A., Vuadens, F., and Beermann, F. (2003) Cell. Mol. Life Sci. 60, 2254-2265[CrossRef][Medline] [Order article via Infotrieve]
  18. Renko, M., Quarto, N., Morimoto, T., and Rifkin, D. B. (1990) J. Cell. Physiol. 144, 108-114[CrossRef][Medline] [Order article via Infotrieve]
  19. Quarto, N., Finger, F.P., and Rifkin, D.B. (1991) J. Cell. Physiol. 147, 311-318[CrossRef][Medline] [Order article via Infotrieve]
  20. Bugler, B., Amalric, F., and Prats, H. (1991) Mol. Cell. Biol. 11, 573-577[Abstract/Free Full Text]
  21. Presta, M., Gualandris, A., Urbinati, C., Rusnati, M., Coltrini, D., Isacchi, A., Caccia, P., and Bergonzoni, L. (1993) Growth Factors 9, 269-278[Medline] [Order article via Infotrieve]
  22. Zhan, X., Hu, X., Frieman, S., and Maciag, T. (1992) Biochem. Biophys. Res. Commun. 188, 982-991[CrossRef][Medline] [Order article via Infotrieve]
  23. Dundr, M., and Misteli, T. (2001) Biochem. J. 356, 297-310[CrossRef][Medline] [Order article via Infotrieve]
  24. Pederson, T. (1998) J. Cell Biol. 143, 279-281[Free Full Text]
  25. Timmers, A.C., Stuger, R., Schaap, P.J., van't Riet, J., and Raué, H. A. (1999) FEBS Lett. 452, 335-340[CrossRef][Medline] [Order article via Infotrieve]
  26. Shu-Nu, C., Lin, C. H., and Lin, A. (2000) FEBS Lett. 484, 22-28[CrossRef][Medline] [Order article via Infotrieve]
  27. Rowland R. R., and Yoo, D. (2003) Virus Res. 95, 23-33[CrossRef][Medline] [Order article via Infotrieve]
  28. Catez, F., Erard, M., Schaerer-Uthurralt, N., Kindbeiter, K., Madjar, J. J., and Diaz, J. J. (2002) Mol. Cell. Biol. 22, 1126-1139[Abstract/Free Full Text]
  29. Ojakian, G. K., Nelson, W. J., and Beck, K. A. (1997) J. Cell Sci. 110, 2781-2794[Abstract]
  30. Piston, D. V., Patterson, G. H., and Knobel, S. M. (1999) Methods Cell Biol. 58, 31-48[Medline] [Order article via Infotrieve]
  31. Thompson, S. A., Protter, A. A., Bitting, L., Fiddes, J. C., and Abraham, J. A. (1991) Methods Enzymol. 198, 96-116[Medline] [Order article via Infotrieve]
  32. Davis, J. C., Venkataraman, G., Shriver, Z., Raj, P. A., and Sasisekharan, R. (1999) Biochem. J. 341, 613-620[Medline] [Order article via Infotrieve]
  33. Sheng, Z., Chang, S. B., and Chirico, W. J. (2003) Protein Expr. Purif. 27, 267-271[CrossRef][Medline] [Order article via Infotrieve]
  34. Sohn, Y. D., Lim, H. J., Hwang, K. C., Kwon, J. H., Park, H. Y., Chung, K. H., Cho, S. Y., and Jang, Y. (2001) Biochem. Biophys. Res. Commun. 284, 931-936[CrossRef][Medline] [Order article via Infotrieve]
  35. Hofmann, K., Bucher, P., Falquet, L., and Bairoch, A. (1999) Nucleic Acids Res. 27, 215-219[Abstract/Free Full Text]
  36. Cokol, M., Nair, R., and Rost, B. (2000) EMBO Rep. 1, 411-415[CrossRef][Medline] [Order article via Infotrieve]
  37. Piñol-Roma, S. (1999) Mol. Biol. Cell 10, 77-90[Abstract/Free Full Text]
  38. Srivastava, M., and Pollard, H. B. (1999) FASEB J. 13, 1911-1922[Abstract/Free Full Text]
  39. Jans, D. A., Xiao, C. Y., and Lam, M. H. C. (2000) BioEssays 22, 532-544[CrossRef][Medline] [Order article via Infotrieve]
  40. Boulikas, T. (1993) Crit. Rev. Eukaryot. Gene Expr. 3, 193-227[Medline] [Order article via Infotrieve]
  41. Fontes, M. R. M., Teh, T., Jans, D., Brinkworth, R. T., and Kobe, B. (2003) J. Biol. Chem. 278, 27981-27987[Abstract/Free Full Text]
  42. Moy, F. J., Seddon, A. P., Campbell, E. B., Bohlen, P., and Powers, R. (1995) J. Biomol. NMR 6, 245-254[Medline] [Order article via Infotrieve]
  43. Presta, M., Statuto, M., Isacchi, A., Caccia, P., Pozzi, A., Gualandris, A., Rustani, M., Bergonzoni, L., and Sarmientos, P. (1992) Biochem. Biophys. Res. Commun. 185, 1098-1107[CrossRef][Medline] [Order article via Infotrieve]
  44. Bailly, K., Soulet, F., Leroy, D., Amalric, F., and Bouche, G. (2000) FASEB J. 14, 333-344[Abstract/Free Full Text]
  45. Bonnet, H., Filhol, O., Truchet, I., Brethenou, P., Cochet, C., Amalric, F., and Bouche, G. (1996) J. Biol. Chem. 271, 24781-24787[Abstract/Free Full Text]
  46. Bouche, G., Baldin, V., Belenguer, P., Prats, H., and Amalric, F. (1994) Cell. Mol. Biol. Res. 40, 547-554[Medline] [Order article via Infotrieve]
  47. Filhol, O., Nueda, A., Martel, V., Gerber-Scokaert, D., Benitez, M. J., Souchier, C., Saoudi, Y., and Cochet, C. (2003) Mol. Cell. Biol. 23, 975-987[Abstract/Free Full Text]
  48. Antoine, M., Reimers, K., Dickson, C., and Kiefer, P. (1997) J. Biol. Chem. 272, 29475-29481[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
P. D. Da Forno, J. H. Pringle, P. Hutchinson, J. Osborn, Q. Huang, L. Potter, R. A. Hancox, A. Fletcher, and G. S. Saldanha
WNT5A Expression Increases during Melanoma Progression and Correlates with Outcome
Clin. Cancer Res., September 15, 2008; 14(18): 5825 - 5832.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Cell. Biol.Home page
L. Meng, Q. Zhu, and R. Y. L. Tsai
Nucleolar Trafficking of Nucleostemin Family Proteins: Common versus Protein-Specific Mechanisms
Mol. Cell. Biol., December 15, 2007; 27(24): 8670 - 8682.
[Abstract]