Specific interaction between adenoviral 55-kDa E1B protein and in vivo produced p53 fusion proteins

Gene, 131 (1993) 231-236 0 1993 Elsevier Science Publishers

B.V. All rights

reserved.

231

0378-l 119/93/$06.00

GENE 07204

Specific interaction between adenoviral55-kDa produced ~53 fusion proteins (Protein-protein expression)

interactions, expression vector, affinity chromatography;

ElB protein and in vivo glutathione-S-transferase;

eukaryotic

Alexey Chumakov and H. Phillip Koeffler Division of Hematology/Oncology, Department of Medicine, Cedars-Sinai Medical Center, UCLA School of Medicine, Los Angeles, CA 90048, USA Received by D.T. Denhardt:

12 October

1992; Revised/Accepted:

12 February/l5

February

1993; Received at publishers:

I5 April 1993

SUMMARY

Several protein fusion systems have been used in recent years to study protein-protein and DNA-protein interactions. Most of them use bacterially produced proteins which have several inherent disadvantages, notably, the absence of correct post-translational modifications and the frequent insolubility of recombinant proteins. We sought to develop a system to study proteins interacting with the nuclear phosphoprotein ~53, which is believed to be a tumor suppressor. To prepare fusions of ~53, we developed a convenient system that permits both in vivo and in vitro production and easy affinity purification of peptides and protein fragments as glutathione-transferase fusions. We placed the coding sequence of the Schistosoma japonica glutathione S-transferase (GST) under the control of the strong CMV/T7 promoter and SV40 splice and polyadenylation signals. An extensive polylinker (MCS) at the 3’ end of the GST gene is preceded by the sequence encoding the cleavage site of the site-specific protease. We cloned the complete coding sequences of human wild-type ~53, as well as p53 mutants representing all four mutational hotspots (codons 141, 175, 248, and 273), into our expression vector. In vitro transcription using the upstream T7 promoter and translation in reticulocyte lysates form an easy way to produce hybrid proteins; affinity purification on a glutathione-agarose column removes proteins that are present in reticulocyte lysates. We have also studied specific in vivo interactions of human p53 with the adenoviral 55-kDa ElB protein by transfecting expression constructs of GST-p53 fusions into human AdS-transformed 293 cells.

INTRODUCTION

In recent years, several protein fusion systems have been constructed that permit the production and purification of peptides in bacteria (Kellerman and Ferenci, 1982; Uhlen et al., 1984; Nilsson et al., 1985; Hochuli et al., 1987; Glass et al., 1990; Johnson et al., 1989; Studier Correspondence to: Dr. A. Chumakov, Division of Hematology/ Oncology, Cedars-Sinai Medical Center/UCLA School of Medicine, 8700 Beverly Blvd., B121, Los Angeles, CA 90048, USA. Tel.: 310-85557729;Fax:310-652-8411. Abbreviations: aa, amino acid(s); Ab, antibody(ies); Ap, ampicillin; bp, base pair(s); CMV, human cytomegalovirus; factor X, blood coagulation factor activated to factor Xa; GSH, glutathione; GST, glutathione

et al., 1990). Glutathione S-transferase (GST; EC 2.5.1.18) fusion proteins produced in the system developed by Smith and Johnson (1988) directly expressed high levels of hybrid proteins in E. coli and offer easy and nondenaturing purification of the polypeptides. However, the bacterial expression systems have several inherent disadvantages, notably the absence of correct modification of the eukaryotic proteins and the frequent insolubility of S-transferase; GST, gene encoding GST; MCS, multiple cloning site (polylinkers); nt, nucleotide(s); ori, origin(s) of DNA replication; ~53, protein of 53 kDa known as a tumor suppressor; PAGE, polyacrylamide-gel electrophoresis; PBS, 10 mM Na,HPO, pH 7.4/150 mM NaCl; PEG, polyethylene glycol; PolIk, Klenow (large) fragment of E. coli DNA polymerase I; Py, polyoma virus; SDS, sodium dodecyl sulfate; Tc, tetracycline; TE, 10mM TrisHCl pH 7.5/l mM EDTA; wt, wild type.

recombinant proteins. In the case of ~53. WCfound that some of the bacterially expressed GST protein fusions with parts of ~53 were insoluble, making them unusable for comparison with soluble fusion proteins. Proteinprotein and protein-DNA interactions can depend greatly on the correct assembly as well as phosphorylation status of several subunits (Abel and Maniatis, 1989; Glass et al., 1989). As previously demonstrated (Milner et al., 1991; Braithwaite and Jenkins, 1989), co-translation may be required to demonstrate the formation of protein complexes, making it difficult to use bacterially produced peptides to study the interactions of proteins. This may occur because additional proteins participate in complex formation (Ellis et ai.* 1989), or only the co-translated nascent peptide chains can interact, and this becomes impossible after the complete protein molecule has assumed its stable conformation. This phenomenon complicates attempts to study relevant interactions of parts of the ~53 molecule, such as localization of its sites of interaction. While proteins overexpressed in either eukaryotic cells or cell extracts can also assume the wrong conformation, they are easy to handle since they can be simply labeled and are soluble. We have developed expression vectors that can be used for both in viva and in vitro expression and that allow the easy purification of proteins which are fused to GST. Such chimeras can be used as a tool to study protein-protein and DNAprotein interactions.

Factor X

CMGX

ATCGAA

AND DISCUSSION

(a) Construction

of pCMG fusion vectors

To study protein-protein interactions with ~53, we developed a vector system that can be used for the expression of cloned fragments as hybrid GST fusion proteins (Fig. 1). The two vectors pCMGT and pCMGX contain the strong early promoter of human CMV, which is active in a variety of cell types (Norton and CofIin, 1985; MacGregor and Caskey, 1989; Alam and Cook, 1990), an SV40 splice acceptor site and polyadenylation site, as well as SV40 or polyoma ori that permit replication of these vectors as extrachromosomal multicopy plasmids in cells containing T-antigens of SV40 and polyoma (Gluzman, 1981; Mellon et al., 1981). The plasmids also have a promoter for T7 RNA-polymerase in the same orientation as the CMV promoter that can be used to produce RNA for in vitro translation using the same plasmid. A cDNA coding for Schistasoma japonica GST (Smith and Johnson, 1988; Smith et al., 1988) was derived from the bacterial expression plasmids pGEX2T and pGEX3X (Smith and Johnson, 1988). The encoded aa sequence contains protease cleavage sites for either acti-

ATC CCC GGG AAT T5T ,, %si+ SmSl

EaJnHl’

GCA GAT ATC CAT! EC&“! Pstl

-CAC ACT G,GC GGC CGC TCG AGC ATG CAT CTA GA ’ L.._zu Not+ Sphl Xba I XhOl

Thrombin

r CMGT

CTG GlT

CCG CGT’GGA

ICC

SamHI’

1 CCG GGA Al?

CTG GAG ATA ICC

ECORI

SllWl

PSI I

EwR

ATC-

V

-ACA CTG GCG GCC GCT CGA GCA TGC ATC TAG A

Fig. 1. Structure EXPERIMENTAL

GGT CZGGG

of the pCMG

vectors,

Map

of pCMGT

expression

vector and nt sequences of pCMGX and pCMGT vectors at the MCS. Construction of eukaryotic expression vectors: The pCDNA- I plasmid (fnvitrogen, San Diego, CA) was used as a backbone for cloning the cDNA fragment of Schistosoma japonicum GST. The coding of Sj26 cDNA was excised from pGEX-2T and pGEX-3X

sequence bacterial

expression vectors (Smith and Johnson, 1988) by cleavage with AseI and treated with Polfk and EcoRI. The 0.7.kb fragment was gel-purified and cloned

between

blunt-ended

BarnHI and EcoRI sites of pcDNA-1.

The resulting plasmid has the coding sequence for 26-kDa GST under control of the strong early promoter of human CMV, a MCS preceded by either factor

X (pCMGX)

sites and followed

or thrombin

(pCMGT)

by the SV40 splice acceptor

protease

cleavage

site and polyadenylation

signal (poly A). The T7 promoter upstream from the GST sequence can be used for in vitro expression of the fusion proteins, The vector backbone also contains for cloning

a set of nri, as well as the supF selectable

marker

in E. rob (Seed, 1983).

vated factor X (pCMGX) or thrombin (pCMGT) located just upstream from the MCS region, The two vectors differ in reading frames for gene fusions. Cloning of GST cDNA into the pcDNA-1 expression vector (Invitrogen, San Diego, CA) also added several useful unique restriction enzyme cleavage sites to MCS. Rare Not1 and XhoI sites can be used for the linearization of plasmids for in vitro transcription. To search for proteins interacting with nuclear phos-

233

kDa

123456789

phoprotein ~53 (Jenkins and Sturzbecher, 1988), we constructed a number of GST-p53 gene fusions, using both a full-length ~53 coding sequence and several fragments (Fig. 2). Plasmids pCMGC1 and pCMGC2 contain the N-terminal (159 aa of ~53) and C-terminal (233 aa of ~53) coding fragments of wt human ~53 cDNA (O’Rourke et al., 1990) respectively, cloned into the pCMGX vector into the BumHI site. Plasmids pCMG53, pCMGl41, pCMG175, pCMG248, and pCMG273 code for fusion proteins of about 80 kDa and contain complete coding sequences of wt human ~53 (pCMG53) as well as a ~53 mutant containing mutations in codons 141, 175, 248, or 273, respectively (Miller et al., 1993). Another plasmid, pCMGETS2, contained the complete coding sequence of human ets-2 proto-oncogene (Watson et al., 1988) and served as a control in protein-binding experiments.

97.4 -

69

-

46

(b) In vitro expression and purification of p53-GST fusion

30

proteins

In vitro transcription/translation is a powerful method for producing proteins in their native conformation. The lysate used in translation reactions contains most of the proteins normally present in reticulocytes. Although the purification of proteins from relatively small-scale reactions can be cumbersome by conventional methods, the availability of specific Ab to the protein of interest can make possible an easy immunopurification. The vector system described here can be used to translate GSTfusion proteins in vitro (Fig. 3) and to purify translated

Fig. 3. SDS-PAGE

separation

of 35S-labelled

products

of in vitro tran-

scription/translation of GST-fusions. Standard procedures were used to transfect plasmids into MC1061/P3 E. coIi host cells (Invitrogen, San

Diego,

(Sambrook

CA),

to grow

bacteria,

and

to purify

et al., 1989). Briefly, E. coli cells containing

plasmid

DNA

pCMG-derived

plasmids were selected in LB-agar plates with IO ug Tc/mI and 25 ug Ap/ml, then grown overnight in liquid culture in 2 x YT medium (Sambrook et al., 1989) supplemented with Tc and Ap. For in vitro transcription, plasmid DNA was purified extensively by PEG precipitation (Sambrook nol/chloroform

et al., 1989), cleaved by XhoI, and extracted by phe(1:l). Uncapped RNA was synthesized by T7 RNA

polymerase according to the protocol WI). Translation in vitro was done

provided by Promega in rabbit reticulocyte

(Madison, lysates as

recommended by Promega. Portions (5 ~1) of 40 ul in vitro translation reactions were analyzed by 0.1% SDS-8% PAGE (Laemmli and Favre, 1973). Proteins pCMGC2 pCMGl75, 1

cw

GST

-L- PCMGETS

b 1

439

Fig. 2. Construction of GST-~53 and GST-ets-2 gene fusions. Human wt and mutant ~53 cDNA fragments were cloned into pCMGX vector as BamHI or BamHI-PstI fragments from GAL4-~53 fusion constructs (O’Rourke et al., 1990). The resulting plasmids code for the N-terminal 159 aa (pCMGCl), C-terminal 233 aa (pCMGCZ), complete wt ~53, or p53 with mutations in the conserved hotspots (codon 141, 175, 248, 273). To make a GST-ers-2 fusion, the human ers-2 cDNA fragment was taken from plasmid pCMETS2. HindIII-digested pCMETS2 DNA was treated with PolIk and XhoI, and a 1.6-kb fragment was cloned between the blunted PstI site and Xhol site of pCMGT. The correct constructs were identified by restriction mapping and sequencing (Tabor

and

Richardson,

1987) using

a primer

specific

to the region

upstream from the MCS in pCMG plasmids (5’-GCGACCATCC TCCAAAATCG). The @S-derived fragments are shown by hatched boxes: the GST-derived fragments are shown by lightly shaded boxes. The ef.s-2 coding sequence is shown by the black box.

were derived

from

(lanes I-4, respectively); pCMG248, pCMG273

pCMG53,

pCMGl41,

pCMGC1,

pCMGX (coding for GST; lane 5); (lanes 6-8), and pCMGETS2 (lane

9). Markers were: myosin (H-chain), 200 kDa; phosphorylase b, 97.4 kDa; bovine serum albumin, 69 kDa; ovalbumin, 46 kDa; carbonic anhydrase, 30 kDa.

proteins free of reticulocyte lysate components. Even big proteins such as GST-Ets-2 fusion protein (about 90 kDa) are efficiently produced and purified using this system. The strong and specific binding of GST fusion proteins to GSH-agarose beads is resistant to high salt concentrations (up to 1.5 M NaCl), which allows us to remove associated proteins from fusion proteins without releasing the fusion proteins from the beads. Thus, these beads can serve as a trap to detect either labelled proteins or DNA fragments that specifically interact with immobilized protein (our unpublished observations). The GST-fusion pro-

234 tcin is released either by the addition of 5 mM free GSH for additional cycles of purification of associated protein or DNA ligands (Smith and Johnson, 1988) or by the addition of denaturing sample buffer and SDS-PAGE (Laemmli and Favre, 1973). The usefulness of this system was shown by demonstrating the interaction of ~53 with the Ad5 55-kDa E I B protein, initially examining in vitro interactions (Fig. 4A). Following in vitro transcription and translation in rabbit reticulocyte extracts, IOO-~1 reactions were diluted with 0.551 ml PBS containing 0.5% NP-40 and protease inhibitors and incubated with 50- 100 11 of GSH-agarose beads at 4 C for 224 h or overnight on a rocking platform. In our experience, GST fusions were not efficiently bound by using GSH-agarose columns. Washing of the bound material was performed either with the same buffer or by applying a salt gradient up to 1.5 M NaCI. Unlabeled fusion proteins attached to GSH-agarose were incubated with whole cell extracts prepared from [35S]Met-labelled human 293 cells. After extensive washing, proteins were eluted by the addition of IO mM GSH. Both wt and mutant (His273) ~53 efficiently bound radiolabelled E 1B protein (lanes 2 and 3, respectively), while the fusion protein containing aa 160 393 of wt human ~53 bound less El B protein (lane I ). No E I B was bound to the fusion protein containing the N-terminal part (aa I-1 59) of ~53 (lane 4). This result confirms previous studies of human p53-ElB interactions (Jenkins and Sturzbecher, 1988; Braithwaite and Jenkins, 1989) which indicated that the smallest EIB-binding product of in vitro translation of human ~53 with an intact C terminus has the size of about 32 kDa. The pGSTC2 construct used in our experiments encodes a ~53 fragment of this size (aa 160.-393) fused to a 27-kDa GST protein. The N-terminal 160 codons may play some role in ~53 binding because much more El B was able to bind to the fulllength ~53 than to the C-terminal half of ~53. (c) In vivo synthesis of GST fusion proteins To demonstrate the interaction of ~53 protein fusions produced in vivo, hybrid GST-p53 plasmids were transiently transfected into human Ad-transformed 293 cells: these cells express high levels of EIB protein. Cells were metabolically labeled by [3sS]Met, lysed, and incubated with GSH-agarose beads to allow the binding of GST~53 complexes with EIB. A sample of each eluate was immunoprecipitated with either anti-ElB Ab to determine the amount of p53-associated El B (Fig. 4B, lanes I-6) or with anti-p53 Ab to identify truncated p53-GST fusions (Fig. 4B, lanes 8. 9). All plasmids that code for either full-size mutant or normal ~53 (lanes 1, 5, 6. 7, Fig. 4C) produced proteins (asterisks) that interacted with equal efficiency with 55-kDa ElB protein (arrow). The protein coded by pCMGC2, which has 160 N-terminal

codons deleted (lane 9. Fig. 4B), was also capable of interacting with E I B, but not as effectively as the entire GST~53 protein (lane 3, Fig. 4C). In contrast, the protein coded by pCMGCl ( 160 N-terminal codons: lane 8. Fig. 4B) failed to interact and bind E I B protein when the labelled cell lysate was incubated with immobilized GSH (lane 2, Fig. 4C). E 1B protein was identified by secondary immunoprecipitation of GSH-affinity purified proteins with pAB 2A6 (Fig. 4B). Two studies have recently located the E I B binding site of murine ~53 to a N-terminal segment between aa 14 and 66 (Kao et al., 1990; Braithwaitc ct al.. 1991). Although in this study we did not precisely locate the E I B-binding site in human ~53, we did not observe E I B binding to aa I-160 of human ~53. Deletion of 159 N-terminal aa decreased, but did not abolish, El B binding to human ~53. One inherent problem in this approach is the presence in most cells of endogenous GSH-binding proteins. which can result in a high background on the autoradiogram. especially when non-fractioned cell lysates are studied. The use of either anti-GST Ab as a second step of purification, fractionation of cell lysates, or use of secondary Ab (as in Fig. 4 C) can solve this problem. Also. the presence of protease cleavage sites for either activated factor X or thrombin in the fusion proteins allows the removal of the purified protein from the GST fragment, which remains bound to the agarose (Smith and Johnson. 1988). (d) Conclusions (I) We have developed a vector system that can be useful both for the production of proteins in vivo and in vitro and for their rapid purification. Fusion proteins produced in vitro can be firmly bound to immobilized GSH for purification, as well as for studies of their interactions with either other proteins or ligands. (2) The transient expression of fusion proteins in eukaryotic cells provides a system to study the in vivo interactions of peptides, even when no Ab arc available. (3) Using this system, we have shown that the C-terminal part (233 aa long) of ~53 is necessary to bind EIB. This is confirmatory of studies of human p53 (Jenkins and Sturzbecher, 1988; Braithwaite and Jenkins, 1989), although studies of murine ~53 (Kao et al., 1990; Braithwaite et al., 1991) had suggested that in vitro only the N-terminal region of murine ~53 bound EIB 55kDa protein. (4) No significant difference was observed between either wt or certain mutant ~53 proteins in their ability to interact with the Ad5 55-kDa El B protein. Specifically, mutations of ~53 at aa 14 I and I75 that affect interactions of ~53 with heat-shock proteins (HSP70) (Jenkins and Sturzbecher, 1988) did not alter the ability of ~53 to bind to El B protein.

235

A kDa

4

3

2

1

-

200

-

97.4

-

69

-

46

-

30

B 1

2

3

4

5

6

7

8

9

*_. lr

.I I

1

2

.

3

4

5

6

7

6

kDa

9 +

Wtp53

N

term

C term

vector

175

246

273

GST-ets

200

+-

97.4

+

69

4

46

Fig. 4. Interaction of p53-GST proteins with 55-kDa EIB. (A) In vitro interaction of p53-GST proteins with 55-kDa EIB. Unlabelled GSTp53 protein fusions were produced in vitro in rabbit reticulocyte lysates and attached to GSH-agarose beads as described by the supplier, except that unlabelled Met was used. Human 293 cells were radiolabelled with [?j]Met (100 l&i/ml, 1000 Ci/mmol) and lysed as described (Sarnow et al., 1982). Extracts were preadsorbed to GSH-agarose for 12 h at 4°C to remove endogenous GSH-binding proteins. GST-p53 agarose beads (100 ~1) were incubated with cellular extract (5 x IO6 cell equivalent for each sample) for 4 h at 4°C and washed 3 times with 10 bed volumes of cell lysis buffer. Proteins were eluted in 2 bed volumes of TE buffer/l0 mM GSH. Samples of eluted protein were separated on 0.1% SDS-8% PAGE, impregnated with sodium salicylate, and exposed to x-ray film for 12 h. EIB protein was detected bound to GSTCZ fusion (lane I; 233 aa of C-terminal part of human p53), full-length wt, and mutant His273 GST-p53 (lanes 2 and 3, respectively). No El B was bound in vitro to GSTCl fusion (lane 4; aa I-159 of ~53). (B) Identification of ElB protein in GSH-affinity purified extracts of transfected 293 cells. Samples of GSH-agarose purified GST-p53/ElB complexes were denatured by the addition of SDS to 0.5% and boiled for 10 min. ElB proteins were then immunoprecipitated by pAB 2A6 (Sarnow et al., 1982) (generous gift of A. Levine, Princeton) and protein A/protein G agarose beads and analysed on 0.1% SDS-8% PAGE. The same samples were used as in C. [35S]Met-labelled extracts were from 293 cells transfected with expression vectors for wt p53 (lane l), GSTCl (aa l-159 of ~53; lane 2), GSTC2 (aa 160-393 of ~53; lane 3), mutant p53 (Tyr14’, Trpz4s, HisZ73; lanes 4-6, respectively). Markers were run in lane 7. Aliquots from the same samples were immunoprecipitated with anti-p53 Ab pAB 1801 (Banks et al., 1986), as recommended by supplier (Oncogene Science), and pAB 421 to identify GSTCl and GSTC2 fusion proteins (lanes 8 and 9, respectively) in GSH-affinity purified fractions, after disruption of El B-p53 interactions by boiling. The gel was impregnated with sodium salycilate, dried, and exposed to x-ray film for 10 days. (C) SDS-PAGE of in vivo expressed p53-GST fusion proteins following transfection into human AdS-transformed cells by &phosphate coprecipitation. Human 293 cells were maintained in alpha culture medium (Gibco BRL, Gaithersburg, MD, USA) supplemented with 5% fetal bovine serum at 5% CO,. At 12 h before transfection, the cells were transferred into IO-cm dishes at 5 x lo5 cells per plate. For transient expression, supercoiled plasmid DNA was purified by PEG precipitation, and transfection was done by Ca,phosphate coprecipitation, essentially as described by Chen and Okayama (1987). Then 10 ug of pCMGP53, pCMGC1, pCMGC2 (lanes l-3, respectively), pCMG175, pCMG248, pCMG273, and pCMGETS2 (lanes 5-8, respectively), or vector plasmid pCMGX (lane 4) were used to transfect cells plated in IO-cm dishes. At 48 h after transfection, the plates were washed with PBS and incubated for 1 h in 5 ml of Met-free IMDM (Iscove’s modified Dulbecco medium; Gibco BRL, Gaithersburg, MD) at 37°C. Cells were washed again and incubated for 3 h at 37°C in 3 ml of the same medium containing [35S]methionine (100 uCi/ml, 1000 Ci/mmol, ICN). After labelling, cells were washed twice with ice cold PBS, scraped off the plates by using rubber policeman, and spun at 2000 x g for 5 min. The cell pellet was either stored at -70°C or lysed immediately (Sarnow et al., 1982). To demonstrate protein-protein interactions, cells were lysed in TNN (50 mM TrisHCl pH 7.5/150 mM NaCI/O.lS% NP-40) buffer supplemented with protease inhibitors. Frozen cells were suspended in TNN, incubated on ice for 30 min, frozen and slowly thawed once, and then repeatedly passed through a high gauge needle using a 1 ml syringe. Insoluble cellular debris was removed by centrifugation for 5 min at 12000 rpm at 4°C in microcentrifuge. Cleared lysate (200 ul fraction) was incubated overnight at 4°C on a rocking platform with 50 ul of GSH-agarose beads (Pharmacia) and washed in the same buffer. After adsorption, the GSTfusion protein complexes were washed three times with TNN at 4°C and eluted in the same buffer containing 5 mM free GSH (Sigma). Electrophoresis buffer was added to a sample of eluted proteins, and they were analyzed by 0.1% SDS-7.5% PAGE. GST fusion proteins are indicated by asterisks; the position of the 55-kDa ElB protein is indicated by an arrow (on the left margin).

236 ACKNOWLEDGEMENTS

We would Margery

like to thank

Goldberg,

secretarial

Kim

Burgin,

and Elaine Epstein

support.

This manuscript

part by US Public

Health

Elisa

Weiss,

for their excellent was supported

Service grants

in

No. DK42792,

DK41936, CA427 IO, CA32737, CA26038-1 I, the Weisz Family Foundation, the Leukemia Fund in memory of Marilyn

Epstein-Levine,

and the Realtors

the Parker

of Real Estate

Hughes

Industry

Division.

first author

is on leave from the Institute

Chemistry,

Russian

Academy

of

Foundation, The

of Bioorganic

Sciences,

Moscow,

Russia.

REFERENCES

Miller. C.W.. Chumakov. A.M., Said, J., Chcn, D.L., Koelller. H.I.: Mutant ~53 proteins have diverse mtracellular abilitlcs to ohgomerize and activate transcription. Oncogene X (IYY?) in press. Milner. J.. Medcalf, sis of wild-type (IYYI) I2 IO.

E.A. and Cook, A.C.: Tumor suppressor ~53: anal>and mutant ~53 complexes. Mol. Cell. Biol. I I

Nigro. J.M.. Baker. S.J., Prclsinger. A.(‘.. et al.: Mutations gene occur in diverse human tumour types. Nature

In the ~53 343 (19X9)

705 70x. Nilsson. B., Abrahmsen.

and purili-

L. and Uhltn.

M.: Immoblhzation

cation ofenzymes with staphylococcal EMBO J. 4 (1985) 1075m 10X0.

Abel. T. and Maniatis. T.: Gene regulation. Nature 341 (1989) 24-25. Alam, J. and Cook. mammalian

Mellon. I’.. Parker. V.. Glurman, 1’. and Maniatls. T.: Itlcntllication (11 DNA sequences required for transcription of the human alpha I-globin gene in a new SV40 ho+vector system. (‘cll 77 t 19~01 77Y 28X.

J.L.: Reporter

gene transcription.

Action

of leucine zippers.

genes: application Anal. Biochem.

I88

to the study

of

(I 990) 245-254.

protein

A gcnc fusion vectors.

Norton. P. and Cofin. J.: Bacterial beta-galactosidasc Rous sarcoma virus gene expression and replication.

as a marker 01 Mol. Cell. Biol.

5 (lYX5) 281 290.

Banks, L., Matlashevski, G. and Crawford, L.: Isolation of human-p53specific monoclonal antibodies and their use in the studies of human

O’Rourke. R., Miller, C.W.. Kato. G.. Simon, K.. Chcn. D.-L.. Dang. C.-V. and Koeffler, H.P.: A potential transcriptional activation clcment in the ~53 protein. Oncogene 5 (1990) 1X29- 1x32.

~53 expression. Eur. J. Biochem. I59 (1986) 529-534. Braithwaite, A.W. and Jenkins. J.R.: Ability of ~53 and the adenovirus

Sambrook,

El b 5%kilodalton protein to form a complex J. Virol. 63 (1989) 1792-1799.

is determined

by ~53.

Braithwaite. A.W., Blair, G.E., Nelson, CC., McGovern, J. and Bellett, A. J.: Adenovirus Elb-5X kD antigen binds to ~53 during infection of rodent cells: evidence for an N-terminal binding site on ~53. Oncogene,

6 (1991) 781-787.

Chen. C. and Okayama, lian cells by plasmid

H.: High-efficiency transformation of mammaDNA. Mol. Cell. Biol. 7 (1987) 2747-2752.

Ellis. R.J., van der Vies, SM. and Hemmingsen, chaperone concept. Biochem. Sot. Symp. 55

S.M.: The molecular

(I 989) 145% 153.

Glass, C.K., Lipkin, SM., Devary, O.V. and Rosenfeld, M.G.: Positive and negative regulation of gene transcription by a retinoic acidthyroid hormone receptor heterodimer. Cell 59 (1989) 697-708. Glass, C.K.. Devary, O.V. and Rosenfeld. M.G.: Multiple cell typespecific proteins differentially regulate target sequence recognition by the alpha retinoic acid receptor. Cell 63 (1990) 729-738. Gluzman. Y.: SV40-transformed simian cells support the replication of early SV40 mutants. Cell 27 (1981) 175-182. Hochuli, E., Dobeli. H. and Schacher, A.: New metal chelate adsorbent selective for proteins and peptides residues. J. Chromatography 41

containing

neighbouring

histidine

I (1987) I77- 184.

Jenkins, J.R. and Sturzbecher, H.-W.: The ~53 oncogene. E.P., Skalka, A.M. and Curran. T. (Eds.), The Oncogene Elaevier, Amsterdam, 1988, pp. 403-423.

In: Reddy, Handbook.

Johnson. KS., Harrison, G.B., Lightowlers, M.W.. O’Hoyk, L.. Cougle, W.G.. Dempster, R.P., Lawrence, S.B., Vinton. J.G.. Heath, D.D., and Richard, M.D.: Vaccination against bovine cysticercosis using a defined recombinant antigen. Nature 338 (1989) 585-587. Kao. C.C.. Yew, P.R. and Berk, A.J.: Domains required for in vitro association between the cellular ~53 and the adenovirus 2 El B 55K proteins. Virology 179 (1990) X06&814. Kellerman. 0. and Ferenci, T.: Maltose-binding protein from Escherichicr co/i. Methods Enzymol. 90 (1982) 459-463. Laemmli, U.K. and Favre. M. (1970) Maturation of the head of bacteriophage T4, 1. DNA packaging events. J. Mol. Biol. 80 (1973) 575-599. Lawrence. S.B., Vinton. J.G.. Heath, D.D. and Richard, M.D.:

J.. Fritsch,

E.F. and

Mania&

T.: Molecular

Clonmg.

A

Laboratory Manual. 2nd cd. Cold Spring Harbor Laboratory Press. Cold Spring Harbor. NY. 1989. Sarnow, P.. Sullivan. C.A. and Le\‘ine, A.J.: A monoclonal antibody detecting the adenovirus ization of the Elb-58Kd

type 5-E I b-58Kd tumor antigen: charactertumor antigen in adenovirua-Infected and

-transformed cells. Virology 20 ( 19X2) 5 IO-5 17. Sarnow, P.. Hearing, P., Anderson, A.N.. Halbert, D.N.. Shenk. T. and Levine, A.J.: Adenovirus early region I B 5X OOO-dalton tumor antigen is physically associated with an early rcglon 3 25000-dalton protein in productively infected cells. J. Virol. 49 (I 9X4) 6Y2- 700. Seed. B.: Purification of genomic sequences from bacteriophage libraries by recombination and selection In viva. Nucleic Acids Rea. I I (IYXZ) 2427-2445. Smith, D.B. and Johnson. K.S.: Single-step purification of polypeptldea expressed in Escherichitr c,o/i as fusions with glutathlonc S-tranafcrase. Gene 67 (198X)3 I 30. Smith, D.B.. Davern, K.M., Board. P.G.. Tiu, W.U.. Garcia. E.D. and Mitchell. G.F.: M-26000 antigen of S~~hi.sros~,n~crjclpc~rlrc,lrm rccognized by resistant WEHI 125.‘j mtcc is a parasite glutathionc Stransferase. Proc. Natl. Acad. Sci. USA X3 (1986) X703-8707. Smith, D.B., Rubira, M.R., Simpson. R.J., Davern, K.M.. Tiu. W-t.. Moard, P.G. and Mitchell, G.F.: Expression of an enzymaticall) active parasite molecule in Esckrrichicl w/i: Schisrosomu ppmiww glutathione S-transferaac. Mol. Biochem. Parasitol. 77 (19Xx) 249 256. Studier, F.W., Rosenberg, A.H., Dunn, J.J. and Dubendorff. J.W.: I:sc of T7 RNA polymerase to direct expression of cloned gcncs. Methods Enzymol. 185 (1990) 60-89. Tabor. S. and Richardson, C.: DNA sequence analysis with a modilied bacteriophage T7 DNA polymeraae. Proc. Natl. Acad. Sci. USA X3 ( 19X7) 4767-4771. Uhl&. M.. Cuss, G., Nilsson, B., Gotz, t. and Lindbcrg. M.: txprcssion of the gene encoding protein A in Srcrph~lococ~cus L!~~~Pu.s and coagulase-negative staphylococci. J. Bacterial. 155, ( 1984) 7 I? ~7IO. Watson. D.K.. McWilliams. M.J.. Lapis, P., Lautenberger. J.A., Schweinfest. C.W. and Papas. T.S.: Mammalian &s-l and rlv-.? genes encode highly conserved proteins. Proc. Natl. Acad. SCI. L’SA X5 ( I9XX) 7X62-7X66.