Characterization of ubiquitin-like polypeptide acceptor protein, a novel pro-apoptotic member of the Bcl2 family Morihiko Nakamura 1 and Yoshinori Tanigawa 2 1 Cooperative Medical Research Center and 2 Department of Biochemistry, Shimane Medical University, Japan Monoclonal nonspecific suppressor factor (MNSF) is a cytokine with antigen nonspecific suppressive activity. MNSFb (a subunit of MNSF) is a 14.5 kDa fusion protein consisting of a protein with 36% identity with ubiquitin and ribosomal protein S30. The ubiquitin-like segment (Ubi-L) may be cleaved from MNSFb in the cytosol. Recently, we have observed that Ubi-L covalently binds to intracellular proteins in mitogen-activated murine T-helper type 2 clone, D.10 cells. In this study, we purified a 33.5 kDa Ubi-L adduct from D.10 cell lysates by sequential chromatography on DEAE, anti-(Ubi-L) Ig–conjugated Sepharose, and hydroxylapatite. MALDI-TOF-MS fingerprinting revealed that this Ubi-L adduct consists of an 8.5 kDa Ubi-L and a Bcl2-like protein, murine orthologue of a previously cloned human BCL-G gene product with pro-apoptotic function. Murine Bcl-G mRNA was highly expressed in testis and significantly in spleen. In addition, the level of Bcl-G mRNA expression was increased in concanavalin A- and inter- feron c-activated D.10 cells. The 33.5 kDa Ubi-L adduct was expressed in spleen but not in testis, even though Bcl-G protein was highly expressed in this tissue. The antisense oligonucleotide to Bcl-G significantly decreased the level of the Ubi-L adduct formation in concanavalin A-activated D.10 cells and the proliferative response of the D.10 cells. These results suggest that the post-translational modification of Bcl-G by Ubi-L might be implicated in T-cell activation. Keywords: Bcl2; T-cell; ubiquitin-like protein. The covalent attachment of ubiquitin to proteins and their subsequent degradation by the 26 S proteasome represents the most commonly ascribed role for the protein ubiquiti- nation system. In this respect, ubiquitin conjugation to target substrates participates in a variety of important eukaryotic processes, such as DNA repair [1], cell cycle control [2], ribosome biogenesis [3], and the inflammatory response [4]. In addition to ubiquitin, it is evident that several ubiquitin-like proteins have been found to be covalently or noncovalently attached to target proteins [5–7]. Monoclonal nonspecific suppressor factor (MNSF), a lymphokine produced by murine T-cell hybridoma, pos- sesses pleiotrophic antigen-nonspecific suppressive func- tions [8]. We have cloned a cDNA encoding a subunit of MNSF, which was termed MNSFb [9]. MNSFb cDNA encodes a protein of 133 amino acids (aa) consisting of a ubiquitin-like protein (36% identity with ubiquitin) fused to the ribosomal protein S30. The ubiquitin-like moiety (Ubi-L) of MNSFb shows MNSF-like biologic activity without cytotoxic action [10]. Interferon c (IFNc)is involved in the mechanism of action of Ubi-L. We have demonstrated that Ubi-L specifically binds to cell surface receptors on mitogen-activated lymphocytes and the T- helper type 2 clone, the D.10 cells [11]. We have also shown that Ubi-L covalently conjugates to acceptor proteins and forms Ubi-L adducts including the 33.5 kDa protein in concanavalin A (Con A)- and IFNc- stimulated D.10 cells [12]. Intracellular function of Ubi-L remains largely unknown. In this study, we isolated and characterized the 33.5 kDa Ubi-L adduct in D.10 cells. Peptide mass fingerprinting using MALDI-TOF MS after in-gel V8 protease digestion revealed that Ubi-L covalently binds to a novel protein, a new member of the Bcl2 family, suggesting that the Ubi-L conjugation might be involved in the mechanism of survival of T-cells. Materials and methods Purification of the 33.5 kDa Ubi-L adduct D.10 G4.1 cells, a murine T-helper clone type 2, were cultured in the presence of 3 lgÆmL )1 Con A (Calbiochem, La Jolla, CA) as described previously [12,13] to a density of 5 · 10 )6 cellsÆmL )1 (total volume 5 L). Cells were collected by centrifugation and solubilized in lysing buffer (0.01 M sodium phosphate buffer, 1% Triton, 0.5% sodium deoxycholate, 0.1% SDS, 0.1 M NaCl, 1 m M EGTA, 10 lgÆmL )1 aprotinin, 10 lgÆmL )1 leupeptin, 2 m M phenyl- methylsulfonyl fluoride). After sonication, cell debris were removed by centrifugation at 28 000 g at 4 °C for 60 min. Supernatants were dialysed against Buffer A (10 m M Tris/ HCl, pH 7.2, 1 m M phenylmethylsulfonyl fluoride) and applied to a column of DEAE equilibrated with the same buffer. The column was washed extensively with Buffer A containing 50 m M NaClandelutedwithBufferA Correspondence to M. Nakamura, Cooperative Medical Research Center, Shimane Medical University, 89-1 Enya-cho, Izumo 693–8501, Japan. Fax: + 81 853 20 2913, Tel.: + 81 853 20 2916, E-mail: nkmr0515@shimane-med.ac.jp Abbreviations: MNSF, monoclonal nonspecific suppressor factor; Ubi-L, ubiquitin-like moiety of MNSF; Con A, concanavalin A; IFNc,interferonc; SUMO, small ubiquitin-related modifier. (Received 5 July 2003, revised 4 August 2003, accepted 12 August 2003) Eur. J. Biochem. 270, 4052–4058 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03790.x containing 75 m M NaCl. The eluates were applied to anti-(Ubi-L) Ig affinity column and Ubi-L adducts were eluted, as described previously [14]. The eluates containing anti-(Ubi-L) Ig reactive proteins were dialysed against 10 m M sodium phosphate (pH 7.3) and applied to a column of hydroxylapatite. The column was eluted with 80 m M sodium phosphate. To obtain the 33.5 kDa Ubi-L adduct, each of the fractions were assayed by immunoblotting with the use of anti-(Ubi-L) Ig as described below. Antibody preparation A peptide corresponding to aa 199–208 of Bcl-G underlined in Fig. 2A was synthesized with the use of the multiple- antigen peptide system as described previously [15]. After purification by reverse-phase HPLC, the multiple-antigen peptide system was used to immunize rabbits. IgG in the serum was purified by the use of protein A-Sepharose. Immunoblotting Cell extracts in SDS sample buffer were subjected to 12.5% SDS/PAGE, and blotted onto polyvinylidene fluoride membranes. The membranes were blocked with 5% ovalbumin in NaCl/P i for 1 h and then washed with NaCl/P i containing 1% Triton X-100 (NaCl/P i /Triton X). Subsequently, the membranes were incubated with anti- (Ubi-L) rabbit IgG (anti-PU1) in the blocking buffer, after which they were incubated with peroxidase-conjugated anti-rabbit IgG. Detection was done according to the enhanced chemiluminescence detection system (Amersham Biosciences). We have previously demonstrated that anti-PU1 does not crossreact with ubiquitin [12]. RT-PCR A mouse multiple tissue cDNA panel used as a template for semiquantitative RT-PCR to confirm tissue-specific expres- sion of Bcl-G was obtained from Clontech (Palo Alto, CA). PCR was performed for 30 cycles according to the manufacturer’s instructions. The PCR primers used to detect Bcl-G mRNA shown in Fig. 3 are as follows: sense, 5¢-CCCAAGCTCTCCAGAACAAG-3¢;antisense,5¢-CT GAGCTCGGATCTCCTTTG-3¢ (213 bp). All short amplified PCR products were isolated and sequenced to verify their identity. PCR products were separated by electrophoresis through 2% agarose gel and stained with ethidium bromide. In some experiments, signals were quantitated by densitometry and optical densities for Bcl-G were normalized to the corresponding values for glyceraldehyde-3-phosphate dehydrogenase. In-gel digestion with trypsin and peptide separation The stained protein band from SDS/PAGE was digested in-gel, and the peptides were extracted essentially according to the methods of Rosenfeld et al. [16]. The peptides were separated by reversed-phase chromatography on C18 column (Nacarai Tesque Inc., Kyoto, Japan) using a linear gradient of acetonitrile (4–40% in 150 min) in formic acid. The flow rate was 5 lLÆmin )1 and detection was at 214 nm. Selected peaks were collected in tubes containing 10 lLÆmin )1 of 30% acetonitrile, 0.1% formic acid and subjected to sequence analysis. In-gel digestion and MALDI-TOF Silver-stained spots were cut out of the gels for in-gel digestion and destained with 1 mL 50 m M sodium thiosul- fate, 15 m M potassium ferricyanide followed by four washes in 1 mL H 2 O. The spots were then equilibrated for 20 min in 500 lLof100m M ammonium bicarbonate and then incubated for 20 min in 500 lL of 50% acetonitrile, 50 m M ammonium bicarbonate. The spots were dried, rehydrated for digestion with 5 lgÆmL )1 V8 protease (Sigma) in 25 m M ammonium bicarbonate, and incubated at 37 °Covernight. The reaction was stopped by adding 1 lL of 88% formic acid. The peptides were extracted from the gel matrix by vortex for 30 min and then concentrated using Zip Tips (Millipore Corp.). Peptide mass fingerprinting was per- formed using a PerkinElmer/PerSeptive Biosystems Voy- ager-DE-RP MALDI-TOF mass spectrometer, operating in delayed reflector mode at an accelerating voltage of 20 kV. The peptide samples were cocrystallized with matrix on a gold-coated sample plate using 0.5 lLmatrix (a-cyano-4-hydroxytranscinnamic acid) and 0.5 lLsample. Cysteines were treated with iodoacetamide to form carboxy- amidomethyl cysteine, and methionine was considered to be oxidized. Mutagenesis and transfection Mutant Bcl-G (K110R) was generated by replacing the codon for lysisne 110 with the codon for arginine by utilizing QuickChange site-directed mutagenesis (Strata- gene). cDNA encoding Bcl-G was subcloned in frame into pQE-31 vector (Qiagen), resulting in the addition of a His 6 tag to the amino terminus. The gene was then subcloned into the vector pcDNA3.1(+) (Invitrogen Corp.). D.10 cells were transfected with 5 lg plasmid DNA as described by Zhang et al.[17]. Oligonucleotide treatments The oligonucleotides were synthesized as phosphorothioate- containing two methoxyethyl modifications at positions 1–5 and 15–20. The antisense oligomer was complementary to nucleotides 318–337 of Bcl-G which encode aa 24–31 of the protein. Sequences were follows: antisense Bcl-G, 5¢-CG TAGAAGGCCAGGATTTTG-3¢;senseBcl-G,5¢-CAA AATCCTGGCCTTCTACG-3¢. The cells were transfected 20 h after Con A-activation using 15 lgÆmL )1 Lipofectin (Invitrogen) and 1 l M Bcl-G antisense or sense oligomer for 6 h. The cultures were washed with serum-free RPMI 1640 three times to remove Lipofectin. Cells were stimulated with additional Con A following transfection for 46 h. Results Purification of a 33.5 kDa Ubi-L adduct from murine T-cell line Previous experiments have shown that Ubi-L conjugates to acceptor proteins in Con A- and IFNc-stimulated T-cells Ó FEBS 2003 Ubiquitin-like polypeptide acceptor protein (Eur. J. Biochem. 270) 4053 and T-cell lines. Con A-activated D.10 cells specifically induces the 33.5 kDa Ubi-L adduct [12]. Thus, we tried to isolate this adduct to elucidate the function of an unknown Ubi-L target molecule in D.10 cells. The Ubi-L adduct was induced in a 1-L shake flask culture for 2 days as described and the 33.5 kDa Ubi-L adduct was purified to homogen- eity by a combination of ion exchange chromatography, anti-(Ubi-L) (anti-PU1) affinity chromatography and hyd- roxylapatite chromatography (Fig. 1). The 33.5 kDa Ubi-L adduct present in each fraction was identified by immuno- blotting using anti-(Ubi-L) Ig. The final preparation gave a single stained band on SDS/PAGE with mobility corres- ponding to 33.5 kDa under reducing conditions (Fig. 1, lane 2). This protein band was the first subjected to N-terminal sequence analysis after electroblotting. Despite repeated attempts, ambiguous signals were obtained from about 100 pmol protein. For internal sequencing the protein band was digested in-gel with trypsin. Selected peptides were subjected to sequence analysis with results as shown in Table 1. MS of 33.5 kDa Ubi-L adduct MALDI-TOF-mass fingerprinting was performed by sep- arating the 33.5 kDa Ubi-L adduct by SDS/PAGE under reducing conditions. Bands corresponding to the Ubi-L adduct were excised and subjected to in-gel digestion with V8 protease as described. Then, the resulting mixtures of peptides were analysed by MALDI-TOF MS. Table 2 shows the peptide masses of observed by MALDI-TOF mass fingerprinting of 33.5 kDa Ubi-L adduct purified from murine D.10 cells. The resulting sets of peptide masses were then used to search the NCBI database for potential matches, confirming the Ubi-L adduct as a Ubi-L–Bcl2-like protein complex. This Bcl2-like protein is a murine ortho- logue of human Bcl-G, a novel pro-apoptotic member of the Bcl2 family. Bcl-G possesses the Bcl2 homology domains (BH2 and BH3) (Fig. 2). Signals were detected at 1241.0 and 1412.1 Da that correspond to aa 38–49 of Ubi-L and 20–32, respectively. Importantly, a pair of signals was detected at 1556.6 and 2934.7 Da (Table 3). These signals correspond to digestion fragments in which aa 104–111 of Bcl-G are covalently linked by an isopeptide bond to aa 67–72 of Ubi-L, and aa 104–124 are linked by an isopeptide Fig. 1. SDS/PAGE of the 33.5 kDa Ubi-L adduct. Purified fractions of Ubi-L adduct analysed by SDS/PAGE (12% polyacrylamide gel) and immunostained for protein. Lanes 1 and 3 contain an aliquot from the anti-PU1 affinity chromatography purification step; lanes 2 and 4 contain an aliquot from the hydroxylapatite purification step; lanes 1 and 2, silver-stained; lanes 3 and 4, immunostained with anti-(Ubi-L). Mobilities of the 33.5 kDa Ubi-L adduct and the molecular mass standards (kDa) are indicated to the right and left, respectively. Table 1. Internal peptide sequences. The amino acid sequence of Ubi-L is shown in bold. Peptide Sequence Residues 1 VACIANR 154–160 2 FEGPCDSK 249–256 3 ALGTWSTDSWTQV 57–69 4 AQELHT 7–12 Table 2. Assignments for peptide fragments from a Staphylococcus V8 protease digest of the 33.5 kDa Ubi-L adduct. The 33.5 kDa Ubi-L adduct was digested by V8 protease and subjected to MALDI-MS analysis. The data in the second column are the mass values obtained experimentally, whereas the results in the third column are those calculated from the V8 protease fragmentation of the gene products of Bcl-G and Ubi-L. The fourth column indicates the number of the first and last amino acid of the identified Bcl-G and Ubi-L peptides, whereas the fifth shows the corresponding amino acid sequences. Protein Mass (MH + ) Residues Sequence Observed Calculated Bcl-G 897.0 897.1 318–325 KILGISHE 930.1 930.1 208–215 QIISKIVE 1283.9 1283.5 173–184 VIHSQGGSKLKE 1397.0 1396.6 112–124 IRAQGPQGPFPVE 1599.0 1598.7 305–317 YFSPWVQQNGGWE 2065.2 2065.4 288–304 NHPMNRMLGFGTKYLRE 2152.8 2153.3 125–142 RQSGFHNQHWPRSLSSVE 2350.1 2349.8 81–100 KNISLGKKKSSWRTLFRVAE Ubi-L 1241.0 1241.4 38–49 DQVVLLAGSPLE 1412.1 1411.6 20–32 TVAQIKDHVASLE 4054 M. Nakamura and Y. Tanigawa (Eur. J. Biochem. 270) Ó FEBS 2003 bond to aa 67–72. Collectively, Ubi-L may conjugate to Bcl- G with a linkage between the C-terminal Gly74 and Lys110. RT–PCR and immunoblotting experiments To investigate mRNA levels of Bcl-G in various organs, we performed PCR on a cDNA reverse transcribed from the mRNA of different organs. A 213 bp PCR product was generated by using the primers within the coding sequence. PCR products were isolated and sequenced to verify their identity (data not shown). Fig. 3A shows that testis had the highest expression as described for human Bcl-G [18]. Low but detectable expression of Bcl-L mRNA was found in some other tissues including spleen. We also tested Bcl-G mRNA levels in D.10 cells incubated with or without Con A and IFNc. As can be seen in Fig. 3B, mRNA level of Bcl-G in D.10 cells was increased by the treatment with Con A and IFNc in good agreement with the previous observations that 33.5 kDa Ubi-L conjugation is increased in activated T-cells [12]. To confirm that Bcl-G covalently conjugates to Ubi-L, we performed immunoblotting of the 33.5 kDa Ubi-L adduct using an antibody against synthetic peptide based on the sequence of Bcl-G. The results of the Fig. 3. RT-PCR analysis of Bcl-G transcripts in mouse tissues. (A) The mouse multiple tissue cDNA panel was subjected to PCR using Bcl-G- specific primers, and the DNA products were analysed by agarose gel electrophoresis. The expected 213 bp band was prominent in cDNA from testis. (B) Total RNA was isolated from cultured D.10 cells. cDNA was synthesized from total RNA and was subjected to PCR as described in (A). The number under each band is the treated/control ratiooftheintensityofeachbandnormalizedtothatofglyceralde- hyde-3-phosphate dehydrogenase measured by densitometry. Fig. 2. Primary structure of Bcl-G. (A) The predicted amino acid sequence of murine Bcl-G is presented with the BH2 and BH3 domains shown in bold and residue numbers indicated. An internal sequence of 10 residues for polyclonal antibody is double underlined. The under- lined amino acid sequences correspond to peptides whose masses were detected by MALDI-TOF mass fingerprinting of the extracted in-gel digest. Lys110 responsible for isopeptide formation is circled. (B) Amino acid sequence of Ubi-L is presented with the C-terminal Gly-Gly doublet shown in bold. Internal sequences obtained following V8 protease digest are underlined. Table 3. Isopeptide bonds between the C terminal of the glycine residue of Ubi-L and the lysine of Bcl-G. The 33.5 kDa Ubi-L adduct was digested by V8 protease and subjected to MALDI-MS analysis. The data in the first column are the mass values obtained experimentally, whereas the results in the second column are those calculated from the V8 protease-fragmented peptide complexes. The third column shows the corresponding amino acid sequences of Ubi-L and Bcl-G (shown in bold). Mass (MH + ) Sequence Observed Calculated 1556.6 1556.9 2934.7 2934.4 Ó FEBS 2003 Ubiquitin-like polypeptide acceptor protein (Eur. J. Biochem. 270) 4055 blotting revealed that this antibody specifically recognized the authentic 33.5 kDa Ubi-L adduct (Fig. 4A), indicating that Ubi-L covalently binds to Bcl-G. Thus, the results of immunoblotting together with the internal peptide sequences in Table 1 and MALDI-TOF analysis show that Ubi-L covalently binds to Bcl-G via an isopeptide bond. To determine how much of endogenous Bcl-G is present as a Ubi-L adduct in D.10 cells, immunoprecipitation experi- ments were performed. As can be seen in Fig. 4B the 33.5 kDa Ubi-L was precipitated with anti-(Bcl-G) Ig from a lysate of Con A-activated D.10 cells, whereas no Ubi-L adduct was precipitated from unstimulated cells. We next determined whether transfection of a K110R mutant of a His-tagged Bcl-G construct would reveal the absence of Ubi-L modified Bcl-G. As shown in Fig. 4C, Ubi-L linked Bcl-G could be immunoprecipitated with anti-Bcl-G anti- body from Con A-activated D.10 cells transfected with wild-type Bcl-G, but not from cells transfected with the mutant Bcl-G. These results were consistent with those of fingerprinting analysis (Table 3). We next carried out immunoblotting analysis to measure 33.5 kDa Ubi-L and Bcl-G levels in different organs of mice. As shown in Fig. 4D the 33.5 kDa Ubi-L adduct formation was repro- ducibly found in the spleen and thymus. Interestingly, high level of the Ubi-L adduct was found consistently in the brain. Unexpectedly, we could not observe Ubi-L adduct in the testis, even though this tissue expressed higher level of Bcl-G (Fig. 3A and Fig. 4D). Antisense oligonucleotide to Bcl-G inhibits the proliferative response of mitogen-stimulated T-cell line We next examined the effect of the antisense oligonucleotide to Bcl-G on T-cell functions. The proliferative response of Con A-activated D.10 cells was measured 52 h after addition of the oligonucleotide. Antisense oligonucleotide to Bcl-G inhibited the response of the mitogen-stimulated cells by 33 ± 5% when compared with control cells treated with Lipofectin alone (Fig. 5A). An equal concentration of the sense oligonucleotide showed no effect. To confirm that the antisense oligonucleotides were effective in depleting Bcl-G protein expression, we examined the levels of Bcl-G protein in unstimulated D.10 cell lysates 42 h after oligo- nucleotide application. As detected by immunoblotting analysis (Fig. 5B), the antisense oligonucleotide effectively decreased the level of Bcl-G when compared with control cells or to cells treated sense oligonucleotide. In addition, we tested the levels of 35.5 kDa Ubi-L–Bcl-G complex in Con A-activated D.10 cell lysates 22 h after oligonucleotide application. The antisense oligonucleotide to Bcl-G signifi- cantly decreased the level of the Ubi-L adduct formation. Thus, it is possible that the post-translational modification of Bcl-G by Ubi-L might be involved in T-cell activation. Discussion We have previously demonstrated that Ubi-L conjugates to several proteins in Con A- and IFNc-stimulated T-cells [12]. MALDI-TOF mass fingerprinting and immunoblotting by the use of anti-Bcl-G antibody demonstrate that Ubi-L covalently binds to Bcl-G via isopeptide bond in activated T-cells. Thus, Bcl-G might be one of the target molecules of Fig. 4. Immunoblotting of the 33.5 kDa Ubi-L adduct. (A) Authentic 33.5 kDa Ubi-L adduct was subjected to SDS/PAGE, and blotted onto nitrocellulose membranes. The membranes were immuno-stained with antibody to Bcl-G: lane 1, pre-immune (IgG); lane 2, anti-(Bcl-G); lane 3, anti-(Ubi-L) IgG (positive control). (B) Cell lysate was prepared from cell cultured with or without Con A for 48 h. An equal amount of protein (50 lg) from each cell lysate was immunoprecipitated (IP) with anti-(Bcl-G), and the immunoprecipitant was Western blotted (WB) for Bcl-G as well as Ubi-L. (C) D.10 cells were transfected with vectors expressing either wild-type Bcl-G or its mutants (K110R). Transfected cells were stimulated with Con A for 48 h. Subsequently, the lysate was immunoprecipitated with anti-(Bcl-G) Ig, and the bound proteins were analysed by Western blot with the antibodies indicated on the left of each panel. (D) Tissue distribution of 33.5 kDa Ubi-L– Bcl-G complex. Tissues homogenate (50 lgofproteineach)obtained from the indicated organs were subjected to immunoblotting analysis using anti-(Ubi-L) Ig as well as anti-(Bcl-G) Ig. Molecular mass is shown in kDa. Anti-actin Ig was used to calibrate the amount of protein loading and efficient protein transfer. 4056 M. Nakamura and Y. Tanigawa (Eur. J. Biochem. 270) Ó FEBS 2003 Ubi-L. Guo et al. demonstrated that Bcl-G is a novel pro- apoptotic member of the Bcl2 family [18]. They showed that Bcl-G induces apoptosis in transfected cells. The present experiments suggest a new role for Ubi-L as an intracellular regulator of the proliferation of mitogen-activated T-cell. Data supporting this conclusion was obtained by antisense study (Fig. 5). It may be inferred that the formation of Ubi-L–Bcl-G complex in the early phase might be respon- sible for T-cell activation. We have previously demonstrated that the level of the Ubi-L adduct was gradually decreased during T-cell proliferation [10]. Further studies are required to more fully define the mechanism of Bcl-G modification by Ubi-L in T-cell survival. Bcl2 family proteins are functionally classified into two groups. Both Bcl2 and Bcl-XL are anti-apoptotic members of the Bcl2 family protein. In contrast, the other group, comprising Bax and its related proteins including Bid and Bad, promotes apoptosis. The activity of Bcl2 family proteins can be regulated by post-translational modifica- tions, including proteolysis and phosphorylation. Cleavage of Bid by caspase 8 results in translocation of the cleaved Bid to the mitochondria where it induces the release of cytochrome c [19]. Bad is phosphorylated by the prosur- vival kinases Akt [20]. Phosphorylation of Bad provides an important link between extracellular survival factors and the intrinsic cell death pathway regulated by Bcl2. In this context, it is interesting that pro-apoptotic Bcl-G can be modified by Ubi-L protein. We showed that Ubi-L formed complex with Bcl-G in spleen, thymus and activated T-cells. Interestingly, the Ubi-L adduct was also observed in brain. It should be noted that parkin with an N-terminal ubiquitin- like domain is important for the survival of the neurons that degenerate in Parkinson’s disease [21,22]. In contrast, we could not observe the complex in testis, although Bcl-G protein is expressed in this organ. One might speculate that enzyme(s) involved in the Ubi-L conjugation may be absent or noninducible in testis. This phenomenon is being characterized and will be the topic of another paper. We could not identify the N-terminal region of Bcl-G by MALDI-TOF analysis (Table 2) and internal peptide sequence analysis (Table 1). It is possible that Bcl-G might be digested by a caspase, because Bcl-G possesses candidate caspase recognition site at its N-terminal region. Further support for this hypothesis was obtained using immuno- blotting. We carried out immunoblotting of the 33.5 kDa Ubi-L adduct using an antibody against synthetic peptide based on sequence of N-terminal of Bcl-G. This antibody recognized Bcl-G but not the 33.5 kDa Ubi-L adduct (data not shown). Indeed, the migrated position of the Ubi-L adduct is somewhat smaller than the expected mass (38 kDa) of Ubi-L–Bcl-G complex. Ubiquitin-like proteins modify intracellular proteins as well as ubiquitin. The activities of a number of important transcription factors, including p53, c-Jun, and androgen receptor, are regulated by small ubiquitin-like modifier-1 (SUMO-1) modification [23,24]. Thus, one question for future study is whether Ubi-L also modifies some of these factors. The nuclear dot protein Sp100 and promyelocytic leukemia proteins are constituents of nuclear domains, known as nuclear dots or PML bodies, and are both covalently modified by SUMO [25]. It is evident that nuclear dots play a role in autoimmunity [26]. Ubi-L/MNSFb is also implicated in the mechanism of autoimmune disease. We showed the presence of Ubi-L in the ascitic fluid of a patient with systemic lupus erythematosus [27]. Interest- ingly, the expression of both nuclear dots and Ubi-L are induced by interferons [12,14,28–30]. Thus, ubiquitin-like proteins may be involved in the pathogenesis of auto- immune disease. References 1. Jentsch, S., McGrath, J.P. & Varshavsky, A. (1987) The yeast DNA repair gene RAD6 encodes a ubiquitin-conjugating enzyme. Nature 329, 131–134. Fig. 5. Effect of Bcl-G antisense oligonucleotide on mitogen-activated D.10 cells. (A) D.10 cells were activated with Con A and the proli- ferative response was determined as described [10]. Cells were exposed to either antisense oligonucleotide or an equal concentration of sense oligonucleotide: C, Lipofectin only; a, antisense oligonucleotide; S, sense oligonucleotide. Percent suppression was calculated by com- parison with the control response. The decrease in the antisense treated cells was statistically significant compared with control (P <0.05). Data are expressed as mean ± SD. (B) Immunoblotting of D.10 cells treated with oligonucleotides as described in (A). Unstimulated D.10 cells were treated with oligonucleotides and blotted with anti-(Bcl-G) Ig (upper panel); Con A-stimulated D.10 cells were treated with oligonucleotide and blotted with anti-(Ubi-L) Ig (middle panel). Equal quantities of protein (15 lg) were loaded on SDS/PAGE gels. The data represent one of three independent experiments with similar results. Anti-actin Ig was used to calibrate the amount of protein loading and efficient protein transfer (lower panel). Ó FEBS 2003 Ubiquitin-like polypeptide acceptor protein (Eur. J. Biochem. 270) 4057 2. Skowyra, D., Koepp, D.M., Kamura, T., Conrad, M.N., Cona- way, R.C., Conaway, J.W., Elledge, S.J. & Harper, J.W. (1999) Reconstitution of G1 cyclin ubiquitination with complexes con- taining SCFgrr1 and Rbx1. Science 1, 662–665. 3. Finley,D.,Bartel,B.&Varshavsky,A.(1989)Thetailsofubi- quitin precursors are ribosomal proteins whose fusion to ubiquitin facilitates ribosome biogenesis. Nature 338, 394–401. 4. Chen, Z.J., Hagler, J., Palombella, V.J., Melandri, F., Scherer, D., Ballard, D. & Maniatis, T. (1995) Signal-induced site-specific phosphorylation targets I kappa B alpha to the ubiquitin-pro- teasome pathway. Genes Dev. 9, 1586–1597. 5. Mahajan, R., Delphin, C., Guan, T., Gerace, L. & Melchior, F. (1997) A smallubiquitin-related polypeptide involved in targeting RanGAP1 to nuclear porecomplex protein RanBP2. Cell 88, 97–107. 6. Mizushima,N.,Sugita,H.,Yoshimori,T.&Ohsumi,Y.(1998)A new protein conjugation system in human. The counterpart of the yeast Apg12p conjugation system essential for autophagy. J. Biol. Chem. 273, 33889–33892. 7. Liu, Y.C., Pan,. J., Zhang, C, Fan, W., Collinge, M., Bender,. J. R. & Weissman, S.M. (1999) A MHC-encoded ubiquitin-like protein (FAT10) binds noncovalently to the spindle assembly checkpoint protein MAD2. Proc.NatlAcad.Sci.USA96, 4313–4318. 8. Nakamura, M., Ogawa, H. & Tsunematsu, T. (1986) Isolation andcharacterization of a monoclonal nonspecific suppressor fac- tor (MNSF). produced by a T cell hybridoma. J. Immunol. 136, 2904–2909. 9. Nakamura, M., Xavier, R.M., Tsunematsu, T. & Tanigawa, Y. (1995) Molecular cloning and characterization of a cDNA encoding monoclonal nonspecific suppressor factor. Proc. Natl Acad. Sci. USA 92, 3463–3467. 10. Nakamura,M.,Xavier,R.M.&Tanigawa,Y.(1996)Ubiquitin- like moiety of the monoclonal nonspecific suppressor factor b is responsible for its activity. J. Immunol. 156, 532–538. 11. Nakamura, M. & Tanigawa, Y. (1999) Biochemical analysis of the receptor for ubiquitin-like polypeptide. J. Biol. Chem. 274, 18026–18032. 12. Nakamura, M. & Tanigawa, Y. (1998) Ubiquitin-like polypeptide conjugates to acceptor proteins in concanavalin A- and interferon c-stimulated T-cells. Biochem. J. 330, 683–688. 13. Nakamura, M., Xavier, R.M. & Tanigawa, Y. (1995) Monoclonal nonspecific suppressor factor beta inhibits interleukin-4 secretion by a type-2 helper T cell clone. Eur. J. Immunol. 25, 2417–2419. 14. Nakamura, M., Tsunematsu, T. & Tanigawa, Y. (1998) TCR-a chain-like molecule is involved in the mechanism of antigen-non- specific suppression of a ubiquitin-like protein. Immunology 94, 142–148. 15. Tam, J.P. (1988) Synthetic peptide vaccine design: synthesis and properties of a high-density multiple antigenic peptide system. Proc.NatlAcad.Sci.USA85, 5409–5413. 16. Rosenfeld, J., Capdevielle, J, Guillemot, J.C. & Ferrara, P. (1992) In-gel digestion of proteins for internal sequence analysis after one- or two-dimensional gel electrophoresis. Anal. Biochem. 203, 173–179. 17. Zhang, D.H., Cohn, L, Ray, P., Bottomly, K. & Ray, A. (1997) Transcription factor GATA-3 is differentially expressed in murine Th1 and Th2 cells and controls Th2-specific expression of the interleukin-5 gene. J. Biol. Chem. 272, 21597–21603. 18. Guo, B., Godzik, A. & Reed, J.C. (2001) Bcl-G, a novel pro-apoptotic member of the Bcl-2 family. J. Biol. Chem. 276, 2780–2785. 19. Li, H., Zhu, H., Xu, C.J. & Yuan, J. (1998) Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell 94, 491–501. 20. Datta, S.R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y. & Greenberg, M.E. (1997) Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91, 231–241. 21. Kitada, T., Asakawa, S., Hattori, N., Matsumine, H., Yamamura, Y., Minoshima, S., Yokochi, M., Mizuno, Y. & Shimizu, N. (1998) Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature 392, 605–660. 22. Kahns, S., Lykkebo, S., Jakobsen, L.D., Nielsen, M.S. & Jensen, P.H. (2002) Caspase-mediated parkin cleavage in apoptotic cell death. J. Biol. Chem. 277, 15303–15308. 23. Muller, S., Berger, M., Lehembre, F., Seeler, J.S., Haupt, Y. & Dejean, A. (2000) c-Jun and p53 activity is modulated by SUMO-1 modification. J. Biol. Chem. 275, 13321–13329. 24. Poukka, H., Karvonen, U., Janne, O.A. & Palvimo, J.J. (2000) Covalent modification of the androgen receptor by small ubiqui- tin-like modifier 1 (SUMO-1). Proc. Natl Acad. Sci. USA 97, 14145–14150. 25. Sternsdorf, T., Jensen, K., Reich, B. & Will H. (1999) The nuclear dot protein sp100, characterization of domains necessary for dimerization, subcellular localization, and modification by small ubiquitin-like modifiers. J. Biol. Chem. 274, 12555–12566. 26. Guldner, H.H., Szostecki, C., Grotzinger, T. & Will, H. (1992) IFN enhance expression of Sp100, an autoantigen in primary biliary cirrhosis. J. Immunol. 149, 4067–4073. 27. Xavier, R.M., Nakamura, M. & Tsunematsu, T. (1994) Isolation and characterization of a human nonspecific suppressor factor from ascitic fluid of systemic lupus erythematosus. Evidence for a human counterpart of the monoclonal nonspecific suppressor factor and relationship to the T cell receptor alpha-chain. J. Immunol. 152, 2624–2632. 28. Lavau, C., Marchio, A., Fagioli, M., Jansen, J., Falini, B., Lebon, P., Grosveld, F., Pandolfi, P.P., Pelicci, P.G. & Dejean, A. (1995) The acute promyelocytic leukaemia-associated PML gene is induced by interferon. Oncogene 11, 871–878. 29. Chelbi-Alix, M.K., Pelicano, L., Quignon, F., Koken, M.H.M., Venturini, L., Stadler, M., Pavlovic, J., Degos, L. & de the, H. (1995) Induction of the PML protein by interferons in normal and APL cells. Leukemia 9, 2027–2033. 30.Grotzinger,T.,Sternsdorf,T.,Jensen,K.&Will,H.(1996) Interferon-modulated expression of genes encoding the nuclear- dot-associated proteins Sp100 and promyelocytic leukemia pro- tein (PML). Eur. J. Biochem. 238, 554–560. 4058 M. Nakamura and Y. Tanigawa (Eur. J. Biochem. 270) Ó FEBS 2003 . Characterization of ubiquitin-like polypeptide acceptor protein, a novel pro-apoptotic member of the Bcl2 family Morihiko Nakamura 1 and Yoshinori Tanigawa 2 1 Cooperative. to a novel protein, a new member of the Bcl2 family, suggesting that the Ubi-L conjugation might be involved in the mechanism of survival of T-cells. Materials