Simian Virus 40 Strains with Novel Properties ... - Journal of Virology

Simian Virus 40 Strains with Novel Properties Generated by Replacing the Viral Enhancer with Synthetic Oligonucleotides Viola Günther, Till Strassen, Uschi Lindert, Patrizia Dagani, Dominique Waldvogel, Oleg Georgiev, Walter Schaffner, and Tobias Bethge Institute of Molecular Life Sciences, University of Zürich, Zürich, Switzerland

Typical enhancers of viral or cellular genes are approximately 100 to 400 bp long and contain several transcription factor binding sites. Previously, we have shown that simian virus 40 (SV40) genomic DNA that lacks its own enhancer can be used as an “enhancer trap” since it reacquires infectivity upon incorporation of heterologous enhancers. Here, we show that SV40 infectivity can be restored with synthetic enhancers that are assembled by the host cell. We found that several oligonucleotides, cotransfected with enhancerless SV40 DNA into host cells, were incorporated into the viral genome via cellular DNA end joining. The oligonucleotides tested included metal response elements (MREs), the binding sites for the transcription factor MTF-1, which induces gene activity in response to heavy metals. These recombinant SV40 strains showed preferential growth on cells overloaded with zinc or cadmium. We also cotransfected enhancerless SV40 DNA with oligonucleotides corresponding to enhancer motifs of human and mouse cytomegalovirus (HCMV and MCMV, respectively). In contrast to SV40 wild type, the viruses with cytomegalovirus-derived patchwork enhancers strongly expressed T-antigen in human HEK293 cells, accompanied by viral DNA replication. Occasionally, we also observed the assembly of functional viral genomes by incorporation of fragments of bovine DNA, an ingredient of the fetal calf serum in the medium. These fragments contained, among other sites, binding sites for AP-1 and CREB transcription factors. Taken together, our studies show that viruses with novel properties can be generated by intracellular incorporation of synthetic enhancer DNA motifs.

I

n higher eukaryotes, including mammals, regulatory DNA sequences for gene transcription can activate transcription over long distances of thousands of base pairs (bp), independent of their orientation and position relative to the transcription unit. These DNA segments, termed enhancers, were first discovered in simian virus 40 (SV40), a member of the Polyomaviridae, and in mouse polyomavirus (2, 6, 18). The first example of a cellular enhancer discovered was the immunoglobulin heavy chain (IgH) enhancer, which was a cell-type-specific regulatory sequence since it was active only in B-lymphocyte-type cells and not in other cell types (1, 9). Thereafter, many more cellular and viral enhancers were described, including a steroid hormone-responsive segment in mouse mammary tumor virus (4), the particularly strong enhancers associated with immediate-early genes of human and mouse cytomegaloviruses (HCMV and MCMV, respectively) (3, 7), and zinc-responsive enhancers associated with human and mouse metallothionein genes (28). The typical enhancers of viral or cellular genes are approximately 100 to 400 bp long and contain binding sites for several DNA-binding transcription factors. For facilitated identification of enhancers, a selection system termed “enhancer trap” can be used (20, 33). One simple approach is to transfect linearized SV40 genomic DNA lacking its own enhancer into monkey host cells together with DNA fragments containing a putative enhancer segment. Whenever an enhancerless SV40 genome acquires an active enhancer DNA via cellular DNA endjoining processes, it confers on that recombinant virus the ability to proliferate (20, 33). Here, we show that this enhancer trap system, which before was used to select preexisting enhancers, is also able to compose new enhancers from a mixture of short DNA sequence motifs, derived either from enhancers of cytomegaloviruses or from metal-inducible genes. From the latter we tested binding sites for the zinc finger protein metal-responsive transcription factor 1 (MTF-1; also referred to as metal response ele-

0022-538X/12/$12.00

Journal of Virology

p. 3135–3142

ment [MRE]-binding transcription factor or metal regulatory transcription factor). MTF-1 induces gene activity in response to a heavy metal load via binding to MREs, short DNA motifs with the consensus sequence TGCRCNC (35). With this approach, enhancers were assembled from multiple MRE-containing oligonucleotides to yield recombinant SV40 viruses with the novel property of preferential growth in heavy metal-loaded cells. Both lines of experiments, with cytomegalovirus-derived motifs and with MREs, show that synthetic enhancers can readily be assembled from double-stranded DNA oligonucleotides containing transcription factor binding sites and that the resulting recombinant viruses can have novel properties clearly distinct from those of the original virus. Interestingly, in a few cases we found enhancer assemblies not originating from the provided synthetic oligonucleotides but from bovine DNA fragments present in the fetal calf serum (FCS) of the cell culture medium. MATERIALS AND METHODS Oligonucleotides and DNA constructs. All oligonucleotides used in this study were synthesized by Microsynth Co., Balgach, Switzerland. Since individual repeat motifs of cytomegalovirus (CMV) enhancers deviate slightly from each other, we derived an idealized consensus sequence for each of them. The oligonucleotide sequences with CMV repeat motifs and with metal response elements are depicted in Fig. 1B and 3A, C, and D,

Received 13 September 2011 Accepted 22 December 2011 Published ahead of print 11 January 2012 Address correspondence to Tobias Bethge, [email protected]. V. Günther and T. Strassen contributed equally to this article. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/JVI.06293-11

jvi.asm.org

3135

Günther et al.

respectively. A single-stranded oligonucleotide representing one copy of the 72-bp repeat from the SV40 wild-type (wt) enhancer (see Fig. 1E) (NCBI accession number NC_001669) as well as the separate single strands of the 90-bp MCMV (Fig. 1B) was also used. The enhancer trap construct, as depicted in Fig. 1A, was described previously (33). Cell culture and transfection. African green monkey kidney cells (CV-1) and human embryonic kidney cells (HEK293) were maintained in Dulbecco’s modified Eagle medium (DMEM; Gibco) supplemented with 5% FCS (Biochrom AG, Berlin, Germany), 2 mM L-glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin (Invitrogen). Cells were seeded into 10-cm petri dishes 1 day before transfection. At the time of transfection, the cell monolayer was approximately two-thirds confluent. Transfections were performed using a standard calcium phosphate coprecipitation method (10). One microgram of enhancer trap DNA was mixed with 1 to 10 ␮g of double-stranded oligonucleotides (see figures and figure legends), representing a 500- to 1,000-fold molar excess over the SV40 genome. Single-stranded oligonucleotides were transfected in a 1,500fold molar excess over the SV40 genome. To standardize the amount of DNA, herring sperm DNA was added to a total amount of 20 ␮g of DNA. Cells were washed 14 to 16 h after transfection with Tris-buffered saline (TBS). Unless indicated otherwise, metal treatments started 1 day after transfection by supplementation of the medium with ZnCl2 to a final concentration of 100 ␮M, which was raised to 200 ␮M after another day. Similarly, cells were conditioned with 3 ␮M CdCl2, which was subsequently raised to 6 ␮M. T-antigen immunofluorescence. Indirect immunofluorescence was performed as described previously (24) with a 1:200 dilution of a monoclonal mouse anti-T-antigen antibody (Ab-2; product number DP02; Oncogene/Calbiochem) and secondary Alexa Fluor 546 goat anti-mouse antibody (A-11030; Molecular Probes/Invitrogen) in a 1:200 dilution. Nuclei were stained by 4=,6-diamidino-2-phenylindol (DAPI) at a concentration of 1 ␮g/ml for 1 min. Analysis of viral DNA. Small, closed circular DNA was extracted from infected CV-1 cells by an alkaline precipitation procedure according to the method of G. Magnusson (Department of Medical Biochemistry and Microbiology, Uppsala University, Biomedical Centre, Uppsala, Sweden). First, cells were washed once with TBS and lysed with a NaOH- and SDScontaining buffer (50 mM glucose, 25 mM Tris-HCl, pH 8, 10 mM EDTA, 0.2 M NaOH, 1% SDS). After neutralization with 3 M potassium acetate, pH 4.8, and centrifugation of the lysates, viral DNA was purified from the supernatants by phenol-dichloromethane extraction and isopropanol precipitation. Purified viral DNA was cloned via a BamHI site into Bluescript plasmid (pBSK). From individual colonies, complete viral genomes were reclaimed by BamHI digestion and transfected into CV-1 cells to verify their infectivity. In parallel, the enhancer inserts were sequenced. For some experiments, only the enhancer region was isolated, cloned, and, to test for its activity, mixed again with enhancerless SV40 DNA. Detection of viral DNA by Southern blotting. Total DNA of HEK293 cells that were transfected with wild-type or recombinant SV40 viral DNA was isolated and purified as described above under “Analysis of viral DNA.” Five micrograms of DNA was digested by EcoRI, and viral DNA was detected by Southern blotting with a 32P-end-labeled oligonucleotide probe (5=-GCACACTCAGGCCATTGTTTGCAGTACATTGCATCAAC ACCAGG-3=). As an internal control, pBluescript-KS was cotransfected and duplicates were run on a parallel gel and detected with the following oligonucleotide: 5=-CCAAGTCATTCTGAGAATAGTGTATGCGGCGA CCGAGTTGCTCT-3=. Signals were visualized using the fluorescent image analyzer FLA-7000. Semiquantitative PCR. Reactions were performed with 28 cycles using the following primers: 5=-GCCAAGCAACTCCAGCCATC-3= and 5=TGAGGAAAGTTTGCCAGGTG-3= to detect SV40 and 5=-GCGGATAA AGTTGCAGGACCAC-3= and 5=-CTGCTGGAAGCCAGTTACCTTC-3= to detect cotransfected pBluescript as a control. EMSA. Electrophoretic mobility shift assays (EMSAs) of transcription factor binding sites were performed as described before (25). Briefly, 10

3136

jvi.asm.org

␮g of nuclear extract from CV-1 cells was mixed with binding buffer (20 mM PIPES [piperazine-N,N=-bis(2-ethanesulfonic acid)], pH 6.8, 50 mM NaCl, 1 mM dithiothreitol [DTT], 0.25 mg/ml bovine serum albumin [BSA], 100 ␮M ZnSO4, 0.05% NP-40, 4% Ficoll) and approximately 5 fmol of 32P-labeled, duplexed oligonucleotide in a final volume of 20 ␮l. Nuclear extracts were prepared by scraping cells off 10 cm-dishes with ice-cold phosphate-buffered saline (PBS) and collecting them in Eppendorf tubes. After centrifugation, cells were swollen in hypotonic buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 2.5 mM DTT) and lysed by addition of NP-40 (final concentration, 0.5%). After centrifugation of the nuclei, proteins were extracted in nuclear extract buffer (20 mM HEPES, pH 7.9, 25% glycerol, 400 mM NaCl, 1 mM EDTA, 2.5 mM DTT). Nuclear extracts with activated CREB were prepared after cells were treated for 1 h with 10 ␮M forskolin (Sigma-Aldrich). Competing, unlabeled oligonucleotides were used in 200-fold excess. After the mixture was incubated for 30 min on ice, samples were separated on a native 4% polyacrylamide gel with 0.25⫻ Tris-borate-EDTA (TBE) as a running buffer. The detection of ␤-decay within the samples was carried out with Fujifilm Imaging Plates and a Fujifilm FLA-7000 Image Plate reader.

RESULTS

Cytomegalovirus-derived oligonucleotides are integrated into the SV40 enhancer trap to restore enhancer activity. The very strong enhancer associated with the immediate-early transcription unit of human cytomegalovirus (HCMV) harbors four major types of transcription factor-binding DNA sequence motifs, referred to as the 17-bp repeat, 18-bp repeat, 19-bp repeat, and 21-bp repeat. These repeats are embedded within several hundred base pairs of nonrepetitive enhancer sequence (3). Similar motifs are present also in the mouse cytomegalovirus (MCMV) enhancer, which is organized into several nested repeat units of approximately 90 bp (7). To find out whether a synthetic transcription enhancer that is able to substitute for SV40’s own enhancer can be generated from double-stranded oligonucleotides, we used the SV40 enhancer trap, an enhancerless viral genome that has been shown to reacquire infectivity upon incorporation of a heterologous viral or cellular enhancer (33). SV40 DNA was liberated from the vector plasmid by cleavage with XbaI and KpnI to yield the enhancerless viral genome of 5,040 bp (Fig. 1A). This genome was cotransfected with either a mixture of all four repeated sequence motifs of the HCMV enhancer or with oligonucleotides representing individual repeats (Fig. 1B). In a parallel experiment, a synthetic 90-bp consensus repeat from the MCMV enhancer, which among other sites contains binding sites for NF-␬B and AP-1, was also tested (Fig. 1B). In all cases, infectious viruses were readily obtained. Various enhancer lengths and compositions result in similar levels of virus propagation. To characterize the synthetic enhancer sequences acquired by enhancerless SV40, viral DNA was isolated from cell lysates and subcloned. After a sequencing step, selected clones were reanalyzed by retransfecting cells to identify opportunistic genomes, which had replicated without having incorporated a functional enhancer of their own. In the case of HCMV oligonucleotides, we found all four enhancer motifs in various combinations and orientations in different isolates (Fig. 1C), which upon retesting were similarly infective (first signs of cytopathic effects [CPE] at around day 10 and full infection at around day 14 after transfection). Thus, although the lengths and compositions of the isolated enhancers were quite variable, our experimental setup apparently selected for viral enhancers with similar activities. Of note, the 18-bp repeat, which harbors a previously described binding site for NF-␬B (23), was inte-

Journal of Virology

Synthetic Enhancers Confer Novel Properties to SV40

FIG 1 Enhancer trap and CMV oligonucleotides. (A) Junctions of enhancerless SV40. Linear SV40 DNA of 5,040 bp was liberated from the pBSK vector plasmid by XbaI and KpnI. Numbers refer to the positions in the SV40 wt genome. Ori, viral origin of replication. (B) Oligonucleotides harboring the four HCMV-derived repeat motifs of 17 bp (blue), 18 bp (yellow), 19 bp (red), 21 bp (green), and MCMV 90 bp (gray). Respective transcription factor binding sites are underlined. (C) Ten representative SV40 isolates and their composition. (D) Example for enhancer assembly by NHEJ. Cotransfected was the 18-bp repeat derived from HCMV with sticky 3= GG and CC overhangs (18 bp; yellow box). (E) Example for enhancer assembly from a single-stranded 72-nt SV40 wild-type enhancer segment (red) by MMEJ. SV40 enhancer trap sequence (black) attaches to the oligonucleotide, noncomplementary overhangs are trimmed (blue), and cellular enzymes fill the gaps and synthesize the complementary strand (green).

grated with the highest frequency: in 10 representative clones with enhancer inserts (Fig. 1C), which on average incorporated 6.9 oligonucleotides, the 18-bp fragment appeared with an average frequency of 2.8 copies, or about 40%. This finding is consistent with previous studies, which showed that NF-␬B is important for enhancer activation in CMV-infected human fibroblasts (23, 29), and suggests that under stress conditions of transfection/viral infection, transcriptional activation by NF-␬B is particularly robust. The second-most-abundant oligonucleotide was the 17-bp repeat, which contains a binding site for the housekeeping factor NF-1 (12, 19) and occurred with a frequency of 1.7 oligonucleotides per insert. The lowest in abundance, with an average of 1.2 oligonucleotides per insert, were the 19-bp repeat with a CREB site (13) and the 21-bp repeat with an Sp1 site (3). The potency of NF-␬B sites was confirmed by offering individual HCMV oligonucleotides, rather than a mixture, to the enhancer trap. When CV-1 cells were transfected with enhancer trap vector and the 18-bp repeat, the first signs of CPE could be observed after an average of 22 days. In direct comparison, this was as efficient as the use of the oligonucleotide mixture with all four sites. An intermediate

March 2012 Volume 86 Number 6

progression of viral infectivity was observed with the 21-bp or 17-bp fragment (33 days or 35 days until first signs of CPE, respectively). Consistent with the integration ratios seen with the oligonucleotide mixture, the poorest enhancer activation and viral growth seemed to be conferred by the 19-bp oligonucleotide with its CREB site, leading, on average, to first signs of CPE after 41 days (data not shown). Regarding the discrepancies of timelines between different experiments, it has to be noted here that the reinfection of cells with isolated viruses leads to a faster progression of viral infection and visible CPE than the transfection with enhancer trap and oligonucleotides since the initial number of functional viruses is much lower in the latter case. Assembly of recombinant viruses by cellular end-joining processes. Even though the oligonucleotides had 5= protruding “sticky ends,” these were not necessarily used by the cell. For example, in head-to-tail or head-to-head fusions of oligonucleotides where the ends did not match each other, the new enhancers were most likely assembled by a trimming and ligation process, termed nonhomologous end joining (NHEJ) (5, 11, 15, 30, 34). An extensively trimmed junction between the XbaI site of the enhancerless

jvi.asm.org 3137

Günther et al.

FIG 2 Cytomegalovirus-derived synthetic enhancers drive SV40 early gene expression and DNA replication in human cells. Recombinant SV40 viral genomes which contain HCMV-derived or MCMV-derived enhancers or its wild-type enhancer (SV40) were used to transfect monkey CV-1 cells or human HEK293 cells. (A) Immunofluorescence stainings using anti-T-antigen antibody; nuclei were stained with DAPI. (B) Columns represent the percentages of T-antigen-positive cells of two independent experiments. (C) Southern blotting with total DNA isolated from HEK293 cells that were transfected 8 days prior to harvesting with a mixture of viral DNA and pBluescript (plasmid) as a control (ctrl), run on parallel gels. (D) Semiquantitative PCR with template DNA isolated from the transfected HEK293 cells described for panel C. ␣, anti.

SV40 and an 18-bp repeat oligonucleotide is depicted in Fig. 1D. Apparently, before or in the course of integration, the 5= end of the oligonucleotide as well as the 3= end of the SV40 genome was trimmed by cellular nucleases, resulting in the removal of the sticky ends and the XbaI site, and this was followed by blunt-end ligation. Interestingly, the enhancer trap system also works with singlestranded substrate oligonucleotides, such as the separate single strands of the 90-bp MCMV enhancer segment (Fig. 1B) or a 72-nucleotide (nt) repeat of the genuine SV40 enhancer (Fig. 1E). However, in these cases incorporation into the viral genome seems to be less efficient than with the use of duplexed oligonucleotides. To achieve assembly of functional viruses, the 72-nt repeat had to be provided in larger molar excess over the SV40 genome (1,500-fold instead of 500- to 1,000-fold double-stranded oligonucleotides). First signs of CPE could be observed after 23 days. Sequence analysis suggested that integration had not happened via NHEJ but by a variant mechanism, termed microhomology-mediated end joining (MMEJ) (30), referring to the fact that prior to repair-ligation, single-stranded DNA or single-stranded overhangs can attach to sticky ends with the help of one or more complementary bases. In the example shown in Fig. 2E, the XbaI site as well as the KpnI site of the enhancer trap was able to pair a G or C with a terminal base of the oligonucleotide, followed by the removal of noncomplementary overhangs (blue sequences), gap filling, and synthesis of the complementary strand (green sequences).

3138

jvi.asm.org

T-antigen expression and viral DNA replication in human cells as a novel property of CMV-SV40 recombinants. To compare the performance of CMV enhancer-containing recombinant viruses with the SV40 wild type (wt), CV-1 cells were transfected with a mixture of viral DNA from cell culture supernatants. While cells transfected with SV40 wt showed the first CPE after 2 days and full infection after 4 days, progression of infection with the CMV segment-containing viruses was delayed by approximately 2 days. Interestingly, although the ability of the recombinant viruses to replicate in monkey CV-1 cells was not equivalent to that of the SV40 wt, they gained an advantage over SV40 wt by acquiring CMV elements. That is, if transfected into immortalized human embryonic kidney (HEK293) cells, SV40 with its genuine enhancer failed to produce T antigen at a detectable level, which is a prerequisite for a successful infection. In contrast, human and even mouse CMV segment-containing viruses readily expressed the early gene region, as determined by T-antigen immunofluorescence (Fig. 2A and B). This is further testimony of the strong promiscuous activity of HCMV and MCMV enhancers (3, 7, 8) and demonstrates that enhancer composition is a determinant of host cell tropism. In agreement with T-antigen expression of the recombinants, Southern blotting and semiquantitative PCR revealed that in HEK293 cells the synthetic CMV-derived enhancers, especially the enhancer from HCMV, boosted viral replication, while only a small amount of SV40 wild-type DNA was produced (Fig. 2C and D).

Journal of Virology


FIG 3 Substitution of the SV40 enhancer by synthetic MREs. (A) Oligonucleotides representing MREs from several metal-inducible genes were cotransfected with enhancer trap DNA into CV-1 cells. Csrp1 MRE2, cysteine and glycine-rich protein 1 MRE2 (green); Ndr1 MRE3,4, N-myc downstreamregulated 1 protein MRE3/MRE4 (blue); Ndr1 MRE2, N-myc downstreamregulated gene 1 protein MRE2 (yellow); MRE-s, synthetic MRE consensus oligonucleotide (red); SepW1 MRE, selenoprotein W1 MRE (gray); Slc30a10, zinc transporter MRE2 (orange). Underlined are the MRE core sequences. (B) Two examples of MRE-derived enhancers. The color for each MRE block is according to the scheme in panel A. (C) MRE(15) and MRE(20), MRE oligonucleotides with 15-bp and 20-bp spacing, respectively, between oligomerized oligonucleotides. (D) Enhancer assembled from a single type of motif, the MREd from the mouse metallothionein 1 promoter (shown without the sticky ends). All motifs were found to be arranged in the same direction, either as shown here or in the opposite orientation (not shown).

MREs can be assembled to generate metal-dependent SV40 recombinant viruses. In a second series of experiments, we used a mixture of oligonucleotides representing established metal response elements (MREs) (21, 27). These were derived from metallothionein genes and other known cellular target genes of MTF-1 and were chosen because of their avid binding to MTF-1 (22, 36). Most of these MRE-containing oligonucleotides had protruding single-stranded ends which did not, however, fit each other (Fig. 3A). Nevertheless, by using a culture medium supple-


mented with 150 ␮M ZnCl2, a number of viruses were readily obtained in independent culture dishes. Upon cloning and sequence analysis, they turned out to have incorporated a mix of different MRE sequences, similar to the situation with the sequence elements from the human cytomegalovirus enhancer (Fig. 3B). Also here, as observed for the CMV-derived segments, there was no obvious preference regarding the orientation of the MREs for enhancer activation. Note that only two MRE oligonucleotides have sticky ends that might facilitate self-ligation: the artificial consensus MREs (Fig. 3B, red) and the MRE of selenoprotein SepW1 (Fig. 3B, gray). The fact that no overrepresentation of multiple tandem arrangements of these two MREs was observed supports the notion that compatible sticky ends are not a prerequisite for the intracellular assembly of enhancers. Again, the aforementioned repair via NHEJ led to the trimming and ligation of incompatible ends. To expand on the observation that the MREmediated activation of transcription was independent not only of the orientation but also of the spacing between individual MREs, we decided to test preformed arrays of tandem oligonucleotides with defined spacings of 15 and 20 bp between the beginning of one core MRE sequence and the next [MRE(15) and MRE(20), respectively] (Fig. 3C). With both spacings, severe cytopathic effects were observed at about day 10 postinfection. Recombinant viruses were sequenced, and in all cases the predetermined spacing was maintained, indicating that spacing is not critical (data not shown). In a further experiment we sought to examine whether a single MRE, the well-characterized MREd (metal response element d) of the mouse metallothionein 1 gene promoter, could, alone or in multiple copies, substitute for the SV40 enhancer. For this experiment, an oligonucleotide with compatible sticky ends was used, resulting in functional viruses that grew well in zinc-supplemented medium (Fig. 3D). To confirm the novel properties of the MRE-containing viral isolates, CV-1 cells were reinfected with viruses from supernatants of previous experiments and kept in either medium supplemented with 200 ␮M ZnCl2, or control medium. Here, a clear preference for viral growth in zinc-containing medium was evident since complete lysis of host cells was observed after 18 days in the presence of ZnCl2, while cells in unsupplemented medium did not even show signs of CPE at day 23 when the experiment was stopped (Fig. 4A). In contrast, the SV40 wild type did not multiply any better in zinc-loaded versus control cells (data not shown). In a separate experiment with another MREcontaining virus, CV-1 cells were infected with either SV40 wt virus or recombinant viruses containing 14 tandem copies of the MRE(15) motif. After treatment with or without 150 ␮M zinc for 17 days, the viral DNAs were extracted and analyzed by gel electrophoresis (Fig. 4B). In the cases of de novo DNA replication, a band was visible that migrated at the speed of a 3-kb linear marker DNA, which represents supercoiled, covalently closed circular viral DNA (Fig. 4B, ccc). As expected, the yield of DNA from MRE-containing recombinant viruses correlated with the zinc treatment (Fig. 4B). The zinc-treated cells, infected with SV40-MRE, showed viral replication (lane 6 to 8), while the untreated cells (lanes 1 to 3) did not produce supercoiled viral DNA. In contrast, the amount of viral DNA from SV40 wt-infected cells was independent of zinc (Fig. 4B, lane 4 versus lane 9). Regardless of the zinc treatment or enhancer, in all lanes with infected cells a band which migrated at approxi-

jvi.asm.org 3139

Günther et al.

FIG 4 Growth of MRE-containing SV40 in the presence of zinc. (A) Reinfection of MREd-containing, recombinant SV40 to compare growth in zinc-supplemented medium (200 ␮M ZnCl2) and normal medium (without zinc). Arrows mark days of inspection. (B) DNA agarose gel after treatment with or without 150 ␮M ZnCl2 of cells with SV40 wt or SV40 containing tandem copies of MRE(15) inserts. Seventeen days after infection of CV-1 cells, DNA was extracted and loaded onto an agarose gel. M, phage ␭ DNA fragments as markers; arrows mark covalently closed circular (supercoiled) SV40 DNA (ccc) and nicked, relaxed circular SV40 input DNA (rc).

mately 5 kb can be observed (Fig. 4B, rc), which most likely originates from viral input DNA. Other than the newly synthesized supercoiled DNA, the input DNA persists as nicked, relaxed circles due to nuclease-mediated damage. Cadmium is a nonessential, toxic heavy metal that, like zinc, strongly induces transcription via MTF-1/MRE motifs. We also tested medium with 6 ␮M cadmium salt in the same series of experiments. While higher concentrations of cadmium are particularly strong inducers of MTF-1 in transient transfection assays, we found that for long-term culture of CV-1 cells, 6 ␮M cadmium was close to the upper tolerated limit. MRE-containing recombinant viruses that proliferated well in zinc also did so in cadmiumsupplemented medium, but the first signs of infection appeared with a delay of a few days (data not shown). Taken together, in both lines of experiments, with cytomegalovirus-derived motifs and with metal response elements, viruses with novel properties could be generated. Spontaneous generation of SV40 recombinants by incorporation of bovine DNA fragments from cell culture medium. Interestingly, in some enhancer trap experiments, we found inserts which did not match with the oligonucleotide sequences provided in that experimental setup. Sequence analysis revealed that the SV40 enhancer trap constructs had acquired 85- to 100-bp-long fragments of bovine DNA. The likely source for these fragments was the fetal calf serum used to supplement the cell culture medium. With the help of the Transcription Element Search System (TESS [http://www.cbil .upenn.edu/cgi-bin/tess/tess), we looked for potential transcription factor binding sites within the bovine sequences to explain the enhancer activity and designed oligonucleotides carrying several of the promising sites (Fig. 5A); among them were a CRE element (Bo1CREB) and an AP-1 site (Bo2-AP-1). By subjecting these oligonucleotides to electrophoretic mobility shift assays (EMSAs), we could verify the binding of the aforementioned transcription factors (Fig. 5B). In the case of Bo1-CREB, the amount of CREB and, hence, CREB binding could be increased by stimulating the CV-1 cells with forskolin (14, 26), while an unlabeled oligonucleotide containing the CREB binding site from the somatostatin promoter (17) or the HCMV

3140

jvi.asm.org

19-bp oligonucleotide (HCMV19) (Fig. 1B) could be used to compete with the binding of Bo1-CREB. Likewise, the Bo2-AP-1 oligonucleotide showed binding to multiple transcription factors, among them AP-1, as demonstrated by competition with oligonucleotides

FIG 5 EMSA with oligonucleotides derived from bovine sequences. (A) Oligonucleotides derived from isolated bovine sequences. Underlined are potential binding sites for CREB and AP-1. (B) EMSA gels. CV-1 ex, nuclear extracts from CV-1 cells; ⫺, no extract added (free probe); ⫹, 10 ␮g of nuclear extract added; forsk, stimulation of cells with 10 ␮M forskolin for 1 h; CREB soma, competition with 200-fold excess of unlabeled oligonucleotide (1 pmol) containing a CREB site from the somatostatin promoter (the other competing oligonucleotides, HCMV19, AP-1, SV40 and AP-1 Ad5, were used in the same amount of excess). Arrows mark the bandshift signals from the respective transcription factors.

Journal of Virology


containing the AP-1 sites from SV40 or adenovirus 5 (Ad5) (16, 31). These findings revealed that in addition to synthetic oligonucleotides, heterologous binding-site-containing DNA fragments from other sources are also able to exert enhancer activity for SV40 transcription. DISCUSSION

The data presented here demonstrate that short sequence motifs, represented by synthetic double-stranded DNA oligonucleotides, can be assembled in vivo to convert an enhancerless SV40 into a functional virus. Novel enhancers are assembled from different nonmatching sequence motifs, most likely via the pathway of nonhomologous end joining (NHEJ) or microhomology-mediated end joining (MMEJ). Preligation of sticky ends of these oligonucleotides facilitates the assembly of a functional virus but is not mandatory, thanks to the efficient cellular repair activities. Functional enhancers were readily assembled from a mixture of synthetic CMV-derived oligonucleotides or metal response elements, representing MREs from a number of target genes of MTF-1. It is of particular interest that the resulting viruses could display a novel property, like preferential growth, in cells loaded with the heavy metals zinc and cadmium. It remains to be seen how such synthetic enhancers with repeated elements might evolve further (see also reference 32). As a side note, we observed the occasional integration of bovine DNA fragments from the calf serum in the culture medium into the enhancerless viral backbone although the DNA in the serum was hardly detectable. Interestingly, three independent isolates of recombinant viruses contained similar but nonidentical fragments of the bovine genome. Taken together, our results show that (i) SV40 harboring synthetic enhancers can readily be generated by transfection of host cells with a mix of enhancerless DNA and synthetic oligonucleotides containing transcription factor binding sites, (ii) the oligonucleotides do not even have to contain ligatable ends since the cellular repair machinery joins them to build an enhancer by trimming and/or filling up singlestranded overhangs, and (iii) the recombinant viruses can have novel properties. With synthetic MREs, viruses could be obtained that grew poorly, if at all, in cells kept in normal medium but grew well in heavy metal-loaded cells. Similarly, viruses containing cytomegalovirus-derived enhancer motifs displayed altered properties in that they acquired the ability of directing early gene expression and viral DNA replication in human HEK293 cells. However, this effect did not result in viruses with a fully expanded host range since substantial early gene expression and DNA replication in human cells were not sufficient to allow for a progressive spread of infection. Among the four different HCMV repeats (17, 18, 19, and 21 bp), stress conditions of transfection and infection clearly favored the 18-bp oligonucleotide with its NF-␬B site in CV-1 cells. By using defined binding sites for transcription factors, this system of combining synthetic components followed by natural selection could be used to create made-to-measure enhancers that fit a particular need, e.g., to obtain viruses with a certain cell type specificity. This might be useful for various applications, such as targeted gene therapy. ACKNOWLEDGMENTS We are indebted to Antonia Manova for technical advice, to Kurt Steiner for analysis of DNA sequences, and to George Hausmann for critical reading of the manuscript.


This work was supported by the Schweizerischer Nationalfonds and the Kanton Zürich.

REFERENCES 1. Banerji J, Olson L, Schaffner W. 1983. A lymphocyte-specific cellular enhancer is located downstream of the joining region in immunoglobulin heavy chain genes. Cell 33:729 –740. 2. Banerji J, Rusconi S, Schaffner W. 1981. Expression of a beta-globin gene is enhanced by remote SV40 DNA sequences. Cell 27:299 –308. 3. Boshart M, et al. 1985. A very strong enhancer is located upstream of an immediate early gene of human cytomegalovirus. Cell 41:521–530. 4. Chandler VL, Maler BA, Yamamoto KR. 1983. DNA sequences bound specifically by glucocorticoid receptor in vitro render a heterologous promoter hormone responsive in vivo. Cell 33:489 – 499. 5. Daza P, et al. 1996. Mechanisms of nonhomologous DNA end-joining in frogs, mice and men. Biol. Chem. 377:775–786. 6. de Villiers J, Schaffner W. 1981. A small segment of polyoma virus DNA enhances the expression of a cloned beta-globin gene over a distance of 1400 base pairs. Nucleic Acids Res. 9:6251– 6264. 7. Dorsch-Hasler K, et al. 1985. A long and complex enhancer activates transcription of the gene coding for the highly abundant immediate early mRNA in murine cytomegalovirus. Proc. Natl. Acad. Sci. U. S. A. 82:8325– 8329. 8. Foecking MK, Hofstetter H. 1986. Powerful and versatile enhancerpromoter unit for mammalian expression vectors. Gene 45:101–105. 9. Gillies SD, Morrison SL, Oi VT, Tonegawa S. 1983. A tissue-specific transcription enhancer element is located in the major intron of a rearranged immunoglobulin heavy chain gene. Cell 33:717–728. 10. Graham FL, van der Eb AJ. 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52:456 – 467. 11. Hagmann M, et al. 1996. Dramatic changes in the ratio of homologous recombination to nonhomologous DNA-end joining in oocytes and early embryos of Xenopus laevis. Biol. Chem. Hoppe Seyler 377:239 –250. 12. Hennighausen L, Fleckenstein B. 1986. Nuclear factor 1 interacts with five DNA elements in the promoter region of the human cytomegalovirus major immediate early gene. EMBO J. 5:1367–1371. 13. Hunninghake GW, Monick MM, Liu B, Stinski MF. 1989. The promoter-regulatory region of the major immediate-early gene of human cytomegalovirus responds to T-lymphocyte stimulation and contains functional cyclic AMP-response elements. J. Virol. 63:3026 –3033. 14. Lamph WW, Dwarki VJ, Ofir R, Montminy M, Verma IM. 1990. Negative and positive regulation by transcription factor cAMP response element-binding protein is modulated by phosphorylation. Proc. Natl. Acad. Sci. U. S. A. 87:4320 – 4324. 15. Lieber MR. 2008. The mechanism of human nonhomologous DNA end joining. J. Biol. Chem. 283:1–5. 16. Mermod N, Williams TJ, Tjian R. 1988. Enhancer binding factors AP-4 and AP-1 act in concert to activate SV40 late transcription in vitro. Nature 332:557–561. 17. Montminy MR, Bilezikjian LM. 1987. Binding of a nuclear protein to the cyclic-AMP response element of the somatostatin gene. Nature 328:175– 178. 18. Moreau P, et al. 1981. The SV40 72 base repair repeat has a striking effect on gene expression both in SV40 and other chimeric recombinants. Nucleic Acids Res. 9:6047– 6068. 19. Niller HH, Hennighausen L. 1991. Formation of several specific nucleoprotein complexes on the human cytomegalovirus immediate early enhancer. Nucleic Acids Res. 19:3715–3721. 20. O’Kane CJ, Gehring WJ. 1987. Detection in situ of genomic regulatory elements in Drosophila. Proc. Natl. Acad. Sci. U. S. A. 84:9123–9127. 21. Palmiter RD. 1994. Regulation of metallothionein genes by heavy metals appears to be mediated by a zinc-sensitive inhibitor that interacts with a constitutively active transcription factor, MTF-1. Proc. Natl. Acad. Sci. U. S. A. 91:1219 –1223. 22. Radtke F, et al. 1993. Cloned transcription factor MTF-1 activates the mouse metallothionein I promoter. EMBO J. 12:1355–1362. 23. Sambucetti LC, Cherrington JM, Wilkinson GW, Mocarski ES. 1989. NF-␬B activation of the cytomegalovirus enhancer is mediated by a viral transactivator and by T cell stimulation. EMBO J. 8:4251– 4258. 24. Saydam N, Georgiev O, Nakano MY, Greber UF, Schaffner W. 2001. Nucleo-cytoplasmic trafficking of metal-regulatory transcription factor 1 is regulated by diverse stress signals. J. Biol. Chem. 276:25487–25495.

jvi.asm.org 3141

Günther et al.

25. Schreiber E, Matthias P, Muller MM, Schaffner W. 1988. Identification of a novel lymphoid specific octamer binding protein (OTF-2B) by proteolytic clipping bandshift assay (PCBA). EMBO J. 7:4221– 4229. 26. Seamon KB, Padgett W, Daly JW. 1981. Forskolin: unique diterpene activator of adenylate cyclase in membranes and in intact cells. Proc. Natl. Acad. Sci. U. S. A. 78:3363–3367. 27. Searle PF, Stuart GW, Palmiter RD. 1985. Building a metal-responsive promoter with synthetic regulatory elements. Mol. Cell. Biol. 5:1480 – 1489. 28. Serfling E, Lubbe A, Dorsch-Hasler K, Schaffner W. 1985. Metaldependent SV40 viruses containing inducible enhancers from the upstream region of metallothionein genes. EMBO J. 4:3851–3859. 29. Stinski MF, Roehr TJ. 1985. Activation of the major immediate early gene of human cytomegalovirus by cis-acting elements in the promoterregulatory sequence and by virus-specific trans-acting components. J. Virol. 55:431– 441. 30. Thode S, Schafer A, Pfeiffer P, Vielmetter W. 1990. A novel pathway of DNA end-to-end joining. Cell 60:921–928.

3142

jvi.asm.org

31. van Dam H, et al. 1990. Differential effects of the adenovirus E1A oncogene on members of the AP-1 transcription factor family. Mol. Cell. Biol. 10:5857–5864. 32. Vinces MD, Legendre M, Caldara M, Hagihara M, Verstrepen KJ. 2009. Unstable tandem repeats in promoters confer transcriptional evolvability. Science 324:1213–1216. 33. Weber F, de Villiers J, Schaffner W. 1984. An SV40 “enhancer trap” incorporates exogenous enhancers or generates enhancers from its own sequences. Cell 36:983–992. 34. Weinstock DM, Richardson CA, Elliott B, Jasin M. 2006. Modeling oncogenic translocations: distinct roles for double-strand break repair pathways in translocation formation in mammalian cells. DNA Repair (Amst.) 5:1065–1074. 35. Westin G, Schaffner W. 1988. A zinc-responsive factor interacts with a metal-regulated enhancer element (MRE) of the mouse metallothionein-I gene. EMBO J. 7:3763–3770. 36. Wimmer U, Wang Y, Georgiev O, Schaffner W. 2005. Two major branches of anti-cadmium defense in the mouse: MTF-1/ metallothioneins and glutathione. Nucleic Acids Res. 33:5715–5727.

Journal of Virology