Address correspondence and reprint requests to Dr Masato Umeda, Department of Molecular Biodynamics, The Tokyo Metropolitan Institute of Medical Science (RINSHOKEN), 3-18-22 Honkomagome, Bunkyo-ku, Tokyo 113–8613, Japan. E-mail: email@example.com
Nadrin is a GTPase-activating protein (GAP) for the rho family of GTPases that controls Ca2+-dependent exocytosis in nerve endings. In this study, three novel splice variants of nadrin were identified and the variants were designated as nadrin-102, -104, -116 and -126 according to their relative molecular masses. All nadrin variants share the GAP domain, coiled-coil domain, serine/threonine/proline-rich domain, SH3-binding motif, and a successive repeat of 29 glutamines. Tissue distribution analyses using polyclonal antibodies that can discriminate each variant showed that the expression of nadrin-102, -104 and -116 was dominant in neuronal tissues and correlates well with the differentiation of neurons while nadrin-126 was strongly expressed in embryonic brain. Expression of nadrin-116 in PC12 cells strongly inhibited NGF-dependent neurite outgrowth and this effect was dependent on its GAP activity. In contrast, no significant effect on either cell morphology or neurite outgrowth was observed with other variants. All variants showed punctate appearance throughout the cytoplasm, while the 66-kDa carboxyl-terminal fragment of nadrin-102 and/or nadrin-116 was localized to the nucleus and its nuclear translocation was accelerated by NGF-induced differentiation of the cells. These results suggested that nadrin variants are different in their ability to regulate rho-mediated signaling and that, in addition to being aGTPase-activating protein, nadrin-102 and -116 have otherdistinct functions in the nucleus of the cell, implying apossible role in the cross-talk between the cytoskeleton andthe nucleus.
Rho GTPases cycle between an active GTP-bound form and an inactive GDP-bound form (Boguski and McCormick 1993). In the GTP-bound form the rho GTPases are able to interact with their downstream effector molecules, which comprise signaling pathways to regulate multiple processes in the morphological development of neurons. The activities of Rho GTPases are regulated by three types of enzymes; guanine nucleotide exchange factors (GEFs), GTPase activating proteins (GAPs) and guanine-nucleotide dissociation inhibitors (GDIs) (Van Aelst and D'Souza-Schorey 1997). GEFs enhance the exchange of bound GDP for GTP and thus activate the rho GTPases, while GAPs stimulate the intrinsic GTP hydrolyzing activity of rho GTPases, thereby inactivating them. In addition, GDIs inhibit the exchange of GDP for GTP and might also serve to regulate the association with membranes (Sasaki and Takai 1998). Although several RhoGEFs have been shown to be important for axon guidance (Gebbink et al. 1997; Awasaki et al. 2000; Penzes et al. 2001; Shamah et al. 2001), less is known about the cellular function of RhoGAPs in the development and function of the nervous system. Among the RhoGAPs that have been described, the function of p190RhoGAP has been extensively studied. P190RhoGAP is highly expressed in the developing and adult mammalian nervous systems and regulates axon outgrowth, guidance, and branch stability (Brouns et al. 2000; Billuart et al. 2001; Brouns et al. 2001). RhoGAPs comprise a family of proteins that share significant sequence homology in a conserved GAP domain, the RhoGAP domain (Boguski and McCormick 1993). To date, more than 20 proteins containing RhoGAP domains have been identified in mammalian cells, and the existence of multiple RhoGAPs suggests that these proteins may act on different downstream effectors in signal transduction events as well as controlling different subcellular pools of RhoGTPases (Van Aelst and D'Souza-Schorey 1997).
We have previously identified a novel neuron-associated developmentally regulated protein, nadrin (Harada et al. 2000). Nadrin has a unique structure; it contains a RhoGAP domain, a putative coiled-coil domain, a serine/threonine/proline rich domain, an SH3-binding motif, a succession repeat of 29 glutamines, and a carboxyl-terminal STAL sequence that binds to a PDZ (PSD-95/DlgA/ZO-1-like) domain-containing protein EBP50 (Reczek and Bretscher 2001). The RhoGAP domain of nadrin activates intrinsic GTPase activities of RhoA, Rac1, and Cdc42 in vitro and nadrin was suggested to control Ca2+-dependent exocytosis, most likely by regulating reorganization of the cortical actin filament network in nerve endings (Harada et al. 2000). Here we report the identification of three novel variants of nadrin with different tissue distribution and timing of expression during brain development. We studied the effect of nadrin variants on NGF-induced neurite outgrowth of PC12 cells and showed that the variants are different in their ability to regulate Rho-mediated signaling. The analyses of cellular localization also suggest that, in addition to being a GTPase-activating protein, some nadrin variants may play a role in the possible cross-talk between the cytoskeleton and the nucleus.
Materials and methods
Wistar rats were obtained from the Japan SLC Inc. (Sizuoka, Japan). They were maintained on laboratory chow and water ad libitum‘The Tokyo Metropolitan Institute of Medical Science Institutional Animal Care and Use Committee according to National Institutes of Health Animal Care and Use protocol’ approved all experimental protocols.
Cloning of nadrin cDNA
A λgt11 cDNA library constructed from the brain of an 8-week-old (8 W) female Wistar rat was screened by colony hybridization using a 400-bp probe, from 1010 nt to 1410 nt of nadrin cDNA (Harada et al. 2000). Positive clones were isolated by successive rounds of plaque purification. cDNA samples were subcloned into a pBluescript vector (Stratagene, La Jolla, CA, USA) and sequenced on an ALFred DNA sequencer (Amersham Parmacia Biotech, Piscataway, NJ, USA) using an AutoCycle sequencing kit (Amersham Pharmacia Biotech). Sequence analyses were performed using Genetyx Version 10.0 (Software Development Co. Ltd, Tokyo, Japan).
RT-PCR reactions was conducted as previously described (Itoh et al. 2000). Briefly, 1 µg of total RNA was isolated from embryonic day 15 (E15) and 8-week-old (8 W) Wistar rat brain and reverse transcribed using SuperScriptTM II reverse transcriptase (Life Technologies, Rochville, MD, USA). PCR was performed by using the pair1 primers 5′-AATCACTCATCCCATACTGGAA-3′, 5′-GCAGCTGTGTGGTCTCGCTTTG-3′, which were primers flanking insert 1, and pair2 primers 5′-GCCACCACCCCCGAATCCAC-3′, 5′-TGAGAACCAAAGTCCAAGTA-3′, which were primers flanking insert 2. Forty cycles of amplification were employed. The PCR products were analyzed by 2.0% agarose gel and visualized by ethidium bromide staining.
Polyclonal antibodies against nadrin were raised in Japanese White female rabbits against the synthetic peptides: Cterm1; CKEVPGHILLDIDNDTESTAL, corresponding to the C-terminal sequences of nadrin-104 and -126, and Cterm2; CRIVTDALPGALTGGEGFQN, corresponding to carboxyl-terminal amino acid sequences of nadrin-102 and -116, respectively, with addition of an extra cysteine to the N-terminal end, according to the method as described (Harada et al. 2000). Briefly, the peptides were conjugated with maleimide-activated keyhole limpet hemocyanin (Pierce, Rockford, IL, USA) according to the manufacturer's protocol. Two rabbits were immunized subcutaneously with 250 µg of synthetic peptide keyhole limpet hemocyanin conjugate in Freund's adjuvant (Difco Laboratories, Detroit, MI, USA), followed by three injections of the antigen in Freund's incomplete adjuvant (Difco) at 4-week intervals. Antibodies were isolated from the immunized sera of the rabbits by affinity chromatography on a synthetic peptide-conjugated 2-fluoro-1-methylpyridinium toluene-4-sulfonate (FMP)-activated column (Seikagaku Corp., Tokyo, Japan).
Organs were isolated from Wistar rats, then homogenized in 10% (w/v) SET buffer (10 mm Tris-HCl (pH 7.4), 250 mm sucrose, 1 mm EDTA pH 7.4, 5 mmN-ethylmaleimide) containing protease inhibitor cocktail (Boehringer Mannheim, Mannheim, Germany) and 1 mm phenylmethanesulfonyl fluoride at 4°C. Protein levels in the homogenates were determined using the BCA system (Pierce). The protein concentrations of the samples were adjusted and the samples were placed in reducing sample buffer [62.5 mm Tris-HCl pH 6.8, 10% glycerol, 1% sodium dodecyl sulfate (SDS), 0.01% bromophenol blue, 1.25% 2-mercaptoethanol]. Immunoblotting was performed as described previously with some modifications (Harada et al. 1999). Proteins from various sources were separated by SDS–polyacrylamide gel electrophoresis and blotted onto polyvinylidene difluoride (PVDF) membrane (Millipore, Bradford, MA, USA) at room temperature using a protein transfer system (ATTO Co., Tokyo, Japan). The membranes were blocked by incubation in blocking buffer (10 mm Tris-HCl buffer (pH 7.4), 150 mm NaCl, 5% skimmed milk, and 0.05% Tween (20) for 1 h at RT with gentle shaking and then incubated with anti-Cterm1 or anti-Cterm2 at a dilution of 1 : 100 for 12 h at 4°C in a blocking buffer. Bound antibodies were detected with horseradish peroxidase-conjugated anti-rabbit IgG (Amersham Pharmacia Biotech) at a dilution of 1 : 1000 and visualized by using an enhanced chemiluminescence (ECL) western blotting detection kit (Pierce).
Subcloned PC12 cells (kindly provided by Dr Y. Fukui, The University of Tokyo) were grown in Dulbecco's modified minimum essential medium (Sigma, St Louis, MO, USA) with 10% horse serum (Life Technologies) and 5% fetal bovine serum (JRH Biosciences, Lenexa, KS, USA). COS7 cells were grown in Dulbecco's modified minimum essential medium with 5% fetal bovine serum (JRH Biosci.).
Construction of expression vectors
The cDNA constructs encoding the N-terminal FLAG-tagged and C-terminal GFP (green fluorescent protein)-tagged variants of nadrin isoforms were generated by PCR using primers that engineered 5′XhoI and 3′XbaI restriction sites into mammalian expression vector pME18S vector (GenBankTM accession number AB009864), and 5′XhoI and 3′BamHI restriction sites into pEGFP-N1 vector (Clontech laboratories, Palo Alto, CA, USA), respectively. To create a nadrin mutant that lacks GAP activity, a point mutation was introduced so as to alter Arg288, which is suggested to be required for the catalytic activity of GAP (Muller et al. 1997; Rittinger et al. 1997) to Ala by overlap PCR (Horton et al. 1989). The PCR products were digested with XhoI and XbaI and ligated into pME18S vector. The nucleotide sequences of PCR products were confirmed by DNA sequencing of both strands.
Expression of recombinant variants in COS7 cells
COS7 cells were plated at 2 × 105 cells/dish and transfected with 2 µg plasmid DNA by calcium chloride method (Graham and vander Eb 1973). Cells were incubated with the transfection mixture for 24 h, rinsed and incubated for further 24 h. Cells were harvested, washed and resuspended in SET buffer. The cells were boiled in reducing sample buffer and were analyzed by immunoblotting as described above.
Transfection and immunofluorescence microscopy
PC12 cells were plated at a density of 6 × 104 cells on a cover slip coated with poly l-lysine (100 µg/mL, Sigma). The cells were transfected with 0.4 µg of nadrin expression vector by using LipofectAMINE according to the manufacturer's instructions (Life Technologies). After 5 h, transfected cells were treated for 2 days with 50 ng/mL nerve growth factor (TaKaRa, Kyoto, Japan) to induce cell differentiation. Immunocytochemistry was performed as described previously (Emoto and Umeda 2000). Briefly, the cells were fixed with 3.7% formaldehyde for 15 min and washed three times with PBS and blocked in blocking buffer for 1 h at room temperature. Anti-Cterm1, anti-Cterm2 or anti-FLAG M2 (Sigma) were used at a dilution of 1 : 50, 1 : 250 and 1 : 1000, respectively. The cells were visualized by incubating 1 h with Alexa Fluor 488-conjugated goat anti-rabbit or anti-mouse secondary antibody (Molecular probe, Eugene, OR, USA) at a dilution of 1 : 2000. F-actin was stained with tetramethylrhodamine B isothiocyanate-labeled phalloidin (Sigma) at a dilution of 1 : 200. The cells were examined with a Confocal imaging system (LSM510; Carl Zeiss, Oberkochen, Germany). To study the effect of nadrin variants on NGF-induced neurite outgrowth, the cells transfected with nadrin expression vectors were identified by anti-FLAG M2. The number of neurite-bearing cells (bearing more than three neurites which were longer than the diameter of their cell body) was determined and the percentage of the neurite-bearing cells among the transfected cells was calculated.
An expression vector for the GAP domain of nadrin and nadrin R288A GAP mutant fused to glutathione S-transferase (GST) was made by subcloning the cDNA fragment encoding amino acid residues 218–482 of nadrin into a pGEX-vector (Amersham Pharmacia Biotech). GST fusion proteins were expressed in DH5αEscherichia coli cells, purified by affinity chromatography on glutathione–sepharose (Amersham Pharmacia Biotechnology) with elution buffer (50 mm Tris-HCl (pH 7.5), 150 mm NaCl, 5 mm MgCl2, 1 mm dithiothreitol, 10 mm glutathione), and dialyzed into 20 mm Tris-HCl (pH 7.5), 150 mm NaCl, 5 mm MgCl2, and 1 mm dithiothreitol. The GAP activity of nadrin was assayed using a GTPase-activating protein assay biochem kit (Cytoskeleton, Denver, CO, USA) according to the manufacturer's protocol. Briefly, recombinant GST-tagged RhoA (1.5 mm) were preloaded 10 min at 30°C with 10 mCi of [γ-32P]GTP (6000 Ci/mmol, NEN Life Science Products) in 5 mm Tris-HCl (pH 7.5), 5 mm EDTA, 0.4 mm dithiothreitol, and 8.3 mm NaCl. After the addition of 25 mm MgCl2, preloaded GTPases (final concentration, 100 nm) were diluted in buffer [10 mm Tris-HCl (pH 7.5), 0.05 mm dithiothreitol, 0.5 mg/mL bovine serum albumin (BSA), 0.5 mm (GTP), and proteins (0, 50, or 250 nm GST–GAP fusion protein)] were added to the reaction mixture. Aliquots were incubated for 5 min at 25°C, and the reaction was stopped by adding 1 mL of ice-cold buffer [50 mm Tris-HCl (pH 7.5), 50 mm NaCl, 10 mm MgCl2] and affinity beads and incubated for 15 min at 4°C. The beads were washed with ice-cold buffer and subjected to scintillation counting.
Isolation of nuclear fraction from PC12 cells
PC12 cell nuclei were isolated according to the method as previously described (Ye et al. 1999). Briefly, 1 × 108 cells were rinsed with phosphate-buffered saline, removed from the dish with a cell scraper, and collected by centrifugation (400 gfor 10 min). The pellet was resuspended and washed twice with cold PBS and then resuspended in a two packed cell volume (PCV) of hypotonic buffer [10 mm HEPES, pH 7.9, 10 mm KCl, 1.5 mm MgCl2, 0.5 mm phenylmethylsulfonyl fluoride (PMSF), 0.5 µg/mL of pepstatin A, and 1 mm dithiothreitol] followed by homogenization using Dounce glass homogenizer with a loose-fitting pestle.Cell lysis was monitored microscopically. The homogenates were centrifuged through a 1.5-mL cushion of buffer (10 mm HEPES–KOH, pH 7.6, 2.4 m sucrose, 15 mm KCl, 2 mm EDTA, 1 mm dithiothreitol, and 0.5 mm PMSF) in a SW55Ti rotor (Beckman Instruments Inc., Palo Alto, CA, USA) at 75 000 gfor 30 min at 4°C. The transparent supernatant above the cushion is the supernatant fraction. The pellet, nuclear fraction, was washed once and resuspended in SET buffer and boiled in reducing sample buffer. SDS–polyacrylamide gel electrophoresis and immunoblotting were performed as described above. The densitometric analysis of signal intensities in immunoblotting was performed using Gel Doc system by Quantity One Program (Bio-Rad, San Francisco, CA, USA).
Results and discussion
Isolation of rat nadrin splicing variants
We have cloned nadrin cDNA by immunological screening of expression libraries constructed from adult rat brains (Harada et al. 2000). Further screening of the cDNA libraries using nadrin 400-bp probe (nucleotide 1010–1410) identified two cDNA clones, N1 and N2, and DNA sequencing of the isolated clones revealed that two extra insert, 234-bp insert (insert 1) in the position of 1576-nucleotide, and 48-bp insert (insert 2) in the position of 2297-nucleotide were present in the clone N1 and N2, respectively (Fig. 1a). Insert 1 was located upstream near the polyglutamine-coding sequence and insertion of insert 2 resulted in a reading frameshift which caused an early termination and exposure of a unique carboxyl-terminal amino acid sequence without the STAL sequence motif. The deduced amino acid sequences of each clone were shown in Fig. 1(b). A comparison with listed sequences of the putative genomic organization of human nadrin homologue (GenBankTM accession number, AC010545 and AC046167) showed that sequence of inserts 1 and 2 fully matched exons 17 and 20, respectively. This indicates that the clone N1 and N2 were derived from alternative splice variants of nadrin.
To address whether mRNA representing the observed variant sequences could be independently demonstrated in rat brain, poly(A)+RNA sequences were amplified by reverse transcription-polymerase chain reaction (RT-PCR) from embryonic day 15 (E15) and 8-week-old (8 W) rat brain using sets of primers to distinguish between the variant sequences by size (Fig. 2a). On agarose gels all resulting PCR products correlated well with the fragment sizes predicted for the variants; two products of 919-bp and 685-bp bands by using pair 1 primers, and 681-bp and 633-bp bands by using pair 2 primers, were detected in both E15 and 8 W samples. Differences in the magnitude of the RT-PCR products were also observed. The 685-bp fragment was almost not detectable in stage E15, but was increased in 8 W. It may reflect a low expression level of the isoform without insert 1 in the embryonic stages that is also suggested by the immunoblotting analyses (Fig. 3a). The low expression could also be caused by a reduced stability ofthe mRNA. These findings indicated that the splice variants of nadrin were actually expressed in rat brain and that the transcription of the splice variants may be developmentally regulated.
To demonstrate that the variant sequences are expressed as distinct protein products, cDNA expression plasmids corresponding to the expected four variants were prepared. The recombinant proteins transiently expressed in COS7 cell cultures were detected by using polyclonal antibodies raised against the synthetic peptides corresponding to the unique carboxyl-terminal sequences of the variants (Fig. 1b). The antibodies were affinity purified from the resulting antiserum and used for immunoblotting and immunofluorescence analyses. Anti-Cterm1 raised against the carboxyl-terminal 20 amino acids common to nadrin variants without insert 2 detected 104- and 126-kDa bands and anti-Cterm2 against the carboxyl-terminal 19 amino acid sequence resulted from the insertion of insert 2 detected 102- and 116-kDa bands (Fig. 2b). The observed relative sizes were fully compatible with the primary structures predicted by the variant, suggesting that each sequence variant encodes a distinct protein product. We designated the variants as nadrin-102 (GenBankTM accession number, AB080637), -104 (GenBankTM accession number, AB042827), -116 (GenBankTM accession number, AB060556), and -126 (GenBankTM accession number, AB060557) according to their relative molecular masses (Fig. 2c).
Developmental stage and tissue-specific expression of nadrin variants
Developmental stage and tissue distribution of nadrin variants were analyzed by using anti-Cterm1 and anti-Cterm2 antibodies. In developing rat brain, nadrin-126, recognized by anti-Cterm1, was abundantly expressed in the embryonic stage and its expression decreased gradually during the postnatal stages. Only marginal amounts of nadrin-102, -104, -116 were detected in the embryonic stages, and their level of expression increased gradually during the postnatal weeks to adulthood (Fig. 3a). In adult tissues, nadrin-102 and -116 were detected in thymus, adrenal gland and spleen, with particularly prominent expression of nadrin-102 in spleen (Fig. 3b). Expression of nadrin-126 was also demonstrated in stomach and adrenal gland, and nadrin-104 was detected in stomach and thymus (Fig. 3b). Nadrin-102 and -116 were detected as a pair in any tested tissues, implying a possible co-operative action between these variant. The expression of nadrin-126 is, however, distinct from those of other variants, implying its specific role in embryogenesis.
Differential effect of nadrin variants on neurite outgrowth from PC12 cells
Rat pheochromocytoma PC12 cells have been used as a model system for neuronal differentiation and neurite outgrowth (Katoh et al. 2000). To determine the possible functional differences between the nadrin variants, the nadrin variants were expressed in PC12 cells and their effect on nerve growth factor (NGF)-induced neurite outgrowth was examined. In this experiment, transfected FLAG-tagged nadrin variants were transiently transfected into PC12 cells and the cells expressing the protein were identified by immunofluorescence with anti-FLAG mAb. Expression of nadrin-116 strongly inhibited neurite outgrowth and more than 80% of the cells expressing the variant failed to display neurites (Figs 4e,f and 5a). The inhibition of neurite outgrowth was also observed with cells expressing the RhoGAP domain of nadrin (Figs 4i,j and 5a). In contrast, significant neurite outgrowth was observed with cells expressing other nadrin variants, nadrin-102, -104 and -126 (Figs 4a–d,g,h). It has been shown that the Arg residue, which is highly conserved in GAP domains, is required for their catalytic activity. The equivalent residue in nadrin is Arg288 and replacement of Arg288 with Ala in the GAP domain reduced the GAP activity to one-fifth of the original activity (Fig. 5b). To confirm that the observed inhibitory effect of nadrin-116 on neurite outgrowth was due, at least in part, to GAP activity of the protein, we replaced Arg288 of nadrin-116 and the effect of the GAP activity-negative nadrin-116 (nadrin-116 GAP mutant) on neurite outgrowth was examined. Expression of the nadrin-116 mutant had no significant inhibitory effect on the neurite outgrowth (Figs 4k,l and 5a), suggesting that GAP activity is required for the inhibitory effect of nadrin-116 on NGF-induced neurite outgrowth.
In PC12 cells, NGF-induced neurite outgrowth is dependent upon activation of Rac1 and Cdc42 (Katoh et al. 2000; Yamaguchi et al. 2001) and suppression of their activities by introducing dominant-negative forms of Rac1 and Cdc42 inhibited neurite outgrowth (Daniels et al. 1998). The following lines of evidence suggest that the inhibitory effect of nadrin-116 on neurite outgrowth was due, at least in part, to the suppression of Rac1 and/or Cdc42 activities through its GAP activity. First, the recombinant GAP domain of nadrin activates the intrinsic GTPase activity of RhoA, Rac1, and Cdc42. Overexpression of nadrin in NIH3T3 cell markedly induced the reorganization of actin cytoskeleton, suggesting that nadrin function as a GAP for rho family proteins (Harada et al. 2000). Second, the inhibitory effect of nadrin-116 was not observed with the GAP activity-negative mutant, suggesting that GAP activity is essential for the effect. Third, Richnau and Aspenstrom (2001) have recently isolated a human rho GTPase-activating protein that interacts with the Src homology 3 (SH3) domain of Cdc42-interacting protein 4 (CIP4). The amino acid sequence of RICH showed 83% identity with nadrin and all the typical domains including SH3 domain are conserved between these proteins, indicating that RICH is a human homologue of nadrin. Since, in vivo, RICH was shown to act as a GAP for Rac1 and Cdc42, it is likely that excessive expression of nadrin-116 in PC12 cells suppressed the activity of Rac1 and/or Cdc42, resulting in the inhibition of NGF-induced neurite outgrowth.
Of considerable interest is the observation that not all thenadrin variants showed inhibitory effects on neurite outgrowth, which implicates functional differences in signaling pathways. Recently, Reczek and Bretscher (2001) reported that nadrin binds to the PDZ (PSD-95/DlgA/ZO-1-like) domain of a cortical scaffolding protein EBP50 (ERM-binding phosphoprotein-50) through its carboxyl-terminal STAL sequence. The carboxyl-terminal STAL sequence is conserved in nadrin-104 and -126, implying a role of these variants in the reorganization of the cortical cytoskeletons. Nadrin-102 and -116 lack the STAL sequence by an early termination of translation caused by the insertion of inset2. Hence, these two variants may be recruited to other functional sites where they down-regulate RhoGTPase-signaling pathways or act as effectors of additional signaling pathways downstream of RhoGTPases (Kozma et al. 1996). Alternatively, the differential effect on neurite outgrowth may be due to differences in the tertiary structures of the nadrin variants, which was suggested by the immunoblotting analyses. We previously raised a rabbit polyclonal antibody against the internal amino acid sequence (amino acids 562–580) of nadrin-104 (Harada et al. 2000). In immunoblotting analyses, this antibody bound strongly to nadrin-102 and -104, but only weakly to nadrin-116 and -126, suggesting that the insertion of 78 amino acids into the middle region of nadrin caused significant conformational change inthe tertiary structure of the protein (data not shown). Theconformational difference may affect the GAP activity or substrate specificity of the nadrin variants. To test thishypothesis, we have tried to purify the recombinant full-length nadrin variants and to measure their in vitro GAP activities. This approach, so far, has not been successful since the recombinant full-length nadrin expressed in either E. coli or yeast was extremely unstable to be purified to a single protein.
Cellular localizations of nadrin variants in PC12 cells
To further investigate the functional differences between the variants, the cellular localization of endogenous nadrin variants in PC12 cells was examined by immunofluorescence staining. Immunoblotting clearly demonstrated that all of the four variants were present in PC12 cells, with a relative abundance of nadrin-116 (Fig. 6a). Figure 6(b) shows a single confocal plane showing the immunofluorescent staining of PC12 cells at 48 h after the NGF-stimulation. Anti-Cterm1, which was reactive with nadrin-104 and -126, exhibited a punctate staining throughout the cytoplasm, and the staining was also evident in the tip of the neurite and peripheral of the cells where an intense staining of actin filaments was observed (Fig. 6b, upper panel). Anti-Cterm2, which was reactive with nadrin-102 and -116, showed the similar immunofluorescence profiles with those observed with anti-Cterm1 antibody, except that an intense speckled staining was evident in the nucleus (Fig. 6b). The intensity of the nuclear staining by anti-Cterm2 was increased 72 h after the NGF-treatment (Fig. 6c). The immunofluorescence staining by anti-Cterm1 and anti-Cterm2 antibodies were completely abolished when the antibodies were preincubated with their antigenic peptides, indicating that the fluorescence images obtained with anti-Cterm1 and anti-Cterm2 represented the localization of the pairs of nadrin variants, nadrin-104 and -126, and nadrin-102 and -116, respectively. The results suggested that nadrin-102 and/or nadrin-116 were localized in the nucleus and their nuclear localization was accelerated by NGF-induced differentiation of the cells.
We further confirmed the immunofluorescent data by immunoblotting of in vitro prepared nuclear (N) and postnuclear (S, supernatant) fractions from PC12 cells. The purity of the nuclear preparation was assessed on the basis of immunocytochemical criteria for the presence of lamin B, a protein marker for nucleus, and tubulin, a protein enriched in cytoplasm (Fig. 7a). In the postnuclear fractions, the pairs of nadrin-104 and -126, and nadrin-102 and -116 were detected by the anti-Cterm1 and anti-Cterm2 antibodies, respectively. In the nuclear fractions, no protein band was detected by anti-Cterm1 antibody, while an intense band corresponding to molecular mass of 66 kDa was detected by anti-Cterm2 antibody (Fig. 7b, right panel). The 66-kDa band was detected before the NGF-treatment and the signal intensity was doubled 72 h after NGF-induced differentiation (Fig. 7c, upper panel). The binding of anti-Cterm2 to the 66-kDa band was completely inhibited by preincubation with the synthetic antigenic peptide (Fig. 7c, middle panel). To further confirm the nuclear localization of the C-terminal fragment of nadrin-116, cellular localization of green fluorescent protein (GFP)-tagged fragments was examined The carboxyl-terminal domain of nadrin with predicted molecular mass of 70 kDa (amino acids 498–830 of nadrin-116; C-half) expressed in PC12 cells localized specifically to the nucleus, while the N-terminal fragment (amino acids 1–498 of nadrin-116; N-half) localized throughout the cytoplasm (Fig. 8). These results raised the possibility that nadrin-102 and/or nadrin-116 are processed by a protease and the resulting carboxyl-domain of the protein is translocated into the nucleus.
Although accumulating data indicate the involvement of RhoGTPases in regulating nuclear signaling (Van Aelst and D'Souza-Schorey 1997; Bar-Sagi and Hall 2000), so far, only one protein that directly exert its effect on RhoGTPases, p190RhoGAP, has been found in the nucleus (Settleman et al. 1992). The presence of a highly consensus 778 amino acid residues with those of glucocorticoid receptor repression factor-1 (GRF-1) in p190RhoGAP implicates its role in transcriptional regulation (Settleman et al. 1992), though the function of p190RhoGAP in the nucleus remains to be elucidated. The enrichment of proline, serine and threonine (P/S/T) residues in the carboxyl-terminal domain and the presence of polyglutamine of the nadrin variants are reminiscent of the structural motif typical for transcription factors (Hoffman et al. 1990; Blank and Andrews 1997). Sequence analysis of nadrin also shows the presence of two stretches of positively charged amino acids in the carboxyl-terminal domain (in nadrin-104, Arg479LysArgPro and Arg552ArgAlaValLysLys) that are similar to the nuclear localization signal (NLS) sequences (Gorlich and Mattaj 1996). The nuclear translocation of the nadrin fragment suggested that, in addition to being a GAP, nadrin may play a unique role in the cross-talk between the cytoskeleton and the nucleus. Further studies are, however, required to elucidate the molecular mechanism and biological significance of the nuclear localization of the nadrin variants.
In summary, we have identified the splice variants of nadrin that have distinct characteristics of cellular localization and ability to regulate rho-mediated signaling pathways. Nuclear localization of the variant implicated their unique roles in carrying signals into the nucleus. Since other alternative splice variants of nadrin homologue have also been found in human tissues (Richnau and Aspenstrom 2001), nadrin forms a unique GAP family that appears to play important roles in the regulation of signaling networks in eukaryotic cells.