Conjugation of ubiquitin-like polypeptide to intracellular acceptor proteins

Ž .Biochimica et Biophysica Acta 1401 1998 319–328

Conjugation of ubiquitin-like polypeptide to intracellular acceptorproteins

Tomoko Nagata a,), Morihiko Nakamura b, Hideyuki Kawauchi a, Yoshinori Tanigawa b

a Department of Otolaryngology, Shimane Medical UniÕersity, Izumo 693, Japanb Department of Biochemistry, Shimane Medical UniÕersity, Izumo 693, Japan

Received 22 July 1997; accepted 7 October 1997

Abstract

Ž .Monoclonal nonspecific suppressor factor MNSF , a lymphokine produced by a murine hybridoma, was originally foundto inhibit the generation of LPS-induced immunoglobulin secreting cells. MNSF comprises of MNSFb, an isoform of

Ž .MNSF, and the other isoform, MNSFa . Ubiquitin-like segment Ubi-L of MNSFb shows MNSF-like activity. Ubi-LŽ . 1257.8 kDa has 36% homology with 8.5 kDa ubiquitin. GST–Ubi-L was labeled with I by the chloramine T method andtested for its conjugation to acceptor proteins in splenocyte lysates. 125I–GST–Ubi-L conjugation on SDS–PAGE showedheterogeneous bands including 95 kDa GST–Ubi-L conjugation in the splenocyte, but not reticulocyte lysates. The Ubi-L

Ž .adduct appeared to be MNSF-related molecule because anti-MNSF monoclonal antibody mAb recognized the 95 kDaband. The pattern of the conjugations was different from that seen in ubiquitination. Unlabeled GST–Ubi-L inhibited theconjugations, while ubiquitin did not. a-Lactalbumin, one of the target proteins for ubiquitination, failed to conjugate toGST–Ubi-L. In addition, covalent conjugation of ubiquitin to reticulocyte lysates was also interfered by GST–Ubi-L. Theseresults suggest that Ubi-L may conjugate to acceptor proteins in a similar, but not in the same way as ubiquitination, andmight play an important role in lymphoid cells. q 1998 Elsevier Science B.V.

Keywords: Ubiquitin-like protein; Ubiquitin; Conjugation

1. Introduction

Ubiquitin is one of the most highly conservedŽproteins in all eukaryotic cells, which is small M sr

.8500 76-amino acid protein, plays an important rolein the degradation of abnormal, misfolded or improp-

w xerly assembled proteins 1,2 . The process is accom-plished through a unique post-translational modifica-

Abbreviations: MNSF: monoclonal nonspecific suppressorfactor; rUbi-L: recombinant ubiquitin-like protein

) Corresponding author. Fax: q81 853 20 2125; E-mail:[email protected]

tion in which the carboxyl terminus Gly–Gly iscovalently ligated to primary amines on target intra-

w xcellular proteins 3 . A major consequence of ubiqui-tin conjugation is the targeting of proteins for degra-

w xdation via multi-enzymes 4–6 , ubiquitin activatingŽ . Ž .enzyme E1 , ubiquitin conjugating enzyme E2 ,

Ž .protein ligase E3 , by ATP-dependent pathway re-w xquiring the 26S protease complex 7 . The carboxyl

terminus Gly–Gly is also conserved in ubiquitinŽ .cross-reactive protein UCRP which is induced by

w xinterferons 8,9 . UCRP conjugates to cellular pro-w xteins within interferon-responsive cells 10,11 . As

recombinant human UCRP is not inhibited its conju-

0167-4889r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0167-4889 97 00131-6

( )T. Nagata et al.rBiochimica et Biophysica Acta 1401 1998 319–328320

gation by ubiquitin, the mode of action of UCRPconjugation is apparently distinct to that of ubiquitinw x12 .

Although several soluble factors with antigen-non-specific suppressive activity in the immune responseshave been reported, the genes that encode them havenot been clearly defined. Some factors are known tobe associated to another protein. The monoclonal

Ž .nonspecific suppressor factor MNSF , a product ofŽ .concanavalin A Con A stimulated murine T cell

hybridoma, inhibits the generation of LPS-inducedimmunoglobulin secreting cells, proliferation of mito-gen activated T and B cells and division in various

w xtumor cell lines of murine origin 13–16 . Like UCRP,interferon gamma is involved in the mechanism of

w xaction of MNSF 14 .Recently we have isolated and cloned the sequence

w xof MNSFb 17 , an isoform of MNSF, and the re-Ž .combinant form of the ubiquitin-like segment rUbi-L

w xof MNSFb has an MNSF-like activity 15,18 . rUbi-L,a 8.5 kDa ubiquitin like polypeptide, has 36% homol-ogy to ubiquitin and possesses the carboxyl terminal

w xGly–Gly doublet 17 . Since Ubi-L may associate tow xthe other subunit before extracellular release 17 , we

studied whether Ubi-L conjugates to some targetŽ .protein s in murine splenocyte lysate, and compared

the mechanism of Ubi-L conjugation with ubiquitina-tion process.

2. Materials and methods

2.1. Animals

Female BALBrc mice, 8 to 16 weeks of age, wereused throughout this study and were obtained from

Ž .Clea Japan Osaka . The Japanese White rabbits andthe mice were reared in the Laboratory of Institutefor Experimental Animals, Shimane Medical Univer-sity.

2.2. Materials

Bovine ubiquitin, bovine a-lactalbumin, bovineserum albumin, aprotinin, leupeptin, prestainedmolecular weight standard mixture were purchased

Ž .from Sigma St. Louis, MO . PhenylmethanesulfonylŽ . X Žfluoride PMSF , EGTA, EDTA, 5,5 -dithiobis 2-

. Ž .nitrobenzoic acid DTNB were purchased fromŽ .Wako Pure Chemical Industries, Dithiothreitol DTT

was purchased from Nakalai Tesque, Kyoto, Japan.ŽMolecular standard marker Low molecular weight

.calibration kit for electrophoresis was obtained fromŽ .Pharmacia Biotech. DEAE-cellulose DE52 was from

Whatman.

2.3. Purification of recombinant GST–Ubi-L and ra-dio-iodination of the proteins

Ž .Murine recombinant Ubi-L rGST–Ubi-L was ex-pressed in Escherichia coli and purified, as described

w xpreviously 18 except for the cleavage with throm-bin. Briefly, rUbi-L was expressed as a fusion proteinwith glutathione-S-transferase using the pGEX-2T

Ž .vector Pharmacia, Uppsala, Sweden , extracted andŽ .purified by Glutathione Sepharose 4B Pharmacia .

rGST–Ubi-L was applied to an anion exchange col-Ž .umn DEAE 5PWrTSK gel equilibrated in 50 mM

Tris-HCl, pH 7.5, and eluted with an increasingconcentration of NaCl. rGST–Ubi-L was eluted at300 mM NaCl and its purity was more than 95%.

The purified GST–Ubi-L, ubiquitin, bovine a-lactalbumin were labeled with 125I by the chloramine

w xT method, as described previously 19 .

2.4. Antibodies

Specific antibody against synthetic peptides corre-Ž .sponding to the ubiquitin-like region PU-1 was

raised in rabbits, and the anti-PU1 Ab was coupled toŽ .Sepharose 4B Pharmacia as described previously

w x Ž .17 . An anti-MNSF monoclonal antibody MO6 wasw xpurified as described previously 14 .

2.5. Preparation of murine and rabbit reticulocytelysates

Reticulocyte rich blood was obtained by phenylhy-drazine induction of adult female rabbits and micew x20 . Reticulocytosis was more than 85%. To depleteendogenous ATP reticulocyte rich blood was washed

Žtwice with cold phosphate-buffered saline PBS;.150 mM NaCl, 10 mM sodium phosphate, pH 7.4

and incubated at 378C for 90 min with the addition of20 mM 2-deoxyglucose and 0.2 mM 2,4-dinitrophenolw x21 . Subsequently, the cells were washed with PBS,

( )T. Nagata et al.rBiochimica et Biophysica Acta 1401 1998 319–328 321

lysed in a solution A; containing 1 mM DTT, 1 mMEDTA, 1 mM EGTA, 1 mM PMSF, and 1 mM leu-peptin. The particulate fraction was removed by cen-

Ž .trifugation 40 000=g, 60 min , and the supernatantwas harvested, and stored at y808C until use.

2.6. Preparation of murine splenocyte lysates

Single cell suspension of spleen was prepared,washed twice with cold PBS and lysed in solution A.Particulate fraction was removed by centrifugationŽ .40 000=g, 60 min . In some experiments, spleno-

Ž 7 .cytes 1.2=10 cellsrml were incubated with ConŽ .A 3 mgrml for 48 h. The activated cells were lysed

in the solution A.

2.7. Preparation of Ubi-L depleted lysates

To deplete endogenous Ubi-L, cell lysates in PBSwere subjected to anti-PU-1 antibody coupled toSepharose 4B and incubated at 48C for overnight, andthe unbound fraction was harvested.

2.8. Fractionation of cell lysates

Ž .Fraction 1 proteins unbound to DE52 and frac-Ž .tion 2 bound to DE52 and eluted with 0.5 M KCl

were prepared from rabbit and murine reticulocytes,w xand murine splenocyte lysate 21 . Protein concentra-

tions in the lysates were determined by the BradfoldŽ .assay Bio-Rad .

All the lysates including both fractions were dia-Žlyzed twice against reaction buffer 50 mM Tris-HCl,.pH 7.5, 0.2 mM DTT, 1 mM PMSF by Centricon-10

Ž .Amicon , and protein concentrations were preparedwithin 5–10 mgrml in splenocyte lysates, 20–30 mgrml for reticulocyte lysates, respectively. Allthe lysates could be stored at y808C for over a yearwithout loss of activity.

2.9. Construction and incubation of the conjugationreaction mixture

Ž .The standard incubation mixture 20 ml contained50 mM Tris-HCl, pH 7.5, 0.2 mM DTT, 1 mM PMSF,20–30 mg of lysate proteins, 0.5 mgrml carrierbovine serum albumin and approximately 40 000 cpm

125 Ž 6 . w xof I-labeled protein 10 cpmrmg 19 . Reaction

mixtures were incubated at 378C for indicated peri-ods, subsequently quenched by the addition of 2=

Ž .SDS sample buffer Laemmli containing b-mer-captoethanol, and dissolved by 10% or 15% SDSpolyacrylamide gel. Detection of 125I-labeled GST–Ubi-L protein conjugates was determined by theSDS–PAGE after 16 h incubation. The conjugationof ubiquitin to reticulocyte lysates, a-lactalbumin,and ubiquitin was determined after 15 min incubation.Conjugated proteins were visualized by autoradiog-raphy of the dried gel. The region of the gel contain-ing the conjugates was excised and estimated by

Ž .Bio-Imaging analyzer BAS-2000II Fuji Photo Film,Ž .and Ultroscan XL enhanced laser densitometer LKB .

2.10. Immunoblotting

Ž .The reaction mixture 60 ml contained 50 mMTris-HCl, pH 7.5, 0.2 mM DTT, 1 mM PMSF, 100 mgof unstimulated splenocyte lysates and approximately2 pmoles of unlabeled GST–Ubi-L, after 6 h incuba-tion at 378C and the resultants were resolved in 10%polyacrylamide gels under reducing conditions. Thegels were transferred to enhanced chemiluminescenceŽ .ECL Hybond membranes, following whichimmunostaining was performed. MO6, a murine

w xmonoclonal antibody to native MNSF 14 , was used

Fig. 1. Conjugation of 125I–GST–Ubi-L to acceptor proteins inŽ .splenocyte lysates. The reaction mixtures 20 ml including vari-

ous cell lysates, 50 mM Tris-HCl, pH 7.5, 0.2 mM DTT, 1 mMPMSF and 1 pmole of 125I–GST–Ubi-L were incubated for 16 hat 378C and then quenched with SDS sample buffer. Lane 1: inthe absence of cell lysates; lane 2: containing 30 mg of crudesplenic cell lysates; lane 3: 30 mg of murine spleen f1; lane 4:30 mg of f2; lane 5: 30 mg each of f1 and f2; lane 6, 60 mg ofcrude lysates. Lane 1: for the control experiments, 125I–GST–Ubi-L was added after quenching with SDS sample buffer.


Fig. 2. Time course and dose dependence of 125I–GST–Ubi-Lconjugation, Panel A: Indicated amounts of splenic cell lysateswere incubated at 378C for 16 h with 125I–GST–Ubi-L in a 20 mlvolume of reaction mixture and quenched by SDS sample buffer.Lane 1: no addition of lysates; lane 2: 5mg; lane 3: 12.5mg;lane4: 25mg; lane 5: 50 mg; lane 6: 75mg; lane 7: 90 mg. PanelB: 25mg of cell extracts and 125I–GST–Ubi-L were incubated at378C for indicated period and then quenched with SDS samplebuffer. Lane 1: no addition of extracts; lane 2: 0 min; lane 3:30 min; lane 4: 90 min; lane 5: 4 h; lane 6: 16 h.

for a first antibody and biotinylated anti-mouse IgGantibody for a second antibody. Peroxidase-con-jugated streptavidin was used for the last step. Detec-tion was done according to the ECL systemŽ .Amersham .

3. Results

3.1. Conjugation of 125I-GST–Ubi-L to intracellularproteins

We speculated that Ubi-L conjugates to some pro-teins in splenic cells, because a specific Ab to Ubi-Limmunoprecipitates several proteins from mouse

w xsplenocyte lysates 17 . This notion is supported by

the fact that the amino acid sequences consideredimportant for the ubiquitination process are con-served in Ubi-L, including the carboxyl terminal

Ž .Gly–Gly doublet. Recombinant Ubi-L rUbi-L tendsto form self-aggregation due to its strong hydropho-

w xbicity 17 . Thus, in this study, we employed GST–Ubi-L to investigate whether or not Ubi-L actuallyconjugates to intracellular proteins. This fusion pro-tein is stable in terms of aggregative and adsorptivecharacteristics, as compared with rUbi-L. It should benoted that GST–ubiquitin substitutes functionally for

w xauthentic ubiquitin in ligation to substrates 22 .GST–Ubi-L was radiolabeled by the method of

chloramine T as described under Section 2. It ispossible that the tyrosine residues of GST were radio-labeled with 125I, because Ubi-L lacks the aromatic

w x 125 Ž .amino acid 17 . I-GST–Ubi-L 1 pmole was incu-Ž .bated with murine splenocyte lysates 30 mgr20 ml

for 16 h at 378C. Although a strong conjugate bandwas developed within 4 h, longer incubation wasperformed to accomplish the conjugation. Subse-quently, the reaction mixtures were resolved in 10%polyacrylamide gels. Heterogeneous bands were ob-

Ž . 125served Fig. 1, lanes 2 and 6 , suggesting that I-GST–Ubi-L conjugates to some intracellular proteinsin the lysates. 125I-GST did not conjugate to any

Ž .proteins in the splenocyte lysates data not shown .Ubiquitin conjugates to murine reticulocyte and

Fig. 3. Inhibition of Ubi-L conjugation. 125I–GST–Ubi-L wasincubated with splenic cell lysates which had been depleted ofendogenous Ubi-L as described under Section 2. Lane 1: in theabsence of Ubi-L depleted lysates; lane 2: in the presence of the

Ž .lysates 25mg ; lane 3: the same to lane 2 but a 100 fold excessamount of unlabeled GST–Ubi-L was added; lane 4: the same tolane 2 but a 100 fold amount of ubiquitin added.


Žsplenocyte lysates in the presence of 2 mM ATP data.not shown . In contrast with ubiquitination, Ubi-L

conjugations occurred in the absence of added ATP.The pattern of the conjugation was same in thepresence of 2 mM ATP. Interestingly, 125I-GST–Ubi-L did not associate with any proteins in murinereticulocyte lysates, presumably due to a lack oftarget proteins andror enzymes involving in the Ubi-

L conjugation. To determine the optimal conditionfor Ubi-L conjugation, 125I-GST–Ubi-L was incu-

Ž Ž ..bated with several amounts of lysates Fig. 2 A forŽ Ž ..various periods Fig. 2 B . The concentration of

25mgr20 ml resulted in a maximal conjugation andwas therefore used in the following experiments. Wealso observed that conjugation was completed after16 h.

Fig. 4. The influence of DTT and DTNB upon Ubi-L conjugation. Panel A: 125I–GST–Ubi-L was incubated with splenocyte lysatesŽ .25mg in the presence of various amounts of DTT. Lane 1: in the absence of lysates and DTT; lane 2: in the absence of DTT; lane 3:0.2 mM DTT; lane 4: 2 mM; lane 5: 5 mM; lane 6: 10 mM; 7, 15 mM; 8, 20 mM. Relative intensity of 95 kDa band estimated by theimaging analyzer was also presented. Panel B: 125I–GST-Ubi-L was incubated with splenic cell lysates in the presence of DTNB. Lane 1:in the absence of lysates; lane 2: 25mg of lysates; lane 3: 25mg lysates and 0.5 mM DTNB. NS indicates nonspecific band.


Fig. 5. Ubi-L conjugation in Con A stimulated splenocyte lysates.125I–GST–Ubi-L was incubated with unstimulated or Con AŽ .48 h - stimulated splenocyte lysates. Lane: 1, in the absence ofcell lysates; lane 2: 20 mg of unstimulated cell lysates; lane 3:30 mg of the cell lysates; lane 4: 30 mg of Con A-stimulatedsplenocyte lysates.

The reaction proceeded very slowly and the amountof Ubi-L conjugations was lesser than that of ubiqui-tination. The splenocyte lysates were fractionated by

Ž .anion exchange chromatography DE52 into fractionŽ . Ž1 f1, flow through and fraction 2 f2, 0.5 M KCl

. w xeluate as described previously 21 . As shown in Fig.1, the pattern of Ubi-L conjugation to acceptor pro-teins in f1 was different from that seen in f2. It hasbeen shown that 125I-ubiquitin does not conjugate toany proteins in f1 due to the lack of enzymes respon-sible for ubiquitination. On the contrary, 125I-GST–Ubi-L associated with several proteins in both thefractions. The conjugations in f1 showed compara-tively low molecular mass, whereas those in f2 in-cluded a strong band migrated at the position ofapproximately 95 kDa. The pattern of the conjuga-tions in crude splenocyte lysates was similar to thatseen in the fraction reconstructed with f1 and f2Ž .compare lanes 2 with 5 . Thus, it seems likely thatthe acceptor proteins for Ubi-L exist in both f1 andf2.

To demonstrate the specificity of these conjuga-tions, we performed inhibition tests. As shown in Fig.3, 125I-GST–Ubi-L conjugations were partly inhibitedby the addition of 100-fold unlabeled GST–Ubi-Lwhen Ubi-L depleted-lysates were employed as de-scribed in Section 2. In contrast, ubiquitin did not

Ž .affect the GST–Ubi-L conjugations Fig. 3, lane 4 .When crude lysates were used the GST–Ubi-L didnot affect at all presumably due to the existence ofendogenous Ubi-L in the lysates.

3.2. The effect of thiol modifying agent on GST–Ubi-Lconjugation

Since the activation of ubiquitin is constituted byw xE1-ubiquitin thiol formation 23 , we investigated the

effect of thiol modifying agent on Ubi-L conjuga-tions. As expected, DTT, a disulfide reducing reagent,abrogated the conjugations in a dose dependent man-

Ž Ž .. Xner Fig. 4 A . We next investigated whether 5,5 -di-Ž . Ž .thiobis 2-nitrobenzoic acid DTNB , a thiol modify-

ing reagent, affects the Ubi-L conjugations. DTNBremarkably inhibited the conjugations including the

Ž Ž . .95 kDa band Fig. 4 B , lane 3 , suggestive of theinvolvement of thiol linkage in the Ubi-L conjuga-tion. However, the Ubi-L adducts do not containdisulfide bond, because these bands are stable forboiling in the presence of 2b-mercaptoethanol.

3.3. GST–Ubi-L conjugation in mitogen actiÕatedcell lysates

We also examined the Ubi-L conjugation in Con AŽ .activated 48 h splenocyte lysates. Interestingly, the

strong band of 95 kDa Ubi-L adduct was decreasedŽ . Ž .approximately 80% Fig. 5, lane 4 . Calculativemolecular mass of GST–MNSF is approximately95 kDa, because the masses of GST and MNSF are25 kDa and 70 kDa, respectively. To determine

Fig. 6. Immunoblot analysis of 95 kDa molecule. In vitro conju-gation and immunoblot analysis were performed as described inSection 2. Lane 1: hybridoma derived native 70 kDa MNSF; lane2: GST–Ubi-L alone; lane 3: unstimulated splenic cell lysates;lane 4: GST–Ubi-L incubated with splenic cell lysates.


Fig. 7. Inhibition of ubiquitination by Ubi-L. 125I-ubiquitin wasincubated with rabbit reticulocyte lysates for 15 min at 378C.Lane 1: in the absence of lysates; lane 2: 20 mg; lane 3: 20 mg oflysates and unlabeled 100-fold amount of ubiquitin; lane 4: 20 mgof lysates and unlabeled 100-fold amount of GST–Ubi-L.

whether the 95 kDa band is MNSF-related molecule,Ž .we employed a monoclonal antibody MO6 to native

w xMNSF. MO6 does not recognize Ubi-L 17 . Im-munoblot experiments showed that MO6 did detect

Ž .the 95 kDa band Fig. 6, lane 4 , clearly demonstrat-ing that the band is GST–MNSF in term of themolecular mass and the antigenicity. There was, how-ever, no detectable band in the lysates incubated

Ž .without GST–Ubi-L Fig. 6, lane 3 . One may specu-late that the conformational change of an unidentified

w x ŽMNSF subunit 17 the conjugation of the subunit to

Fig. 8. 125I-a-lactalbumin conjugates to ubiquitin but not toGST–Ubi-L. 125I-a-lactalbumin was incubated with Fraction 2Ž .F2 from Ubi-L depleted-murine reticulocyte lysates and ubiqui-

Ž . Ž .tin lane 3 or GST–Ubi-L lane 4 for 15 min at 378C. Lane 1:125I-a-lactalbumin alone; lane 2: 125I-a-lactalbumin plus thereticulocyte lysates F2. Bracket indicates the conjugated adducts.

.Ubi-L should be necessary to generate the anti-genicity.

3.4. GST–Ubi-L affects ubiquitination

We prepared mouse reticulocytes by phenylhydra-zine induction, since the mode of action of Ubi-L

w x 125shows species-specificity 15 . I-GST–Ubi-L didnot conjugate to any proteins in murine reticulocytelysates, indicative of the specificity of the Ubi-Lconjugation in splenocyte lysates. Ubiquitination torabbit reticulocyte lysates was inhibited by the addi-

Ž .tion of the unlabeled GST–Ubi-L Fig. 7, lane 3 , butŽ .not by GST data not shown . Ubiquitin conjugated

to 125I-a-lactalbumin in Fraction 2 from murine Ubi-LŽ .depleted-reticulocyte lysates Fig. 8, lane 3 , whereas

Ž .GST–Ubi-L did not Fig. 8, lane 4 , suggesting thatthe acceptor proteins for Ubi-L may differ from thosefor ubiquitin. Collectively, the mechanism of Ubi-Ladduct formation might be distinct from ubiquitina-

Ž .tion, although a common pathway s may exist.

4. Discussion

In the current study, we used splenocyte lysates forUbi-L conjugation experiments, because 70 kDaMNSF, a Ubi-L adduct, has been observed in the cell

w xlysates 17 . Ubi-L conjugation to intracellular pro-Ž .teins was specific and reproducible, because; i cold

GST–Ubi-L remarkably inhibited the conjugation;Ž .ii GST itself did not conjugate to any proteins in the

Ž . 125lysates; iii I-GST–Ubi-L did not conjugate to anyŽ .proteins in mouse reticulocyte lysates; iv the Ubi-L

Ž .conjugation occurred dose- and time-dependently; vthe conjugation could not be seen when damaged

Ž .lysates were used not shown .Ubiquitin is highly conserved in eukaryotic cells,

w xwhereas Ubi-L shows species specificity 15 . Ubi-LŽpresents more than 50% homology with ubiquitin if

.conservative amino acids are considered , and pos-sesses Gly–Gly doublet essential for the ubiquitina-tion. Thus, we speculate that the mechanism of Ubi-Lconjugation may be similar but not identical to ubiq-uitination process. This notion may be supported by

Žthe fact that GST–Ubi-L inhibited ubiquitination Fig.


.7 , although the pattern of Ubi-L adduct formationwas quite different from that of ubiquitination.

Covalent conjugation of ubiquitin to intracellularprotein is a signal for degradation by the 26S pro-

w xtease 24,25 . Sequential action of ubiquitin conjugat-ing enzyme forms a covalent thiol ester with ubiqui-tin as part of its catalytic cycle. Ubiquitination typi-cally occurs through an intermolecular enzyme-

w xubiquitin thiol ester cascade 24,26 . At least, theactivation of ubiquitin requires ATP and then thesequential conjugation proceeds through, whileGST–Ubi-L conjugation accomplishes in the absenceof added ATP. A similar phenomenon has beenreported for extracts from green plant tissues and

w xprokaryote 27–29 . However, it is possible that cellextracts might include a small amount of ATP. It hasbeen reported that many E1-like enzymes have verylow Km’s for ATP. Thus, further investigations are

Ž .underway in our laboratory to identify the enzyme sinvolved in the Ubi-L conjugations.

In spite of the sequence homology, Ubi-L does notŽ 29 48have some important residues Lys , Lys and

63.Lys in ubiquitin which are critical amino acids forw x 48multi-ubiquitin chain formation 30–32 . The Lys

of ubiquitin is responsible for ATP dependent degra-dation and Arg74 represents a major contribution inthe initial binding of the polypeptide to E1. Theseamino acids are replaced with Ser 46 and Leu72, re-spectively, in the Ubi-L sequence. It should be notedthat Ubi-L lacks the amino acids, Thr7 and Leu8, in

w xubiquitin 17 . Thus, it is possible that Ubi-L may befree from the function of protein degradation. Indeed,ubiquitin conjugated to a-lactalbumin, while Ubi-L

Ž Ž ..did not Fig. 8 A . Although the ubiquitination wasinhibited by the addition of unlabeled GST–Ubi-LŽ Ž ..Figs. 7 and 8 B , the GST–Ubi-L conjugation was

Ž .not inhibited by excess ubiquitin Fig. 3 .The amount of the 95 kDa Ubi-L adduct in Con

A-stimulated murine splenocyte lysates was drasti-cally decreased. We have observed the strong expres-

Ž .sion of mRNA for Ubi-L in Con A stimulated 24 hw xmurine splenocytes 17 . We also described several

Ž .Ubi-L conjugations in Con A stimulated 48 hsplenocytes with the use of a specific antibody toUbi-L. The strong 70 kDa band was observed in bothcultured cell lysates and conditioned media. Thus, weconceive that the low level of Ubi-L conjugationwithin the Con A-activated cells may be due to the

result of the extracellular release of 70 kDa MNSF,the Ubi-L adduct. Another possibility is that theincrease of the intracellular Ubi-L might competewith 125I-GST–Ubi-L, because notable augmentationof Ubi-L mRNA is evident in Con A-activatedsplenocytes, as described above.

We have reported that the products of MNSFb

cDNA may be cleaved in the cytoplasm, as has beendemonstrated for the rat Fau protein in reticulocyte

w xlysates 33 , and only the ubiquitin-like segmentŽ . Ž .7.8 kDa associated with other protein s is released

Žas a complex of higher molecular mass 70 kDa. w xMNSF 17 . In the immunoprecipitation experi-

ments, several bands including the 70 kDa MNSFwere observed in mouse splenocyte lysates. Judgingfrom the molecular mass, the 95 kDa protein ap-peared to be GST–MNSF. We confirmed that the

Ž .95 kDa band is GST–MNSF Fig. 6 , with the use ofŽ .a monoclonal antibody to MNSF MO6 , which does

not react with Ubi-L.The Ubi-L conjugation was affected by thiol modi-

fying reagents DTNB and DTT. Ubi-L possesses onecysteine residue at the position of 57. These thiolmodifying reagents also affect the activity of glyco-

Ž . w xsylation inhibitory factor GIF 34 , a cytokine withimmunosuppressive function. Modification of inac-

Ž .tive human recombinant GIF hrGIF containing 3cysteine residues with DTNB results in active form.GIF is ubiquitously expressed and lacks a signal

w xpeptide as well as MNSF 34 . The activity of thisbioactive hrGIF and native GIF is lost by the treat-ment with DTT. In contrast, the Ubi-L conjugation isinhibited by the addition of DTT and DTNB.

Ž .Ubiquitin cross-reactive protein UCRP is also aubiquitin-like protein. Interestingly, this cytokine can

w xbe induced by IFNs as well as MNSF 9,10 . Viralinfection of mammalian cells also induces the expres-sion of UCRP, as a part of the interferon responsew x9,35 . Most recently, it has been shown that anotherubiquitin-like protein, sentrin, is a nuclear protein andconjugates to other proteins in a process analogous to

w xubiquitination 36,37 . Therefore, one may speculatethat ubiquitin-like proteins may be closely implicatedin the immune responses and that their mode ofaction would be similar to ubiquitination. Furtherinvestigations of Ubi-L conjugation are currently inprogress to clarify the underlying mechanism of ac-tion of MNSF.


Acknowledgements

The contents of this report serve as the basis of thedissertation of T. Nagata for the degree of Doctor ofMedical Science.

References

w x1 J.D. Etlinger, A.L. Goldberg, A soluble ATP-dependentproteolytic responsible for the degradation of abnormal pro-

Ž .teins in reticulocytes, Proc. Natl. Acad. Sci. USA 74 197754–58.

w x2 A. Hershko, H. Heller, D. Ganoth, A. Ciechanover, in: H.L.Ž .Segal, D.S. Doyle Eds. , Protein Turn over and Lysosome

Function, Academic Press Inc., New York, 1978, pp. 149–169.

w x3 A. Hershko, A. Ciechanover, H. Heller, A.L. Haas, I.A.Rose, Proposed role of ATP in protein breakdown: Conjuga-tion of proteins with multiple chains of the polypeptide ofATP-dependent proteolysis, Proc. Natl. Acad. Sci. USA 77Ž .1980 1783–1786.

w x4 A. Hershko, A. Ciechanover, The ubiquitin system for pro-Ž .tein degradation, Annu. Rev. Biochem. 61 1992 761–807.

w x5 A. Hershko, H. Heller, E. Eytan, G. Kaklij, I.A. Rose, Roleof the a-amino group of protein in ubiquitin-mediated pro-

Ž .tein breakdown, Proc. Natl. Acad. Sci. USA 81 19847021–7025.

w x6 A. Hershko, H. Heller, S. Elias, A. Ciechanover, Compo-nents of ubiquitin-protein ligase system, J. Biol. Chem. 258Ž .1983 8206–8214.

w x7 M. Rechsteiner, L. Hoffman, W. Dubiel, The multicatalyticŽ .and 26S proteases, J. Biol. Chem. 268 1993 6065–6068.

w x8 P.J. Farrell, R.J. Broeze, P. Lengyel, Accumulation of anmRNA and protein in interferon- treated Ehrlich ascites

Ž . Ž .tumor cells, Nature London 279 1979 523–525.w x9 B.D. Korant, D.C. Blomstrom, G.J. Jonak, E. Knight Jr.,

Ž .Interferon induced proteins, J Biol. Chem. 259 198414835–14839.

w x10 K.R. Loeb, A.L. Haas, The interferon-inducible 15 kilodal-ton ubiquitin homolog conjugates to intracellular proteins, J.

Ž .Biol. Chem. 267 1992 7806–7813.w x11 K.R. Loeb, A.L. Haas, Conjugates of ubiquitin cross-reac-

tive protein distribute in a cytoskeletal pattern, Mol. Cell.Ž .Biol. 14 1994 8408–8419.

w x12 J. Narasimhan, J.L. Potter, A.L. Haas, Conjugation of the 15kilodalton interferon-induced ubiquitin homologue distinct

Ž .from that of ubiquitin, J. Biol. Chem. 271 1996 324–330.w x13 M. Nakamura, H. Ogawa, T. Tsunematsu, Isolation and

characterization of a monoclonal nonspecific suppressor fac-Ž .tor MNSF produced by a T cell hybridoma, J. Immunol.Ž .136 1986 2906–2909.

w x14 M. Nakamura, H. Ogawa, T. Tsunematsu, CharacterizationŽ .of monoclonal nonspecific suppressor factor MNSF with

Ž .the use of a monoclonal antibody, J. Immunol. 138 19871799–1803.

w x15 M. Nakamura, R.M. Xavier, Y. Tanigawa, Ubiquitin-likemoiety of the monoclonal nonspecific suppressor factor b is

Ž .responsible for its activity, J. Immunol. 156 1996 533–538.w x16 M. Nakamura, H. Ogawa, T. Tsunematsu, IFN-g enhances

the expression of cell surface receptors for monoclonalŽ .nonspecific suppressor factor, Cell. Immunol. 139 1992

131–138.w x17 M. Nakamura, R.M. Xavier, Y. Tanigawa, Molecular cloning

and characterization of a cDNA encoding monoclonal non-specific suppressor factor, Proc. Natl. Acad. Sci. USA 92Ž .1995 3463–3467.

w x18 M. Nakamura, T. Nagata, R.M. Xavier, Y. Tanigawa, Ubiq-uitin-like polypeptide inhibits the IgE response of

Ž .lipopolisaccharide-activated B cells, Int. Immunol. 8 19961659–1665.

w x19 A. Ciechanover, H. Heller, S. Elias, A.L. Haas, A. Hershko,ATP-dependent conjugation of reticulocyte proteins with thepolypeptide required for protein degradation, Proc. Natl.

Ž .Acad. Sci. USA 77 1980 1365–1368.w x20 M. Ravinovitz, J.M. Fisher, Characteristics of the inhibition

of hemoglobin synthesis in rabbit reruculocytes by threo-a-amino-b-chlorobutyric acid, Biochim. Biophys. Acta 91Ž .1964 313–322.

w x21 A. Ciechanover, Y. Hod, A. Hershko, A heat stablepolypeptide component of an ATP-dependent proteolyticsystem from reticulocytes, Biochem. Biophys. Res. Com-

Ž .mun. 81 1978 1100–1105.w x22 M. Scheffner, J.M. Huibregtse, R.D. Vierstra, P.M. Howley,

The HPV-16 E6 and E6-AP complex functions as a ubiqui-Ž .tin-protein ligase in the ubiquitination of p53, Cell 75 1993

495–505.w x23 A. Haas, I.A. Rose, The mechanism of ubiquitin activating

Ž .enzyme, J. Biol. Chem. 257 1982 10329–10337.w x24 A. Ciechanover, The ubiquitin proteasome proteolytic path-

Ž .way, Cell 79 1994 13–21.w x25 A. Ciechanover, A.L. Schwartz, The ubiquitin-mediated pro-

teolytic pathway: mechanisms of recognition of the prote-olytic substrate and involvement in the degradation of native

Ž .cellular proteins, FASEB J. 8 1994 182–191.w x26 M. Scheffner, U. Nuber, J.M. Huibregtse, Protein ubiquiti-

nation involving an E1-E2-E3 enzyme ubiquitin thiolesterŽ . Ž .cascade, Nature London 373 1995 81–83.

w x27 V. Agustini, T. McIntosh, L. Malek, Ubiquitination andŽ .ATP levels in garden pea seeds, Physiol. Plant. 97 1996

463–468.w x28 R.D. Vierstra, Demonstration of ATP-dependent, ubiquitin-

conjugating activities in higher plants, Plant Physiol. 84Ž .1987 332–336.

w x29 B. Veierskov, I.B. Ferguson, Conjugation of ubiquitin toŽ .protein from green plant tissues, Plant Physiol. 96 1990

4–9.w x30 J. Spence, S. Sadis, A.L. Haas, D. Finley, A ubiquitin

mutant with specific defects in DNA repair and multiubiqui-Ž .tinaion, Mol. Cell. Biol. 15 1995 1265–1273.

w x31 E.S. Johnson, B. Bartel, W. Seufert, A. Varshavsky, Ubiqui-Ž .tin as a degradation signal, EMBO J. 11 1992 497–505.


w x32 E.S. Johnson, P.C.M. Ma, I.M. Ota, A. Varshavsky, Aproteolytic pathway that recognizes ubiquitin as a degrada-

Ž .tion signal, J. Biol. Chem. 270 1995 17442–17456.w x33 C.V. Beveren, S. Enami, T. Curran, I.M. Verma, FBR

Ž .murine osteosarcoma virus, Virology 135 1984 229–243.w x34 T. Nakano, H. Watarai, Y.C. Liu, Y. Oyama, T. Mikayama,

K. Ishizaka, Conversion of inactive glycosylation inhibitingfactor to bioactive derivatives by modification of a SH

Ž .group, Proc. Natl. Acad. Sci. USA 94 1997 202–207.w x35 A.L. Haas, P. Ahrens, P.M. Bright, H. Ankel, Interferon

induces a 15 kilodalton protein exhibiting marked homologyŽ .to ubiquitin, J. Biol. Chem. 262 1987 11315–11323.

w x36 T. Okura, L. Gong, T. Kamitani, T. Wada, I. Okura, C.F.Wei, H.M. Chang, E.T.H. Yeh, Protection againstFasrAPO-1- and tumor necrosis factor-mediated cell death

Ž .by a novel protein, sentrin, J. Immunol. 157 1996 4277–4281.

w x37 T. Kamitani, H.P. Nguyen, E.T.H. Yeh, Preferential modifi-cation of nuclear proteins by a novel ubiquitin-like molecule,

Ž .J. Biol. Chem. 272 1997 14001–14004.

Documents

Conjugation of ubiquitin-like polypeptide to intracellular acceptor proteins