Biochemical Analysis of the Receptor for Ubiquitin-like Polypeptide*

Monoclonal nonspecific suppressor factor (MNSF), a lymphokine produced by murine T cell hybridoma, possesses pleiotrophic antigen-nonspecific suppressive functions. A cDNA clone encoding MNSF-β, an isoform of the MNSF, has been isolated and characterized. MNSF-β cDNA encodes a fusion protein consisting of a ubiquitin-like segment (Ubi-L) and ribosomal protein S30. Ubi-L appears to be cleaved from the ribosomal protein and released extracellularly in association with T cell receptor-like polypeptide. In the current study we have characterized the biochemical nature of the Ubi-L receptor on D.10 G4.1, a murine T helper clone type 2. Biotinylated Ubi-L bound preferentially to concanavalin A-stimulated but not to unstimulated D.10 cells. Detergent-extracted membrane proteins were applied to an immobilized Ubi-L column. SDS-polyacrylamide gel electrophoresis of eluted fraction revealed a band ofM r = 82,000. Biotinylated Ubi-L specifically recognized this band, confirming that the 82-kDa protein is the Ubi-L receptor. A complex of M r = 90,000 was visualized by immunoprecipitation of 125I-Ubi-L cross-linked to the purified receptor followed by SDS-polyacrylamide gel electrophoresis and autoradiography. In addition, a 105-kDa protein was coimmunoprecipitated by anti-Ubi-L receptor (82-kDa polypeptide) antibody, indicative of the association of this protein with the Ubi-L receptor complex. Amino acid sequence analysis of the 82-kDa polypeptide revealed that the Ubi-L receptor may be a member of a cytokine receptor family.

The monoclonal nonspecific suppressor factor (MNSF) is a product of a concanavalin A (ConA)-activated murine hybridoma that inhibits the generation of lipopolysaccharide (LPS)induced immunoglobulin-secreting cells, proliferation of mitogen-activated T and B cells, and interleukin (IL)-4 secretion by bone marrow-derived mast cells (10,11). We have cloned a cDNA encoding a subunit of MNSF (12). The subunit, termed MNSF-␤, encodes a protein of 133 amino acids consisting of a ubiquitin-like protein (36% identity with ubiquitin) fused to the ribosomal protein S30. We have reported evidence showing that the ubiquitin-like segment of MNSF-␤ (Ubi-L) is responsible for its activity (13). Ubi-L inhibits IgE and IgG1 production by LPS-activated B cells and division in various tumor cell lines of murine origin (14).
Most recently, we have demonstrated that Ubi-L covalently conjugates to intracellular acceptor proteins in vitro (15) and in vivo (16). MNSF-␣, a subunit of MNSF, was identified as an acceptor protein for Ubi-L. It is probable that Ubi-L might be released in a posttranslationally modified form (i.e. conjugation of Ubi-L to MNSF-␣) because it lacks a signal peptide (12). Partially purified isopeptidase dissociates Ubi-L from MNSF-␣, suggesting that the COOH-terminal Gly-Gly doublet of Ubi-L may covalently ligate to the lysine residue in MNSF-␣ (17).
Several other ubiquitin-like proteins have been isolated and characterized. Sentrin (also called SUMO-1), for instance, preferentially modifies nuclear proteins (18). Like ubiquitin, the ubiquitin cross-reactive protein (UCRP) conjugates to a number of intracellular proteins (19). Interestingly, UCRP and Ubi-L are subjected to induction by interferon (IFN) (11,20). Furthermore, they show type specificity for IFN and have immunoregulatory properties (14,21), although they have opposite functions (22). Accordingly, ubiquitin-like proteins may be involved in many biological reactions such as immune responses.
We have presented evidence that mitogen-activated T and B cells, and murine lymphoid cell lines, may have an MNSF receptor (23). Although further biochemical and functional analysis of this receptor protein had been prevented because of the lack of a recombinant ligand, Ubi-L enabled us to isolate and characterize the receptor for MNSF. In the present study, we describe how Ubi-L specifically binds to cell surface receptors on mitogen-activated lymphocytes and the T helper type 2 clone, the D.10 cell. Studies were also performed to characterize the biochemical nature of Ubi-L receptor protein.
Preparation of T Cells and B Cells-T cells were enriched from splenocytes by Lymphoprep centrifugation, passage over nylon wool, and treatment with a mixture of monoclonal Abs against B220 and MAC-1, and complement. All Ab depletions resulted in Ͼ99% of T cells as assayed by flow cytometry. For B cell preparation, splenocytes were incubated with a mixture of monoclonal Abs to Thy-1.2, L3T4, Lyt-2, MAC-1, and a granulocyte-specific antigen, and complement. Live cells were separated from dead cells and erythrocytes on Histopaque. With this procedure, the T-depleted population was consistently more than 95% B cells (B220 ϩ ).
MNSF-␤ and Ubi-L-Recombinant MNSF-␤ and Ubi-L were obtained as described previously (12). Briefly, either MNSF-␤ or Ubi-L was expressed as a fusion protein with glutathione S-transferase (GST) using the pGEX-2T vector (Amersham Pharmacia Biotech). Ubiquitinlike segment was cleaved from the fusion partner by thrombin and purified by using anti-PU1 Ab coupled to Sepharose 4B (Amersham Pharmacia Biotech).
Biotinylation of Ubi-L-500 g of Ubi-L was mixed with a 20-fold molar excess of sulfo-N-hydroxysulfosuccinimide-biotin in 1 ml of 0.01 M sodium phosphate, 0.15 M NaCl, and 0.1 mM phenylmethylsulfonyl fluoride, pH 7.2. Incorporation of biotin into Ubi-L was determined by the 2-(4Ј-hydroxyazobenzene)benzilic acid assay (24), performed according to the manufacturer's instructions (Pierce, Rockford, IL). In some experiments, MNSF-␤ was also biotinylated as described for Ubi-L. Each molecule of Ubi-L was labeled with three molecules of biotin, albeit self-aggregation occurred during the procedure.
Binding Experiments of Biotinylated Ubi-L to Target Cells-Binding experiments were performed at 24°C. Biotinylated Ubi-L (bio-Ubi-L; 20 nM) and the cells (2 ϫ 10 6 ) to be tested were incubated in 200 l of the binding medium (RPMI 1640 containing 1% bovine serum albumin and 20 mM Hepes, pH 7.4) for 2 h. Nonspecific binding, unless stated otherwise, was determined by measuring binding in the presence of 100-fold molar excess unlabeled Ubi-L. Specific binding was determined by subtracting nonspecific binding from total binding. The cells were washed with binding medium five times and incubated with 200 l of streptavidin-horseradish peroxidase conjugate in the same medium for an additional hour. The cells were then washed five times with 0.01 M sodium phosphate, 0.15 M NaCl, pH, 7.4; the substrate, turbo-3,3Ј,5,5Јtetramethylbenzidine (100 l; Pierce), was added for 5 min. The color reaction was halted by the addition of 1 M phosphoric acid (100 l), and bound biotinylated proteins were quantified by measuring the absorbance of the reaction mixture at A 450 nm using a spectrometer. Bio-Ubi-L to cells was qualified by converting absorbance units into mol of bio-Ubi-L using a standard curve made relating the absorbances of known amounts of bio-Ubi-L.
Labeling of Cell Surface Receptors by Biotinylation-Unstimulated and ConA-stimulated D.10 cells (5 ϫ 10 7 ) were washed three times with PBS and incubated in 2 ml of PBS containing 1 mg/ml sulfo-N-hydroxysulfosuccinimide-biotin at 4°C for 1 h. Thereafter, cells were washed three times with cold PBS containing 15 mM glycine to quench the reaction. The packed cells were suspended in homogenization buffer containing 10 mM Tris-HCl, pH 7.5, 1% Nonidet P-40, 0.1% SDS, 150 mM NaCl, 10 g/ml aprotinin, 5 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. The cells were homogenized in a glass homogenizer. Detergent-insoluble materials was removed by centrifugation at 12, 000 ϫ g for 30 min at 4°C. The supernatant fraction was either subjected immediately to affinity chromatography or stored at Ϫ80°C.
Preparation of Cell Membranes-All procedures were performed at 4°C. D.10 cells (2 ϫ 10 8 cells) were washed with Hanks' balanced salt solution and suspended in a hypotonic buffer (10 mM Tris-HCl, pH 7.5, 1 mM MgCl 2 , 1 mM CaCl 2 , 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin). The suspension was mixed vigorously on a Vortex mixer and allowed to stand for 5 min before being spun at three different stages: 700 ϫ g for 5 min, 3,500 ϫ g for 10 min, and 40,000 ϫ g for 1 h. The 40,000 ϫ g precipitate was collected. Membranes were suspended in a solubilization buffer (1% Triton X-100, 10 mM 4-Hepes, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, pH 7.5) and left at 4°C for 1 h with occasional shaking. The mixture was then spun (10,000 ϫ g, 15 min, 4°C), and the supernatant was collected and spun in an ultracentrifuge (100,000 ϫ g, 60 min, 4°C). The supernatant was subjected to ligand blot assay as described below.
Affinity Column Chromatography-Ubi-L was coupled to CNBr-activated Sepharose 4B (Amersham Pharmacia Biotech) according to a procedure provided by the manufacturer. Detergent extract of membrane fraction (13 mg of protein) was applied to the Ubi-L column and incubated overnight under gentle agitation at 4°C. The beads were then packed into a column and washed with sodium phosphate buffer, pH 7.5, containing 100 mM NaCl and 0.1% Triton X-100, and then the receptor was eluted with 1 M NaCl. The elutes were concentrated in Centricon-10 (Amicon, Beverly, MA) and subjected to reverse phase HPLC (Cosmosil, Nacalai tesq, Kyoto, Japan).
To obtain small peptides suitable for amino acid sequence analysis, purified Ubi-L receptor was digested with trypsin at an enzyme:substrate molar ratio of 1:100 at 37°C for 4 h. Tryptic peptides were separated by reverse phase HPLC using a C18 column equilibrated with 5% acetonitrile in 0.1% trifluoroacetic acid. The column was developed with a 5-60% linear gradient of acetonitrile for 120 min at 25°C with a flow rate of 1.0 ml/min. Peptides were sequenced directly from polyvinylidene difluoride membranes using an ABI477 protein sequencer (Applied Biosystems Inc.).
Immunoblotting-The biotinylated cell surface proteins (250 g of protein) were applied to the Ubi-L column and incubated as described above. The receptor was eluted with a 2 ϫ SDS sample buffer, subjected to 10% SDS-PAGE, and blotted onto nitrocellulose membranes. The membranes were blocked with 5% bovine serum albumin in PBS for 1 h and then washed three times with PBS containing 0.5% Tween 20 (PBS/Tween 20). Subsequently, the membranes were incubated with streptavidin conjugated to horseradish peroxidase (1:1,000) in PBS/ Tween 20 for 45 min. Membranes were washed five times in PBS/Tween 20 and developed using the ECL reagents (Amersham Pharmacia Biotech). In some experiments, MNSF-␤ or ubiquitin was coupled to CNBractivated Sepharose 4B and used for affinity chromatography as described above.
Ligand Blot Assay-Detergent-extracted membrane proteins were prepared from antigen-stimulated D.10 cells that were not labeled with biotin. The proteins (730 g) were blotted onto nitrocellulose sheets as described above. The nitrocellulose sheets were incubated in buffer containing 50 mM Tris-HCl, pH 7.5, 0.1% Tween 20, and 50 g/ml biotin-labeled Ubi-L or 200 g/ml biotin-labeled ubiquitin for 16 h at 4°C. The sheets were washed twice with binding buffer without the Ubi-L for 5 min and then incubated with streptavidin conjugated to horseradish peroxidase in PBS/Tween 20 for 30 min and subsequently washed three times for 5 min with the same buffer. Detection of labeled proteins was performed with ECL reagents.
Immunization of Mice-A protein band (82 kDa) was excised from polyacrylamide slab gels, minced, and injected subcutaneously together with complete Freund's adjuvant. Additional immunizations were given 3 and 4 weeks later.
Immunoprecipitation of a Cross-linked 125 I-Ubi-L Receptor Complex with Anti-receptor Serum-Ubi-L was labeled with 125 I as described previously for GST-Ubi-L (15). Aliquots of the purified receptor (500 ng) were mixed with 125 I-Ubi-L and left for 1 h at 24°C. Disuccinimidyl suberate was added to a final concentration of 0.3 mM. The cross-linking was stopped after 30 min at 4°C by the addition of Tris-HCl buffer, pH 7.5, to a final concentration of 20 mM. Mouse anti-receptor serum or control serum was added and incubated for 2 h at 24°C. The antigenantibody complex was adsorbed on protein A-Sepharose and analyzed by SDS-PAGE followed by autoradiography. In some experiments, ConA-activated D.10 cells were biotinylated and solubilized as described under "Labeling of Cell Surface Receptors by Biotinylation." To determine whether proteins other than the 82-kDa polypeptide are associated with Ubi-L receptor, lysates were immunoprecipitated with anti-Ubi-L receptor Ab.
Affinity Chromatography on a Wheat Germ Agglutinin (WGA) Column-Ubi-L receptor purified by a Ubi-L column was incubated at 4°C for 6 h with 5 ml of WGA-Sepharose equilibrated in 50 mM Tris-HCl, pH 7.4, 140 mM NaCl, 10% glycerol, and 0.1% Triton X-100. After incubation, the mixture was transferred to a column, the beads allowed to settle, and the column was washed with 50 ml of 0.5 M KCl, 50 mM Tris-HCl, pH 7.4, 140 mM NaCl, and 0.1% Triton X-100. The absorbed Ubi-L receptor was then eluted with 10 ml of 0.5 M N-acetyl-D-glucosamine, 50 mM Tris-HCl, pH 7.4, 0.5 M NaCl, 10% glycerol, and 0.1% Triton X-100.
Deglycosylation of Ubi-L Receptor Proteins-The membrane fraction of ConA-activated D.10 cells was acidified by the addition of sodium citrate buffer, pH 6.2, to a final concentration of 50 mM. Neuraminidase (0.7 unit) was added, and the mixture was incubated at 37°C for 4 h. SDS sample buffer was added to deglycosylated samples and boiled for 5 min. SDS-PAGE and immunostaining with anti-Ubi-L receptor Ab were carried out as described above.
Neutralizing Tests-Determination of IgE production by LPS-stimulated B cells was done as described previously (14). Briefly, the purified B cells (5 ϫ 10 5 /ml) were cultured with 20 g/ml LPS. Ubi-L, recombinant IFN-␥, and anti-Ubi-L receptor serum (IgG) were added at the initiation of the cultures. Supernatants were harvested 7 days after initiation of the cultures, and IgE production was detected by enzymelinked immunosorbent assay.

Binding Experiments of Recombinant Ubi-L to Target Cells-
Previous experiments have shown that Ubi-L acts on murine helper T cell clone, D.10 cells (25). To investigate whether or not Ubi-L would bind specifically to D.10 cells, a binding assay was performed. Purified recombinant Ubi-L enabled acquisition of bio-Ubi-L with a high specific activity comparable to that of unlabeled Ubi-L. As shown in Fig. 1, bio-Ubi-L bound to ConA-activated (48 h), but not to unstimulated D.10 cells. Bio-Ubi-L bound to the activated D.10 cells most rapidly at 24°C (data not shown). We also tested the possibility that 14.5-kDa MNSF-␤ (Ubi-L-ribosomal protein S30), which shows Ubi-L-like activity (26), might bind to the D.10 cells. It should be noted that the COOH-terminal Gly-Gly doublet of Ubi-L is followed by ribosomal protein S30. Despite of the blocked COOH terminus, biotinylated MNSF-␤ bound to ConA-activated D.10 cells (Fig. 1), as did Ubi-L. The pattern of MNSF-␤ binding was much the same. In contrast, ribosomal protein S30 did not bind to the cells (data not shown). These findings indicate that Ubi-L may not bind to its receptors via the COOH-terminal glycyl doublet responsible for ubiquitination process. Because Ubi-L shows a 36% homology with ubiquitin (12), we tested the possibility of the binding of ubiquitin to D.10 cells. Fig. 1 shows that 10 mM biotinylated ubiquitin, 500-fold the amount of Ubi-L used in the binding experiments, bound slightly to antigen-stimulated D.10 cells probably because of the homology. We also examined whether Ubi-L would recognize mitogen-activated lymphocytes. T cells and B cells were separated from splenocytes as described under "Experimental Procedures." As shown in Fig. 2, exposure of the T cells to 3 g/ml ConA for 2 days led to maximal binding of bio-Ubi-L on the cell surface. Bio-Ubi-L bound to 20 g/ml LPS-activated B cells almost in the same manner as ConA-activated T cells. In contrast, neither unstimulated (day 0) T cells nor B cells showed any significant bio-Ubi-L binding. These results are consistent with those of binding experiments of native MNSF (23).
Ubi-L Receptor Expression Is Limited to Lymphoid Cells-We next investigated the cellular distribution of the Ubi-L receptor. Binding experiments were carried out using various cells of murine and human origin. Among a series of mouse cell lines tested, EL4 and MOPC-31C cells apparently carried the Ubi-L receptors (Table I). Of note, both cell lines are sensitive to Ubi-L in terms of inhibition of proliferation (13). D.10 cells were stimulated with antigen (conalbumin) and ConA, as described previously (25). Ubi-L bound to both stimulated D.10 cells. Together, the expression of Ubi-L receptor should be limited to lymphoid cells. In contrast, Ubi-L did not bind to any human cell lines such as Jurkat, K562, MOLT-3, Namalwa, HL60, U937, Detroit 562 (data not shown), suggestive of the species specificity of Ubi-L action.
Specificity of Ubi-L Binding to Its Receptor-To clarify the specificity of the Ubi-L binding, competitive assay experiments were performed by adding to the experimental system the nonlabeled Ubi-L, MNSF-␤, ubiquitin, and other suppressive cytokines such as IFN-␥ and IL-10. As can be seen in Fig. 3, Ubi-L and MNSF-␤ exclusively inhibited the binding of bio-Ubi-L to ConA-activated D.10 cells. However, the irrelevant ligands (IFN-␥ and IL-10) did not show any competition, indicative of the specificity for Ubi-L binding. Ubiquitin showed a slight but significant inhibition of the Ubi-L binding in agreement with previous finding that ubiquitin inhibits Ubi-L-induced suppression (13).
Affinity Chromatography and Immunoblotting-We next investigated biochemical nature of the receptor protein for Ubi-L. Cross-linking experiments with the use of recombinant Ubi-L were insufficient for identification of Ubi-L receptor protein(s). We observed heterogeneous bands on SDS-PAGE (data not shown), which seemed to be self-aggregation of Ubi-L probably because of its strong hydrophobicity. We also employed 125 I-GST-Ubi-L, which is a stable fusion protein. It should be noted that the activity of GST-Ubi-L is lower than that of Ubi-L. 2 A trace amount of protein (approximately 120 kDa) was reproducibly cross-linked by 125 I-GST-Ubi-L (data not shown). Therefore, we decided to use affinity chromatography on an immobilized Ubi-L column as the main step of Ubi-L receptor purification. D.10 cells were stimulated with ConA, biotinylated, and lysed. Biotinylated membrane proteins (250 g/0.1 ml) were incubated with Ubi-L-Sepharose, and bound proteins were eluted as described under "Experimental Procedures." The eluates were subjected to SDS-PAGE, blotted onto nitrocellulose membranes, and visualized by ECL system. As depicted in Melanoma a Specifically bound Ubi-L was determined as described in the legend to Fig. 1. b Less than 0.05 pmol/10E6 cells. c D.10 cells were stimulated with conalbumin (100 g/ml) for 48 h. d D.10 cells were costimulated with 2 g/ml ConA and 0.5 unit/ml IL-1␤ for 48 h. ( Fig. 1) and the previous observations that antigen-activated, but not unstimulated, D.10 cells are sensitive to Ubi-L (25). Additionally, we made MNSF-␤ and ubiquitin affinity columns to isolate Ubi-L receptor protein(s). The same band of 82 kDa was obtained from MNSF-␤ column, whereas no significant amount of protein band could be recovered from the ubiquitin column (data not shown). To confirm that the 82-kDa protein is a Ubi-L receptor, a ligand blot assay was carried out. Detergent-extracted membrane proteins were prepared from ConAstimulated D.10 cells and blotted onto nitrocellulose membranes. Bio-Ubi-L specifically recognized the 82-kDa band (Fig.  4B, lane 1), suggesting that it should be a ligand for this receptor protein. In contrast, Ubi-L did not bind to any membrane proteins from unstimulated D.10 cells (Fig. 4B, lane 3). Interestingly, ubiquitin slightly bound to the Ubi-L receptor (Fig. 4B, lane 2). We also examined whether the 82-kDa Ubi-L receptor binds to lectins. Ubi-L receptor purified by a Ubi-L column was incubated with WGA-Sepharose. After 6 h, the absorbed Ubi-L receptor was eluted with N-acetyl-D-glucosamine. The eluate was electrophoresed and silver stained. As can be seen in Fig. 4C, the 82-kDa Ubi-L receptor was recovered in the eluate fraction, suggesting that Ubi-L receptor may bind to WGA.
We next performed immunoblotting analysis by the antireceptor Ab. Detergent extract of ConA-activated D.10 cells was resolved by SDS-PAGE and subjected to electrophoretic transfer and immunochemical staining for Ubi-L receptor. As shown in Fig. 5A, a major protein band at 82 kDa and a minor band at 65 kDa under nonreducing conditions were detected by the anti-receptor Ab (lane 2). This finding indicates sequence homology between 82-kDa and 65-kDa proteins. We speculate that a protease, for instance, endoproteinase, has cleaved up to 17 kDa from the intact 82-kDa receptor, which results in the generation of receptor fragment that has lost the ability to bind Ubi-L but not the capacity to interact with the anti-receptor Ab. The migrated positions of these bands were unchanged under reducing conditions (lane 3). The same results were obtained from antigen (conalbumin)-stimulated D.10 cells (not shown). In contrast, the Ubi-L receptor was not seen in detergent extract of unstimulated D.10 cells (lane 4). We next performed deglycosylation analysis of Ubi-L receptor proteins because the 82-kDa protein may be glycosylated (Fig. 4C). The membrane fraction of ConA-activated D.10 cells was treated with neuraminidase and subjected to SDS-PAGE and immunoblotting as described above. Deglycosylation converted 82-kDa and 65-kDa molecules to 74-kDa and 60-kDa molecules, respectively (Fig.  5B). The mobility on a gel was changed further by additional treatment with N-glycanase, albeit a product was very faint (data not shown). These findings indicate that Ubi-L receptor protein may be a glycoprotein.
Immunoprecipitation of Ubi-L Receptor Complex-We next determined whether antiserum prepared against the 82-kDa protein band would immunoprecipitate the cross-linked complex of 125 I-Ubi-L (8 kDa) and its receptor. A band of 90 kDa was obtained (Fig. 6A), although several bands including 98-, 106-, 114-kDa were also observed presumably because of selfaggregation of 125 I-Ubi-L. Control serum did not immunoprecipitate any complexes. To investigate whether proteins other than the 82-kDa polypeptide are associated with Ubi-L receptor complex, extracts of biotinylated D.10 cells were immunoprecipitated with anti-Ubi-L receptor Ab. As can be seen in Fig.  6B, the 105-kDa protein was coimmunoprecipitated by the anti-receptor (82-kDa polypeptide) Ab. We speculate that this 105-kDa protein may be involved in the signal transduction of Ubi-L, although it lacks an ability to bind Ubi-L itself.
Neutralization of Ubi-L Activity by Anti-Ubi-L Receptor Ab-To demonstrate that the 82-kDa protein is involved in Ubi-L-mediated signal transduction, a neutralizing test was carried out by the use of specific Ab (IgG) to this protein. As shown in Table II, this Ab could neutralize the Ubi-L activity, confirming that the 82-kDa protein is bioactive Ubi-L receptor. Interestingly, this Ab abolished the synergism between Ubi-L and IFN-␥. This finding supports the notion that IFN-␥ might enhance the expression of Ubi-L receptor protein on the target cells (27).
Determination of the Amino Acid Sequence of Ubi-L Receptor Protein-We next attempted to analyze the amino acid sequence of the 82-kDa Ubi-L receptor protein (Table III). The NH 2 -terminal amino acid sequence of the receptor protein could not be determined probably because of acetylation. Thus this protein was digested with trypsin, separated by reverse phase HPLC, and tryptic peptides were sequenced. As shown below, four of the five peptide sequences derived from Ubi-L receptor are in alignment with a related sequence found in the open reading frame predicted by the DNA sequence of the cDNA encoding mouse IL-11 receptor (28). The similarity of Ubi-L receptor peptide sequences to sequences in mouse IL-11 receptor as well as the similarity in size of the two proteins suggests that the Ubi-L receptor might be an another closely related protein. It should be noted that tryptic peptide 2 con-tains a WSXWS motif commonly seen in a cytokine receptor family (29). DISCUSSION Several lines of evidence have emerged from our studies indicating that a ubiquitin-like polypeptide specifically binds to target cells such as D.10 cells. We employed D.10 cells throughout the experiments because the response to Ubi-L had been well characterized in our previous study (25). The binding of Ubi-L to its receptor might be specific for the following reasons. (i) Ubi-L bound to antigen-stimulated but not to unstimulated D.10 cells. (ii) The binding of bio-Ubi-L was inhibited by the addition of unlabeled Ubi-L but not by irrelevant proteins. (iii) The binding showed species specificity. (iv) The molecular mass of affinity-purified Ubi-L receptor was the same as that of the protein obtained from ligand blot assay. (v) The antigenicity of these proteins was much the same. Finally (vi) Ubi-L activity was neutralized by anti-Ubi-L receptor Ab. On SDS-PAGE under nonreducing conditions, the apparent molecular weight of the Ubi-L receptor was 82,000. This apparent size was unchanged by the disulfide reducing agent dithiothreitol, indicating a lack of disulfide bonds to other proteins. This 82-kDa protein may be associated with the 105-kDa protein lacking an ability to bind Ubi-L (Fig. 6B). Characterization of this 105-kDa protein is under way to clarify its function. Scatchard analysis of Ubi-L receptor was unsuccessful because of an aggregative property of recombinant Ubi-L. Of note, Ubi-L has a secondary structure containing the tandem largely hydrophobic structural units. This likely accounts for our consistently poor yield of Ubi-L after cleavage from the fusion partner GST. It is difficult to handle a small amount of Ubi-L even in the presence of a carrier protein such as bovine serum albumin. In contrast, ubiquitin has a stable secondary structure containing only a low percentage of ␣-helix or ␤-sheet (30).
It has been claimed that suppressor factors consist of effector and accessory molecules (31). Like antigen-specific suppressors, hybridoma-derived (native) 70-kDa MNSF consists of 8-kDa Ubi-L (MNSF-␤) and 62-kDa MNSF-␣, a polypeptide serologically related to TCR-␣ chain (17). We speculate that the Ubi-L may possibly be a candidate for an effector molecule. Support for this theory is the fact that the Ubi-L itself bound directly to the target cells. On the other hand, MNSF-␣ is necessary for the extracellular release and stability of Ubi-L. Whether MNSF-␣ might contribute to the Ubi-L binding to its target cells remains to be determined. To date, isolation of cDNA encoding MNSF-␣ has been unsuccessful because of a transient and faint expression of the mRNA.
It is evident that several cell surface receptor proteins are ubiquitinated. For instance, platelet-derived growth factor receptor is a ubiquitin acceptor (8,9). The intracellular domain of   frame of cloned murine IL-11 receptor All alignments are between sequences of tryptic peptides from Ubi-L receptor with internal sequences of the cloned murine gene. Ubi-L receptor peptide fragments were generated by trypsin treatment as described under "Experimental Procedures." The WSXWS motif is shown in bold.

1.
Ubi-L R this receptor is ubiquitinated, which leads to intracellular signaling. Similarly, cytosolic regions of the TNF-␣ receptor (32), -subunit of the TCR-CD3 complex (6), and high affinity IgE receptor (7) are ubiquitinated. Interestingly, Ab specific for ubiquitin detected antigenic determinants on the cells expressing lymphocyte horming receptor, indicating that ubiquitin covalently binds to the extracellular domain of this receptor (33,34). Although Ubi-L also recognizes cell surface receptor proteins, it seems unlikely that it binds covalently to the receptors in a manner similar to ubiquitination because MNSF-␤, which lacks the free COOH-terminal glycyl doublet responsible for isopeptide bond formation, bound to target cells (Fig. 1).
Ubiquitin is not the only molecular tag for protein modification. Sentrin (18) and UCRP (19) have been shown to be conjugated to other proteins in a process analogous to ubiquitination. The COOH termini of UCRP and sentrin are processed efficiently, which allows for subsequent protein conjugation. Similarly, Ubi-L is cleaved from a fusion partner, ribosomal protein S30, in cytosol. Like UCRP, Ubi-L conjugates to intracellular proteins in vitro (15) and in vivo (16). Ubi-L conjugates to intracellular acceptor proteins including MNSF-␣. The Ubi-L conjugation is similar but not identical to the ubiquitination process because acceptor proteins for Ubi-L are different from those for ubiquitin (15). The Ubi-L conjugation is thought to occur via isopeptide bond formation because isopeptidase prepared from murine livers dissociates MNSF-␣ from Ubi-L (17). In this context, we speculate that the intracellular and extracellular mode of actions of Ubi-L may differ.
A number of biotinylated cytokines have been employed for receptor analysis instead of iodination (35)(36)(37). Bio-Ubi-L shows biological activity similar to that of unlabeled Ubi-L. 3 Nevertheless, cross-linking experiments with the use of biotinylated or iodinated Ubi-L were unsuccessful because of the nature of self-aggregation, as described above. Therefore, affinity chromatography on an immobilized Ubi-L column was used for the identification of Ubi-L receptor protein. This Ubi-L column allowed us to purify the Ubi-L receptor and prepare antiserum against it. We are attempting to investigate whether or not soluble Ubi-L receptor may affect the immune responses in vitro and in vivo and isolate a cDNA encoding this receptor protein. Recombinant receptor may pave the way for further clarification of the mechanism of action of ubiquitin-like polypeptide(s).
Peptide sequence analysis suggests that this 82-kDa protein should be a member of a cytokine receptor family. Ubi-L receptor peptide sequences are similar to those in IL-11 receptor (Table III), although the mode of action of Ubi-L is different from that of IL-11. Most recently we observed that Ubi-L receptor ligation rapidly induces protein tyrosine phosphorylation of distinct proteins in D.10 cells. 2 We have reported previously that IFN-␥ is involved in the mode of action of Ubi-L (14). However, neither IFN-␥ nor IL-10, an immunosuppressive cytokine, shared the Ubi-L receptor (Fig. 3). Interestingly, ubiquitin slightly but significantly bound to the receptor probably because of its homology with Ubi-L. Indeed, we have observed recently that high dose ubiquitin (100 M) showed a Ubi-L-like activity in vitro and that low dose (10 nM) ubiquitin, which is insufficient for the suppression of the antibody response, could inhibit the Ubi-L-suppressive activity (13). In the current study, we showed evidence that the inhibition is caused, at least in part, by a competitive action at receptor level. Most recently, we observed that Ubi-L exerts its suppressor activity in vivo (38). High dose administration of ubiquitin (5 g/body) showed a mimetic activity. Pancré et al. (39) mentioned that ubiquitin shows an inhibitory effect on IgE-induced platelet cytotoxicity. Additionally, ubiquitin is thought to be involved in the mechanism of hypersensitivity to Hymenoptera venom and aspirin-sensitive asthma (40). Taken together, ubiquitin-like protein(s) and ubiquitin may play an important role in keeping the immune system in balance.