NA14 Is a Novel Nuclear Autoantigen with a Coiled-coil Domain*

The serum from a patient with Sjögren’s syndrome (RM serum) was used to screen a human testis cDNA expression library. A cDNA of 865 base pairs containing the entire coding sequence for a novel protein was isolated. The 14-kDa predicted protein contains an acidic domain (amino acids 6–80) with a high frequency of heptad repeats characteristic of α-helices that form dimeric coiled-coil structures and an alkaline carboxyl-terminal domain (amino acids 81–119). It seems to be widely expressed, but its expression level varies depending on tissues. A protein of apparent molecular mass of 14 kDa was immunoprecipitated from cell lysates by the autoimmune serum, and it was recognized by rabbit antibodies raised to a recombinant bacterial fusion protein generated from the cDNA clone. Conventional and confocal immunofluorescence microscopy on HeLa and 3T3 cells transiently transfected with a tagged form of the protein showed numerous punctate structures scattered throughout the nucleus. This novel protein has been termed NA14 for NuclearAutoantigen of 14 kDa.

The serum from a patient with Sjö gren's syndrome (RM serum) was used to screen a human testis cDNA expression library. A cDNA of 865 base pairs containing the entire coding sequence for a novel protein was isolated. The 14-kDa predicted protein contains an acidic domain (amino acids 6 -80) with a high frequency of heptad repeats characteristic of ␣-helices that form dimeric coiled-coil structures and an alkaline carboxylterminal domain (amino acids 81-119). It seems to be widely expressed, but its expression level varies depending on tissues. A protein of apparent molecular mass of 14 kDa was immunoprecipitated from cell lysates by the autoimmune serum, and it was recognized by rabbit antibodies raised to a recombinant bacterial fusion protein generated from the cDNA clone. Conventional and confocal immunofluorescence microscopy on HeLa and 3T3 cells transiently transfected with a tagged form of the protein showed numerous punctate structures scattered throughout the nucleus. This novel protein has been termed NA14 for Nuclear Autoantigen of 14 kDa.
Human autoantibodies are valuable reagents for identification and study of novel cellular constituents since they usually recognize highly conserved components, which otherwise are difficult to detect due to their low immunogenicity. In addition, the autoantibodies usually react with the active sites of the antigen, and therefore, they are often capable of inhibiting its functional activity (1).
In a number of rheumatic diseases, the presence of the socalled antinuclear antibodies is a dominant feature. These diseases include systemic lupus erythematosus, mixed connective disease, polymyositis, scleroderma, and Sjögren's syndrome (2).
RM serum, from a patient with Sjögren's syndrome, contains high titer autoantibodies to the Golgi apparatus and low titer autoantibodies to the nucleus (11). Western blotting and immunoprecipitation experiments demonstrated that this serum shows multiple autoreactivities and recognizes three major bands of 210, 130, and 45 kDa (p210, p130, and p45, respectively) in HeLa cell extracts. p210 was shown to be localized at the Golgi complex (11). We have recently cloned and characterized the cDNA encoding p210, now called GMAP-210, by immunoscreening of a HeLa cell cDNA library. 1 In the present study, we used RM serum to screen a human testis oligo(dT) cDNA library, and we isolated a clone encoding a novel protein of 14 kDa, termed NA14, that localizes to the nucleus. The analysis of its amino acid sequence revealed the presence of a domain probably involved in the formation of a coiled-coil structure and sequence similarity with a number of proteins with coiled-coil motifs.

EXPERIMENTAL PROCEDURES
Cell Culture and Lysis-Jurkat cells were grown in RPMI 1640 medium (Bio-Whittaker) supplemented with 10% heat-inactivated FCS. 2 HeLa cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% FCS. Mouse NIH3T3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% newborn calf serum. 2 mM L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin were included in all culture media. All cells were maintained in a 5% CO 2 humidified atmosphere at 37°C.
Antibodies-RM serum of the patient suffering from Sjögren's syndrome (12,13) was aliquoted, had sodium azide added, and then stored at Ϫ70°C. IgG fraction (10 mg/ml final concentration) was purified from whole serum on protein A-Sepharose columns and stored in 50% glycerol at Ϫ70°C. Specific antibodies from RM serum were affinity purified on nitrocellulose strips using immunoreactive proteins from total cell extracts as described (14).
cDNA Cloning and Sequence Analysis-RM serum diluted 1:1500 in TBST (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Tween 20) containing 5% non-fat dry milk was used for immunoscreening of 5 ϫ 10 5 recombinants from a human testis oligo(dT)-primed cDNA library in gt11. Positive clones were rescreened and plaque-purified. Two positive clones (5 and 10) were isolated and subcloned into pTZ19R plasmid. The entire 10 EcoRI insert was used as a probe to hybridize against a HeLa random primed cDNA ZAPII library. The clones obtained were isolated and subcloned in vivo into pBluescript plasmid using R408 helper phage as described in the manufacturer instructions (Stratagene). The nucleotide sequence was determined by using an automated sequencer from Pharmacia Biotech Inc.
Nucleic acids and protein sequences were analyzed by the University * This work was supported by the Junta de Andalucía, by the Ministerio de Educación y Ciencia of Spain (SAF96 -0275), and by Institut Curie, France. 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) Z96932.
§ Contributed equally to this work. ¶ Supported by a return grant of the TMR program from the European Union.
ʈ Supported by a Fundación Cá mara fellowship from the University of Seville.
of Wisconsin Genetics Computer Group sequence analysis software package version 8.1 for UNIX computers (15). Comparisons to known sequences were performed by BLAST (16) on the Internet server. Secondary structure analysis was conducted with the software programs Coils (17), and PeptideStructure (18 -20).
Preparation of GST Fusion Proteins and Rabbit Anti-autoantigen Fusion Protein Antibodies-A fusion protein containing the complete EcoRI insert from 10 was constructed at the carboxyl terminus of GST using the prokaryotic expression vector pGEX-4T-1. Plasmids were transformed in Escherichia coli strain DH5␣. Expression of the GST fusion proteins was induced by the addition of 1 mM isopropyl-␤-Dthiogalactoside, and the fusion proteins were isolated from bacterial lysates by affinity chromatography with glutathione-agarose beads (Sigma). NA14 was released from the fusion protein by treatment with 1% (w/w) thrombin in cleavage buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2.5 mM CaCl 2 ) for 2 h at 25°C. A polyclonal antibody was generated in rabbit using the entire GST-NA14 fusion protein. The antiserum obtained after the fourth immunization, named anti-NA14, was used in the experiments described.
Northern Blotting-A poly(A) ϩ RNA blot from CLONTECH was hy-bridized with the 32 P-labeled cDNA insert from clone 10 and washed as described in the manufacturer instructions. Electrophoresis and Immunoblot Analyses-Proteins were separated by SDS-PAGE on 15% acrylamide gels (21) and stained with Coomassie Brilliant Blue. Gels were electrophoretically transferred to nitrocellulose filters (22), and immunoblot analyses were carried out as described (11).
Immunoprecipitation-HeLa and Jurkat cell lysates (1.5 ϫ 10 7 cells) were incubated with 5 l of RM serum, normal human serum, or an irrelevant human autoimmune serum for 2 h at 4°C, with agitation. 50 l of a 50% solution of protein A-Sepharose was added, and incubation was continued for 1 h. Then the protein A-Sepharose beads were collected and washed five times with Nonidet P-40-buffer. Proteins were released from the beads by boiling in SDS-PAGE sample buffer and analyzed by immunoblotting.
Transient Transfection of HeLa and 3T3 Cells-The full-length open reading frame of na14 (EcoRI insert from 10) was cloned in-frame with the HA epitope (23) into the eukaryotic expression vector pECE (24) to obtain an HA epitope-tagged NA14. The resulting plasmid, pECE-HA-NA14, was purified with a plasmids kit from Qiagen followed by phenol Clones 5 and 10 were isolated from a gt11 human testis cDNA library by immunoscreening with autoimmune RM serum. A ZAPII HeLa cDNA library was screened with 10 yielding clones Z2 and Z20. The open reading frame is boxed. B, cDNA and predicted amino acid sequence of na14. A putative leucine zipper motif and the polyadenylation signal AATAAA are both underlined. Putative motifs predicted by PROSITE are indicated as follows: N-glycosylation site is underlined, casein kinase II phosphorylation sites are in bold, and PKC phosphorylation site is in bold and is italicized. The sequence is available from EMBL/GenBank TM under accesion number Z96932. C, amino acid sequence alignment of human and mouse NA14. A mouse EST sequence (AA274144) homologous to human na14 was obtained by searching the Gen-Bank TM data base with the FASTA program. A BESTFIT comparison between human NA14 and the predicted mouse amino acid sequence is shown. The percent of identity is 96.6. extraction and ethanol precipitation. HeLa or 3T3 cells were split 24 -48 h before transfection so that they were 60 -80% confluent on the day of transfection. 5 ϫ 10 6 cells per assay were resuspended in 200 l of 15 mM HEPES-buffered serum-containing medium, mixed with 50 l of 210 mM NaCl containing 20 -40 g of plasmid DNA, and electroporated at 240 V and 950 microfarads using a Bio-Rad Gene Pulser. 12 h after electroporation, medium was replaced by fresh medium, and cells were processed for Western blotting or immunofluorescence staining after 24 h.
Immunofluorescence Microscopy-Cells were grown on culturetreated slides for 24 or 48 h before an experiment. Cells were rinsed twice with phosphate-buffered saline and incubated in methanol at Ϫ20°C for 6 min to simultaneously fix and permeabilize the cells. After methanol treatment, cells were processed as described (11). Conventional immunofluorescence analysis was performed on a Leica epifluorescence microscope. Confocal laser scanning microscopy was performed using a Leica CLSM instrument based on a Leitz Diaplan microscope with a plan-Neofluar objective ϫ 100. A series of optical sections at 0.25 m intervals were recorded.

RESULTS AND DISCUSSION
Cloning of na14 cDNA-When serum RM was used to screen 5 ϫ 10 5 plaques from a human testis gt11 oligo(dT)-primed cDNA expression library, two clones (5 and 10) were isolated. The cDNA inserts were subcloned into pTZ19R and sequenced. The two clones contained the poly(A) tail as expected and gave identical sequences in the overlapping regions, clone 5 being 14 bp longer than 10 and containing an 865-bp insert. Analysis of the sequence showed that it was not part of GMAP-210 cDNA. 1 In addition, affinity purified antibodies from p130 and p45 did not show staining of the phage plaques. From these data, we concluded that the clones obtained did not correspond to any of the three major proteins recognized by the autoimmune RM serum (11). To obtain an additional sequence, a screening of a ZAPII HeLa random primed cDNA library was performed using the insert from clone 10 as a probe. Two clones (Z2 and Z20) were obtained, and the cDNA inserts were subcloned into pBluescript and sequenced. Both inserts were contained in the sequence present in 5, and no new sequence was obtained (Fig. 1A).
Northern analysis using the entire 10 insert as a probe showed a single band that fits well with the size of the cDNA contained in 5 (Fig. 2, see below).
Sequence Analysis-As shown in Fig. 1B, the entire sequence comprises 865 bp and contains an open reading frame of 360 bp. The first ATG, at positions 47-49, is in excellent agreement with Kozak's sequence for eukaryotic translation initiation (25). The open reading frame is terminated by a stop codon at positions 404 -406. A polyadenylation signal AATAAA (26) is found at positions 837-842 in the 3Ј-untranslated region, and a poly(A) tail is present at the end of the sequence. These data confirm that the complete coding sequence was present in 5 and 10.
The predicted protein sequence consists of 119 amino acids and has a calculated molecular mass of 13,656 daltons and a pI of 5.18. This protein was named NA14.
The PROSITE Dictionary of Protein Sites and Patterns revealed consensus motifs for N-glycosylation, casein kinase II phosphorylation, and PKC phosphorylation (Fig. 1B). It is not known at present how many of these motifs are utilized, but N-glycosylation site would not be used since NA14 is nuclear (see below). Analysis of the NA14 amino acid sequence with the Coil program showed that a significant portion of the protein had a high probability of forming an extended coil (amino acids 6 -80). The secondary structure and charge display of NA14 was predicted using the programs PeptideStructure and Pepstats (not shown). According to these analyses, the protein can be divided in a coiled-coil domain of pI 4.74 and a carboxylterminal basic domain of pI 10.22. Coiled-coil domains have been noted in many other autoantigens, including the nuclear mitotic antigen (NuMA) (27), the nuclear autoantigenic sperm protein (NASP) (28), lamin B (29), myosin heavy chain (30), SS-A/Ro (31), and Ku86 (6). These domains have also been described in a number of other prokaryotic and eukaryotic proteins. A sequence motif similar to the leucine zipper pattern of eukaryotic transcription factors is also found between amino acids 8 and 22 (Fig. 1B). This motif promotes dimerization through ␣-helical coiled-coil formation. Certain proteins that contain this motif are able to bind to DNA and are thought to be regulatory in their functions (32); however, these structures can also mediate dimerization or oligomerization of proteins that do not bind to DNA (33)(34)(35)(36). These observations suggest that NA14 could participate in protein-protein interactions.
The cDNA sequence was used to search the EMBL and GenBank TM data bases for homologous sequences. No overall homology was found with known genes. When the deduced amino acid sequence for NA14 was used to search the SwissProt data bank, 20 -28% identity was found with myosin heavy chain and other proteins that contain coiled-coil domains.
Interestingly, several homologous mouse EST sequences, some of which contained complete coding sequence for na14, were found. The identity between the human and the murine cDNA sequences was higher than 85%. The high degree of conservation at the amino acid level between the human and the predicted mouse NA14 protein should be noted. Both proteins are 96.6% identical (Fig. 1C).
A search on the TIGR Human Gene Index using the BLAST program revealed many human EST sequences similar to na14 cDNA. Table I lists those being more than 95% identical to na14 cDNA, indicating the region of na14 to which homology was shared, the percent of identity, and the tissues or cell strains used to prepare the cDNA libraries from which the sequences were cloned. na14 mRNA is present in both fetal and adult normal tissues, as well as in cancer tissues and transformed cell lines.
Tissue Distribution of NA14 -To investigate the expression of na14, Northern blot analysis of mRNAs from different human tissues was performed with 10 insert as a probe. As seen in Fig. 2, the highest level of expression is observed in testes although na14 can be detected in all the tissues assayed. This result agrees with the data summarized in Table I.
Polyclonal Antiserum Production and Analysis of Expression of NA14 by Western Blotting and Immunoprecipitation-The entire insert from 10 clone was subcloned into a pGEX-4T-1 bacterial expression vector. A GST fusion protein of the expected molecular mass of 41 kDa was produced. This protein was recognized by RM autoimmune serum (Fig. 3A, left panel).
To show the specificity of the recognition, the fusion protein was cut with thrombin to generate GST and NA14. RM serum only recognized the NA14 moiety (Fig. 3A, left panel).
A polyclonal antiserum was raised in rabbit against the GST fusion protein. The resulting immune rabbit serum (anti-NA14) reacted strongly with the fusion protein and with both GST and NA14 thrombin products by immunoblotting (Fig. 3A,  right panel).
The expression of NA14 was assessed in three laboratory cell lines by Western blotting with the rabbit polyclonal serum. Nonidet P-40 lysates from HeLa (human epithelial cells), Jurkat (human T lymphocytes), and NIH3T3 (murine fibroblasts) were resolved by SDS-PAGE, transferred to a nitrocellulose filter, and revealed with the polyclonal serum anti-NA14. The endogenous NA14 was not observed in 3T3 or HeLa cell lysates, but it could be detected in lysates from Jurkat cells after long blot exposures, suggesting that NA14 is present at low levels in these cells (Fig. 3b, left panel). The band of 66 kDa observed in all lanes was revealed by the preimmune rabbit serum (not shown).
HeLa and Jurkat cell lysates were also incubated with RM serum or other irrelevant sera, and complexes were precipitated with protein A-Sepharose. Immunoprecipitates were then analyzed by Western blotting with anti-NA14 polyclonal serum (Fig. 3B, right panel). NA14 was detected in RM immunoprecipitates from Jurkat cells but not HeLa cells or when a normal human serum or another autoimmune serum was used. These results show that (i) NA14 is an autoantigen specifically recognized and immunoprecipitated by RM autoimmune serum, (ii) NA14 seems to be more expressed in Jurkat than in HeLa cells although this protein should also be present in this cell line since two clones were isolated from a HeLa cDNA library (Fig. 1), and (iii) the size of the endogenous protein (Fig.  3B) fits well with the size of the cloned protein as observed after thrombin cleavage of GST-NA14 (Fig. 3A), confirming that the entire cDNA for NA14 was cloned.
Transfection and Subcellular Localization-Subcellular localization of endogenous NA14 was examined in several cell lines by indirect immunofluorescence with anti-NA14 serum, but no staining pattern could be easily recognized (not shown), consistent with the low level of the protein shown by immunoblot analysis (Fig. 3). This prompted us to investigate the subcellular location of NA14 in transfected cells. An HA-tagged form of NA14 was subcloned into the eukaryotic expression vector pECE. Plasmid pECE-HA-NA14 was introduced into human HeLa cells and mouse 3T3 cells by electroporation. Expression of the protein was assessed by Western blotting (Fig. 4, a and f). HeLa and 3T3 cells transiently transfected with HA-NA14 were examined by indirect immunofluorescence microscopy with anti-HA antibodies, anti-NA14 polyclonal an-  3. Immunoblot analysis. A, analysis of bacterial fusion protein with autoimmune serum and anti-NA14 rabbit serum. A NA14 GST fusion protein was produced in bacteria and purified on glutathioneagarose beads. 1 g of GST-NA14 (Ϫ) and GST plus thrombin-released NA14 (ϩ) were electrophoresed, blotted, and probed with autoimmune serum (left panel) or anti-NA14 rabbit serum (right panel). B, immunoblot analysis of cell extracts and autoimmune serum immunoprecipitates. Left panel, Nonidet P-40 lysates from 5 ϫ 10 5 HeLa, Jurkat, and NIH3T3 cells were electrophoresed on a 15% polyacrylamide gel, transferred to nitrocellulose membranes, and immunoblotted with anti-NA14 polyclonal antibodies using the ECL system. Exposure time was 2 h. Right panel, Nonidet P-40 extracts from 1.5 ϫ 10 7 HeLa and Jurkat cells were used for immunoprecipitation with the indicated sera. The immunocomplexes were separated by SDS-PAGE, transferred to nitrocellulose filters, and assayed for their content in NA14 by Western blotting with anti-NA14 rabbit serum. Exposure time after ECL was 10 min. ASerum, autoimmune RM serum; ␣NA14, anti-NA14 rabbit serum; NHS, normal human serum; IHS, irrelevant human autoimmune serum; IgG, IgG heavy chain. tibodies, and RM serum (Fig. 4). All the antibodies, in both HeLa (Fig. 4, b-e) and 3T3 cells (Fig. 4, g-j), revealed the same dotted pattern that localized exclusively to the nucleus. This stain was completely removed by treatment with 0.5% Triton X-100 for 1 min at 37°C (not shown).
A more detailed analysis of NA14 distribution was performed in 3T3 transfected cells by confocal immunofluorescence microscopy. Six optical cuts of a permeabilized cell stained with anti-HA antibody are presented (Fig. 5). NA14 staining appeared concentrated in numerous well defined structures scattered throughout the entire nucleus and that were not associated with the nuclear envelope (as can be seen in bottom panels).
Although no nuclear localization signal was found in the sequence, immunofluorescence studies of HA-NA14 transfected cells demonstrated that this protein is restricted to the nucleus. The remarkable small size of this protein (the smallest coiledcoil protein described to our knowledge) could allow its free diffusion through the nuclear pore. This fact together with the ability to dimerize through the ␣-helical domain suggests that the interaction with other nuclear proteins may be responsible for its nuclear localization despite the absence of a nuclear localization signal. Consistent with this, NA14 was absent from Jurkat nuclear extracts prepared by addition of a high-salt buffer to isolated nuclei, to release soluble proteins without lysing the nuclei (not shown). The well defined pattern of NA14 distribution additionally support that it could be integrated in a complex with other nuclear proteins, probably through hydrophobic interactions since the protein is extracted by detergents but not by high salt concentrations.
Autoantibodies directed against nuclear components are frequently found in the sera of patients with autoimmune diseases. In this study, we have used the antibodies from a patient with Sjögren's syndrome, a disease characterized by lymphocytic infiltrates and fibrosis of exocrine glands, to clone and characterize a novel nuclear autoantigen.
Interestingly, one of the EST sequences identical to na14 cDNA (Table I, accession number U46389) was identified as one of many distinct sequences that are differentially expressed in pancreatic cancer (37). This raises the possibility that na14 could be a potential oncogene or a cancer-associated marker. These data, together with the different levels of expression observed in different tissues or cell lines, suggest that the expression of NA14 might be regulated in normal or pathological conditions. Further experiments will be necessary to ascertain these views and to find the role of this new nuclear protein. The expression of the tagged protein was assessed by Western blotting with anti-NA14 rabbit serum, autoimmune RM serum, and anti-HA monoclonal antibodies (A and F). The only band detected corresponds to HA-NA14. Exposure time after ECL incubation was 1 min. The localization of the transfected protein was analyzed by immunofluorescence microscopy by staining with rhodamine-linked anti-HA antibodies (C, E, H, J). The same cells were also stained with anti-NA14 serum (B, G) or autoimmune RM serum (D, I) and with fluorescein-linked anti-rabbit and anti-human IgG secondary antibodies, respectively. A speckled nuclear pattern is observed.