Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Tumor rejection antigen gp96/grp94 is an ATPase: implications for protein folding and antigen presentation.

Tumor rejection antigen gp96/grp94 is an ATPase: implications for protein folding and antigen... The EMBO Journal vol.12 no.8 pp.3143-3151, 1993 Tumor rejection antigen is an ATPase: gp96/grp94 implications for protein folding and antigen presentation Zihai Li and Pramod K.Srivastava1 from BALB/c spleen However, sequencing of gp96 cDNAs A and CMS5 did not reveal any and fibrosarcomas Meth Departments of Pharmacology and Microbiology, Box 1215, Mount mutations Srivastava tumor-specific, individually distinct (see Sinai School of Medicine, New York, NY 10029, USA observations raise a as to and Maki, 1991). These question 1Corresponding author of of gp96. the structural basis specific immunogenicity M.Feldman to known Communicated by Comparison of the gp96 sequence sequences heat shock protein revealed significant homology with the gp%/grp94 heat shock proteins Immunization of mice with (HSP) hsp9O and possible identity with the glucose-related (HSPs) elicits tumor-specific cellular immunity to the Mazzarella and protein grp94 (Lee et al., 1984; Green, tumors from which gp96 is isolated. However, the cDNA Maki 1993). 1987; Srivastava etal., 1987; etal., 1990, among tumors and normal sequence of gp96 is identical bind a diverse range of cellular Hsp9O has been reported to tissues. This raises the question regarding the structural has been attributed proteins and a protein chaperoning role basis of the specific immunogenicity of gp96. As HSPs et Wiech et A role to it (Shaknovich al., 1992; al., 1992). bind a wide array of molecules including peptides, we in folding has also been suggested for gp96/grp94 protein have proposed that gp% may not be immunogenic per se, basis of the observation that it is induced by on the but antigenic peptides. Furthermore, may chaperone in the endoplasmic accumulation of misfolded proteins is localized in the lumen of the gp96 predominantly et al., 1988). In the absence of reticulum (ER) (Kozutsumi reticulum suggesting that it may act endoplasmic (ER) in between tumors and sequence differences gp96 genes as a and as accessory to peptide loading peptide acceptor of the antigenicity of normal tissues in spite tumor-specific of I We demonstrate here that MHC class molecules. and Srivastava gp96, we proposed (Srivastava Heike, 1991; molecules contain ATP-binding cassettes, bind ATP gp96 and Maki, 1991) that gp96 may not be immunogenic per se, and possess an Mg2+-dependent ATPase activity. Gp96 In view of the but may be a carrier of antigenic peptides. are also observed to contain tightly bound preparations localization of in the ER and Koch, predominant gp96 (Booth extraction. These peptides, which can be eluted by acid immunity elicited by gp96 1989) and our observation that the of with its proposed roles properties gp96 are consistent T lymphocytes (Udono and is mediated through CD8+ in and in facilitaing MHC chaperoning antigenic peptides that acts as a Srivastava, 1993), we proposed gp96 peptide in the ER lumen. We present class I-peptide assembly to the ER and enables acceptor for peptides transported a model to how interaction of gp96 with MHC explain of MHC class I. peptide loading class I result in transfer of peptides to the latter. may In of the fact that peptide charging of MHC class light Key words: cancer immunity/chaperone/hsp9O/MHC class I an (Levy et al., 1991; Luescher is ATP-dependent process I/peptide et we have examined the ability of gp96 to al., 1992), and ATP. We demonstrate that are bind peptides peptides associated with gp96 and that gp96 binds ATP and indeed observations make gp96 a logical is an ATPase. These Introduction for facilitating peptide charging of MHC class candidate of tumors was first demonstrated con- Immunogenicity a structural basis for the tumor-specific I and provide in sarcomas of inbred mice vincingly chemically induced of gp96. immunogenicity Prehn and 1957; Klein (Gross, 1943; Foley, 1953; Main, et Old et al., 1962). A search for the molecules al., 1960; which mediate this individually distinct immunogenicity led Results to identification of gp96 molecules as tumor rejection in sarcomas is an antigens (TRAs) methylcholanthrene-induced Gp96 ATP-binding protein of inbred BALB/c mice (Srivastava et al., 1986; Palladino The deduced amino acid sequence of gp96 was screened Mice et and Srivastava, 1993). for the of motifs which are associated with al., 1987; Feldweg presence became immune immunized with 5-10 of purified gp96 ATP binding. The ATP-binding consensus sequences, types jig with the tumor from which et al. and refined to gp96 A and B as proposed by Walker (1982) subsequent challenges but not to with antigenically distinct One A 217- was isolated, challenges by Chin et al. (1988), were used. type (aa of a with an anti- B 231-241 tumors. Immunodepletion gp96 preparation 224) and two downstream type sequences (aa to it of were identified Within the A gp96 antibody gp96 depletes antigenicity (Srivastava and 303-313) (Figure 1). et Rabbit antisera revealed the presence of gp96 there are three amino acid residues between al., 1986). type sequence, in normal tissues as well et al., 1986). However, and instead of four as proposed in the (Srivastava Gly217 Gly221, does not elicit tumor B isolated from normal tissues A consensus sequence. The two type sequences gp96 type in and These observations have one mismatch each, the hydrophobic regions. Srivastava, 1993). immunity (Udono an additional are mutations in of tumors Flaherty et al. (1991) have suggested sequence that there gp96 genes suggested as a mutations differ from one tumor to another. fingerprint and that these (ILV)X(ILVC)DXG(TSG)(TSG)XX(RKC) Oxford Press University Z.Li and P.K.Srivastava A B Type Type Consensus (G/A)X4(G/A)(HEKIR)Xo-s(T/S/K/R/H) (HEKfR)X5-s8Xcb2(D/EE) GP96: 217-224 T - NTL GR G T 231-241 K Y D EEASD L eL 303-313 K V dW E KVEKT W Adenylate Kinase: 15-23 G GPGS GK G T 108-119Y3 'K L L D IGQPTL RING 4: C-terminal G PNGS GK S T C-terminal K PCVLILDD a S tL D Protease La: 355-362 G PPGV GK T 411-422 VGVKNP L F LL D GroEL: 164-172 A MDKV 241-251 AGKD L IA E LI Fig. 1. Sequence homology among the putative ATP domains of binding gp96 and other ATP-binding proteins. The consensus ATP binding sites proposed by Walker et al. (1982) and modified by Chin et al. are (1988) presented. The putative ATP binding sequences of are gp96 aligned with sequences of four other ATP rabbit muscle binding proteins: adenylate kinase (Kuby et human 1 al., 1984), transporter-associated protein (RING 4) (Trowsdale et al., 1990), E.coli chaperonin groEL (Hemmingsen et al., 1988) and protease La (Chin et amino al., 1988). Single acid codes are used here. F stands for a hydrophobic amino acid (I, V, L, M, Y, W, X F); indicates any amino acid. Lower case letters represent non-homologous residues. indicates a gap. The highly conserved residues are boxed. characteristic for on nucleotide-binding proteins, based the structure of the heat shock hsc7O. This crystal protein motif is in conserved hsc7O, grp78 and dnaK from Escherichia but is not found in the coli, ATP-binding protein groEL. This sequence was not detected in gp96. ATP was binding by gp96 measured experimentally by photoaffinity labeling of purified gp96 by [Ly-32P]8-azido ATP In ATP (Figure 2, lane 1). parallel, of binding hsp90 a lane and an (as positive control, unrelated 96 kDa 2) protein (as a negative control, lane 3) was also tested. Gp96 ATP was observed to bind (lane 1) and this was binding of mM ATP inhibited by inclusion cold 10 8-azido (lane 4), ATP but not AMP GTP (lane 5), cyclic (lane 6), (lane 7) Fig. 2. Gp96 binds ATP in vitro. Binding of gp96 (lane 1), hsp9O or CTP (lane 8). (lane 2) or p96 (a Meth A-derived protein unrelated to gp96, lane 3) to was carried out as described in To determine whether interacts with ATP in [_y-32P]8-azido Materials and gp96 vivo, methods. Lanes 4-8: binding was carried out in the presence of EL-4 with 32P cells were labeled and was immuno- gp96 10 mM unlabeled 8-azido ATP (lane 4), ATP (lane 5), cyclic AMP from whole cell of such cells. precipitated lysates A (lane 6), GTP (lane 7) and CTP (lane 8). phosphate-labeled band was observed gp96 (Figure 3A). These conditions should detect experimental only phos- phorylated or or with ADP-ribosylated proteins, proteins bound nucleotides. In two other tightly fact, Gp96 elutes within a broad range of salt phosphoproteins, concentration MHC class I and were also from these hsp70, precipitated (0.45-0.6 M NaCl), instead of a sharp peak. However, extracts with the antibodies and were detected respective by the protein material in each peak comprises apparently In the case of a number of bands homogeneous as autoradiography. hsp7O, gp96 judged by silver stained gels of these other than the 70 kDa band were detected in the immuno- fractions (Figure 4B). The fractions were individually precipitate; these proteins are associated with assayed for ATPase presumably activity as described in Materials and hsp70 because the used is methods. ATPase anti-hsp7O antibody quite specific activity was clearly detected in the gp96 and detects only hsp70 on Western blots. fractions (Figure 4C). The activity in this experiment is not To determine whether the 32p label in derives linear with the concentration of gp96 gp96 because this protocol from associated ATP, or whether it is a result of a measures net activity rather post- than the initial velocity of translational modification, radioactive adenosine nucleotides reaction. For subsequent characterization of the ATPase were eluted from specifically the activity of gp96, fraction immunoprecipitates by 6 was used. large excess of cold ADP and AMP. It was ATP, observed To confirm that the ATPase activity was derived from (Figure 3B) that and but not MHC class gp96 hsp70, I, gp96 molecule itself, fraction 6 (from the experiment in contain tightly bound ADP and AMP. ADP is the Figure was ATP, 4A) applied sequentially a number of times, to major component eluted from and ATP an gp96 hsp70; anti-gp96 monoclonal antibody column and depleted of contributes to only a minor fraction of the total eluted gp96. The initial and depleted fractions were characterized nucleotides. This is consistent with the observations made by SDS -PAGE and assayed for ATPase activity. It was with dnaK and the crystallographic analysis of hsp70 observed that the loss of ATPase activity was concomitant (Flaherty et al., ATP is not in 1990). detected abundance with loss of gp96 (Figure 5A, B and C). Application of presumably because of its rapid hydrolysis. fraction 6 through an unrelated immunoaffinity column did not deplete the preparation of gp96, nor of ATPase Gp96 is an ATPase activity (data not shown). These experiments show that the 4A shows the Figure chromatographic profile of the last ATPase step activity of gp96 resides in the gp96 itself and is of of Meth A purification gp96 on a Mono FPLC column. not a Q contaminant. 3144 Tumor an ATPase rejection antigen gp96 is /If: , I.' ;u *Ill, A ,:, 'I-N in :...nher ....: t P, KI)9 ,_ ue- P s Ar-T*R., -4-- AII ., W. | ADII Fig. 3. Interaction of gp96 with adenosine nucleotide in vivo. (A) Lmmunoprecipitation of phosphate-labeled gp96 from [32P]orthophosphate-labeled EL-4 cells. Immunoprecipitation was done with a polyclonal rabbit serum against MHC class I whole molecule 4r2 X I I I I I .4 i (K270, obtained from Dr Per Petersen), rat anti-gp96 monoclonal Y\1P 40-- antibody (clone 9G10) and rat anti-hsp7O monoclonal antibody (clone 7.10), respectively, followed by secondary rabbit anti-rat IgG and protein A-Sepharose. (B) PEI thin layer chromatography of gp96-associated nucleotides. Immunoprecipitates were washed and Fig. 4. ATPase co-purifies with gp96. (A) Chromatographic activity associated nucleotides were eluted and separated by PEI TLC plates profile of gp96 on MonoQ FPLC. Proteins bound to concanavalin (see Materials and methods). yu of 100 ATP, ADP and AMP A-Sepharose column were dialyzed against 5 mM sodium phosphate jtM were spotted and localized with the aid of a short-wave UV lamp and buffer, pH 7.0, and applied to MonoQ FPLC column. Bound proteins the are positions indicated by arrows. The spot seen in the MHC class were eluted with a linear 0-1 M NaCl gradient. Gp96 elutes at NaCl I lane on the TLC of the in represents <5% counts seen the gp96 or concentration between 0.45 and 0.6 M. (B) Silver stained hsp70 lanes. The films were exposed to a non-linear degree in order SDS-polyacrylamide gel of MonoQ FPLC-purified fractions shown in to show the ATP spots clearly. (A). Fractions were applied to 10% SDS-PAGE and silver stained. Lanes 1-10 represent 3 1d aliquot of each 1 ml fraction. (C) In vitro ATPase activity of purified gp96. 10 of gradient fractions shown in A1 Effect of temperature, pH and divalent cations on the (A) were assayed for ATPase activity using 2 [a-32P]ATP (see liM Materials and methods). Reaction products were analyzed by PEI thin ATPase activity of gp96 layer chromatography and autoradiography. The positions of ATP, ATPase activity of gp96 was assayed at different temperatures ADP and AMP are shown. at a pH of 7.2 and was found to be optimal at 42°C (Figure 6A). A temperature optimum higher than 37°C has been noted for some other HSPs as well; order kinetics for example, dnaK, a (Figure 7 and inset), suggesting that gp96 has bacterial hsp7O, shows optimal a ATPase activity at 50-55°C single ATPase active site. No evidence of co-operativity (Liberek et al., 1991). Furthermore, was the ATPase activity of observed. Under the conditions of this experiment, the gp96 was observed to be ATPase has optimal at acidic pH (Figure 6B). activity of gp96 a Km of 8 and the turnover AM The divalent cation dependence of rate is -0.08 mol/min/mol. These gp96 ATPase was tested characteristics suggest and the activity was found to be that ATP a dependent on exogenous hydrolysis by gp96 role plays regulatory (see but not on Mg2+ Ca2+ (data not shown). This was con- Discussion). firmed by the use of the divalent cation chelators EDTA and EGTA in the reaction mixture. EDTA, which is a Protein and modulators of ATPase general peptide activity of divalent cation chelator, had a and significantly stronger gp96 hsp70 inhibitory effect on the ATPase activity of gp96 than EGTA, Members of the hsp70 family, including grp78, undergo which is a Ca2+-specific chelator (Figure 6C). The other conformational and substrate release ATP changes during ER luminal HSP, grp78, is similar to gp96 in its requirement hydrolysis (Carlino et al., 1992; and Gething Sambrook, for divalent cations (Kassenbrock and Kelly, 1989): grp78 1992). Moreover, ATP hydrolysis is stimulated by requires Mg2+ but not Ca2+ for its ATPase activity and in- exogenous peptides et As and (Flynn al., 1989, 1991). gp96 the are the two ER we deed, activity is inhibited by the presence of Ca2 . grp78 luminal HSPs, investigated whether A substrate ATPase of is also modulated saturation experiment was done to determine activity gp96 by peptides. Two the kinetics of ATPase of A and which were used for ATPase activity gp96. When the ATPase peptides B, stimulation activity was assayed in the presence of with et al. were added in increasing con- assays grp78 by Flynn (1989), the centrations of the ATPase with As shown in the ATP, velocity was found to follow first assay gp96. Figure 8A, 3145 ZLi and P.K.Srivastava 16 - .. , 12 .14 .;,A i. * 8~ 9 , 10W )' 10 20 30 50 60 70 4+-- AIP Temperature (OC) *. ...,=e", = 30- .: E 25 20- 3 15 t3 ., ' i ;. L an I' iI 3 4 5 6 7 8 9 in molecule Flg. 5. ATPase activity resides the itself. gp96 pH (A) Depletion of gp96 using column. Monoclonal immunoaffinity antibody was coupled to protein column. The same G-Sepharose volume of original materials (-0.5 and materials from the eluting Ag) column each time were to 10% SDS-PAGE and silver C applied stained. (B) Samples before (0.5 and after were 22 - depletion assayed itg) for ATPase activity by with 20 for 30 min. incubating [ca-32P]ATP /tM in (C) Quantitative analysis of results 30 ATP were 20- (B). pmol 0.5 before the hydrolyzed by gp96 immunodepletion and the A jig relative ATPase was defined as 1. activity oa :0 16 - o9 ATPase activity of gp96 was not stimulated and was indeed h. el 14 - inhibited in the presence of either of the two exogenous I- peptides. Fifteen additional from 7 to 20 peptides ranging 12 - in residues length were also found to have no stimulation effect. An hsp7O preparation isolated from Meth A cells was 10- also tested for ATPase activity and its stimulation S by peptides A and B under our experimental conditions. In contrast to .0001 .001 .01 .1 1 10 100 lack of the stimulation of gp96 ATPase by peptides, hsp70 Concentration ATPase was - of EDTA/EGTA (mM) reproducibly found to be stimulated 1.5-fold addition of by peptides (Figure 8B). Flynn et al. have Fig. 6. Effect of temperature, pH and divalent cations on the ATPase observed a maximal 4-fold stimulation of the ATPase activity activity associated with gp96. Effect of on (A) temperature the rate of of grp78 by the same under similar peptides conditions. The ATP hydrolysis by gp96. (B) Optimization of pH. 570 ng purified difference between our results stems perhaps from the fact gp96 were assayed for ATPase in the of activity presence 20 mM that we are using a mixed of molecules pool hsp70 rather NaCl, 2 mM MgCl2 and 20 mM of various buffers 4.0 acetate, (pH than a pH 5.5-6.0 MES, pH 7.0-7.2 HEPES, pH purified grp78 preparation. 7.5-8.8 Tris). ATPase is (C) activity inhibited by divalent cation chelators. Purified It has been suggested that the stimulation of ATPase was for gp96 assayed ATPase activity in the presence of 20 mM of activity grp78 by peptides reflects the fact that peptides MES, pH 20 mM NaCl 6.0, and increasing concentrations of EDTA may mimic certain features of the newly synthesized un- or EGTA (solid line) (dotted line). folded proteins, whose folding is facilitated by grp78. The stimulation of ATPase activity would thus or provide energy so-called 'folding intermediates' or 'molten globules' (for for the folding process. However, in contrast to grp78, see the reviews, Ang et al., 1991; Gething and Sambrook, bacterial chaperonin groEL (which also hydrolyzes ATP) 1992). Thus, groEL does not recognize either completely has been shown to interact only with partially folded proteins folded or totally unfolded polypeptides, nor does it bind 3146 Tumor rejection antigen gp96 is an ATPase O. _ ~~~o C', 00- _> -5 0~~~~~~~~ I-10 50 60 C 100 200 300 400 500 600 <t ATP -10 Concentration (gM) 400 800 1200 1600 Peptide Concentration (IM) Fig. 7. Kinetic study of gp96 ATPase activity. Substrate saturation experiment was carried out by incubating purified gp96 (0.862 per jig reaction) with increasing concentrations of [ax-32P]ATP in 20 mM HEPES buffer, pH 7.2, containing 20 mM NaCl and 2 mM MgCl2 at 37°C for h. Background values in the absence of gp96 have been _ 30 subtracted. Double reciprocal plot of the same data is shown in inset. cm The Km calculated from this experiment is 8 jiM. The that the ATPase activity of gp96 peptides. possibility Z 20 is modulated by similar interactions was therefore pursued and casein was tested for this dephosphorylated purpose. <' 15 ._1 Casein is a soluble with a number of exposed protein hydrophobic surfaces and has been used as a competing o 10 substrate for groEL binding (Langer et al., 1992). It was observed that similar to the stimulation of hsp70 ATPase x 5 casein stimulates the ATPase of by peptides, activity gp96 IL I- -2-fold by (Figure 9). 400 600 800 1000 1200 1400 0 200 is associated with Gp96 peptides Peptide Concentration (jiM) To identify peptides associated with gp96, 109 Meth A cells labeled were metabolically with [35S]methionine of ATPase of and Fig. 8. Modulation activity gp96 hsp70 by peptides. and was isolated. The purified (200 gp96 stimulate ATPase of 1 of ACi/ml) (A) Peptides do not activity gp96. jig preparation was extracted with 10% acetic acid and the low was for ATPase in the of purified gp96 assayed activity presence concentrations of either A or molecular weight material isolated by centricon centrifuga- increasing peptide (A-A) peptide B at 370C for 1 h. Both peptides A and B are 15mers and (O-----0) tion which will allow of molecules (centricon-10, passage are derived from vesicular stomatitis virus glycoprotein (peptide A, of 10 000 Da or as described in Materials and less), KRQIYTDLEMNRLGK; peptide B, LSSLFRPKRRPIYKS). Peptide- methods. This material was applied to a reverse phase C18 ATP of 50 was subtracted. independent hydrolysis pmol/h/ljg HPLC column and eluted on an acetonitrile gradient. stimulate ATPase of The same (B) Peptides activity hsp70. peptide ATP preparation used in (A) was used. Peptide-independent hydrolysis Individual fractions were collected and counted. A number of 85 has been subtracted. pmolh/jtg of [35S]methionine-labeled distinct peptide peaks were obtained (Figure 10A). However, as this material was insufficient for quantitative analysis, we purified 3 mg gp96 for acid extraction or it may be stored in this high salt buffer from mammalian liver and acid-extracted it as before. The for several hours in the presence of high concentrations of mM peptides eluted from it were applied to a column. protease inhibitors (2 PMSF, 10 leupeptin) at C18 AM before acid extraction. In either we Figure lOB shows that a number of peptide peaks were 4°C, case, routinely the in obtained. Some of the peaks have been partially characterized centrifuged purified preparations the high salt buffer a acid extraction by mass spectroscopy and have been determined to be through centricon-1O immediately preceding masses between and did not find in the low molecular peptides of heterogeneous size (molecular any peptides weight not An fraction. contains tightly bound peptides, 400 and 2000 atomic mass units, data shown). Thus, gp96 of the eluted apparently in the same manner and by the same criteria as approximate quantitative analysis peptides that and are in the MHC class I and class II molecules et suggests gp96 peptides present roughly (Falk al., 1990; that the Rotzschke et al., 1990; Rotzschke and Falk, 1991). equimolar stoichiometry, assuming average peptide molecular mass of 1000 Da. It should be noted that has a the eluted are not associated with peptides merely loosely Discussion but are bound to it. The purified gp96 gp96, tightly ATPase of and other HSPs was obtained after a number of that include activity gp96 preparation steps is the most abundant of the elution on a salt on which elutes at - 0.5 M Gp96/grp94 single component gradient gp96 of the but no in vivo function or final be used lumen ER, enzymatic NaCl. The gp96 preparation may immediately 3147 Z.Li and P.K.Srivastava I 0 4) 1- co 0 1 2 3 4 5 6 7 Casein Concetrtion (mni) fractions Fig. 9. Casein stimulates ATPase activity of gp96. ATPase activity of was in the gp96 assayed presence of increasing concentrations of casein. ATPase with alone well activity gp96 as as the minor background ATP hydrolysis associated with casein itself are subtracted from each point. U-) has been to it as activity assigned yet (Pelham, Our 1986). results show novel of this molecule and properties suggest rn -I4 for it a crucial role in MHC antigen presentation by class 60 I molecules and in tumor We demonstrate that immunity. binds ATP and is an ATPase. gp96 Mg2+-dependent The deduced amino acid of contains sequence gp96 ATP-binding LI) sn o motifs and gp96 binds ATP in vitro and in vivo 2 (Figures The Km of the ATPase was calculated as and 3). activity 8 and the maximum as 50 velocity (Figure ytM pmol/h/tg 7). This to a turnover rate of 0.08 mol/min/mol. corresponds 0 8 16 24 32 40 48 These characteristics of the ATPase fall within the FRACTION NUMBER gp96 range of for the ATPases. corresponding parameters hsp70 Fig. 10. is For example, the ATPase of the E. coli dnaK Gp96 associated with heterogeneous peptides. Purified activity hsp70 gp96 was extracted with acetic acid. 10% Peptides eluted from gp96 were has a Km value of 20 and a turnover rate of /.M separated through and centricon-10 separated on a C18 HPLC column, 0.2 mol/min/mol (Liberek et an al., 1991). Grp78, hsp70 as described in Materials and methods. (A) Peptides eluted from gp96 member of the ER, has a Km of 0.1 tM and a turnover derived from 1 x 109 Meth A cells metabolically labeled with rate of -0.34 mol/min/mol (Kassenbrock and Kelly, 1989). [35S]methionine. (B) Peptides isolated from liver macaque gp96. A comparison of the turnover rates of ATPase activities of of the ER. We suggest that the two molecules use the gp96, dnaK and grp78 with the values for energy corresponding from hydrolysis of ATP in divergent ways, as a non-HSP such as kinase is The discussed adenylate revealing. below. turnover rate of the ATPase of rabbit muscle activity adenylate kinase is 2 106 mol/min/mol from (calculated and in Role in Hampton Slotin, 1975); comparison, the turnover rate of gp96 folding/assembly of proteins in the ER of HSP as ATPases, described above, is several orders of Grp78 is generally thought to correct catalyze folding of magnitude lower. This indicates that ATP hydrolysis by newly synthesized proteins in the lumen of the ER. The role HSPs has a of in significantly dissimilar biological purpose from gp96/grp94 this process has remained largely obscure. ATP hydrolysis by non-HSPs such as Similar to adenylate kinase and grp78, grp94 has been shown to be induced by the transporter the of accumulation of proteins; specifically, ATPase activity malfolded proteins in the ER (Kozutsumi HSPs may mediate a regulatory rather than a et al., 1988) suggesting that transporting gp96 acts as a chaperone. A or modifying function. number of observations have now begun to show this A of comparison the ATPase activities of the two HSP formally. In our analysis of the assembly of MHC class I ATPases of the ER lumen, with (32 grp78 and gp96, reveals critical microglobulin and peptides, we have observed differences. The ATPase of association of activity grp78 is stimulated by MHC class I heavy chain with gp96 (Z.Li peptides, while that of gp96 is inhibited or and in unaffected. P.K.Srivastava, preparation). In earlier studies, Furthermore, ATPase activity of grp78 is unaffected by the Navarro et al. (1991) observed that grp78 and gp96 associate presence of other proteins such as casein, while the ATPase with aberrant forms of glycoprotein b of the herpes simplex activity of gp96 is stimulated by casein. These differences virus 1. Similar observations were reported for newly hint towards different perhaps complementary roles for the synthesized immunoglobulin chains by Melnick et al. (1992). two proteins in the ER lumen. Clairmont et al. (1992) have Schaiff et al. (1992) showed that MHC class II (HLA-DR) demonstrated that and are the grp78 gp96 two major oa and chains associate with gp96 in the ER in the absence recipients of the pool of ATP translocated into the lumen of invariant chain. This association does not occur in the 3148 Tumor rejection antigen gp96 is an ATPase presence of invariant chain. The observations reported in this paper, that gp96 molecules bind peptides and ATP and have a regulatory ATPase activity, provide a missing piece of the puzzle and furnish an enzymatic basis for the emerging evidence for the role of gp96 as a bonafide chaperone. There are two main paradigms known for the mechanism -1 AD- AM by which HSP chaperones catalyze protein folding or assembly (Rothman, 1989; Gething and Sambrook, 1992). In one paradigm, typified by grp78, cycles of binding and release of grp78 with short peptide regions of the substrate protein are propelled by the energy provided by ATP Fig. 11. A model for the role of gp96 in MHC class I peptide hydrolysis and a polypeptide chain emerges fully folded after assembly. It is postulated that the interaction of MHC class I and gp96 having gone through a number of transient associations. In stimulates the ATPase activity of gp96. The free energy (AG0) released due to ATP hydrolysis induces a conformational change in concordance with this paradigm, ATPase activity of grp78 gp96, leading to transfer of peptides from gp96 to MHC class I. TAP is stimulated by peptides, but not by folded intact proteins denotes transporter-associated proteins. (Flynn et al., 1989, 1991). In another paradigm, typified by the groE proteins, protein substrates bind to the groEL oligomer at a number of sites and the folding proceeds at Gp96-accessory to peptide-loading of MHC class I? different portions of the polypeptide substrate by a We demonstrate that gp96 is associated with a wide array progressive, ATP hydrolysis-dependent release of different of peptides (Figure 10). In the context of our observation regions of the substrate from the oligomeric groEL (Gething of co-immunoprecipitation of gp96 and MHC class I from and Sambrook, 1992). Another groE protein, groES, metabolically labeled cells (Z.Li and P.K.Srivastava, in modulates the ATPase activity of groEL. In concordance preparation) and in light of the ATPase activity of gp96, we with this paradigm, the ATPase activity of groEL is propose (Figure 11) that (i) gp96 acts as a repository of the stimulated by interaction with some proteins, but is not peptides transported into the lumen of the ER, (ii) interaction stimulated by peptides. of gp96 with MHC class I stimulates its ATPase activity, The ATPase activity of gp96 appears to be distinct from resulting in hydrolysis of gp96-bound ATP, (iii) the energy both paradigms, although it shares selected features with released from ATP hydrolysis is used for transfer of peptides each. Gp96 molecules may exist as dimers or tetramers from gp96 to MHC class I. (Srivastava and Das, 1984; unpublished) and like the groEL, An ER luminal chaperone which will facilitate charging and in contrast to grp78, the gp96 ATPase is not stimulated ofMHC class I with peptides has been invoked by Rothman, by peptides, but is stimulated by casein-a good model for Cresswell, Townsend, Kvist and their colleagues (Alexander a et partially folded protein. It would appear that the ATPase al., 1989, 1990; Rothman, 1989; Townsend et al., 1990; activity of gp96, like that of dnaK and groEL, is modulated Levy et al., 1991). In view of the ability of gp96 to bind by protein-protein rather than protein-peptide interactions. peptides and ATP, its ATPase activity, and its localization It is conceivable ER in the ER that the lumen contains other proteins, lumen (where it is the major component), it is which modulate the ATPase activity of gp96 in the same reasonable to suggest this role for gp96. manner as An ER groES modulates the activity of groEL, or DnaJ chaperonin p88 -IP90-calnexin has recently been and GrpE modulate the activity of dnaK. Gp96 and grp78 identified (Degen and Williams, 1991; Wada et al., 1991; also may conceivably modulate and collaborate with each Galvin et al., 1992; Hochstenbach et al., 1992). It is a other. Sequential collaboration of the two chaperones dnaK membrane-bound (as opposed to luminal) protein and is and in associated groEL the folding process is a precedent for such with partially but not fully assembled com- an of T effort (Langer et al., 1992). plexes immunoglobulins, cell receptors and MHC. While and have shared roles in the Hochstenbach et al. that gp96 grp78 may folding (1992) suggest p88 might participate the of the two ATPases in of multi-subunit It is that process, divergent regulation assembly complexes. possible 8 and an role is involved in of MHC class I (Figures 9) may suggest additional, unique p88 folding heavy chain, but for is the most evolved HSPs. there is no evidence of or ATPase gp96. Gp96 among recently peptide binding activity It is not found in or and to have of p88 and it does not to be an to yeasts Drosophila appears appear accessory peptide a of the loading of MHC class I. emerged relatively recently by duplication cytosolic hsp90. The major structural differences between and hsp90 the amino-terminal and Implications for immune to cancer gp96, signal peptide carboxy- response terminal KDEL mediate ER and reten- The observations that gp96 is associated with binds sequence, targeting peptides, tion. The additional difference between the two ATP and is an ATPase lends to the only strong support suggestion of molecules is in the ATPase that of activity gp96: although hsp90 tumor-specific immunogenicity gp96 (and perhaps would that from but from binds it is not an ATPase. It other HSPs) does not derive ATP, appear gp96 directly gp96 function in the associated with it. The source of the evolved to fulfill an peptides immunogenic ATPase-dependent ER, In of the from tumor cells lies in altered which may be unique to higher organisms. light peptides clearly (mutated) effects of vaccination with or cellular and the of immunological gp96 (Srivastava mis-expressed proteins specificity and Srivastava et Palladino et result from randomness of mutations Das, 1984; al., 1986; immunogenicity may al., et and and our observation et Szikora 1987; Feldweg Srivastava, 1993) (Sibille al., 1990; Srivastava, 1990; al., 1990; I Structural that gp96 associates with MHC class (Z.Li and Van den Eynde et al., 1991; Boon, 1992). that one of of eluted from in we the characterization antigenic P.K.Srivastava, preparation), suggest peptide(s) gp96 of MHC class from distinct tumors will functions of involves preparations antigenically specific gp96 charging provide further for this I molecules with support hypothesis. peptides. 3149 Z.Li and P.K.Srivastava Immunogenetics, 31, 169-178. Materials and methods Ang,D., Liberek,K., Skowyra,D., Zylicz,D. and Georgopoulos,C. (1991) J. Bio. Chem., 266, 24233-24236. Reagents, mice and tumors Boon,T. (1992) Adv. Cancer Res., 58, 177-210. All chemicals were purchased from Sigma except where specified. Booth,C. and Koch,G.L.E. (1989) Cell, 59, 729-737. Radiochemicals were from ICN. Cell culture media and reagents were mainly Carlino,A., Toledo,H., Skaleris,D., Delisio,R., Weissbach,H. and Brot,N. obtained from GIBCO-BRL. Monoclonal antibodies against gp96 and hsp7o (1992) Proc. Natl Acad. Sci. USA, 89, 2081-2085. were purchased from Stressgen. Peptides were synthesized on an Applied Chin,D.T., Goff,S.A., Webster,T., Smith,T. and Goldberg,A.L. (1988) Biosystems Model 430A peptide synthesizer using FMOC chemistry. Inbred J. Biol. Chem., 263, 11718-11724. BALB/c mice were from our mouse colonies. EL-4 cells were obtained Clairmont,C.A., De Maio,A. and Hirschberg,C.B. (1992) J. Biol. Chem., obtained from ATCC. 267, 3983-3990. Degen,E. and Williams,D.B. (1991) J. Cell Biol., 112, 1099-1115. and Purification of gp96 hsp7O Downward,J., Graves,J.D., Warne,P.H., Rayter,S. and Cantrell,D.A. Gp96 purification has been described by Srivastava et al. (1986). Hsp,70 (1990) Nature, 346, 719-723. was purified essentially as described by Welch and Feramisco (1985). Falk,K., Rotzschke,O. and Rammensee,H.G. (1990) Nature, 348, 248-251. Feldweg,A.M. and Srivastava,P.K. (1993) J. Cell. Biochem., 17D, Abstract ATP binding no. NZ206. Purified gp96, hsp90 or other protein (10 yg) in 10 mM Tris, pH 7.4, 5 mM Flaherty,K.M., DeLuca-Flaherty,C. and McKay,D.B. (1990) Nature, 346, and 2 mM ATP in a final volume of 100 1l was CaCl2 [y_-32P]8-azido 623-628. irradiated with a short wavelength (254 nM) UV source from a distance Flaherty,K.M., McKay,D.B., Kabsch,W. and Holmes,K.C. (1991) Proc. of 8-10 cm for 4 min (Hobson et al., 1984). Samples were then Natl Acad. Sci. USA, 88, 5041-5045. microcentrifuged for 5 min to remove large complexes and analyzed Flynn,G.C., Chappell,T.G. and Rothman,J.E. (1989) Science, 245, by SDS-PAGE. 385-390. Flynn,G.C., Pohl,J., Flocco,M.T. and Rothman,J.E. (1991) Nature, 353, A TPase assay 726-730. ATPase activity was measured by the method of Flynn et al. (1989, 1991). Foley,E.J. (1953) Cancer Res., 13, 835-837. Typically, 1 of purified gp96 or hsp7o was incubated with 20,M Ag Galvin,K. (1992) Proc. Natl Acad. Sci. USA, 89, 8452-8456. in a reaction volume of 20 containing 20 mM HEPES, pH [a-32P]ATP Al Gething,M.-J. and Sambrook,J. (1992) Nature, 355, 33-45. 7.2, 20 mM NaCl and 2 mM MgCl2 at 37°C for 1 h. 1 1l of the reaction Gross,L. (1943) Cancer Res., 3, 323-326. mixture was then spotted onto a polyethyleneimine (PEI) cellulose plate. Hampton,A. and Slotin,L.A. (1975) Biochemistry, 14, 5438-5444. Thin layer chromatography was performed against 1: 1 ratio of 1 M LiCl Hemmingsen,S.M., Woolford,C., Van der Vies,S.M., Tilly,K., and 1 M HCOOH. The plate was then dried, exposed to film and Dennis,D.T., Georgopoulos,C.P., Hendrix,R.W. and Ellis,R.J. (1988) corresponding radioactive spots were excised and counted. ATPase activity Nature, 333, 330-334. was determined from the amount of [32P]ADP and [32P]AMP generated Hobson,A.C., Weatherwax,R. and Ames,G.F.-L. (1984) Proc. NatlAcad. from [a-32P]ATP, i.e. the percentage of ATP hydrolyzed calculated as Sci. USA, 81, 7333-7337. + + AMP + x 100%. Background ATP [ADP AMP]/[ATP ADP] Hochstenbach,F., David,V., Watkins,S. and Brenner,M.B. (1992) Proc. hydrolysis lacking purified gp96 or hsp70 was subtracted. NatlAcad. Sci. USA, 89, 4734-4738. Kassenbrock,C.K. and Kelly,R.B. (1989) EMBO J., 8, 1461-1467. immunoprecipitation Phosphate labeling and Klein,G., Sjogren,H.O., Klein,E. and Hellstrom,K.E. (1960) CancerRes., Labeling cells with 2p was done according to Downward et al. (1990). 20, 1561-1572. was carried out using the Townsend protocol Immunoprecipitation Kozutsumi,Y., Segal,M., Normington,K., Gething,M.-J. and Sambrook,J. nucleotides, immunoprecipitates were (Townsend et al., 1990). To elute 332, 462-464. (1988) Nature, with 2 mM 2 mM 0.5 mM ATP, incubated EDTA, DTT, 0.2% SDS, Kuby,S.A., Palmieri,R.H., Frischat,A., Fischer,A.H., Wu,L.H., Maland,L. ADP mM AMP for of 0.5 mM and 0.5 at 68'C 20 min. Separation eluted and Manship,M. (1984) Biochemistry, 23, 2393-2399. nucleotides was on PEI-cellulose chromatography. plates Lu,C., Echols,H., Flanagan,J., Hayer,M.K. and Hartl,F.U. Langer,T., (1992) Nature, 356, 683-689. - silver and SDS PAGE, Western blotting, staining Lee,A.S., Bell,J. and Ting,J. (1984) J. Biol. Chem., 259, 4616-4621. immunoaffinity protein G columns Levy,F., Gabathuler,R., Larsson,R. and Kvist,S. (1991) Cell, 67, 265 -274. Western was done to the HRP color developing system blotting according Liberek,K., Marszalek,J., Ang,D., Georgopoulos,C. and Zylicz,M. (1991) Silver and G immunoaffmiity chromatography (Bio-Rad). staining protein Proc. Natl Acad. Sci. USA, 88, 2874-2878. were carried out using standard protocols. Luescher,I.F., Loez,J.A., Malissen,B. and Cerottini,J.C. (1992) J. Immunol., 148, 1003-1011. extraction and separation Peptide Maki,R.G., Old,L.J. and Srivastava,P.K. (1990) Proc. NatlAcad. Sci. USA, Isolation of from purified gp96 was done by the acid extraction peptides 87, 5658-5662. technique (Van Bleek and Nathenson, 1990). The low molecular weight Maki,R.G., Eddy,R., Byers,M., Shows,T.P. and Srivastava,P.K. (1993) materials were then enriched by separating out the high molecular weight Som. Cell Mol. Genet., in press. proteins using a centricon-10 spin column (MW cut off 10 000 Da, Amicon) Mazzarella,R.A. and Green,M. (1987) J. Biol. Chem., 262, 8875 -8883. according to the manufacturer's instructions. The low MW material isolated Melnick,J., Aviel,S. and Argon,Y. (1992) J. Biol. Chem., 267, from gp96 was analyzed by directly injecting into HPLC (Applied 21303-21306. BioSystems) and further resolved using a column and acetonitrile C18 Navarro,D., Qadri,I. and Pereira,L. (1991) Virology, 184, 253-264. 1 B, gradient with a flow rate of ml/min (buffer A, 0.05% TFA; buffer Old,L.J., Boyse,E.A., Clarke,D.A. and Carswell,E.A. (1962) Ann. NY the is linear from 0 to 100% buffer B). 1 ml 100% acetonitrile, gradient Acad. Sci., 101, 80-106. fractions were collected. Palladino,M.A., Srivastava,P.K., Oettgen,H.F. and Deleo,A.B. (1987) Cancer 47, 5074-5079. Res., Acknowledgements Pelham,H.R.B. (1986) Cell, 46, 959-961. Prehn,R.T. and Main,J.M. (1957) J. Natl Cancer Inst., 18, 769-778. We thank Nathalie Blachere for purification of the macaque gp96, Anya Rothman,J.E. (1989) Cell, 59, 591-601. for with the in ATP experiments and Dr Reza Harry help vitro binding Rotzschke,O. and Falk,K. (1991) Immunol. Today, 12, 447-455. Green for the was supported by critically reading manuscript. This work Rotzschke,O., Falk,K., Deres,K., Schild,H., Norda,M., Metzger,J., NIH CA44786. P.K.S. is an Investigator of the Cancer Research grant Jung,G. and Rammensee,H.G. (1990) Nature, 348, 252-254. New York, and an Irma T.Hirschl Scholar. Institute, Schaiff,W.T., Hruska,K.A., Jr, McCourt,D.W., Green,M. and Schwartz,B.D. (1992) J. Exp. Med., 176, 657-666. Shaknovich,R., Shue,G. and Kohtz,D.S. (1992) Mol. Cell. Biol., 12, References 5059-5068. Sibille,C., Chomez,P., Wildman,C., vanPel,A., DePlaen,E., MaXyanski,J., and and Boon,T. Alexander,J., Payne,A., Murray,R., Frelinger,J. Cresswell,P. (1989) DeBergeyck,V. (1990) J. Exp. Med., 172, 35-45. 380-388. Curr. Immunogenetics, 29, Srivastava,P.K. (1990) Opin. Immunol., 3, 654-659. and and Intl Alexander,J., Payne,A., Shikegawa,B., Frelinger,J. Cresswell,P. (1990) Srivastava,P.K. Das,M.R. (1984) J. Cancer, 33, 417-422. 3150 Tumor rejection antigen gp96 is an ATPase and (1991) Semin. Immunol., 3, 57-64. Srivastava,P.K. Heike,M. Srivastava,P.K. and Maki,R.G. (1991) Curr. Top. Microbiol. Immunol., 167, 109-123. Srivastava,P.K., Deleo,A.B. and Old,L.J. (1986) Proc. Natl Acad. Sci. USA, 83, 3407-3411. Srivastava,P.K., Chen,Y.-T. and Old,L.J. (1987) Proc. Natl Acad. Sci. USA, 84, 3807-3811. Szikora,J.P., vanPel,A., Brichard,V., Andre,M., VanBaren,N., Henry,P. and Boon,T. (1990) EMBO J., 9, 1041-1050. Townsend,A., Elliot,T., Cerunodo,V., Foster,L., Barber,B. and Tse,A. (1990) Cell, 62, 285-295. Trowsdale,J., Hanson,I., Mockridge,I., Beck,S., Townsend,A. and Kelly,A. (1990) Nature, 348, 741-744. Udono,H. and Srivastava,P.K. (1993) J. Cell. Biochem., 17D, Abstract no. NZ225. Van Bleek,G.M. and Nathenson,S.G. (1990) Nature, 348, 213-216. Van den Eynde,B., Lethe,B., VanPel,A., DePlaen,E. and Boon,T. (1991) J. Exp. Med., 173, 1373-1384. Wada,I. et al. (1991) J. Biol. Chem., 266, 19599-19610. Walker,J.E., Saraste,M., Runswick,M.J. and Gay,N.J. (1982) EMBO J., 8, 945-951. Welch,W.J. and Feramisco,J.R. (1985) Mol. Cell. Biol., 5, 1229-1237. Wiech,H., Buchner,J., Zimmenmann,R. and Jakob,U. (1992) Nature, 358, 169-170. Received on February 15, 1993; revised on April 20, 1993 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Tumor rejection antigen gp96/grp94 is an ATPase: implications for protein folding and antigen presentation.

Li, Z.; Srivastava, P.K.
The EMBO Journal , Volume 12 (8) – Aug 1, 1993
Download PDF
9 pages

Loading next page...
 
/lp/springer-journals/tumor-rejection-antigen-gp96-grp94-is-an-atpase-implications-for-LlcvNssPiT

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1993
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1993.tb05983.x
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal vol.12 no.8 pp.3143-3151, 1993 Tumor rejection antigen is an ATPase: gp96/grp94 implications for protein folding and antigen presentation Zihai Li and Pramod K.Srivastava1 from BALB/c spleen However, sequencing of gp96 cDNAs A and CMS5 did not reveal any and fibrosarcomas Meth Departments of Pharmacology and Microbiology, Box 1215, Mount mutations Srivastava tumor-specific, individually distinct (see Sinai School of Medicine, New York, NY 10029, USA observations raise a as to and Maki, 1991). These question 1Corresponding author of of gp96. the structural basis specific immunogenicity M.Feldman to known Communicated by Comparison of the gp96 sequence sequences heat shock protein revealed significant homology with the gp%/grp94 heat shock proteins Immunization of mice with (HSP) hsp9O and possible identity with the glucose-related (HSPs) elicits tumor-specific cellular immunity to the Mazzarella and protein grp94 (Lee et al., 1984; Green, tumors from which gp96 is isolated. However, the cDNA Maki 1993). 1987; Srivastava etal., 1987; etal., 1990, among tumors and normal sequence of gp96 is identical bind a diverse range of cellular Hsp9O has been reported to tissues. This raises the question regarding the structural has been attributed proteins and a protein chaperoning role basis of the specific immunogenicity of gp96. As HSPs et Wiech et A role to it (Shaknovich al., 1992; al., 1992). bind a wide array of molecules including peptides, we in folding has also been suggested for gp96/grp94 protein have proposed that gp% may not be immunogenic per se, basis of the observation that it is induced by on the but antigenic peptides. Furthermore, may chaperone in the endoplasmic accumulation of misfolded proteins is localized in the lumen of the gp96 predominantly et al., 1988). In the absence of reticulum (ER) (Kozutsumi reticulum suggesting that it may act endoplasmic (ER) in between tumors and sequence differences gp96 genes as a and as accessory to peptide loading peptide acceptor of the antigenicity of normal tissues in spite tumor-specific of I We demonstrate here that MHC class molecules. and Srivastava gp96, we proposed (Srivastava Heike, 1991; molecules contain ATP-binding cassettes, bind ATP gp96 and Maki, 1991) that gp96 may not be immunogenic per se, and possess an Mg2+-dependent ATPase activity. Gp96 In view of the but may be a carrier of antigenic peptides. are also observed to contain tightly bound preparations localization of in the ER and Koch, predominant gp96 (Booth extraction. These peptides, which can be eluted by acid immunity elicited by gp96 1989) and our observation that the of with its proposed roles properties gp96 are consistent T lymphocytes (Udono and is mediated through CD8+ in and in facilitaing MHC chaperoning antigenic peptides that acts as a Srivastava, 1993), we proposed gp96 peptide in the ER lumen. We present class I-peptide assembly to the ER and enables acceptor for peptides transported a model to how interaction of gp96 with MHC explain of MHC class I. peptide loading class I result in transfer of peptides to the latter. may In of the fact that peptide charging of MHC class light Key words: cancer immunity/chaperone/hsp9O/MHC class I an (Levy et al., 1991; Luescher is ATP-dependent process I/peptide et we have examined the ability of gp96 to al., 1992), and ATP. We demonstrate that are bind peptides peptides associated with gp96 and that gp96 binds ATP and indeed observations make gp96 a logical is an ATPase. These Introduction for facilitating peptide charging of MHC class candidate of tumors was first demonstrated con- Immunogenicity a structural basis for the tumor-specific I and provide in sarcomas of inbred mice vincingly chemically induced of gp96. immunogenicity Prehn and 1957; Klein (Gross, 1943; Foley, 1953; Main, et Old et al., 1962). A search for the molecules al., 1960; which mediate this individually distinct immunogenicity led Results to identification of gp96 molecules as tumor rejection in sarcomas is an antigens (TRAs) methylcholanthrene-induced Gp96 ATP-binding protein of inbred BALB/c mice (Srivastava et al., 1986; Palladino The deduced amino acid sequence of gp96 was screened Mice et and Srivastava, 1993). for the of motifs which are associated with al., 1987; Feldweg presence became immune immunized with 5-10 of purified gp96 ATP binding. The ATP-binding consensus sequences, types jig with the tumor from which et al. and refined to gp96 A and B as proposed by Walker (1982) subsequent challenges but not to with antigenically distinct One A 217- was isolated, challenges by Chin et al. (1988), were used. type (aa of a with an anti- B 231-241 tumors. Immunodepletion gp96 preparation 224) and two downstream type sequences (aa to it of were identified Within the A gp96 antibody gp96 depletes antigenicity (Srivastava and 303-313) (Figure 1). et Rabbit antisera revealed the presence of gp96 there are three amino acid residues between al., 1986). type sequence, in normal tissues as well et al., 1986). However, and instead of four as proposed in the (Srivastava Gly217 Gly221, does not elicit tumor B isolated from normal tissues A consensus sequence. The two type sequences gp96 type in and These observations have one mismatch each, the hydrophobic regions. Srivastava, 1993). immunity (Udono an additional are mutations in of tumors Flaherty et al. (1991) have suggested sequence that there gp96 genes suggested as a mutations differ from one tumor to another. fingerprint and that these (ILV)X(ILVC)DXG(TSG)(TSG)XX(RKC) Oxford Press University Z.Li and P.K.Srivastava A B Type Type Consensus (G/A)X4(G/A)(HEKIR)Xo-s(T/S/K/R/H) (HEKfR)X5-s8Xcb2(D/EE) GP96: 217-224 T - NTL GR G T 231-241 K Y D EEASD L eL 303-313 K V dW E KVEKT W Adenylate Kinase: 15-23 G GPGS GK G T 108-119Y3 'K L L D IGQPTL RING 4: C-terminal G PNGS GK S T C-terminal K PCVLILDD a S tL D Protease La: 355-362 G PPGV GK T 411-422 VGVKNP L F LL D GroEL: 164-172 A MDKV 241-251 AGKD L IA E LI Fig. 1. Sequence homology among the putative ATP domains of binding gp96 and other ATP-binding proteins. The consensus ATP binding sites proposed by Walker et al. (1982) and modified by Chin et al. are (1988) presented. The putative ATP binding sequences of are gp96 aligned with sequences of four other ATP rabbit muscle binding proteins: adenylate kinase (Kuby et human 1 al., 1984), transporter-associated protein (RING 4) (Trowsdale et al., 1990), E.coli chaperonin groEL (Hemmingsen et al., 1988) and protease La (Chin et amino al., 1988). Single acid codes are used here. F stands for a hydrophobic amino acid (I, V, L, M, Y, W, X F); indicates any amino acid. Lower case letters represent non-homologous residues. indicates a gap. The highly conserved residues are boxed. characteristic for on nucleotide-binding proteins, based the structure of the heat shock hsc7O. This crystal protein motif is in conserved hsc7O, grp78 and dnaK from Escherichia but is not found in the coli, ATP-binding protein groEL. This sequence was not detected in gp96. ATP was binding by gp96 measured experimentally by photoaffinity labeling of purified gp96 by [Ly-32P]8-azido ATP In ATP (Figure 2, lane 1). parallel, of binding hsp90 a lane and an (as positive control, unrelated 96 kDa 2) protein (as a negative control, lane 3) was also tested. Gp96 ATP was observed to bind (lane 1) and this was binding of mM ATP inhibited by inclusion cold 10 8-azido (lane 4), ATP but not AMP GTP (lane 5), cyclic (lane 6), (lane 7) Fig. 2. Gp96 binds ATP in vitro. Binding of gp96 (lane 1), hsp9O or CTP (lane 8). (lane 2) or p96 (a Meth A-derived protein unrelated to gp96, lane 3) to was carried out as described in To determine whether interacts with ATP in [_y-32P]8-azido Materials and gp96 vivo, methods. Lanes 4-8: binding was carried out in the presence of EL-4 with 32P cells were labeled and was immuno- gp96 10 mM unlabeled 8-azido ATP (lane 4), ATP (lane 5), cyclic AMP from whole cell of such cells. precipitated lysates A (lane 6), GTP (lane 7) and CTP (lane 8). phosphate-labeled band was observed gp96 (Figure 3A). These conditions should detect experimental only phos- phorylated or or with ADP-ribosylated proteins, proteins bound nucleotides. In two other tightly fact, Gp96 elutes within a broad range of salt phosphoproteins, concentration MHC class I and were also from these hsp70, precipitated (0.45-0.6 M NaCl), instead of a sharp peak. However, extracts with the antibodies and were detected respective by the protein material in each peak comprises apparently In the case of a number of bands homogeneous as autoradiography. hsp7O, gp96 judged by silver stained gels of these other than the 70 kDa band were detected in the immuno- fractions (Figure 4B). The fractions were individually precipitate; these proteins are associated with assayed for ATPase presumably activity as described in Materials and hsp70 because the used is methods. ATPase anti-hsp7O antibody quite specific activity was clearly detected in the gp96 and detects only hsp70 on Western blots. fractions (Figure 4C). The activity in this experiment is not To determine whether the 32p label in derives linear with the concentration of gp96 gp96 because this protocol from associated ATP, or whether it is a result of a measures net activity rather post- than the initial velocity of translational modification, radioactive adenosine nucleotides reaction. For subsequent characterization of the ATPase were eluted from specifically the activity of gp96, fraction immunoprecipitates by 6 was used. large excess of cold ADP and AMP. It was ATP, observed To confirm that the ATPase activity was derived from (Figure 3B) that and but not MHC class gp96 hsp70, I, gp96 molecule itself, fraction 6 (from the experiment in contain tightly bound ADP and AMP. ADP is the Figure was ATP, 4A) applied sequentially a number of times, to major component eluted from and ATP an gp96 hsp70; anti-gp96 monoclonal antibody column and depleted of contributes to only a minor fraction of the total eluted gp96. The initial and depleted fractions were characterized nucleotides. This is consistent with the observations made by SDS -PAGE and assayed for ATPase activity. It was with dnaK and the crystallographic analysis of hsp70 observed that the loss of ATPase activity was concomitant (Flaherty et al., ATP is not in 1990). detected abundance with loss of gp96 (Figure 5A, B and C). Application of presumably because of its rapid hydrolysis. fraction 6 through an unrelated immunoaffinity column did not deplete the preparation of gp96, nor of ATPase Gp96 is an ATPase activity (data not shown). These experiments show that the 4A shows the Figure chromatographic profile of the last ATPase step activity of gp96 resides in the gp96 itself and is of of Meth A purification gp96 on a Mono FPLC column. not a Q contaminant. 3144 Tumor an ATPase rejection antigen gp96 is /If: , I.' ;u *Ill, A ,:, 'I-N in :...nher ....: t P, KI)9 ,_ ue- P s Ar-T*R., -4-- AII ., W. | ADII Fig. 3. Interaction of gp96 with adenosine nucleotide in vivo. (A) Lmmunoprecipitation of phosphate-labeled gp96 from [32P]orthophosphate-labeled EL-4 cells. Immunoprecipitation was done with a polyclonal rabbit serum against MHC class I whole molecule 4r2 X I I I I I .4 i (K270, obtained from Dr Per Petersen), rat anti-gp96 monoclonal Y\1P 40-- antibody (clone 9G10) and rat anti-hsp7O monoclonal antibody (clone 7.10), respectively, followed by secondary rabbit anti-rat IgG and protein A-Sepharose. (B) PEI thin layer chromatography of gp96-associated nucleotides. Immunoprecipitates were washed and Fig. 4. ATPase co-purifies with gp96. (A) Chromatographic activity associated nucleotides were eluted and separated by PEI TLC plates profile of gp96 on MonoQ FPLC. Proteins bound to concanavalin (see Materials and methods). yu of 100 ATP, ADP and AMP A-Sepharose column were dialyzed against 5 mM sodium phosphate jtM were spotted and localized with the aid of a short-wave UV lamp and buffer, pH 7.0, and applied to MonoQ FPLC column. Bound proteins the are positions indicated by arrows. The spot seen in the MHC class were eluted with a linear 0-1 M NaCl gradient. Gp96 elutes at NaCl I lane on the TLC of the in represents <5% counts seen the gp96 or concentration between 0.45 and 0.6 M. (B) Silver stained hsp70 lanes. The films were exposed to a non-linear degree in order SDS-polyacrylamide gel of MonoQ FPLC-purified fractions shown in to show the ATP spots clearly. (A). Fractions were applied to 10% SDS-PAGE and silver stained. Lanes 1-10 represent 3 1d aliquot of each 1 ml fraction. (C) In vitro ATPase activity of purified gp96. 10 of gradient fractions shown in A1 Effect of temperature, pH and divalent cations on the (A) were assayed for ATPase activity using 2 [a-32P]ATP (see liM Materials and methods). Reaction products were analyzed by PEI thin ATPase activity of gp96 layer chromatography and autoradiography. The positions of ATP, ATPase activity of gp96 was assayed at different temperatures ADP and AMP are shown. at a pH of 7.2 and was found to be optimal at 42°C (Figure 6A). A temperature optimum higher than 37°C has been noted for some other HSPs as well; order kinetics for example, dnaK, a (Figure 7 and inset), suggesting that gp96 has bacterial hsp7O, shows optimal a ATPase activity at 50-55°C single ATPase active site. No evidence of co-operativity (Liberek et al., 1991). Furthermore, was the ATPase activity of observed. Under the conditions of this experiment, the gp96 was observed to be ATPase has optimal at acidic pH (Figure 6B). activity of gp96 a Km of 8 and the turnover AM The divalent cation dependence of rate is -0.08 mol/min/mol. These gp96 ATPase was tested characteristics suggest and the activity was found to be that ATP a dependent on exogenous hydrolysis by gp96 role plays regulatory (see but not on Mg2+ Ca2+ (data not shown). This was con- Discussion). firmed by the use of the divalent cation chelators EDTA and EGTA in the reaction mixture. EDTA, which is a Protein and modulators of ATPase general peptide activity of divalent cation chelator, had a and significantly stronger gp96 hsp70 inhibitory effect on the ATPase activity of gp96 than EGTA, Members of the hsp70 family, including grp78, undergo which is a Ca2+-specific chelator (Figure 6C). The other conformational and substrate release ATP changes during ER luminal HSP, grp78, is similar to gp96 in its requirement hydrolysis (Carlino et al., 1992; and Gething Sambrook, for divalent cations (Kassenbrock and Kelly, 1989): grp78 1992). Moreover, ATP hydrolysis is stimulated by requires Mg2+ but not Ca2+ for its ATPase activity and in- exogenous peptides et As and (Flynn al., 1989, 1991). gp96 the are the two ER we deed, activity is inhibited by the presence of Ca2 . grp78 luminal HSPs, investigated whether A substrate ATPase of is also modulated saturation experiment was done to determine activity gp96 by peptides. Two the kinetics of ATPase of A and which were used for ATPase activity gp96. When the ATPase peptides B, stimulation activity was assayed in the presence of with et al. were added in increasing con- assays grp78 by Flynn (1989), the centrations of the ATPase with As shown in the ATP, velocity was found to follow first assay gp96. Figure 8A, 3145 ZLi and P.K.Srivastava 16 - .. , 12 .14 .;,A i. * 8~ 9 , 10W )' 10 20 30 50 60 70 4+-- AIP Temperature (OC) *. ...,=e", = 30- .: E 25 20- 3 15 t3 ., ' i ;. L an I' iI 3 4 5 6 7 8 9 in molecule Flg. 5. ATPase activity resides the itself. gp96 pH (A) Depletion of gp96 using column. Monoclonal immunoaffinity antibody was coupled to protein column. The same G-Sepharose volume of original materials (-0.5 and materials from the eluting Ag) column each time were to 10% SDS-PAGE and silver C applied stained. (B) Samples before (0.5 and after were 22 - depletion assayed itg) for ATPase activity by with 20 for 30 min. incubating [ca-32P]ATP /tM in (C) Quantitative analysis of results 30 ATP were 20- (B). pmol 0.5 before the hydrolyzed by gp96 immunodepletion and the A jig relative ATPase was defined as 1. activity oa :0 16 - o9 ATPase activity of gp96 was not stimulated and was indeed h. el 14 - inhibited in the presence of either of the two exogenous I- peptides. Fifteen additional from 7 to 20 peptides ranging 12 - in residues length were also found to have no stimulation effect. An hsp7O preparation isolated from Meth A cells was 10- also tested for ATPase activity and its stimulation S by peptides A and B under our experimental conditions. In contrast to .0001 .001 .01 .1 1 10 100 lack of the stimulation of gp96 ATPase by peptides, hsp70 Concentration ATPase was - of EDTA/EGTA (mM) reproducibly found to be stimulated 1.5-fold addition of by peptides (Figure 8B). Flynn et al. have Fig. 6. Effect of temperature, pH and divalent cations on the ATPase observed a maximal 4-fold stimulation of the ATPase activity activity associated with gp96. Effect of on (A) temperature the rate of of grp78 by the same under similar peptides conditions. The ATP hydrolysis by gp96. (B) Optimization of pH. 570 ng purified difference between our results stems perhaps from the fact gp96 were assayed for ATPase in the of activity presence 20 mM that we are using a mixed of molecules pool hsp70 rather NaCl, 2 mM MgCl2 and 20 mM of various buffers 4.0 acetate, (pH than a pH 5.5-6.0 MES, pH 7.0-7.2 HEPES, pH purified grp78 preparation. 7.5-8.8 Tris). ATPase is (C) activity inhibited by divalent cation chelators. Purified It has been suggested that the stimulation of ATPase was for gp96 assayed ATPase activity in the presence of 20 mM of activity grp78 by peptides reflects the fact that peptides MES, pH 20 mM NaCl 6.0, and increasing concentrations of EDTA may mimic certain features of the newly synthesized un- or EGTA (solid line) (dotted line). folded proteins, whose folding is facilitated by grp78. The stimulation of ATPase activity would thus or provide energy so-called 'folding intermediates' or 'molten globules' (for for the folding process. However, in contrast to grp78, see the reviews, Ang et al., 1991; Gething and Sambrook, bacterial chaperonin groEL (which also hydrolyzes ATP) 1992). Thus, groEL does not recognize either completely has been shown to interact only with partially folded proteins folded or totally unfolded polypeptides, nor does it bind 3146 Tumor rejection antigen gp96 is an ATPase O. _ ~~~o C', 00- _> -5 0~~~~~~~~ I-10 50 60 C 100 200 300 400 500 600 <t ATP -10 Concentration (gM) 400 800 1200 1600 Peptide Concentration (IM) Fig. 7. Kinetic study of gp96 ATPase activity. Substrate saturation experiment was carried out by incubating purified gp96 (0.862 per jig reaction) with increasing concentrations of [ax-32P]ATP in 20 mM HEPES buffer, pH 7.2, containing 20 mM NaCl and 2 mM MgCl2 at 37°C for h. Background values in the absence of gp96 have been _ 30 subtracted. Double reciprocal plot of the same data is shown in inset. cm The Km calculated from this experiment is 8 jiM. The that the ATPase activity of gp96 peptides. possibility Z 20 is modulated by similar interactions was therefore pursued and casein was tested for this dephosphorylated purpose. <' 15 ._1 Casein is a soluble with a number of exposed protein hydrophobic surfaces and has been used as a competing o 10 substrate for groEL binding (Langer et al., 1992). It was observed that similar to the stimulation of hsp70 ATPase x 5 casein stimulates the ATPase of by peptides, activity gp96 IL I- -2-fold by (Figure 9). 400 600 800 1000 1200 1400 0 200 is associated with Gp96 peptides Peptide Concentration (jiM) To identify peptides associated with gp96, 109 Meth A cells labeled were metabolically with [35S]methionine of ATPase of and Fig. 8. Modulation activity gp96 hsp70 by peptides. and was isolated. The purified (200 gp96 stimulate ATPase of 1 of ACi/ml) (A) Peptides do not activity gp96. jig preparation was extracted with 10% acetic acid and the low was for ATPase in the of purified gp96 assayed activity presence concentrations of either A or molecular weight material isolated by centricon centrifuga- increasing peptide (A-A) peptide B at 370C for 1 h. Both peptides A and B are 15mers and (O-----0) tion which will allow of molecules (centricon-10, passage are derived from vesicular stomatitis virus glycoprotein (peptide A, of 10 000 Da or as described in Materials and less), KRQIYTDLEMNRLGK; peptide B, LSSLFRPKRRPIYKS). Peptide- methods. This material was applied to a reverse phase C18 ATP of 50 was subtracted. independent hydrolysis pmol/h/ljg HPLC column and eluted on an acetonitrile gradient. stimulate ATPase of The same (B) Peptides activity hsp70. peptide ATP preparation used in (A) was used. Peptide-independent hydrolysis Individual fractions were collected and counted. A number of 85 has been subtracted. pmolh/jtg of [35S]methionine-labeled distinct peptide peaks were obtained (Figure 10A). However, as this material was insufficient for quantitative analysis, we purified 3 mg gp96 for acid extraction or it may be stored in this high salt buffer from mammalian liver and acid-extracted it as before. The for several hours in the presence of high concentrations of mM peptides eluted from it were applied to a column. protease inhibitors (2 PMSF, 10 leupeptin) at C18 AM before acid extraction. In either we Figure lOB shows that a number of peptide peaks were 4°C, case, routinely the in obtained. Some of the peaks have been partially characterized centrifuged purified preparations the high salt buffer a acid extraction by mass spectroscopy and have been determined to be through centricon-1O immediately preceding masses between and did not find in the low molecular peptides of heterogeneous size (molecular any peptides weight not An fraction. contains tightly bound peptides, 400 and 2000 atomic mass units, data shown). Thus, gp96 of the eluted apparently in the same manner and by the same criteria as approximate quantitative analysis peptides that and are in the MHC class I and class II molecules et suggests gp96 peptides present roughly (Falk al., 1990; that the Rotzschke et al., 1990; Rotzschke and Falk, 1991). equimolar stoichiometry, assuming average peptide molecular mass of 1000 Da. It should be noted that has a the eluted are not associated with peptides merely loosely Discussion but are bound to it. The purified gp96 gp96, tightly ATPase of and other HSPs was obtained after a number of that include activity gp96 preparation steps is the most abundant of the elution on a salt on which elutes at - 0.5 M Gp96/grp94 single component gradient gp96 of the but no in vivo function or final be used lumen ER, enzymatic NaCl. The gp96 preparation may immediately 3147 Z.Li and P.K.Srivastava I 0 4) 1- co 0 1 2 3 4 5 6 7 Casein Concetrtion (mni) fractions Fig. 9. Casein stimulates ATPase activity of gp96. ATPase activity of was in the gp96 assayed presence of increasing concentrations of casein. ATPase with alone well activity gp96 as as the minor background ATP hydrolysis associated with casein itself are subtracted from each point. U-) has been to it as activity assigned yet (Pelham, Our 1986). results show novel of this molecule and properties suggest rn -I4 for it a crucial role in MHC antigen presentation by class 60 I molecules and in tumor We demonstrate that immunity. binds ATP and is an ATPase. gp96 Mg2+-dependent The deduced amino acid of contains sequence gp96 ATP-binding LI) sn o motifs and gp96 binds ATP in vitro and in vivo 2 (Figures The Km of the ATPase was calculated as and 3). activity 8 and the maximum as 50 velocity (Figure ytM pmol/h/tg 7). This to a turnover rate of 0.08 mol/min/mol. corresponds 0 8 16 24 32 40 48 These characteristics of the ATPase fall within the FRACTION NUMBER gp96 range of for the ATPases. corresponding parameters hsp70 Fig. 10. is For example, the ATPase of the E. coli dnaK Gp96 associated with heterogeneous peptides. Purified activity hsp70 gp96 was extracted with acetic acid. 10% Peptides eluted from gp96 were has a Km value of 20 and a turnover rate of /.M separated through and centricon-10 separated on a C18 HPLC column, 0.2 mol/min/mol (Liberek et an al., 1991). Grp78, hsp70 as described in Materials and methods. (A) Peptides eluted from gp96 member of the ER, has a Km of 0.1 tM and a turnover derived from 1 x 109 Meth A cells metabolically labeled with rate of -0.34 mol/min/mol (Kassenbrock and Kelly, 1989). [35S]methionine. (B) Peptides isolated from liver macaque gp96. A comparison of the turnover rates of ATPase activities of of the ER. We suggest that the two molecules use the gp96, dnaK and grp78 with the values for energy corresponding from hydrolysis of ATP in divergent ways, as a non-HSP such as kinase is The discussed adenylate revealing. below. turnover rate of the ATPase of rabbit muscle activity adenylate kinase is 2 106 mol/min/mol from (calculated and in Role in Hampton Slotin, 1975); comparison, the turnover rate of gp96 folding/assembly of proteins in the ER of HSP as ATPases, described above, is several orders of Grp78 is generally thought to correct catalyze folding of magnitude lower. This indicates that ATP hydrolysis by newly synthesized proteins in the lumen of the ER. The role HSPs has a of in significantly dissimilar biological purpose from gp96/grp94 this process has remained largely obscure. ATP hydrolysis by non-HSPs such as Similar to adenylate kinase and grp78, grp94 has been shown to be induced by the transporter the of accumulation of proteins; specifically, ATPase activity malfolded proteins in the ER (Kozutsumi HSPs may mediate a regulatory rather than a et al., 1988) suggesting that transporting gp96 acts as a chaperone. A or modifying function. number of observations have now begun to show this A of comparison the ATPase activities of the two HSP formally. In our analysis of the assembly of MHC class I ATPases of the ER lumen, with (32 grp78 and gp96, reveals critical microglobulin and peptides, we have observed differences. The ATPase of association of activity grp78 is stimulated by MHC class I heavy chain with gp96 (Z.Li peptides, while that of gp96 is inhibited or and in unaffected. P.K.Srivastava, preparation). In earlier studies, Furthermore, ATPase activity of grp78 is unaffected by the Navarro et al. (1991) observed that grp78 and gp96 associate presence of other proteins such as casein, while the ATPase with aberrant forms of glycoprotein b of the herpes simplex activity of gp96 is stimulated by casein. These differences virus 1. Similar observations were reported for newly hint towards different perhaps complementary roles for the synthesized immunoglobulin chains by Melnick et al. (1992). two proteins in the ER lumen. Clairmont et al. (1992) have Schaiff et al. (1992) showed that MHC class II (HLA-DR) demonstrated that and are the grp78 gp96 two major oa and chains associate with gp96 in the ER in the absence recipients of the pool of ATP translocated into the lumen of invariant chain. This association does not occur in the 3148 Tumor rejection antigen gp96 is an ATPase presence of invariant chain. The observations reported in this paper, that gp96 molecules bind peptides and ATP and have a regulatory ATPase activity, provide a missing piece of the puzzle and furnish an enzymatic basis for the emerging evidence for the role of gp96 as a bonafide chaperone. There are two main paradigms known for the mechanism -1 AD- AM by which HSP chaperones catalyze protein folding or assembly (Rothman, 1989; Gething and Sambrook, 1992). In one paradigm, typified by grp78, cycles of binding and release of grp78 with short peptide regions of the substrate protein are propelled by the energy provided by ATP Fig. 11. A model for the role of gp96 in MHC class I peptide hydrolysis and a polypeptide chain emerges fully folded after assembly. It is postulated that the interaction of MHC class I and gp96 having gone through a number of transient associations. In stimulates the ATPase activity of gp96. The free energy (AG0) released due to ATP hydrolysis induces a conformational change in concordance with this paradigm, ATPase activity of grp78 gp96, leading to transfer of peptides from gp96 to MHC class I. TAP is stimulated by peptides, but not by folded intact proteins denotes transporter-associated proteins. (Flynn et al., 1989, 1991). In another paradigm, typified by the groE proteins, protein substrates bind to the groEL oligomer at a number of sites and the folding proceeds at Gp96-accessory to peptide-loading of MHC class I? different portions of the polypeptide substrate by a We demonstrate that gp96 is associated with a wide array progressive, ATP hydrolysis-dependent release of different of peptides (Figure 10). In the context of our observation regions of the substrate from the oligomeric groEL (Gething of co-immunoprecipitation of gp96 and MHC class I from and Sambrook, 1992). Another groE protein, groES, metabolically labeled cells (Z.Li and P.K.Srivastava, in modulates the ATPase activity of groEL. In concordance preparation) and in light of the ATPase activity of gp96, we with this paradigm, the ATPase activity of groEL is propose (Figure 11) that (i) gp96 acts as a repository of the stimulated by interaction with some proteins, but is not peptides transported into the lumen of the ER, (ii) interaction stimulated by peptides. of gp96 with MHC class I stimulates its ATPase activity, The ATPase activity of gp96 appears to be distinct from resulting in hydrolysis of gp96-bound ATP, (iii) the energy both paradigms, although it shares selected features with released from ATP hydrolysis is used for transfer of peptides each. Gp96 molecules may exist as dimers or tetramers from gp96 to MHC class I. (Srivastava and Das, 1984; unpublished) and like the groEL, An ER luminal chaperone which will facilitate charging and in contrast to grp78, the gp96 ATPase is not stimulated ofMHC class I with peptides has been invoked by Rothman, by peptides, but is stimulated by casein-a good model for Cresswell, Townsend, Kvist and their colleagues (Alexander a et partially folded protein. It would appear that the ATPase al., 1989, 1990; Rothman, 1989; Townsend et al., 1990; activity of gp96, like that of dnaK and groEL, is modulated Levy et al., 1991). In view of the ability of gp96 to bind by protein-protein rather than protein-peptide interactions. peptides and ATP, its ATPase activity, and its localization It is conceivable ER in the ER that the lumen contains other proteins, lumen (where it is the major component), it is which modulate the ATPase activity of gp96 in the same reasonable to suggest this role for gp96. manner as An ER groES modulates the activity of groEL, or DnaJ chaperonin p88 -IP90-calnexin has recently been and GrpE modulate the activity of dnaK. Gp96 and grp78 identified (Degen and Williams, 1991; Wada et al., 1991; also may conceivably modulate and collaborate with each Galvin et al., 1992; Hochstenbach et al., 1992). It is a other. Sequential collaboration of the two chaperones dnaK membrane-bound (as opposed to luminal) protein and is and in associated groEL the folding process is a precedent for such with partially but not fully assembled com- an of T effort (Langer et al., 1992). plexes immunoglobulins, cell receptors and MHC. While and have shared roles in the Hochstenbach et al. that gp96 grp78 may folding (1992) suggest p88 might participate the of the two ATPases in of multi-subunit It is that process, divergent regulation assembly complexes. possible 8 and an role is involved in of MHC class I (Figures 9) may suggest additional, unique p88 folding heavy chain, but for is the most evolved HSPs. there is no evidence of or ATPase gp96. Gp96 among recently peptide binding activity It is not found in or and to have of p88 and it does not to be an to yeasts Drosophila appears appear accessory peptide a of the loading of MHC class I. emerged relatively recently by duplication cytosolic hsp90. The major structural differences between and hsp90 the amino-terminal and Implications for immune to cancer gp96, signal peptide carboxy- response terminal KDEL mediate ER and reten- The observations that gp96 is associated with binds sequence, targeting peptides, tion. The additional difference between the two ATP and is an ATPase lends to the only strong support suggestion of molecules is in the ATPase that of activity gp96: although hsp90 tumor-specific immunogenicity gp96 (and perhaps would that from but from binds it is not an ATPase. It other HSPs) does not derive ATP, appear gp96 directly gp96 function in the associated with it. The source of the evolved to fulfill an peptides immunogenic ATPase-dependent ER, In of the from tumor cells lies in altered which may be unique to higher organisms. light peptides clearly (mutated) effects of vaccination with or cellular and the of immunological gp96 (Srivastava mis-expressed proteins specificity and Srivastava et Palladino et result from randomness of mutations Das, 1984; al., 1986; immunogenicity may al., et and and our observation et Szikora 1987; Feldweg Srivastava, 1993) (Sibille al., 1990; Srivastava, 1990; al., 1990; I Structural that gp96 associates with MHC class (Z.Li and Van den Eynde et al., 1991; Boon, 1992). that one of of eluted from in we the characterization antigenic P.K.Srivastava, preparation), suggest peptide(s) gp96 of MHC class from distinct tumors will functions of involves preparations antigenically specific gp96 charging provide further for this I molecules with support hypothesis. peptides. 3149 Z.Li and P.K.Srivastava Immunogenetics, 31, 169-178. Materials and methods Ang,D., Liberek,K., Skowyra,D., Zylicz,D. and Georgopoulos,C. (1991) J. Bio. Chem., 266, 24233-24236. Reagents, mice and tumors Boon,T. (1992) Adv. Cancer Res., 58, 177-210. All chemicals were purchased from Sigma except where specified. Booth,C. and Koch,G.L.E. (1989) Cell, 59, 729-737. Radiochemicals were from ICN. Cell culture media and reagents were mainly Carlino,A., Toledo,H., Skaleris,D., Delisio,R., Weissbach,H. and Brot,N. obtained from GIBCO-BRL. Monoclonal antibodies against gp96 and hsp7o (1992) Proc. Natl Acad. Sci. USA, 89, 2081-2085. were purchased from Stressgen. Peptides were synthesized on an Applied Chin,D.T., Goff,S.A., Webster,T., Smith,T. and Goldberg,A.L. (1988) Biosystems Model 430A peptide synthesizer using FMOC chemistry. Inbred J. Biol. Chem., 263, 11718-11724. BALB/c mice were from our mouse colonies. EL-4 cells were obtained Clairmont,C.A., De Maio,A. and Hirschberg,C.B. (1992) J. Biol. Chem., obtained from ATCC. 267, 3983-3990. Degen,E. and Williams,D.B. (1991) J. Cell Biol., 112, 1099-1115. and Purification of gp96 hsp7O Downward,J., Graves,J.D., Warne,P.H., Rayter,S. and Cantrell,D.A. Gp96 purification has been described by Srivastava et al. (1986). Hsp,70 (1990) Nature, 346, 719-723. was purified essentially as described by Welch and Feramisco (1985). Falk,K., Rotzschke,O. and Rammensee,H.G. (1990) Nature, 348, 248-251. Feldweg,A.M. and Srivastava,P.K. (1993) J. Cell. Biochem., 17D, Abstract ATP binding no. NZ206. Purified gp96, hsp90 or other protein (10 yg) in 10 mM Tris, pH 7.4, 5 mM Flaherty,K.M., DeLuca-Flaherty,C. and McKay,D.B. (1990) Nature, 346, and 2 mM ATP in a final volume of 100 1l was CaCl2 [y_-32P]8-azido 623-628. irradiated with a short wavelength (254 nM) UV source from a distance Flaherty,K.M., McKay,D.B., Kabsch,W. and Holmes,K.C. (1991) Proc. of 8-10 cm for 4 min (Hobson et al., 1984). Samples were then Natl Acad. Sci. USA, 88, 5041-5045. microcentrifuged for 5 min to remove large complexes and analyzed Flynn,G.C., Chappell,T.G. and Rothman,J.E. (1989) Science, 245, by SDS-PAGE. 385-390. Flynn,G.C., Pohl,J., Flocco,M.T. and Rothman,J.E. (1991) Nature, 353, A TPase assay 726-730. ATPase activity was measured by the method of Flynn et al. (1989, 1991). Foley,E.J. (1953) Cancer Res., 13, 835-837. Typically, 1 of purified gp96 or hsp7o was incubated with 20,M Ag Galvin,K. (1992) Proc. Natl Acad. Sci. USA, 89, 8452-8456. in a reaction volume of 20 containing 20 mM HEPES, pH [a-32P]ATP Al Gething,M.-J. and Sambrook,J. (1992) Nature, 355, 33-45. 7.2, 20 mM NaCl and 2 mM MgCl2 at 37°C for 1 h. 1 1l of the reaction Gross,L. (1943) Cancer Res., 3, 323-326. mixture was then spotted onto a polyethyleneimine (PEI) cellulose plate. Hampton,A. and Slotin,L.A. (1975) Biochemistry, 14, 5438-5444. Thin layer chromatography was performed against 1: 1 ratio of 1 M LiCl Hemmingsen,S.M., Woolford,C., Van der Vies,S.M., Tilly,K., and 1 M HCOOH. The plate was then dried, exposed to film and Dennis,D.T., Georgopoulos,C.P., Hendrix,R.W. and Ellis,R.J. (1988) corresponding radioactive spots were excised and counted. ATPase activity Nature, 333, 330-334. was determined from the amount of [32P]ADP and [32P]AMP generated Hobson,A.C., Weatherwax,R. and Ames,G.F.-L. (1984) Proc. NatlAcad. from [a-32P]ATP, i.e. the percentage of ATP hydrolyzed calculated as Sci. USA, 81, 7333-7337. + + AMP + x 100%. Background ATP [ADP AMP]/[ATP ADP] Hochstenbach,F., David,V., Watkins,S. and Brenner,M.B. (1992) Proc. hydrolysis lacking purified gp96 or hsp70 was subtracted. NatlAcad. Sci. USA, 89, 4734-4738. Kassenbrock,C.K. and Kelly,R.B. (1989) EMBO J., 8, 1461-1467. immunoprecipitation Phosphate labeling and Klein,G., Sjogren,H.O., Klein,E. and Hellstrom,K.E. (1960) CancerRes., Labeling cells with 2p was done according to Downward et al. (1990). 20, 1561-1572. was carried out using the Townsend protocol Immunoprecipitation Kozutsumi,Y., Segal,M., Normington,K., Gething,M.-J. and Sambrook,J. nucleotides, immunoprecipitates were (Townsend et al., 1990). To elute 332, 462-464. (1988) Nature, with 2 mM 2 mM 0.5 mM ATP, incubated EDTA, DTT, 0.2% SDS, Kuby,S.A., Palmieri,R.H., Frischat,A., Fischer,A.H., Wu,L.H., Maland,L. ADP mM AMP for of 0.5 mM and 0.5 at 68'C 20 min. Separation eluted and Manship,M. (1984) Biochemistry, 23, 2393-2399. nucleotides was on PEI-cellulose chromatography. plates Lu,C., Echols,H., Flanagan,J., Hayer,M.K. and Hartl,F.U. Langer,T., (1992) Nature, 356, 683-689. - silver and SDS PAGE, Western blotting, staining Lee,A.S., Bell,J. and Ting,J. (1984) J. Biol. Chem., 259, 4616-4621. immunoaffinity protein G columns Levy,F., Gabathuler,R., Larsson,R. and Kvist,S. (1991) Cell, 67, 265 -274. Western was done to the HRP color developing system blotting according Liberek,K., Marszalek,J., Ang,D., Georgopoulos,C. and Zylicz,M. (1991) Silver and G immunoaffmiity chromatography (Bio-Rad). staining protein Proc. Natl Acad. Sci. USA, 88, 2874-2878. were carried out using standard protocols. Luescher,I.F., Loez,J.A., Malissen,B. and Cerottini,J.C. (1992) J. Immunol., 148, 1003-1011. extraction and separation Peptide Maki,R.G., Old,L.J. and Srivastava,P.K. (1990) Proc. NatlAcad. Sci. USA, Isolation of from purified gp96 was done by the acid extraction peptides 87, 5658-5662. technique (Van Bleek and Nathenson, 1990). The low molecular weight Maki,R.G., Eddy,R., Byers,M., Shows,T.P. and Srivastava,P.K. (1993) materials were then enriched by separating out the high molecular weight Som. Cell Mol. Genet., in press. proteins using a centricon-10 spin column (MW cut off 10 000 Da, Amicon) Mazzarella,R.A. and Green,M. (1987) J. Biol. Chem., 262, 8875 -8883. according to the manufacturer's instructions. The low MW material isolated Melnick,J., Aviel,S. and Argon,Y. (1992) J. Biol. Chem., 267, from gp96 was analyzed by directly injecting into HPLC (Applied 21303-21306. BioSystems) and further resolved using a column and acetonitrile C18 Navarro,D., Qadri,I. and Pereira,L. (1991) Virology, 184, 253-264. 1 B, gradient with a flow rate of ml/min (buffer A, 0.05% TFA; buffer Old,L.J., Boyse,E.A., Clarke,D.A. and Carswell,E.A. (1962) Ann. NY the is linear from 0 to 100% buffer B). 1 ml 100% acetonitrile, gradient Acad. Sci., 101, 80-106. fractions were collected. Palladino,M.A., Srivastava,P.K., Oettgen,H.F. and Deleo,A.B. (1987) Cancer 47, 5074-5079. Res., Acknowledgements Pelham,H.R.B. (1986) Cell, 46, 959-961. Prehn,R.T. and Main,J.M. (1957) J. Natl Cancer Inst., 18, 769-778. We thank Nathalie Blachere for purification of the macaque gp96, Anya Rothman,J.E. (1989) Cell, 59, 591-601. for with the in ATP experiments and Dr Reza Harry help vitro binding Rotzschke,O. and Falk,K. (1991) Immunol. Today, 12, 447-455. Green for the was supported by critically reading manuscript. This work Rotzschke,O., Falk,K., Deres,K., Schild,H., Norda,M., Metzger,J., NIH CA44786. P.K.S. is an Investigator of the Cancer Research grant Jung,G. and Rammensee,H.G. (1990) Nature, 348, 252-254. New York, and an Irma T.Hirschl Scholar. Institute, Schaiff,W.T., Hruska,K.A., Jr, McCourt,D.W., Green,M. and Schwartz,B.D. (1992) J. Exp. Med., 176, 657-666. Shaknovich,R., Shue,G. and Kohtz,D.S. (1992) Mol. Cell. Biol., 12, References 5059-5068. Sibille,C., Chomez,P., Wildman,C., vanPel,A., DePlaen,E., MaXyanski,J., and and Boon,T. Alexander,J., Payne,A., Murray,R., Frelinger,J. Cresswell,P. (1989) DeBergeyck,V. (1990) J. Exp. Med., 172, 35-45. 380-388. Curr. Immunogenetics, 29, Srivastava,P.K. (1990) Opin. Immunol., 3, 654-659. and and Intl Alexander,J., Payne,A., Shikegawa,B., Frelinger,J. Cresswell,P. (1990) Srivastava,P.K. Das,M.R. (1984) J. Cancer, 33, 417-422. 3150 Tumor rejection antigen gp96 is an ATPase and (1991) Semin. Immunol., 3, 57-64. Srivastava,P.K. Heike,M. Srivastava,P.K. and Maki,R.G. (1991) Curr. Top. Microbiol. Immunol., 167, 109-123. Srivastava,P.K., Deleo,A.B. and Old,L.J. (1986) Proc. Natl Acad. Sci. USA, 83, 3407-3411. Srivastava,P.K., Chen,Y.-T. and Old,L.J. (1987) Proc. Natl Acad. Sci. USA, 84, 3807-3811. Szikora,J.P., vanPel,A., Brichard,V., Andre,M., VanBaren,N., Henry,P. and Boon,T. (1990) EMBO J., 9, 1041-1050. Townsend,A., Elliot,T., Cerunodo,V., Foster,L., Barber,B. and Tse,A. (1990) Cell, 62, 285-295. Trowsdale,J., Hanson,I., Mockridge,I., Beck,S., Townsend,A. and Kelly,A. (1990) Nature, 348, 741-744. Udono,H. and Srivastava,P.K. (1993) J. Cell. Biochem., 17D, Abstract no. NZ225. Van Bleek,G.M. and Nathenson,S.G. (1990) Nature, 348, 213-216. Van den Eynde,B., Lethe,B., VanPel,A., DePlaen,E. and Boon,T. (1991) J. Exp. Med., 173, 1373-1384. Wada,I. et al. (1991) J. Biol. Chem., 266, 19599-19610. Walker,J.E., Saraste,M., Runswick,M.J. and Gay,N.J. (1982) EMBO J., 8, 945-951. Welch,W.J. and Feramisco,J.R. (1985) Mol. Cell. Biol., 5, 1229-1237. Wiech,H., Buchner,J., Zimmenmann,R. and Jakob,U. (1992) Nature, 358, 169-170. Received on February 15, 1993; revised on April 20, 1993

Journal

The EMBO Journal – Springer Journals

Published: Aug 1, 1993

There are no references for this article.