Identification of Simian Virus 40 Promoter DNA ... - Journal of Virology

Download PDF
7 downloads 20312 Views 336KB Size Report
Comment
repeat domains have been shown to generate a DNase I-hy- persensitive site but not .... stranded DNA and also contained unique BglII and MluI restriction sites at its 5 end. ... wild-type SV40 chromosomes which is nucleosome free. The early ...
JOURNAL OF VIROLOGY, June 1996, p. 3416–3422 0022-538X/96/$04.0010 Copyright q 1996, American Society for Microbiology

Vol. 70, No. 6

Identification of Simian Virus 40 Promoter DNA Sequences Capable of Conferring Restriction Endonuclease Hypersensitivity RAEJEAN HERMANSEN, MARK A. SIERRA, JAMES JOHNSON, MICHAEL FRIEZ,

AND

BARRY MILAVETZ*

Department of Biochemistry and Molecular Biology, University of North Dakota, School of Medicine, Grand Forks, North Dakota 58202 Received 9 November 1995/Accepted 6 March 1996

The simian virus 40 (SV40) DNA sequences found in the enhancer domain, nucleotides (nt) 103 to 177, and the early domain, nt 5149 to 5232, of the SV40 promoter have been analyzed for their ability to confer restriction endonuclease hypersensitivity in SV40 chromatin by using an SV40-based recombinant reporter system. The reporter system consists of a polylinker of various unique restriction endonuclease recognition sequences introduced into SV40 at nt 2666. We observed that the introduction of the enhancer domain at one end of the reporter and the early domain at the other end of the reporter resulted in a 20% increase in nuclease sensitivity within the reporter. In the enhancer domain, an element capable of conferring hypersensitivity was found between nt 114 and 124 with the sequence 5*CTGACTAATTG3*, which has previously been shown to be the SV40 AP-1 binding site. In the early domain, an element capable of conferring hypersensitivity was localized to nt 5164 to 5187 and had the sequence 5*CATTTGCAAAGCTTTTTGCAAAAGC3*. nucleosomes by specific interactions with the histone core of the nucleosome (32). In order to identify the specific SV40 DNA sequences responsible for phasing nucleosomes and forming an NFR within the SV40 promoter in vivo, we have developed an SV40 reporter system. The reporter consists of a polylinker of unique restriction endonuclease recognition sequences cloned at nt 2666 into an SV40 recombinant lacking the T intron and one of the copies of the enhancer. Sequences of interest were then introduced at one or both ends of the reporter, and the ability of each or both introduced sequences to change the restriction endonuclease sensitivity within the reporter was measured. Since the reporter sequence does not change, any effects on its nuclease sensitivity must result from changes in chromatin structure caused by the introduced sequence(s). Using this analysis, we have analyzed two domains implicated in the formation of the SV40 NFR in previous studies, the enhancer and the early domain, and present evidence for the presence of specific DNA sequences in each which appear to phase nucleosomes and may function in the generation of the SV40 NFR.

A large fraction of simian virus 40 (SV40) chromosomes found in lytically infected cells lack a nucleosome positioned over their promoter region between nucleotides (nt) 5150 and 400 (see Fig. 1). This nucleosome-free region (NFR), which can be observed both by electron microscopy (12, 31) and by an increased sensitivity to DNase I and restriction endonucleases (30, 33, 34, 39), is characteristic of transcriptionally activated eukaryotic chromatin (6, 8). On the basis of initial studies, the DNA sequence information necessary for defining the SV40 NFR has been shown to reside within the sequences present in the promoter (see Fig. 1) (7, 11, 13, 15, 41). The whole SV40 promoter has been shown to be sufficient to direct the formation of an NFR, and the boundaries of the NFR appear to lie within promoter sequences, notably, the early domain (around nt 5200), the enhancer domain (around nt 150), and the late domain (around nt 400). The enhancer domain alone and the 21-bprepeat domains have been shown to generate a DNase I-hypersensitive site but not an NFR, as judged by electron microscopy; the NFR also requires the presence of the early domain (15). Neither the origin of replication (15) nor a Tantigen binding site I (17) is necessary for the formation of the NFR. Over the past few years it has become increasingly clear from other eukaryotic transcription systems that specific short DNA sequences are capable of acting as targets for signaling the remodeling of chromatin structure associated with transcriptional activation and as organizing centers for phasing nucleosomes. A number of DNA sequences bound by inducible transcriptional activators have been shown to be targets for the SWI-SNF complex, which remodels nucleosome structure during transcriptional activation (4, 23, 26). The SWI-SNF complex, which contains a DNA-dependent ATPase activity, functions to disrupt the structure of the nucleosome, apparently by destabilizing the H2A-H2B dimers present in the nucleosome. Other sequences, including the GRF2 binding site (5), are involved in nucleosome phasing independent of SWI-SWF (23). In addition, certain DNA sequences phase

MATERIALS AND METHODS Cells and infections. BSC-1 cells obtained from the American Type Culture Collection were used for the preparation of SV40 reporter viruses and SV40 chromatin. Cells were maintained at 378C in 5% CO2 in Eagle’s minimum essential medium (GIBCO) containing 10% fetal bovine serum (GIBCO) and 100 mg of gentamicin (Schering) per ml. Subconfluent monolayers of cells were infected with reporter SV40 virus as previously described (17). Infected cells were maintained at 378C in Eagle’s minimum essential medium containing 2% fetal bovine serum and 100 mg of gentamicin per ml. At 46 h postinfection, cells were radiolabeled for 3 h with (per flask) 0.75 mCi of [3H]thymidine in 3 ml of Eagle’s minimum essential medium. Isolation and purification of SV40 chromatin. SV40 reporter chromosomes were isolated from infected nuclei and purified as described previously (19), with modifications. Generally, a single 75-cm2 T flask was used for each virus. Infected nuclei were extracted with 0.2 ml of nucleus extraction buffer (10 mM HEPES [N-2-hydroxyethylpiperazine-N9-2-ethanesulfonic acid] [pH 7.5], 1 mM EDTA, 0.5 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride). SV40 reporter chromosomes were separated from virus and cellular debris by sedimentation on a glycerol step gradient containing 1 ml of 10% glycerol in buffer C (10 mM HEPES [pH 7.5], 5 mM KCl, 1 mM EDTA, 0.2 mM MgCl2, 0.5 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride) on a cushion of 0.1 ml of 50% glycerol in buffer C. Step gradients were centrifuged in a TLA 100.3 rotor at 50,000 rpm for 35 min in a Beckman TLA 100 ultracentrifuge. Fractions (0.2 ml)

* Corresponding author. Phone: (701) 777-4708. Fax: (701) 777-3894. 3416

VOL. 70, 1996

NUCLEASE-SENSITIVE SV40 DNA

were collected from the top to the bottom. The peak of the SV40 chromosomes was found in fraction 4. Restriction endonuclease analysis of SV40 chromatin. Aliquots (12.5 ml) from fractions 3 and 4 from glycerol step gradients were digested at 378C for 30 min with saturating amounts of the appropriate restriction endonuclease following adjustment to reaction conditions with a 1/10 volume of buffer A (100 mM Tris-HCl [pH 7.5], 100 mM MgCl2, 10 mM dithiothreitol, 10 mg of bovine serum albumin per ml). Enzymes used included ApaLI (10,000 U/ml), BglI (10,000 U/ml), BglII (8,000 U/ml), MluI (10,000 U/ml), NheI (5,000 U/ml), SalI (20,000 U/ml), and XhoI (20,000 U/ml). All restriction endonucleases were obtained from New England Biolabs. Electrophoresis and fluorography. Following restriction endonuclease digestions of SV40 chromatin, the products were deproteinized, separated electrophoretically in submerged 1% agarose gels, and prepared for fluorography (21). The extent of conversion of form I and II intact SV40 DNA to form III linear SV40 DNA by each endonuclease was quantitated by scanning densitometry with a Beckman DU640 instrument. Preparation of reporter constructs. The parental reporter pBM129, which consists of SV40 mutant strain in(or)1411 (a gift from Thomas Shenk [36]) with a polylinker at nt 2666 containing unique restriction endonuclease sites for MluI, ApaLI, PmlI, NciI, and BglII, was prepared by using a multistep strategy and standard procedures (20, 28). First, an M13mp18-SV40 recombinant containing SV40 strain 776 cloned into the EcoRI site of M13mp18 was prepared. Next, the PflMI fragment from the mutant (nt 4558 to 1007) was substituted for the corresponding wild-type fragment in the recombinant. Single-stranded DNA from this recombinant was cleaved specifically at the HpaI site at nt 2666 by using oligonucleotide-directed restriction endonuclease digestion (20) with the oligonucleotide 59GCAATAAACAAGTTAACAACAACAAT39. The linearized single-stranded DNA was then made double stranded for ligation and transformation by priming with the oligonucleotide 59GAGATCTACGCGTGCAACAAC AACAATTGCA39, which was complementary to the 39 end of the singlestranded DNA and also contained unique BglII and MluI restriction sites at its 59 end. The SV40 portion of the recombinant was then transferred into the EcoRI site of a pBR322 derivative lacking the tetracycline resistance gene. This pBR322-SV40 recombinant was cleaved with MluI, and the polylinker was introduced by ligating the complementary oligonucleotides 59CGCGTGAGTGCA CGTGTCCGGGT39 and 59CGCGACCCGGACACGTGCACTCA39 into the MluI site. DNA sequences from the SV40 promoter were prepared either as PCR products or as complementary oligonucleotides and introduced into either the BglII site or the MluI site adjacent to the reporter. DNA sequences introduced into the BglII site contained a BglII and an XhoI site at one end and a BamHI and a SmaI site at the other end. Similarly, DNA sequences introduced into the MluI site contained a BssHII and a NheI site at one end and an MluI and a SalI site at the other end. All oligonucleotides were synthesized on an Applied Biosystems model 391 oligonucleotide synthesizer. PCR amplifications were performed in a Perkin-Elmer model 480 thermal cycler. The oligonucleotides used for the preparation of the various constructs were as follows: early-region BglII site, 59GAAGATCTCGAGAAGCCGCCTCGGC CTC39, 59GAGGATCCCGGGTAAAACTTTATCCATC39, 59GAAGATCTC GAGAGTGAGGAGGCTTTT39, and 59GAGGATCCCGGGCCTAGGCCTC CAAAA39; nt 5164-to-5187 BglII site, 59GATCTCGAGCTTTTGCAAAAAGC TTTGCAAAGCCCGG39 and 59GATCCCGGGCTTTGCAAAGCTTTTTGC AAAAGCTCGA39; enhancer MluI site, 59TGACACGCGTCGACGGTGTGG AAAGTCCCC39, 59TGACGCGCGCTAGCACTATGGTTGCTGACT39, 59T GACGCGCGCTAGCGAGATGCATGCTTTGC39, 59TGACACGCGTCGAC GCAGAAGTATGCAAAGC39, 59TGACACGCGTCGACTCCCCAGCAGGC AGAAT39, 59TGACGCGCGCTAGCCTGCTGGGGAGCCTGG39, 59TGACA CGCGTCGACGCATGCATCTCAATTA39, and 59TGACGCGCGCTAGCCT GACTAATTGAGATGC39; AP-1, nt 114-to-124 MluI site, 59CGCGCTAGCTG ACTAATTGAGTCGA39 and 59CGCGTCGACTCAATTAGTCAGCTAG39. The sequences of all reporter constructs were confirmed by DNA sequencing (29).

RESULTS Construction of recombinants and overall strategy. Sequences present in the SV40 promoter which may play a role in generating an NFR were identified by their ability to confer nuclease hypersensitivity upon an adjacent region of DNA in an SV40-derived reporter construct. The parental recombinant used for the preparation of the SV40 reporter derivatives is shown in Fig. 1. It consists of pBR322 sequences without the tetracycline resistance gene and SV40 lacking one copy of the enhancer and the large-T-antigen intron to which has been added a linker of unique restriction sites at nt 2666 as the reporter. The strategy which we followed for the preparation of reporter constructs was to introduce sequences of interest at

3417

FIG. 1. Schematic representation of the parental SV40 reporter construct. Both the SV40 promoter (above) and reporter (below) have been expanded to show their organization and relevant restriction endonuclease sites. The major structural elements of the SV40 promoter are indicated, along with the region in wild-type SV40 chromosomes which is nucleosome free. The early and late domains correspond to the approximate positions of the RNA polymerase binding sites for early and late transcription, respectively. The restriction endonuclease sites which are hypersensitive in the wild-type SV40 chromosomes, the BglI, KpnI, and NgoMI sites, are indicated with asterisks. The positions of the deleted copy of the enhancer and the T-antigen intron are also indicated. The sizes of the various parts of the parental reporter are not drawn to scale.

one or both ends of the reporter in the recombinant, to excise the SV40 portion of the recombinant, and to prepare virus and subsequently chromatin from the SV40 portion. The effects of the inserted sequences on nuclease sensitivity in the adjacent reporter were measured by quantitating the proportion of the SV40 chromosomes cleaved by each of the restriction endonucleases recognizing a site within the reporter. Restriction endonucleases which recognize palindromic sequences require both strands of the target DNA to be available in order to cleave the DNA (37). In a population of SV40 chromosomes, the proportion of the chromosomes which is cleaved by a particular restriction endonuclease will depend upon the extent to which the DNA sequence in each of the chromosomes in the population is free of associated proteins (37). Because the sequence of the reporter is not changed in any of the constructs analyzed, changes in the restriction endonuclease sensitivity in the different constructs must reflect changes in chromatin structure around the reporter. In this study, a sequence was judged to confer hypersensitivity if it was able to increase the amount of cleavage by greater than 15% for most of the restriction sites in the reporter. Although the nuclease sensitivity within the reporter generally changed upon the introduction of test sequences, the changes observed were small (between 5 and 10%) for most of the test sequences and large (between 20 and 25%) for only a few combinations of sequences. In preliminary studies, we determined the optimal digestion conditions to be incubation with high concentrations of enzyme at 378C for 30 min with the low-ionic-strength buffer which we have previously used (17). We observed that under

3418

HERMANSEN ET AL.

J. VIROL.

FIG. 2. Representative autoradiograms of [3H]thymidine-radiolabeled DNA from partially purified SV40 chromosomes cleaved with restriction endonucleases recognizing sites within the reporter of the parental construct pBM129-1 (odd-numbered lanes) and a construct containing the early domain and enhancer, pBM140-15 (even-numbered lanes). Lanes 1 and 2, undigested; lanes 3 and 4, BglII; lanes 5 and 6, ApaLI; lanes 7 and 8, MluI. The positions of form I, II, and III SV40 DNA are indicated.

these conditions, with the exception of PmlI, all of the enzymes cleaved purified DNA completely; none of the enzymes overdigested the DNA; and, consistent with previous investigations (17, 37), nucleosome sliding did not occur during the course of the digestion (data not shown). Moreover, the conditions chosen have been previously shown (17) to minimize any effects on restriction endonuclease sensitivity associated with the disruption of virions (21, 22, 36) or the varying amounts of structural protein vp-1 present (2). A typical example of the results obtained by restriction endonuclease digestion of chromosomes is shown in Fig. 2. Comparing digestion of pBM129-1, the parental construct, with digestion of pBM140-15, a construct containing the enhancer and the early domain (see below), we observed very little conversion of form I and II circular DNA to form III linear DNA in the absence of added restriction endonuclease in chromosomes from either construct. The nicked form II DNA is apparently a result of radiochemical damage (17). With added restriction endonuclease, however, a significant proportion of the DNA originally present in the chromosomes was converted to linear form III. The amount of conversion depended upon both the restriction endonuclease used and the construct tested. However, for each enzyme we observed more digestion of the pBM140-15 construct than of the pBM129-1 construct. Analysis of constructs containing the early domain and enhancer. In order to determine whether the combination of enhancer and early domain could confer restriction endonuclease hypersensitivity on the reporter, as expected from previous studies, we prepared a series of constructs containing either the enhancer, the early domain, or combinations of the two domains (Fig. 3A). In order to determine the proportion of the SV40 chromosomes in the population from each construct which was digested by a particular restriction endonuclease, data like those shown in Fig. 2 were quantitated by densitometry. Averages from at least nine separate analyses with each restriction endonuclease of chromosomes from at least three separate infections were then calculated.

FIG. 3. Effects of the early domain and enhancer on restriction endonuclease sensitivity within the reporter. (A) Schematic representation of reporter constructs containing a copy of the SV40 early domain, nt 5149 to 5232, and/or a copy of the SV40 enhancer, nt 103 to 178. The relative orientation of each domain is indicated. The position of each restriction endonuclease site used in the analysis is indicated: M, MluI; S, SalI; N, NheI; A, ApaLI; B, BglII; X, XhoI. (B) The percentage of the SV40 DNA which was cleaved by digestion of SV40 chromosomes from each construct with restriction endonuclease was determined by densitometric quantitation of autoradiograms as shown in Fig. 2. The averages are based upon at least nine separate analyses of SV40 chromosomes from at least three separate infections for each restriction site per construct.

Since the parental construct pBM129-1 was used for comparison with other constructs, it was characterized for restriction endonuclease sensitivity in its promoter and reporter regions. At the BglI site within the promoter of this construct, approximately 63% 6 5% of the chromosomes were cleaved. This value is very close to the values which we previously reported for BglI digestion within the promoters of the wildtype viruses 776 (58% 6 8%), SVS (63% 6 6%), and VA45-54 (57% 6 7%) (17) and is also similar to other reported values (19, 33, 35, 37), indicating that the assay functioned reliably. Next, the nuclease sensitivities for each of the sites within the reporter were measured. As shown in Fig. 3B, at the ApaLI site approximately 18% of the chromosomes were cleaved, at the BglII site approximately 47% were cleaved, and at the MluI site approximately 32% were cleaved. With totally random spacing of nucleosomes one would expect that approximately 30 to 35% of the SV40 chromosomes would be cleaved at any

VOL. 70, 1996

of the sites (35, 37). This expected percentage of digestion was based upon the observation that 145 bp of DNA is found associated with the nucleosome in SV40 chromatin and approximately 60 bp of DNA is found in the internucleosomal linker. The probability that a sequence of DNA will be found within an internucleosomal linker and be available for digestion was calculated by dividing 60 by 205 (the nucleosome repeat length) (35, 37). Although the expected level of digestion was obtained at the MluI site, it appears that the ApaLI site is somewhat hyposensitive while the BglII site is somewhat hypersensitive in this construct. The structures of the constructs containing the enhancer and the early domain are shown in Fig. 3A, and the nuclease sensitivities at each site in the reporter are shown in Fig. 3B. Maximal conferred restriction endonuclease hypersensitivity was obtained from pBM140-15, which contains the enhancer at the MluI site and the early domain introduced at the BglII site in the same relative orientation as found in the SV40 promoter. This construct showed at least an 11% increase in restriction endonuclease sensitivity at each of the reporter sites in comparison with values for the construct without any inserted domains, pBM129-1. For example, we observed a 16% increase in the proportion of chromosomes cleaved at the BglII site (from 47 to 63% digestion), an 11% increase at the ApaLI site (from 18 to 29%), and a 24% increase at the MluI site (from 32 to 56%). We also observed a moderately high level of sensitivity at the restriction sites introduced at a position adjacent to the two new domains. At the XhoI site the sensitivity was 54%, at the NheI site the sensitivity was 35%, and at the SalI site the sensitivity was 45%. Introduction of the enhancer domain in the opposite orientation relative to that of the early domain (pBM140-5) or the enhancer by itself (pBM143-13) resulted in levels of conferred restriction endonuclease hypersensitivity within the reporter lower than those in the construct pBM140-15. The proportion of the chromosomes cleaved at each of the restriction sites was elevated by approximately 6 to 15% above the levels in the parental construct pBM129-1 but were about 5 to 10% less than in pBM140-15. No conferred restriction endonuclease hypersensitivity was observed in a construct, pBM131-1, containing only the early domain introduced at the BglII site. Analysis of constructs containing deletions in the enhancer. In order to determine which sequences in the enhancer were necessary to confer restriction endonuclease hypersensitivity when they were combined with the early domain, the enhancer was subjected to a deletional analysis. A series of constructs in which various portions of the enhancer were deleted (Fig. 4A) were prepared and subsequently analyzed. As shown in Fig. 4B, maximal conferred hypersensitivity was obtained in those constructs in which the sequences from nt 113 to 133 were present. Constructs containing nt 113 to 133, including pRH115, pRH8-5, pRH9-12, and pRH11-6, conferred levels of nuclease sensitivity which were very similar to or greater than the levels obtained from the parental construct pBM140-15. In contrast, the proportion of chromosomes cleaved by restriction endonucleases in constructs lacking the sequences from nt 113 to 133, pBM174-13, pBM177-54, and pRH2-9, was approximately 10% lower than that in the pBM140-15 construct. The principal DNA binding sequence located within nt 113 to 133 was the AP-1 site located from nt 114 to 124. In order to determine whether this sequence was capable of conferring nuclease hypersensitivity, a construct containing only the AP-1 site and the early domain was prepared. This construct, pRH18-9, conferred levels of nuclease hypersensitivity which were similar to the values obtained with the other constructs

NUCLEASE-SENSITIVE SV40 DNA

3419

FIG. 4. Effects of deletions in the enhancer on the restriction endonuclease sensitivity within the reporter in constructs containing both an early domain and an enhancer. (A) Schematic representation of reporter constructs containing a complete copy of the SV40 early domain and various deletions within the SV40 enhancer. The region deleted within the enhancer for each construct is indicated. The position of each restriction endonuclease site used in the analysis is indicated: M, MluI; S, SalI; N, NheI; A, ApaLI; B, BglII; X, XhoI. (B) The percentage of the SV40 DNA which was cleaved by digestion of SV40 chromosomes from each construct with restriction endonuclease was determined by densitometric quantitation of autoradiograms as shown in Fig. 2. The averages are based upon at least nine separate analyses of SV40 chromosomes from at least three separate infections for each restriction site per construct.

3420

HERMANSEN ET AL.

FIG. 5. Effects of deletions and insertions within the early domain on the restriction endonuclease sensitivity within the reporter in constructs containing an early domain and an enhancer. (A) Schematic representation of reporter constructs containing deletions within the SV40 early domain and a complete copy of the SV40 enhancer. The region deleted within the early domain is indicated for each construct. The position of each restriction endonuclease site used in the analysis is indicated: M, MluI; S, SalI; N, NheI; A, ApaLI; B, BglII; X, XhoI. (B) The percentage of the SV40 DNA which was cleaved by digestion of SV40 chromosomes from each construct with restriction endonuclease was determined by densitometric quantitation of autoradiograms as shown in Fig. 2. The averages are based upon at least nine separate analyses of SV40 chromosomes from at least three separate infections for each restriction site per construct.

containing nt 113 to 123 at all the sites except the SalI site, for which the value was about 10% lower. Analysis of constructs containing deletions in the early domain. In order to determine whether sequences in the early domain were also involved in conferring nuclease hypersensitivity, we prepared and analyzed a series of reporter constructs containing an intact enhancer and either deletions or insertions of sequences found in the early domain (Fig. 5A). As shown in Fig. 5B, deletion of the sequences from nt 5149 to 5186 but not deletion of the other sequences in the construct pBM164-1 had a moderate effect upon the conferred restriction endonuclease hypersensitivity. We observed a 9 to 14% reduction in the proportion of chromosomes cleaved with re-

J. VIROL.

striction endonucleases at each site in the reporter in this construct relative to values for pBM140-15. These values for restriction endonuclease sensitivity were also below the values obtained for the reporter construct pBM143-13, which contained only a single copy of the enhancer at the MluI site of the reporter. However, the observed restriction endonuclease sensitivities were still about 4 to 7% above the values obtained for the parental reporter pBM129-1. Constructs pBM159-7 and pBM155-6, which contained nt 5149 to 5186, resulted in chromatins whose restriction endonuclease sensitivities within their reporters were similar to those found in the parental construct pBM140-15, although there were some differences, notably at the XhoI site in pBM155-6 and at the SalI site in pBM159-7. These results suggest that the sequences from nt 5149 to 5186 may also play a role in conferring the restriction endonuclease hypersensitivity seen with pBM140-15, since they had the greatest effect upon nuclease hypersensitivity within the reporter. However, the smaller changes observed at some of the sites in pBM155-6 suggest that sequences in this region may also be involved in generating optimal nuclease hypersensitivity. In order to confirm that the sequence responsible for the cooperative formation of the nuclease-hypersensitive reporter was found between nt 5164 to 5187 and to further delineate the minimal functional sequence, we analyzed a set of constructs containing the enhancer at one end of the reporter and a sequence from nt 5164 to 5187 in either orientation at the other end. This sequence was chosen because it has been suggested on the basis of its proximity to T-antigen site I and its tandemly duplicated sequence to be involved in the formation of the SV40 NFR (17). When this sequence was introduced in the correct orientation (pBM176-1), we observed that the reporter was approximately as hypersensitive as it was in pBM140-15 at the BglII, MluI, and SalI sites but not at the XhoI and NheI sites. When the sequence from nt 5164 to 5187 was introduced in the opposite orientation, we observed very low levels of nuclease sensitivity within the reporter. At the BglII site, the nuclease sensitivity, 33%, was significantly lower than the value obtained from pBM129-1, the reporter without added sequences. Interestingly, the level of nuclease sensitivity at the XhoI site in these two constructs was the lowest observed for any construct containing an XhoI site. Taken together, these results suggest that the sequence from nt 5164 to 5187 in conjunction with a sequence(s) within the enhancer may play an important role in conferring restriction endonuclease hypersensitivity on the reporter. Analysis of constructs containing only nt 5164 to 5187 or nt 114 to 124. In order to determine whether the sequence from nt 114 to 124 or nt 5164 to 5187 was capable of conferring nuclease hypersensitivity by itself, constructs containing each of the sequences were prepared (Fig. 6A) and chromatin from each of the constructs was analyzed for nuclease sensitivity. Because of the difference in nuclease sensitivity which we observed with the constructs containing the enhancer and nt 5164 to 5187 in different orientations, we prepared constructs containing nt 5164 to 5187 in each orientation. As shown in Fig. 6B, similar levels of nuclease hypersensitivity were conferred upon the reporter by either the AP-1 site, nt 114 to 124, or the early sequence, nt 5164 to 5187, in either orientation. In each case there was an approximately 15% increase in the nuclease sensitivity within the reporter. The nuclease sensitivity within the reporter of the construct containing only the AP-1 site was similar to that of the construct containing the AP-1 site and the early region, and as in this construct a reduced sensitivity was conferred on the SalI site. Interestingly, the levels of nuclease sensitivity at the XhoI

VOL. 70, 1996

NUCLEASE-SENSITIVE SV40 DNA

FIG. 6. Effects on the restriction endonuclease sensitivity within the reporter in constructs containing inserts of either nt 114 to 124 or nt 5164 to 5187. (A) Schematic representation of reporter constructs containing either the AP-1 site (nt 114 to 124) or the early sequence (nt 5164 to 5187). The position of each restriction endonuclease site used in the analysis is indicated: M, MluI; S, SalI; N, NheI; A, ApaLI; B, BglII; X, XhoI. (B) The percentage of the SV40 DNA which was cleaved by digestion of SV40 chromosomes from each construct with restriction endonuclease was determined by densitometric quantitation of autoradiograms as shown in Fig. 2. The averages are based upon at least nine separate analyses of SV40 chromosomes from at least three separate infections for each restriction site per construct.

site in constructs containing nt 5164 to 5187 alone were over 20% higher than they were in those constructs containing the sequence and the enhancer (compare pBM178-7 and pBM178-1 [Fig. 6B] with pBM176-1 and pBM176-4 [Fig. 5B]). At the other sites the levels were similar to those for other constructs containing this sequence. DISCUSSION In this study, DNA sequences from two domains of the SV40 promoter, the enhancer and the early domain, were analyzed for their ability to confer restriction endonuclease hypersensitivity upon a reporter with unique restriction endonuclease sites. Within the enhancer, the AP-1 site found between nt 114 and nt 124, 59CTGACTAATTG39, was shown to confer high levels of hypersensitivity either alone or in combination with an intact early domain, while in the early domain, a sequence from nt 5164 to 5187, 59CATTTGCAAAGCTTTTTGCAAA AGC39, was shown to confer moderate levels of nuclease hypersensitivity either alone or in combination with the intact enhancer. The AP-1 site at nt 114 to 124 is the binding site for the jun/fos oncogenes (42) and is located in a domain which has previously been shown to be necessary for the generation of an

3421

NFR and a nuclease-hypersensitive site (11, 15). The AP-1 site occurs in the region in the enhancer, around nt 120, which is hypersensitive following a region of chromatin resistant to digestion with DNase I (30); is a hot spot for psoralen crosslinking at nt 120 (22); and is located at a boundary for the NFR around nt 120 (33). It is not clear how the AP-1 sequence causes the phasing of nucleosomes which results in the observed conferred nuclease hypersensitivity. AP-1 could function by disrupting a preformed nucleosome via the SWI-SNF pathway, as has been suggested for a number of inducible genes (reviewed in references 4, 23, and 26), or by phasing nucleosomes immediately following replication, as has been suggested for GRF2 in the GAL4 promoter (9). This phasing could result from the presence of a factor binding to AP-1 which acts as a barrier to the formation of a nucleosome over that site or by the inherent ability of the DNA to bend (32). We favor a mechanism by which factor binding results in the generation of a physical barrier to nucleosome formation. This model is consistent with the observation that binding of jun/fos to the AP-1 site results in a structure in which the proteins extend both above and below the plane of the DNA and the fact that the DNA is bent following the binding of jun/fos (16). Either of these two events would be likely to create a physical barrier to positioning a nucleosome over the site. Since the direction of bending of the AP-1 site was shown to differ depending on whether jun homodimers or jun/fos heterodimers were used (16), it is possible that nucleosome positioning around an AP-1 site is more dependent on the type of bend than simply on the physical occupation of a binding site. If the jun/fos oncogenes are responsible for generating the observed nucleosome phasing, this result would suggest that the jun/fos oncogenes might function in part as a ‘‘master’’ transcriptional activator by causing the formation of an NFR within a promoter previously silenced by the presence of a nucleosome. The second sequence identified in this analysis was located between nt 5164 and 5187 within the early domain and appeared to act in concert with the enhancer sequences. This sequence is located in the region of the early domain previously shown to be a boundary for the NFR between nt 5100 and 5200 (33) and at the beginning of the nuclease-hypersensitive site at nt 5212 (30), consistent with its suggested role in generating an NFR. Although the sequence is similar to the OCT sequence found in the enhancer (17), the difference in sequence is significant since it is sufficient to prevent binding to this sequence by the OCT-binding protein (3). Either the sequence is interacting with a different protein to cause nucleosome phasing as described above for the AP-1 site or this AT-rich sequence may be capable of causing bends in the DNA which could affect the positioning of nucleosomes (10, 18, 38). Although no biological activity has yet been identified for this sequence, it appears to be required for viability. We were not able to isolate a SV40 virus which contained a deletion of only this sequence, despite a number of attempts (data not shown). The sequences identified in this study are not homologous to any sequences previously shown to be involved in the generation of nuclease-hypersensitive sites. In yeasts, nuclease hypersensitivity has been shown to be generated by the sequence 59YNNYYACCCG39 in the GALI-GAL10 promoter (9) and by the DNA binding sequences for PHO2, 59GAATAGGCAA 5TCTCTAAA39, and PHO4, 59AATTATCACGTTTTCGCA TA39 and 59GCACTCACACGTGGGA39 (5), found in the PHO5 promoter. In the mouse mammary tumor virus promoter, nuclease hypersensitivity is generated by binding of the glucocorticoid hormone receptor to the sequence 59TGTTC T39 (27, 40). Similarly, in the rat tyrosine aminotransferase

3422

HERMANSEN ET AL.

gene nuclease hypersensitivity is generated upon binding of the glucocorticoid receptor to the sequence 59TGTTCT39 in its promoter (14, 24, 25). These results also indicate that the ability to confer nuclease hypersensitivity is sequence specific, since only one 20-bp sequence in the early domain and one 10-bp sequence in the enhancer appeared to be responsible for conferring maximal nuclease hypersensitivity. Sequence specificity is also shown by the fact that constructs with similar sizes have very different abilities to confer nuclease hypersensitivity. The sequence specificity observed in the enhancer is particularly interesting because the enhancer is known to contain a number of binding sites for transcriptional activators (42). Although one might intuitively expect that any DNA sequence capable of binding a transcriptional factor might be able to efficiently phase nucleosomes and generate an NFR, our results indicate that not all of these binding sites for transcriptional activators actually function to confer nuclease hypersensitivity. It should also be noted that the generation of nuclease hypersensitivity did not require the presence of other DNA sequences involved in transcription, such as the TFIID binding site or SP-1 binding sites normally present in the SV40 promoter. ACKNOWLEDGMENT This project was supported by NSF EPSCOR grant OSR 9108770 to the State of North Dakota. REFERENCES 1. Almer, A., H. Rudolph, A. Hinnen, and W. Horz. 1986. Removal of positioned nucleosomes from the yeast PHO5 promoter upon PHO5 induction releases additional upstream activating DNA elements. EMBO J. 5:2689– 2696. 2. Ambrose, C., V. Blasquez, and M. Bina. 1986. A block in initiation of simian virus 40 assembly results in the accumulation of minichromosomes containing an exposed regulatory region. Proc. Natl. Acad. Sci. USA 83:3287–3291. 3. Baumraker, T., R. Sturm, and W. Herr. 1988. OBP100 binds remarkably degenerate octamer motifs through specific interactions with flanking sequences. Genes Dev. 2:1400–1413. 4. Carlson, M., and B. C. Laurent. 1994. The SNF/SWI family of global transcriptional activators. Curr. Biol. 6:396–402. 5. Chasman, D. I., N. F. Lue, A. R. Buchman, J. W. LaPointe, Y. Lorch, and R. D. Kornberg. 1990. A yeast protein that influences the chromatin structure of UASG and functions as a powerful auxiliary gene activator. Genes Dev. 4:503–514. 6. Cremisi, C. 1981. The appearance of DNase I hypersensitive sites at the 59 end of the late SV40 genes is correlated with the transcriptional switch. Nucleic Acids Res. 9:5949–5964. 7. Gerard, R. D., M. Woodworth-Gutai, and W. A. Scott. 1982. Deletion mutants which affect the nuclease-sensitive site in simian virus 40 chromatin. Mol. Cell. Biol. 2:782–788. 8. Gross, D. S., and W. T. Garrard. 1988. Nuclease hypersensitive sites in chromatin. Annu. Rev. Biochem. 57:159–197. 9. Guo, Z.-S., and M. L. DePamphillis. 1992. Specific transcription factors stimulate simian virus 40 and polyomavirus origins of DNA replication. Mol. Cell. Biol. 12:2514–2524. 10. Harrington, R. E. 1992. DNA curving and bending in protein-DNA recognition. Mol. Microbiol. 6:2549–2555. 11. Innis, J. W., and W. A. Scott. 1984. DNA replication and chromatin structure of simian virus 40 insertion mutants. Mol. Cell. Biol. 4:1499–1507. 12. Jakobovits, E. B., S. Bratosin, and Y. Aloni. 1980. A nucleosome-free region in SV40 minichromosomes. Nature (London) 285:263–265. 13. Jakobovits, E. B., S. Bratosin, and Y. Aloni. 1982. Formation of a nucleosome-free region in SV40 minichromosomes is dependent upon a restricted segment of DNA. Virology 120:340–348. 14. Jantzen, H.-M., U. Strahle, B. Gloss, F. Stewart, W. Schmid, M. Boshart, R. Miksicek, and G. Schultz. 1987. Cooperativity of glucocorticoid response elements located far upstream of the tyrosine aminotransferase gene. Cell 49:29–38.

J. VIROL. 15. Jongstra, J., T. L. Reudelhuber, P. Oudet, C. Benoist, C.-B. Chae, J.-M. Jeltsch, D. J. Mathis, and P. Chambon. 1984. Induction of altered chromatin structures by simian virus 40 enhancer and promoter elements. Nature (London) 307:708–714. 16. Kerppola, T. K., and T. Curran. 1991. Fos-Jun heterodimers and jun homodimers bend DNA in opposite orientations: implications for transcriptional factor cooperativity. Cell 66:317–326. 17. Kube, D., and B. Milavetz. 1989. Generation of a nucleosome-free promoter region in SV40 does not require T-antigen binding to site I. Virology 172: 100–105. 18. Laundon, C. H., and J. D. Griffith. 1988. Curved helix segments can uniquely orient the topology of supertwisted DNA. Cell 52:545–549. 19. Milavetz, B. 1986. Analysis of the origin-specific nucleosome-free region in SV40 encapsidation intermediates. Virology 153:310–313. 20. Milavetz, B. 1989. Oligonucleotide-directed restriction endonuclease digestion. Nucleic Acids Res. 17:3322. 21. Milavetz, B., and T. Hopkins. 1982. Simian virus 40 encapsidation: characterization of early intermediates. J. Virol. 43:830–839. 22. Ostrander, E. A., R. A. Karty, and L. M. Hallick. 1988. High resolution psoralen mapping reveals an altered DNA helical structure in the SV40 regulatory region. Nucleic Acids Res. 16:213–227. 23. Peterson, C. L., and J. W. Tamkun. 1995. The SWI-SNF complex: a chromatin remodeling machine? Trends Biochem. Sci. 20:143–146. 24. Pina, B., U. Bruggermeier, and M. Beato. 1990. Nucleosome positioning modulates accessibility of regulatory proteins to the mouse mammary tumor virus promoter. Cell 60:719–731. 25. Reik, A., G. Schultz, and A. F. Stewart. 1991. Glucocorticoids are required for establishment and maintenance of an alteration in chromatin structure: induction leads to a reversible disruption of nucleosomes over an enhancer. EMBO J. 9:2569–2576. 26. Richard-Foy, H. 1994. A way into the packaging. Nature (London) 370:417– 418. 27. Richard-Foy, H., and G. L. Hager. 1987. Sequence-specific positioning of nucleosomes over the steroid-inducible MMTV promoter. EMBO J. 6:2321– 2328. 28. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 29. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. 30. Saragosti, S., S. Cereghini, and M. Yaniv. 1982. Fine structure of the regulatory region of simian virus 40 minichromosomes revealed by DNase I digestion. J. Mol. Biol. 160:133–146. 31. Saragosti, S., G. Moine, and M. Yaniv. 1980. Absence of nucleosomes in a fraction of SV40 chromatin between the origin of replication and the region coding for the late leader RNA. Cell 20:65–73. 32. Schrader, T. E., and D. M. Crothers. 1989. Artificial nucleosome positioning sequences. Proc. Natl. Acad. Sci. USA 86:7418–7422. 33. Scott, W. A., C. F. Walter, and B. L. Cryer. 1984. Barriers to nuclease Bal31 digestion across specific sites in simian virus 40 chromatin. Mol. Cell. Biol. 4:604–610. 34. Scott, W. A., and D. J. Wigmore. 1978. Sites in simian virus 40 chromatin which are preferentially cleaved by endonucleases. Cell 15:1511–1518. 35. Shelton, E. R., P. M. Wassarman, and M. L. DePamphillis. 1980. Structure, spacing, and phasing of nucleosomes on isolated forms of mature simian virus 40 chromosomes. J. Biol. Chem. 255:771–782. 36. Shenk, T. 1978. Construction of a viable SV40 variant containing two functional origins of DNA replication. Cell 13:791–798. 37. Tack, L. C., P. M. Wassarman, and M. L. DePamphillis. 1981. Chromatin assembly. Relationship of chromatin structure to DNA sequence during simian virus 40 replication. J. Biol. Chem. 256:8821–8828. 38. Travers, A. A. 1990. Why bend DNA? Cell 60:177–180. 39. Varshavsky, A. J., O. H. Sundin, and M. J. Bohn. 1978. SV40 viral minichromosomes: preferential exposure of the origin of replication as probed by restriction endonucleases. Nucleic Acids Res. 5:3469–3477. 40. Vogel, K., W. Horsz, and A. Hinnen. 1989. The two positively acting regulatory proteins PHO2 and PHO4 physically interact with PHO5 upstream activation regions. Mol. Cell. Biol. 9:2050–2057. 41. Wigmore, D. J., R. W. Eaton, and W. A. Scott. 1980. Endonuclease-sensitive regions in SV40 chromatin from cells infected with duplicated mutants. Virology 104:462–473. 42. Wildman, A. G. 1988. Regulation of SV40 gene expression. Biochem. Cell Biol. 66:567–577.