Repression of Polyoma Virus DNA Replication by 5”Flanking Region of Mouse DNA Polymerase Gene Containing Transcriptional Silencer Elements*

Dual cis-acting sil.encer elements are located upstream of the mouse DNA polymerase @ gene (Yama- guchi, M., Hayashi, ‘Y., and Matsukage, A. (1989) J. Biochem. (Tokyo) 105, 79-83). In order to examine possible involvement of transcriptional silencer elements in the regulation of DNA replication, we have utilized a transient replication system of the plasmid DNA carrying replicattion origin of polyoma virus DNA in mouse MOP8 cells, which is constitutively producing polyoma virus large T-antigen. The polyoma virus or- igin of DNA replicati’on is composed of three cis-acting genetic elements callNed a, 8, and core, in which a and @ elements correspond to enhancer domains. When the 5”flanking regions of the DNA polymerase @ gene containing silencer e!lements were placed at the late gene border of a element, they effectively repressed the DNA replication. However, when placed at the early gene border of core element, it only marginally repressed the DNA replication. These results suggest that the silencer elements at cis position repress po- lyoma virus DNA rep:lication by impeding the enhancer function that activates the DNA replication.

@ elements correspond to enhancer domains. When the 5"flanking regions of the DNA polymerase @ gene containing silencer e!lements were placed at the late gene border of a element, they effectively repressed the DNA replication. However, when placed at the early gene border of core element, it only marginally repressed the DNA replication. These results suggest that the silencer elements at cis position repress polyoma virus DNA rep:lication by impeding the enhancer function that activates the DNA replication.
The functional origin (ori) of DNA replication in eukaryotic genome frequently contains transcriptional regulatory elements as its integral components (1, 2). In simian virus (SV) 40, the ori-core is flanked by two auxiliary sequences that can enhance DNA replication severalfold (3, 4). One of these auxiliary sequences resides within the promoter region of early gene (3, 4), and SV40 enhancer can compensate the effect of this auxiliary element when it is juxtaposed to the AT-rich side of SV40 ori-core (3-6). At least one of the enhancer elements is essential to activate ori-core sequence in polyoma virus DNA replication (7,8). Furthermore, cellular proteins called NF-I and NF-111, both of which are required for the replication of adenovirus DNA, can bind specifically to promoter and/or enhancer sequences of certain cellular genes (9, 10). Involvement of c-myc protein in replication and transcription of its own gene was also suggested (11, 12). Thus, it is likely that complex circuits between replication and transcription actually function in eukaryotic cells.
Recently, we have identified the dual cis-acting negative regulatory elements in the upstream region of mouse DNA I: To whom correspondence should be addressed. Tel.: 052-762-polymerase p gene. These elements act on the promoters of heterologous genes, relatively independently to the distance from the promoter and to the orientation relative to the promoter (13). These properties are very similar to those of the so-called transcriptional silencer element (14). S' mce one of the sub-elements of yeast silencer that is involved in repression of mating type loci is reported to function as an autonomous replication sequence (15), it is probable that a silencer element regulates not only transcription but also DNA replication in higher eukaryote.
In the present study, we have utilized the transient replication system of polyoma virus DNA in MOP8 cells, which has been extensively characterized (16), to examine the effect of DNA polymerase p gene silencer elements on the DNA replication. The results indicate that silencer elements-containing sequence can efficiently repress polyoma virus DNA replication presumably by impeding the reactions taking place in the late gene side of the replication origin.

MATERIALS AND METHODS
Plusmid Constructions-A 1533-base pair (hp)' DNA fragment containing DNA polymerase p gene silencer elements (nucleotide position -1860 to -327, where +1 denotes the position of 5"mOSt proximal transcription initiation site) (see Fig. 1) was isolated from one of the 3'-end deletion derivatives of pUC19-MGBOl (17), hluntended, and inserted into the blunt-ended unique XhoI site of the plasmid pPyLCAT (18). The obtained plasmids containing the silencer elements in both normal and reverse orientations were designated as pPyLpNRE(N)CAT and pPyLPNRE(R)CAT, respectively (see Fig. 2). The same silencer fragment attached with BamHI linkers at both ends was inserted into a unique BamHI site of the plasmid pPyECAT (19) in both normal and reverse orientations to create the plasmids pPyEPNRE(N)CAT and pPyEpNRE(R)CAT, respectively. The same fragment attached with BamHI linkers at both ends was inserted into a unique BamHI site of the plasmid pUC19 to create the plasmid pUC19pNRE. These plasmids were propagated in Escherichia coli HBlOl and isolated by standard procedures (20). Isolated plasmid DNA was further purified through 2 cycles of ethidium bromide/CsCl density gradient centrifugation (20). The validity of the obtained recombinant plasmids was examined by restriction endonuclease mapping (21) and dideoxy sequencing method (22,23) using a synthetic 17-mer primer (5'-TTTAGTTAAGTCAGACA-3') that hybridized to nucleotide position -456 to -440 of DNA polymerase p gene.
DNA Transfection-MOP8 cells (16) were cultured in Dulbecco's modified minimal essential medium supplemented with 10% (v/v) fetal calf serum. The DEAE-dextran method with chloroquine treatment (24) was used for DNA transfection into cells at 20 h after  . &-Acting sequences that function as the ori (a, @, and core) are indicated. Shaded areas designate the minimum required sequence for these elements. The transition points from discontinuous to continuous DNA synthesis that define the origin of bi-directional replication are indicated by arrows protruding from core that point in the direction of DNA synthesis (32). Translation initiation sites for early and late genes are indicated by ATG. The solid burs in the plasmids indicate the polyoma sequence cloned into CAT vector and the striped bars indicate the DNA polymerase @ gene sequence between the nucleotide position -1860 and -327 (see Fig. 1). fluctuation in transfection efficiency.
DNA Replication Assay-72 h after DNA transfection, low molecular weight plasmid DNA was isolated from MOP8 cells by the Hirt (25) extraction method. After the sedimentation of high molecular weight DNA, 1.5 ml of the cleared lysate was mixed with 3 ml of TE buffer (10 mM Tris-hydrochloride, pH 8.0, 1 mM EDTA) and extracted once with buffer-saturated phenol and once with chloroform/ isoamyl alcohol (24:l). Nucleic acids were precipitated by adding 2.5 volumes of ethanol. Precipitates were collected by centrifugation and suspended in 50 p1 of TE buffer. A 15-pl portion of the DNA sample was digested with BamHI and D p d for the plasmid pPyECAT and its derivatives and with EcoRI and B p~1 for the plasmid pPyLCAT and its derivatives. The digested plasmid DNAs were subjected to electrophoresis through 1.0% agarose gels, and DNA fragments were transferred to Genescreen Plus filters (Du Pont-New England Nuclear). The filters were hybridized with 32P-labeled pPyLPNRE(N)CAT DNA (5 X 108cpm/pg). After washing, the filters were dried and autoradiographed for 3 to 16 h with Kodak XAR-5 x-ray film and Du Pont Lightning Plus intensifying screen. The intensities of radioactive bands were quantified by densitometry. Each recombinant plasmid DNA was assayed in duplicate on three separate occasions, and the duplicate samples were processed separately.

RESULTS
Effect of DNA Polymerase @ Gene Silencer Elements on the Polyoma Virus DNA Replication-By analyzing a set of 5'end deletion mutants in the 5"flanking region of the mouse DNA polymerase @ gene, we have identified the two tandemly aligned, negatively acting regions for the gene expression that functioned in mouse NIH/3T3 cells (Fig. 1)  A DNA fragment containing DNA polymerase @ gene silencer elements (nucleotide position -1860 to -327) was inserted in both normal and reverse orientations into a unique BamHI site of the plasmid pPyECAT. The site resides immediately upstream of the late gene border of the polyoma enhancer (Fig. 2). The same fragment was also inserted into a unique XhoI site of the plasmid pPyLCAT in both normal and reverse orientations. This site resides close to the large T-antigen binding site C (Fig. 2). Constructs of these four plasmids as well as the polyoma virus control region for replication and transcription are illustrated in Fig. 2. In obtained recombinant plasmids, the functional origin of replication (a, @, and core elements) is kept intact, allowing the replication of plasmid DNAs in mouse cells when the polyoma T-antigen is supplied in trans. These plasmids were transfected into MOP8 cells (16), which originated from NIH/3T3 cells and constitutively express polyoma T-antigen from an integrated ori-polyoma genome, and their replicative capacities were measured by the conversion of DNA that was  (lanes a to f and i). Five pg each of pPyLCAT DNA and pUC19BNRE DNA were co-transfected into MOP-8 cells (lanesg and h). DNA replication was assayed 72 h after transfection as described under "Materials and Methods." After digestion with EcoRI and DpnI, the plasmid DNAs were subjected to electrophoresis through a 1% agarose gel and transferred to GeneScreen Plus filters. The DNA band was visualized after hybridization and autoradiography. The two adjacent lanes represent duplicate independent transfections with each plasmid DNA. Average values (S.D., %) obtained from four independent experiments are given as relative DNA amount to that of pPyLCAT. 25 pg ( l a n e j ) , 50 pg (lane k), and 250 pg (lane I ) of EcoRI-digested pPyLCAT DNA were applied as standards. sensitive to cleavage with DpnI restriction endonuclease into DNA that was resistant to the nuclease (27). This MOP8 cell system has been proven to support the transient episomal replication of the plasmid DNA carrying polyoma ori (16,35). Low molecular weight DNA recovered from MOP8 cells were first cut with EcoRI or BamHI to convert all forms of monomeric DNA into two DNA fragments, one of which was derived from the original vector DNA region. All used plasmids containing multiple DpnI sites that were methylated after growth in a dam+ E. coli strain and rendered sensitive to cleavage by DpnI. Mammalian cells do not contain the dam-methylase. Therefore, when the plasmid DNA is replicated in MOP8 cells, it turns into a hemimethylated form after 1 cycle of replication and an unmethylated form after 2 or more cycles of replication. Both hemimethylated and unmethylated forms are insensitive to DpnI. Therefore, the amount of DpnI-resistant DNA represents the extent of replication.
The results with the plasmid pPyECAT and its derivatives are shown in Fig. 3. The plasmid pOLCAT which did not carry polyoma ori did not replicate at all (Fig. 3A, lane g), while the plasmid pPyECAT carrying polyoma ori replicated efficiently (Fig. 3A, lanes a and b), indicating that the DNA replication is dependent on the polyoma ori sequence. This fact also indicates that the DpnI-resistant bands are not produced by repair type DNA synthesis. These data confirm the previous reports showing the polyoma ori dependency of the transfected plasmid DNA replication in MOP8 cells (16,35). The previous reports also show the polyoma large Tantigen dependency of the plasmids carrying polyoma ori in MOP8 cells (16,35). The derivatives carrying DNA polymerase /3 gene silencer elements replicated less efficiently than the pPyECAT and resulted in the reduction of the amount of DpnI-resistant fragments by about 70% (Fig. 3A, lanes c to  f). As shown in Fig. 3B, the amount of total plasmid DNA recovered from the cell without DpnI-digestion is almost same as DpnI-resistant DNA, indicating that the plasmid DNAs not replicated are lost from cells within 72 h after transfection.
The results with the plasmid pPyLCAT and its derivatives are shown in Fig. 4. The 4.5-kb fragment containing silencer elements migrated more slowly than the 3.0-kb fragment derived from original vector DNA. The amounts of 2.1-kb fragments derived from the vector region of the plasmid DNA that replicated in MOP8 cells were compared. Average values (&S.D.) from these results and those from two other independent experiments are given at the top of each lane. In contrast to the results shown in Fig. 3, DNA polymerase /3 gene silencer elements affected only marginally the DNA replicative capacity of the plasmid pPyLCAT DNA in cis and not at all in trans (Fig. 4). Almost no DpnI-digested fragments were detected in the low molecular weight region of the gel, indicating the quick loss of unreplicated DNA from MOP8 cells, such as the case with the plasmid pPyECAT and its derivatives.

DISCUSSION
The 5"flanking region of the DNA polymerase /3 gene containing the transcriptional silencer elements efficiently repressed the polyoma virus DNA replication when it was placed in the late side border of the enhancer region (Fig. 3).
In polyoma virus, either the a or /3 element must be juxtaposed next to core element to form a functional origin of DNA replication (1, 8, 19). These a and /3 elements correspond to enhancer elements A and B + C, respectively (Fig. 2), and they function redundantly to each other to activate transcription (1, 8, 19). The a and /3 elements activate replication independently of their orientation to the core element. However, unlike enhancer elements, they cannot activate the DNA replication when placed a long distance from the core element (19). Thus, even if the initial binding of the trans-acting regulatory factor(s) to the enhancer element(s) is common for activation of both transcription and DNA replication, the subsequent pathways used for these two processes might be different (19). It is suggested that a and /3 elements facilitate binding of T-antigen to core element either directly through interaction with the initiation complex for DNA replication containing T-antigen or indirectly by exposing the ori sequence that is normally present in inactive chromatin structures open for DNA replication enzymes (28). Since DNA polymerase @ gene silencer elements can dominate a number of enhancers such as SV40, Rous sarcoma virus, and polyoma virus to repress transcription, it is probable that the silencer elements repress DNA replication by impeding the activation of the core element by a and /3 elements.
Two of three sub-elements of yeast MAT locus silencer cooperate to exhibit centromere-like segregation activity, suggesting that silencer function involves an attachment of the DNA to the nuclear membrane or lamina (15, 29). Therefore, it might be possible that the DNA polymerase /3 gene silencer elements repress DNA replication by tagging the ori-containing plasmid DNA to the special sub-nuclear compartment where the replication enzymes are not available.
It is not known why the DNA polymerase /3 gene silencer elements can only marginally repress DNA replication when placed in the early gene side of the core element. One possible explanation is that tight binding of T-antigen to high affinity sites A, B, and C (Fig. 2) might block the effect of the silencer sequence. Interestingly, a and /3 elements cannot activate core element when they were juxtaposed to the early side of core element (30). Asymmetry of ori function also showed that the synthesis of the first nascent DNA chain takes place exclusively on the early mRNA template strand of ori and progresses toward the direction of the early genes (31,32). Thus, some unknown mechanisms might exist to prevent the action of the silencer elements when they are placed in the early side of the ori.