mutation of zinc finger motif - BioMedSearch

Download PDF
1 downloads 0 Views 1MB Size Report
Comment
In pPYL-CAT viral. DNA has the opposite orientation in relation to the cat gene with ..... Fritz,H.-J. (1989) Nucleic Acids Res., 17, 4441-4454. 27. McKay,R.D.G. ...
1990

Nucleic Acids Research, Vol. 18, No. 9 2715

Oxford University Press

Loss of DNA-binding and new transcriptional transactivation function in polyomavirus large T-antigen with mutation of zinc finger motif Anders Bergqvist, Mats Nilsson, Kdre Bondeson and Goran Magnusson* Department of Medical Virology, Uppsala University Biomedical Center, Box 584, S-751 23 Uppsala, Sweden Received January 3, 1990; Revised and Accepted April 9, 1990

ABSTRACT A putative zinc finger in polyomavirus large T-antigen was investigated. We were unable to demonstrate unequivocally a requirement for zinc in specific DNAbinding using the chelating agent 1, 1 0-phenanthroline. An involvement of the putative zinc finger in specific DNA-binding was nevertheless suggested by the properties of a mutant protein with a cys-ser replacement in the finger motif. Probably as a result of the defective DNA-binding, the mutant protein had lost its activity in initiation of viral DNA-replication and in negative regulation of viral early transcription. However, the trans-activation of the viral late promoter was normal. The analysis also revealed a previously unrecognized activity of large T-antigen. The mutant protein trans-activated the viral early promoter. In the wild-type protein this activity is probably concealed by the separate, negative regulatory function. INTRODUCTION The polyomavirus early proteins, the T-antigens,

promote the formation of virion components. This is achieved by the induction of the host cell to enter S-phase (1) and by the stimulation of viral DNA-replication (2) and late transcription (3). Only the large T-antigen has been demonstrated to interact directly with the viral genome. The protein binds to several sites in the segment of viral DNA that contains the origin of DNA-replication and the promoters of the early and late genes (4, 5). Specific DNAbinding of large T-antigen is necessary for its function in the initiation of viral DNA-replication (6, 7) and probably also for its inhibition of early transcription (8). The segment of the polypeptide chain that binds to DNA has been tentatively identified by studying mutant proteins. Substitutions at amino acid residues 286-300 led to relaxed specificity or inhibition of high affinity binding to polyomavirus DNA (6, 7). Based on these observations and on studies of the homologous region in the related simian virus 40 protein (9, 10 and references therein), the DNA-binding domain of polyomavirus large T-antigen is thought to be located around amino acid residue 300. The structure of the polypeptide chain at this site has not been *

To whom correspondence should be addressed

determined and its amino acid sequence does not bear any obvious similarity to DNA-binding structures of other proteins. It was noted (11) that large T-antigen contains a sequence motif at amino acid residues 452 -472 that might form a zinc finger (Fig. 1). This type of structure was originally observed in the transcription factor lIIA (12). More recently, it has been found in several other DNA-binding proteins (13). In these proteins the zinc fingers probably interact with DNA in the major groove of the helix (14). Since the zinc finger motif is conserved in the amino acid sequence of large T-antigens produced by several different polyomaviruses, its function might be important and possibly related to DNA-binding. The function of the putative zinc finger of simian virus 40 large T-antigen has been analyzed in detail using a series of mutant proteins with single amino acid substitutions (9). All the cysteine and histidine residues that might participate in Zn2+-binding (see Fig. 1 for the structure of the polyomavirus zinc finger motif) were found to be essential for the activity of the protein in viral DNA-synthesis. In addition, substitution of some of the amino acid residues within the putative zinc finger resulted in an altered activity of large T-antigen. The DNA-binding properties of these mutant proteins were not reported, but in another study of simian virus 40 large T-antigen produced in Escherichia coli similar mutants were reported to be defective in specific DNA-binding (10). Paradoxically, a mutant protein with a truncation of the C-terminal segment, including the zinc finger motif, retains the binding activity (10). Thus, the structural integrity of the zinc finger motif appears to be important for DNA-binding even though the putative finger structure might not contribute directly to the formation of a stable complex. The large T-antigens of simian virus 40 and polyomavirus are closely related in structure and function, but the two proteins also have distinct properties. One such property is the activity of the simian virus 40 protein in cellular transformation that was affected by structural alterations of the putative zinc finger (9). To investigate the function of the putative zinc finger in polyomavirus large T-antigen, the protein was treated with a chelating agent that complexes zinc ions much more strongly than magnesium ions, and was then analyzed for DNA-binding activity. In addition, a mutant was constructed that encoded a protein with

2716 Nucleic Acids Research, Vol. 18, No. 9

,R- L-0

Enhancer Region

I

S

K

"'

I

I N,

2' , Zn

ol

% 52

X

Major late cap sites

K N

\

TATA

100

co

w

E K

Ori

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 5297/1 5200 5 00

H

E

Early cap sites

Figure 2. Map of the polyomavirus regulatory region. A schematic representation of the DNA segment containing the enhancer region, the core of the origin of DNA-synthesis and the early and late promoters is shown. The arrows denote a base sequence with dyad symmetry at the origin of DNA-replication and the location of the 5'-ends of RNA. The ends of the segment are at the BclI and BstXI (XhoI) sites used for linkage of the cat gene.

R

tI

K

447

2

Figure 1. Hypothetical zinc finger structure in polyomavirus large T-antigen. The amino acid sequence of large T-antigen was deduced from the published base sequences of viral DNA (20, 33).

the first cysteine residue in the zinc finger motif replaced by

a

serine residue.

MATERIALS AND METHODS For details on the preparation and analysis of DNA previous publications should be consulted (15, 16). Chemicals and enzymes All chemicals were of reagent grade. 1, 7-Phenathroline and 1, 10-phenanthroline were obtained from Aldrich Chem. Co. (Milwaukee, Wisconsin) and from Serva (Heidelberg, FRG), respectively. Radiochemical were from NEN Research Products (DuPont Scandinavia AB, Stockholm, Sweden). Most restriction enzymes were obtained from Pharmacia (Uppsala, Sweden). The enzyme DpnI was supplied by Promega Corp. (Madison, Wisconsin).

Cells, cell culture and virus Human HeLa cells (obtained from W. Schaffner, University of Zurich) and murine 3T6 cells were maintained in Dulbecco modified minimal essential medium (Flow Labs., Irvine, Scotland) supplemented with 10% newborn calf serum and 10% horse serum, respectively. The Spodoptera frugiperda cell line Sf9 (ACTCC, accession number CRL171 1) was grown in TNMFH medium (Nord-Vacc, Stockholm, Sweden) supplemented with 10% foetal bovine serum. Infection of Sf9 cells with the recombinant baculovirus vEV5 1 LT was performed as described (17). HeLa cells were transfected with viral DNA using the calcium phosphate co-precipitation method (18). In analyses of viral DNAreplication, transfecting DNA was adsorbed to mouse 3T6 cells in the presence of DEAE-dextran.To stimulate uptake of DNA, the cells were then treated with chloroquine diphosphate (19). Plasmids and mutagenesis Polyomavirus DNA was from a derivative of the A2 strain. Nucleotide numbers refer to the sequence of Soeda et al. (20) with later modifications (21, 22). Viral DNA was propagated in Escherichia coli cells as recombinants of plasmid pBR322 or pML (23). The plasmid pPYdllO61 has a deletion of nucleotide

713-765 (24) that prevents the formation of mRNA encoding small and middle T-antigen. The plasmid pPYAE has a deletion of nucleotide 175-3612 that includes all the T-antigen-encoding sequences. The plasmids pPYE-CAT and pPYL-CAT have the regulatory segment of polyomavirus DNA and the chloramphenicol acetyl transferase gene (cat, ref. 25) inserted in the tetracyclin resistance gene of pML. In pPYE-CAT the cat gene is under the control of the viral early promoter with the junction at nucleotide 174 of viral DNA. In pPYL-CAT viral DNA has the opposite orientation in relation to the cat gene with the junction at nucleotide 5023 (Fig 2). For oligonucleotide directed mutagenesis, a segment of viral DNA was cloned in plasmid pMA5 -8 (26). Mutagenesis was carried out as described (26). To reconstruct full length viral DNA, a segment of the mutated DNA (nucleotide 1600-1976) was ligated in vitro to the remainder of the viral genome cloned in plasmid pML.

Analysis of specific DNA-binding Wild-type large T-antigen was prepared from vEV5 ILT infected Sf9 cells at 36 h postinfection. Cells were washed in Tris-buffered saline and then lysed in a buffer consisting of 0.1 M Tris-HCl (pH 7.8), 0.1 M NaCl, 0.001 M dithiothreitol, 0.5% Nonidet P-40 and 50 Ag phenylmethylsulfonyl fluoride per ml. After incubation for 30 min at 0°C, the extracts were centrifuged at 10,000 xg for 10 min. DNA-binding reactions (4, 27) were set up with HpaII digested polyomavirus DNA labelled at the ends with 32p. The binding reactions and analysis by polyacrylamide gel electrophoresis were performed as described before (15). Each 0.2 ml protein extract of ca 3 x 104 cells was mixed with 1.0 ml of 0.02 M Na2HPO4 (pH 7.0), 0.1 M NaCl, 0.001 M EDTA, 0.002 M dithiothreitol, 18% dimethylsulfoxide, 0.05% Nonidet P-40, 5 ,tg of poly(dI-dC). The samples, also containing 1, 10-phenanthroline or 1, 7-phenanthroline as indicated, were incubated for 20 min. This incubation and all subsequent manipulations were done at 20°C. Five ng of radioactively labelled polyomavirus DNA was added and the incubation was continued for 60 min. Fifty kd of a 20% suspension of Staphylococcus aureus cells was added and the cells were sedimented by centrifugation. The supernatant was collected and mixed with 15 ,ul of the monoclonal antibody, LT1 (28). After 1 h, 15 Ml of Staphylococcus aureus cell suspension was added and after an additional hour immune complexes adsorbed to the cells were collected by sedimentation. The sediment was washed 4 times in 0.02 M Tris-HCl (pH 7.6), 0.002 M dithiothreitol, 0.001 M EDTA, 0.5% Nonidet P40, 10 ,.tg of double-stranded calf thymus DNA per ml and after elution of immune complexes, they were analyzed by polyacrylamide gel electrophoresis.

Nucleic Acids Research, Vol. 18, No. 9 2717 Mutant bclO81 large T-antigen was prepared from HeLa cells transfected with plasmid pPYbcl08l/d1l061. Cells were lysed at 45 h posttransfection in a solution of 0.05 M Tris-HCl (pH 8.0), 0.15 M NaCl, 2% Nonidet P-40 and 50 ,tg phenylmethylsulfonyl fluoride per ml. The analysis of specific DNA-binding was carried out as desciribed above, except for the final concentration of dimethylsulfoxide that was 3%.

Analysis of viral DNA-synthesis and gene expression Synthesis of large T-antigen was analysed by transfection of HeLa cell cultures (ca 5 X 105 cells) with 10 ltg of plasmid DNA. The cultures were labelled with [35S]methionine and at the end of the labeling period, the cells were lysed with 0.1 M Tris-HCl (pH 9.0), 0.1 M NaCl, 0.001 M dithiothreitol, 0.5% Nonidet P-40 and 0.2 units of aprotinin (Sigma Chemical Co., St. Louis, Missouri) per ml. After extraction for 30 min at 0°C, the pH was adjusted to 7.6. Immunoprecipitations were done with the monoclonal antibody LTl (28), followed by adsorption to Protein G-Sepharose (Pharmacia, Uppsala, Sweden). The precipitated immunocomplexes were washed 4 times with NET buffer (28) and subjected to SDS-polyacrylamide gel electrophoresis (29). The effect of large T-antigens on the polyomavirus promoters was tested by co-transfection of HeLa cell cultures (5 x 105 cells) with pPYE-CAT and pPYL-CAT, respectively, mixed with a second recombinant plasmid. The cells were lysed in 0.1 ml of 0.01 M Tris-HCl (pH 7.9), 0.15 M NaCl, 0.0015 M MgCl2, 0.65% Nonidet P40. CAT activity was analysed in 0.04 ml portions. The assays, including analysis by thin layer chromatography, were performed exactly as described by Gorman et al. (25). In analyses of viral DNA-replication, mouse 3T6 cells were transfected with viral DNA that had been excised from recombinant plasmids and then recircularized. Lowmolecular-weight DNA was selectively extracted at 42 h posttransfection and cleaved with the restriction endonucleases DpnI and BcI. After agarose gelelectrophoresis, DNA was transferred to a hybridization membrane and annealed to 32plabelled polyomavirus DNA. Viral DNA was identified by autoradiography.

RESULTS AND DISCUSSION Treatment of polyomavirus large T-antigen with a metal ion chelator The putative zinc finger in large T-antigen is located in the Cterminal half of the protein and is therefore more than 100 amino acid residues downstream of the minimal segment of the polypeptide chain that is required for specific and stable DNAbinding (6, 7). In an initial experiment to investigate whether a zinc finger is an auxiliary element in specific DNA-binding, we employed a previously described (30, 31) method for removal of zinc ions by chelation with 1, 10-phenanthroline. This compound is a strong chelator of Zn2+ ions, whereas the binding to Mg2+ ions is weak (32). As a source of large Tantigen an extract of Sf9 cells infected with the recombinant baculovirus vEV5 1LT was used. This was a much richer source of large T-antigen than extracts of polyomavirus infected mouse cells. The large T-antigen formed in insect and mouse cells has the same general properties in DNA-binding and initiation of viral DNA-synthesis (17, Bondeson and Magnusson, unpublished observations). Large T-antigen was incubated with the chelating compound in the DNA-binding reaction buffer including poly(dI-dC), but

1418 _ I 127 882 70 2 4 o O.. * 376-

^

^

a Aba

275

I1 l2

a b c d e f g h i j k Figure 3. Effect of phenanthroline on specific DNA-binding of large T-antigen. Protein extracts containing polyomavirus large T-antigen was prepared from insect Sf9 cells infected with the recombinant baculovirus vEV5lLT. The T-antigen preparations were treated with 1, 10-phenanthroline (lanes b-e), or 1, 7-phenanthroline (lanes g-k) at the indicated final concentrations. Polyomavirus DNA-fragments labelled with 32p were added and large T-antigen was immunoprecipitated. Co-precipitated DNA was subjected to polyacrylamide gel electrophoresis and analyzed by autoradiography. The phenanthroline concentrations were 10 mM, lanes b, g; 3 mM, lanes c, h; 1 mM, lanes d, i; 0.3 mM, lanes e, j; 0.1 mM, lanes f, k. Lane a: marker DNA-fragments. The length of the fragments measured in in base pairs is indicated.

without polyomavirus DNA. After 20 min the radioactively labelled polyomavirus DNA fragments were added and the incubation was continued for 60 min. DNA fragments bound to the protein were then isolated, using antibodies directed against large T-antigen, and analyzed by polyacrylamide gel electrophoresis. Figure 3 (lanes b-f) shows autoradiograms of the gels. At a concentration of 3 mM, 1,10-phenanthroline apparently inhibited stable and specific binding of large T-antigen to DNA. Large T-antigen was not irreversibly inhibited by the chelator, since protein incubated in 3 mM 1, 10-phenanthroline for 20 min, and then equilibrated with the same buffer containing 0.3 mM chelator, bound to polyoma DNA (data not shown). In this experiment no Zn2+ ions were added. The association constant for the chelator metal complex is close to 107 M-' (32). Hence the result suggests that large T-antigen forms an extremely stable complex with Zn2+, or that DNA binding was inhibited by a mechanism other than chelation of zinc ions, e.g. hydrophobic interactions (32). The latter alternative was supported by the observation that the isomer, 1, 7-phenanthroline, that does not chelate Zn2+, had the same inhibitory effect as 1, 10-phenanthroline (Fig. 3, lanes g-k). In another experiment, specific DNA-binding of large T-antigen was assayed in the presence of 0.5 mM 1, 10-, or 1, 7-phenanthroline, varying the amount of protein over a 30-fold range. The signal of specifically bound DNA was proportional to the amount of added large Tantigen, from the lowest amount which was just above the detection limit of the assay, to the highest. No effect on DNAbinding of either phenanthroline compound was observed at any large T-antigen concentration (data not shown). With RNA polymerase from Euglena gracilis 1, 10-phenanthroline, but not 1, 7-phenanthroline, in the 1-10 mM concentration range inhibits enzyme activity (30). Thus, a straightforward interpretation of the data is that specific and stable DNA-binding of polyomavirus large T-antigen does not require the presence of a zinc finger. An alternative explanation is that the zinc ion is not accessible to chelation by 1, 10-phenanthroline.

2718 Nucleic Acids Research, Vol. 18, No. 9 This explanation is unattractive, since a zinc finger that makes contact with DNA is expected to be located at the surface of the protein.

Mutation of the zinc finger motif The function of the putative zinc finger structure in large Tantigen was further studied by the construction of a mutant that had cysteine residue 452 (Fig. 1) substituted by a serine. Cysteine residue 452 has been predicted to be an essential element of the co-ordinate binding of Zn2+ ions. Hence the amino acid substitution would prevent the formation of a zinc finger. The mutant was constructed by oligonucleotide directed mutagenesis (nucleotide 1914 G- C). The base sequence of the mutagenized DNA-segment was determined and found to conform with one of the published base sequences (33) of wild-type polyomavirus DNA, except for the expected single base substitution. The mutant recombinant plasmid was designated pPYbc1081. To express only the large T-antigen, the double mutant pPYbclO81/dllO61 was constructed. The dllO61 deletion precludes the formation of mRNA encoding small and middle T-antigen.

Synthesis and specific binding to DNA of bc1081 large T-antigen The synthesis of the mutant protein was analyzed in HeLa cells transfected with plasmid pPYbclO81. Protein was labelled with [35S]methionine for 10 h, large T-antigen was extracted, immunoprecipitated and then analyzed by electrophoresis in a SDS-polyacrylamide gel (Fig. 4). The mutant protein appeared to be present in larger quantities than the wild-type protein, showing that it was not labile. To test the specific DNA-binding of the bclO81 large Tantigen, protein extracts were prepared from transfected HeLa cells and were then mixed with 32P-labelled polyoma DNAp. w

Transcriptional trans-activation by bcMM81 large T-antigen Polyomavirus early (8) and late (3) transcription is regulated by large T-antigen. The effect of the protein on early transcription is inhibitory, resulting in autoregulation of T-antigen synthesis (34). Data from in vitro experiments with simian virus 40 show

1418 I 127

-9 4

702

-6 7

3

--4 3

275

-3 0s b

Activity of bc108l large T-antigen in viral DNA-synthesis The activity of large T-antigen in the initiation of viral DNAreplication is believed to be mediated, at least in part, by DNAbinding. For this reason alone the bc1O81 mutant was expected to be replication defective. To test the replication properties of the mutant, mouse 3T6 cells were transfected with mutant and wild-type DNA that had been excised from the recombinant plasmids and recircularized. Low-molecular-weight DNA was isolated, cleaved with DpnI and Bcll and then analyzed. The result of the experiment (Fig. 6) showed that viral DNA was formed in cells transfected with genomes expressing wild-type large Tantigen, but not in cells transfected with genomes expressing the bc1O81 mutant protein. In the latter cells only DpnI sensitive input DNA was detected. In a control experiment mutant bc1O81 DNA was able to replicate in the presence of a wild-type helper (data not shown), demonstrating that the mutant genome had no cisacting defect and that the mutant protein did not have a dominant negative effect (15).

4_

882

LT-ag

a

fragments. The mutant protein, in contrast to the wild-type control, did not form a stable complex with any polyoma DNAfragment (Fig. 5). This result shows that the amino acid substitution in the putative zinc finger directly or indirectly decreased the affinity of the protein to DNA.

c

Figure 4. Synthesis of large T-antigen. Human HeLa cells were transfected with 10 ytg of recombinant plasmid DNA. Cells were labelled with [35S]methionine at 32-42 h posttransfection and large T-antigen was then extracted. Material immunoprecipitated with the monoclonal LT1 antibody was subjected to SDSpolyacrylamide gelelectrophoresis and was then analyzed by fluorography. Protein extracts were from cells transfected with pPYwt, lane a; pPYbclO81, lane b; no plasmid DNA, lane c. The position of molecular weight markers is indicated.

4~

-

12

a

b

c

d

Figure 5. DNA-binding of mutant large T-antigen. HeLa cells were transfected with recombinant plasmids and large T-antigen was extracted. Analysis of specific binding of large T-antigen was performed as described in the legend to Fig. 2. After polyacrylamide gel electrophoresis, 32P-labelled DNA was identified by autoradiography. Protein extracts were from cells transfected with pPYwt, b; pPYbc1O81, c; no plasmid DNA, d. Lane a: Total HpaII digested polyomavirus DNA. The length of the fragments are indicated.

Nucleic Acids Research, Vol. 18, No. 9 2719 that the inhibition of the early promoter is mediated by DNAbinding of large T-antigen (35, 36). The late promoter of polyomavirus, in contrast to the early promoter, is activated by

I

0,

a

b

d

c

C

Figure 6. Replication of mutant viral DNA. Viral DNA was excised from recombinant plasmids and then recircularized by treatment with DNA ligase. Mouse 3T6 cells were transfected with wild-type or mutant DNA. At 40 h posttransfection, low-molecular-weight DNA was selectively extracted and partially purified. The samples were incubated with restriction endonucleases Bcll and DpnI. After agarose gel electrophoresis, DNA was transferred to a hybridization membrane, annealed to 32P-labelled polyomavirus DNA and analyzed by autoradiography. Viral DNA used for tranfection was: a, 0.2 jig wild-type; b, 0.2 Ag dlIO61; c, 0.2 jig bc1081; d, 0.2 jig bc1O81/dllO61; e, No viral DNA. Letter L indicates the position of linear viral DNA visible in the gel after staining with ethidium bromide.

large T-antigen, probably as a result of activation of cellular transcription factors (3, 37, 38, 39). The effect of large T-antigen on viral transcription was tested using a co-transfection assay. A plasmid containing the regulatory region of polyomavirus DNA adjacent to the cat gene was used. The cat gene substituted for either the early (pPYE-CAT), or the late viral genes (pPYL-CAT). The trans-activation of the late promoter in polyomavirus DNA by large T-antigen was tested by co-transfection of HeLa cell cultures with pPYL-CAT mixed with pPYAE, pPYbc1O81/dl 1061 or pPYdllO61, respectively. Human HeLa cells were chosen because they are nonpermissive for viral DNA-replication. Hence, the results with the different plasmids do not have to be corrected for the effect of viral DNAreplication. In HeLa cells (M. Nilsson and G. Magnusson, unpublished observations) and other human cells (40), the polyomavirus transcriptional promoters have been shown to respond to the same stimuli as in mouse cells . Transfected cells were lysed after 48 h and CAT activity in the resulting protein extracts was assayed. The result shows (Fig. 7A) that both the wild-type and bc1O81 large T-antigens increased the activity of the late polyomavirus promoter to a moderate extent, suggesting that the putative zinc finger structure is not important for the activity of large T-antigen in transcriptional trans-activation. However, the activities of the two proteins can not be assessed, since the intracellular levels may be different. When HeLa cells were transfected with pPYE-CAT and pPYdllO61 the resulting CAT activity in cell extracts prepared at 40 h posttransfection was significantly lower than the activity in the absence of large T-antigen (Fig. 7B). This effect was probably a result of the inhibition of the early polyomavirus promoter by large T-antigen. The mutant bc1O81 large T-antigen had an opposite effect on the early promoter. In cells cotransfected with pPYE-CAT and pPYbc1O81/dllO61 the CAT activity at 40 h posttransfection was considerably higher than in the control cells. Our interpretation of this result is that large T-antigen has two opposing effects on the early promoter; activation and inhibition. The trans-activation depends on the

2 0-

30

A

B

L.

0

w

20

1 0a..ci

Cu

0. t)

c.us 0.=

C;

a b c d

10

a b c d

Figure 7. Trans-activation of the polyomavirus promoters. HeLa cells were transfected with 1 jig of pPYL-CAT (panel A) or 1 jig of pPYE-CAT (panel B) mixed with 2 Ag of a second plasmid expressing no T-antigen, wild-type large T-antigen, or bc1O81 large T-antigen. At 42 h posttransfection the cells were lysed and CAT activity in the extracts was assayed. The values were normalized to the protein concentration of the extracts. All data represent the average values of 2 separate experiments with less than 10% variation of each value. Plasmids used for co-transfection were: Lanes a, pPYAE; lanes b, pPYwt; lanes c, pPYdllO61; lanes d,

pPYbc1081/dl1061.

2720 Nucleic Acids Research, Vol. 18, No. 9 presence of DNA-sequences in the enhancer region (data not shown) and is probably mediated by the stimulation of a cellular transcription factor. This activity of the wild-type large T-antigen on the early promoter is normally not observed, probably because it is obscured by the second inhibitory activity. The inhibition is believed to be mediated by high affinity binding of large Tantigen to the promoter region and in the bclO81 large T-antigen only the inhibitory activity is reduced. With wild-type large Tantigen stimulation of polyomavirus early gene expression might occur under conditions when the repressor function is not operating, e. g. at low concentrations of large T-antigen. In this situation the stimulatory activity of large T-antigen would participate in a positive feed back loop that regulates early

transcription. Does the polyomavirus large T-antigen contain a zinc finger that participates in DNA-binding? The importance of the zinc finger motif for DNA-binding of large T-antigen was demonstrated by the the properties of mutant bclO81. The 452cys-ser substitution in the large T-antigen abolished high affinity DNA-binding and, probably as a consequence, the activity in the initiation of viral DNA-synthesis. There is no direct information about the structure of the mutant protein. However, its stability and activity in transcriptional trans-activation suggest that the amino acid substitution did not lead to totally changed conformation. In simian virus 40 large T-antigen, mutation of the zinc finger motif also abolish the high affinity DNA-binding. This result is paradoxical, since the C-terminal segment, including the zinc finger motif, is dispensable for the binding activity (10). Thus, the structural integrity of the zinc finger motif appears to be important for DNA-binding, even though the putative finger structure might not contribute directly to the formation of a stable complex. Our data show that large T-antigen subjected to conditions known to remove Zn2+ from proteins contaning zinc fingers, still bound specifically to DNA. This result suggests that the effect of the cys452- ser substitution was not simply due to the collapse of a zinc finger. The question then arises as to whether the polyomavirus large T-antigens contain a zinc finger. The amino acid sequence of the zinc finger motif in polyomavirus and simian virus 40 large T-antigen has approximately the right length and spacing of cysteine and histidine residues, but does not match very well with a consensus sequence of a large number of zinc finger motifs (41). However, in only a few of the proteins used for the compilation of the consensus sequence have zinc finger structures actually been demonstrated. So far the strongest evidence for the presence of a zinc finger in large T-antigen is the properties of mutant simian virus 40 proteins. Mutation of the amino acid residues that might be involved in binding of zinc ions led to loss of function, whereas mutation of most intervening amino acid residues had little effect (9).

ACKNOWLEDGEMENTS We thank Stanley Burnett and Per Elias for comments on the manuscript. This work was supported by the Swedish Cancer Society and the National Board for Technical Development.

REFERENCES 1. Schlegel,R. and Benjamin,T.L. (1978) Cell, 14, 587-599. 2. Francke,B. and Eckhart,W. (1973) Virology, 55, 126-135. 3. Kern,F.G., Pellegrini,S., Cowie,A. and Basilico,C. (1986) J. Virol., 60, 275-285.

4. Cowie,A. and Kamen,R. (1984) J. Virol., 52, 750-760. 5. Pomerantz,B.J., Mueller,C.R. and Hassell,J.A. (1983) J. Virol., 47, 600- 610. 6. Cowie,A., deVilliers,J. and Kamen,R. (1986) Mol. Cell. Biol., 6, 4344-4352. 7. Tang,W.-J. and Folk,W.R. (1989) J. Virol., 63, 242-249. 8. Farmerie, W.G. and Folk,W.R. (1984) Proc. Natl. Acad. Sci. USA, 81, 6919-6923. 9. Loeber,G., Parsons,R. and Tegtmeyer,P. (1989) J. Virol., 63, 94-100. 10. Arthur,A.K., Hoss,A. and Fanning, E. (1988) J. Virol., 62, 1999-2006. 11. Berg,J.M. (1986) Science, 232, 485-487. 12. Miller,J., McLachlan,A.D. and Klug,A. (1985) EMBO J., 4, 1609-1614. 13. Evans,R.M. and Hollenber,S.M. (1988) Cell, 52, 1-3. 14. Berg,J. (1988) Proc. NatI. Acad. Sci. USA, 85, 99-102. 15. Nilsson,S.V. and Magnusson,G (1984) J. Virol., 51, 768-775. 16. Martens,I., Nilsson,S.V., Linder,S. and Magnusson,G. (1989) J. Virol., 63, 2126-2133. 17. Rice,W.C., Lorimer,H.E., Prives,C. and Miller,L.K. (1987) J. Virol., 61, 1712-1716 18. Wigler,M., Pellicer,A., Silverstein,S. and Axel,R. (1978) Cell, 14, 725-73 1. 19. Luthman,H. and Magnusson,G. (1983) Nucleic Acids Res., 11, 1295-1308. 20. Soeda,E., Arrand,J.R., Smolar,N., Walsh,J.E. and Griffin,B.E. (1980) Nature, 283, 445 -453. 21. Tyndall,C., La Mantia,G., Thacker,C.M., Favaloro,J. and Kamn, R. (1981) Nucleic Acids Res., 9, 6231-6250. 22. Katinka,M. and Yaniv, M. (1982) Mol. Cell. Biol., 2, 1238-1246. 23. Lusky,M. and Botchan,M. (1981) Nature, 293, 79-81. 24. Nilsson,S.V. and Magnusson,G. (1983) EMBO J., 2, 2095-2101. 25. Gorman,C.M., Moffat,L.F. and Howard,B.H. (1982) Mol. Cell. Biol., 2, 1044-1051. 26. Stanssens,P., Opsomer,C., McKeown,Y.M., Kramer,W., Zabeau,M. and Fritz,H.-J. (1989) Nucleic Acids Res., 17, 4441-4454. 27. McKay,R.D.G. (1981) J. Mol. Biol., 145, 471-488. 28. Dilworth,S.M. and Griffin,B.E. (1982) Proc Natl. Acad. Sci. USA, 79, 1059-1063. 29. Laemmli,U.K. (1970) Nature, 227, 680-685. 30. Mazus,B., Falchuk,K.H. and Vallee,B.L. (1986) Biochemistry, 25, 2941- 2945. 31. Westin,G. and Schaffner,W. (1988) Nucleic Acids Res., 16, 5771-5781. 32. Auld,D.S. (1988) Meth. Enzymol., 158, 110-114. 33. Deininger,P.L., Esty,A., LaPorte,P., Hsu,H. and Friedmann,T. (1980) Nucleic Acids Res., 8, 855 - 860. 34. Cogen,B. (1978) Virology, 85, 222-230. 35. Rio,D., Robbins,A., Myers,R. and Tjian,R. (1980) Proc. Natl. Acad. Sci. USA, 77, 5706-5710. 36. Hansen,U., Tenen,D.G., Livingston,D.M. and Sharp,P.A. (1981) Cell, 27, 603-612. 37. Keller,J. and Alwine,J.C. (1984) Cell, 36, 381-389. 38. Brady,J., Loeken,M.R. and Khoury,G. (1985) Proc. Natl. Acad. Sci. USA, 82, 7299-7303. 39. Bourachot,B., Yaniv,M. and Herbomel,P. (1989) J. Virol., 63, 2567-2577. 40. Yamaguchi,Y., Satake,M. and Ito,Y. (1979) J. Virol., 63, 1040-1048. 41. Gibson,T.J., Postma,J.P.M., Brown,R.S. and Argos, P. (1988) Protein Engineering, 2, 209-218.