Analysis of Human Cytomegalovirus oriLyt Sequence
http://www.100md.com
病菌学杂志 2005年第6期
Virus Cell Interaction Group, ZAMED, Medical Faculty, Martin-Luther-University of Halle-Wittenberg, Halle (Saale), Germany
ABSTRACT
During the lytic phase of infection, replication of herpesvirus genomes initiates at the lytic origin of replication, oriLyt. Many herpesviruses harbor more than one lytic origin, but so far, only one oriLyt has been identified for human cytomegalovirus (HCMV). Evidence for the existence of additional lytic origins of HCMV has remained elusive. On the basis of transient replication assays with cloned viral fragments, HCMV oriLyt was described as a core region of 1.5 kbp (minimal oriLyt) flanked by auxiliary sequences required for maximal replication activity (complete oriLyt). It remained unclear whether minimal oriLyt alone can drive the replication of HCMV in the absence of its accessory regions. To investigate the sequence requirements of oriLyt in the context of the viral genome, mutant genomes were constructed lacking either minimal or complete oriLyt. These genomes were not infectious, suggesting that HCMV contains only one lytic origin of replication. Either minimal or complete oriLyt was then ectopically reinserted into the oriLyt-depleted genomes. Only the mutant genomes carrying complete oriLyt led to infectious progeny. Remarkably, inversion of the 1.5-kbp core origin relative to its flanking regions resulted in a replication-defective genome. Mutant genomes carrying minimal oriLyt plus the left flanking region gave rise to minifoci, but genomes harboring minimal oriLyt together with the right flanking region were noninfectious. We conclude that the previously defined minimal lytic origin is not sufficient to drive replication of the HCMV genome. Rather, our results underline the importance of the accessory regions and their correct arrangement for the function of HCMV oriLyt.
INTRODUCTION
Human cytomegalovirus (HCMV), like all herpesviruses, has at least two different modes of infection, either lytic replication or latency. Whereas the replication mode of the HCMV genome during latency is rather elusive, substantial progress has been made in dissecting the molecular steps that occur during lytic DNA replication (35, 39; for a review, see reference 41). Following cell entry, HCMV capsids are transported to the nuclear pores, where they release the linear genomes into the cell nucleus. Then, the linear genomes are rapidly circularized and serve as templates for transcription and replication. It has been proposed that the initial steps in DNA replication of herpesviruses take place by a theta-like mechanism, followed by a rolling-circle mechanism of replication at later stages of the lytic infection cycle (reviewed in references 36 and 46). This view of herpesvirus DNA replication has recently been challenged by the observation that genome circularization seems not to be a prerequisite for replication of herpes simplex virus (HSV) and the suggestion that linear genomes serve as replication templates (31). It remains to be shown whether the same applies to HCMV and other herpesviruses as well.
Regardless of which mechanism is used, lytic replication of herpesviruses initiates at defined regions of the viral genomes, the so-called lytic origins of replication (oriLyt). One oriLyt has been defined for HCMV (1, 2, 27, 37). By transient-replication assays with cloned subgenomic fragments, oriLyt was located upstream of the gene encoding the single-stranded DNA binding protein (UL57) approximately in the middle of the unique long (UL) segment (1, 37). In contrast to the origins of alphaherpesviruses, HCMV oriLyt is remarkable for its apparent size and complexity, and it seems to lack any obvious sequence and evolutionary homology to other viral DNA replication origins. Despite the missing sequence homology, however, there is a striking resemblance in the structural organization among the origins of human, chimpanzee, rhesus, rat, and mouse CMVs, e.g., an asymmetric base distribution with an A+T-rich region located at one end and an extended G+C-rich segment at the other end (20, 28, 38, 53). HCMV oriLyt contains numerous reiterated and inverted sequence elements, including known consensus transcription factor recognition sites. Although the analysis of these sequence elements in either oriLyt function or local transcription regulation has just begun (33), their distribution resembles the organization of some complex transcriptional enhancers, and the existence of these binding sites suggests the presence of putative promoter regions. Transcriptional control elements and enhancers are indeed common components of many viral and eukaryotic origins of DNA replication (21, 42, 48, 50, 51). Given the complex structure of HCMV oriLyt, the lack of an apparent trans-acting factor comparable to the HSV origin-binding protein (41), and the recent identification of intraorigin RNA transcripts that in part form stable and persistent RNA-DNA hybrid structures within oriLyt (45), the mechanism of replication initiation in HCMV may be different from that employed by other herpesviruses.
Several efforts have been undertaken to define the sequence requirements and boundaries of the HCMV replicator (1, 2, 37, 59). Using transient replication assays, Masse et al. mapped a 1.5-kbp core region of oriLyt (i.e., minimal oriLyt) to a region situated between nucleotides (nt) 92210 and 93715 of the HCMV genome but stated that full replicator activity requires additional sequences of 1,100 bp to the left and 700 bp to the right of the origin core (37). The complete lytic origin of HCMV was thus determined to be a segment with a size of 3.3 kbp. Insertion and deletion studies combined with transient replication assays performed by Anders and colleagues and Zhu et al. described a core region of HCMV oriLyt that is similar to that defined by Masse et al. but somewhat shifted to the left, namely, between nt 91751 and 93299 (1, 59). Since the introduction of mutations into the accessory sequences flanking either side of this core had little or no effect on replication efficiency (59), the role of these sequences flanking oriLyt requires further investigation.
Some herpesviruses, including HSV and Epstein-Barr virus, harbor more than one lytic origin of replication (26, 46), whereas to date only one lytic origin has been described for HCMV. However, the existence of other lytic origins in HCMV has not been excluded. Although the significance of multiple origins remains to be determined, this raises the question of whether the HCMV genome contains an additional lytic origin(s).
In order to define the sequence requirements for a functional oriLyt of HCMV, we decided to analyze the significance of the core origin and its flanking regions in the context of the HCMV genome. Also, we asked whether the HCMV genome contains additional lytic origins. Using a bacterial artificial chromosome (BAC)-cloned HCMV genome (6), we constructed a series of mutant HCMV genomes in which we first deleted either the minimal or the complete lytic origin and then, in subsequent rounds of mutagenesis, reinserted minimal or complete oriLyt at an ectopic position of the genomes with the origin deleted. Our results imply that the genome of HCMV strain AD169 contains only one lytic origin. We also provide evidence that the accessory regions flanking the origin core are essential for initiation of DNA replication at HCMV oriLyt. Interestingly, the left flanking region seems to be of greater importance for HCMV oriLyt function than the sequences to the right, since upon transfection, mutant genomes with a fragment encompassing minimal oriLyt plus sequences to the left led to the formation of small foci of green fluorescent protein (GFP)-expressing cells, whereas the sequences to the right plus minimal oriLyt did not.
MATERIALS AND METHODS
Virus and cells. The parental HCMV strain used in this report is the BAC-derived recombinant virus HCMV-GFP (8) originating from the BAC-cloned HCMV AD169 genome (6). Culture of human foreskin fibroblasts (HFF), preparation of virus stocks, growth curves, and determination of virus titers were essentially done as described previously (6). The adenovirus 5 recombinant dl1014—defective for most of the E4 region (11)—was propagated on the E4-complementing cell line W162 (56) (both kindly provided by Gary Ketner, Johns Hopkins Bloomberg School of Public Health, Baltimore, Md.). W162 cells were cultured in minimal essential medium (Invitrogen) supplemented with 10% fetal calf serum, 2 mM glutamine, 100 U of penicillin per ml, and 100 μg of streptomycin sulfate per ml. For preparation of replication-deficient, UV-inactivated adenovirus particles, dl1014 was isolated from infected W162 cells by three cycles of freeze-thawing, purified by two rounds of CsCl gradient centrifugation, and inactivated by treatment with psoralen and UV irradiation as described previously (17).
Plasmids. The nucleotide positions of the AD169 genome provided here refer to the GenBank entry with accession number X17403 (13). For construction of pOri6K-Kan1 and pOri6K-Kan4, which contain a kanamycin resistance marker plus one FLP recognition target (FRT) site in different orientations, and the bacterial origin of replication R6K (34), a 1.0-kbp fragment encompassing the kanamycin resistance gene from transposon Tn903 plus one wild-type FRT site was amplified by PCR using the primers Kanori6k.for (5'-CGC AGA TCT GGT GAC CAC GTC GTG GAA TGC CTT-3') and Kanori6k.rev (5'-CCC AGA TCT CGA TTT ATT CAA CAA AGC CAC GTT-3') with pSLFRTKn (3) as a template. The resulting PCR product was digested with BglII and ligated to a 0.5-kbp fragment comprising the R6K origin isolated from plasmid pGP704 (40) (kindly provided by Holger Rüssmann, University of Munich, Munich, Germany) by BamHI digestion. To generate plasmid pOri6K-F5, which harbors the kanamycin resistance cassette flanked by mutant FRT sites, PCR was performed with pSLFRTKn as a template and primers FRT-F5.for (5'-CGC AGA TCT GAA GTT CCT ATA CCT TTT GAA GAA TAG GAA CTT CGC CAG TGT TAC AAC CAA TTA ACC-3') and FRT-F5.rev (5'-CGA GAA TTC AGA AGT TCC TAT TCT TCA AAA GGT ATA GGA ACT TCC GAT TTA TTC AAC AAA GCC ACG T-3'), which provided mutant FRT sites (49). The PCR product was digested with EcoRI and BglII and cloned into the EcoRI and BglII sites of pGP704. For ectopic insertion of minimal oriLyt, plasmids p6K-ori1.5-3 and p6K-ori1.5-5 were constructed. A 1.5-kbp EcoRI-SacI fragment comprising minimal oriLyt (nt 92210 to 93715 of the AD169 genome) (13) was obtained from pBC-oriLyt (9), which carries the HCMV origin region from nt 92210 to 94860 derived from pCM1029 (25). The fragment was treated with the Klenow fragment of Escherichia coli DNA polymerase I and cloned into pOri6K-Kan1 and pOri6K-Kan4, which were cut with EcoRV. The sequence of the 1.5-kbp fragment was determined and compared to the AD169 DNA sequence. We did not find any alterations, except in a region located upstream of the vRNA-2 start site, which was previously reported to exhibit a heterogeneous restriction pattern (45). The sequences equivalent to AD169 nt 93092 to 93218 were missing, and the sequences equivalent to nt 93219 to 93299 were repeated two times. Since the minimal oriLyt is derived from plasmid pBC-oriLyt, which was shown to replicate efficiently in transient replication assays (9), the differences apparently reflect a naturally occurring polymorphism of HCMV oriLyt sequences.
Plasmids p6K-ori3.7-1 and p6K-ori3.7-4, which carry the complete lytic origin, were generated by releasing a 3.7-kbp fragment (nt 91165 to 94860) from HCMV cosmid clone pCM1029 (25) by digestion with DraI and KpnI, followed by insertion into pOri6K-Kan1 and pOri6K-Kan4, which were cut with KpnI and EcoRV. p6K-ori-right, containing the 1.5-kbp origin core plus the right flanking sequences, was constructed by inserting a 2.6-kbp EcoRI-KpnI-fragment (nt 92210 to 94860) derived from pBC-oriLyt into the EcoRV and KpnI restriction sites of pOri6K-Kan-4 after blunting the EcoRI ends using Klenow fragment, and p6K-ori-left originated from pOri6K-Kan-1, which was cut with EcoRV and SacI, followed by insertion of a 2.5-kbp DraI-SacI-fragment (nt 91165 to 93715) obtained from pCM1029. All plasmids containing the R6K origin were propagated in the E. coli strain PIR1 (Invitrogen), which provides the phage lambda protein.
For construction of the shuttle plasmid pST-ori-invers, the oligonucleotides ori-oligo.for (5'-CAG CTG CAC GAT ATC GTG CA-3') and ori-oligo.rev (5'-CGA TAT CGT GCA GCT GTG CA-3') were annealed and cloned into the PstI site of pACYC177 (New England Biolabs). The resulting plasmid, pACYC-O, was digested with EcoRV and PvuII, and a 6.1-kbp PvuII-EcoRV fragment (nt 89795 to 95967) isolated from pCM1029 was inserted, leading to pACYC-ori. This plasmid was cut with EcoRI and SacI, releasing DNA sequences (nt 92210 to 94636) that contain the minimal lytic origin, and a 920-bp EcoRI-SacI fragment (encompassing nt 93717 to 94636) amplified by PCR using the primers ori-sac.for (5'-CGC GAA TTC GTT AAC GGG AGC GGG GTC GAC CGC GA-3') and ori-sac.rev (5'-CCG GAG CTC AGA ACG GCA TGA-3') was reinserted, creating pACYC-ori-PCR. An HpaI cloning site introduced together with the PCR fragment and located directly adjacent to the EcoRI site was used to insert the minimal oriLyt sequences in inverse orientation. The EcoRI-SalI fragment (nt 92210 to 93727) carrying the minimal oriLyt was obtained from pBC-oriLyt (9) and treated with Klenow polymerase. Finally, the complete insert carrying the inverted minimal lytic origin plus the flanking sequences was isolated as an EcoRV-PvuII fragment and cloned into the SmaI site of shuttle plasmid pST76-KSR (10), resulting in pST-ori-invers.
BAC mutagenesis. Mutant HCMV BACs pHG-1, pHG-1.5, and pHG-3.7 were constructed in E. coli DH10B by the ET mutagenesis method (58), which is based on homologous recombination between a linear fragment and the HCMV BAC. The linear fragment encodes antibiotic resistance and is generated by PCR using primers with a size of 85 bp; 60 nt at the 5' end of the primers was homologous to the intended integration site in the BAC-cloned CMV genome, and 25 nt at the 3' end was specific for the antibiotic resistance gene. The recombination functions were provided by plasmid pKD46, which carries the phage lambda recombination genes red , ?, and under the control of an inducible promoter (19). For construction of pHG-1, a PCR fragment was used that contained a tetracycline resistance marker flanked by wild-type FRT sites and that was generated using pCP16 (16) as a template and primers UL1-10ko.for (5'-CAT GGG CAT GCA ATG CAA CAC TAA ATT GTT ATT GCC AGT CGC ACT AAT ACC GGT TGT AAT CAC CAG GGT TTT CCC AGT CAC GAC GT-3') and UL1-10ko.rev (5'-TCT CAG GTT CCG TAG GGA CTG TCC CTA TAT TCA AAC AGT GGG TTT GAA ACC GGA GAT ATC CAC AGG AAA CAG CTA TGA CCA TGA-3'). To generate mutant BACs pHG-1.5 and pHG-3.7, a PCR fragment encoding kanamycin resistance was used, which was obtained by PCR with plasmid pOri6K-F5 and primers ori15ko.for (5'-TGA CAA GCC TAC GTC ACA CTC GCG TGA CCA CAC CCA CTC CGG ATA TAC GTC ATC CTG TGG CCC CGA AAA GTG CCA CCT GCA GAT-3') and ori15ko.rev (5'-GCC GCC GAG AGG GAG CCG CCG CGA GAC CCG GAA GCC GTC GCG GTC GAC CCC GCT CCC GGA GTG ACA CAG GAA CAC TTA ACG GCT GA-3') and ori37ko.for (5'-CCC CCG ACA CGT ACT CGA AGA CCT TAC AGT TTA CGA GTC AAT AAA ACA GGA AAA GAT CCG ACC CCG AAA AGT GCC ACC TGC AGA T-3') and ori37ko.rev (5'-GCG CCC TCT AGG TAG CAT CAC GGA ACA GGC GGC GAC CGC GGA CTC GCC TCG GCC AGG GGG GTG ACA CAG GAA CAC TTA ACG GCT GA-3'), respectively. Excision of the FRT-flanked antibiotic resistance markers was achieved through FLP-mediated recombination in E. coli as described previously (7, 16). HCMV BAC pHG-10 was constructed by a two-step replacement procedure using the shuttle plasmid pST-ori-invers and HCMV BAC pHG-1.5 essentially as described previously (6). In brief, E. coli bacteria containing pHG-1.5 were transformed with the shuttle plasmid and incubated at 30°C on agar plates containing 17 μg of chloramphenicol/ml and 25 μg of kanamycin/ml. Clones harboring cointegrates were selected by streaking the bacteria on agar plates containing chloramphenicol and kanamycin and incubation at 43°C. The bacteria were then restreaked on agar plates containing chloramphenicol only and grown at 37°C overnight. Selection of clones that had resolved the cointegrates was done on agar plates containing chloramphenicol plus 5% sucrose at 30°C. Finally, BAC DNA was isolated from 10-ml cultures and checked for the presence of the inverted sequences by restriction analysis.
Mutant BACs pHG-2 and pHG-3 were generated by FLP recombinase-driven integration of plasmids p6K-ori1.5-3 and p6K-ori1.5-5. For this purpose, bacteria harboring HCMV BAC pHG-3.7 were transformed with plasmid pCP20 (16), which encodes FLP. Bacterial colonies were selected on agar plates containing 17 μg of chloramphenicol/ml and 50 μg of ampicillin/ml at 30°C. The bacteria containing both pCP20 and the HCMV BAC were then transformed with either p6K-ori1.5-3 or p6K-ori1.5-5, and clones harboring recombinant BACs were selected by incubation on agar plates containing 17 μg of chloramphenicol/ml and 30 μg of kanamycin/ml at 43°C. BACs pHG-4 and pHG-5 were generated in the same way, i.e., by integration of p6K-ori1.5-3 and p6K-ori1.5-5 into BAC pHG-1.5, and BAC genomes pHG-6, pHG-7, pHG-8, and pHG-9 were obtained by FLP-mediated insertion of p6K-ori3.7-1, p6K-ori3.7-4, p6K-ori-left, and p6K-ori-right into pHG-3.7. The correct integration of the R6K-based plasmids was verified by PCR analysis and sequencing.
Analysis of virus and BAC DNAs. HCMV BAC DNA was isolated from 10-ml overnight E. coli cultures by the alkaline lysis procedure (47) and characterized by restriction analysis followed by agarose gel electrophoresis as described previously (6). Large preparations of HCMV BACs were obtained from 100-ml bacterial cultures using Nucleobond PC 100 columns (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions. Viral DNA was isolated from virions purified from the supernatant of infected cells as described previously (7), and Southern blot analysis of BAC and virus DNAs was performed as reported previously (6). The DNA fragments used for hybridization were either obtained by PCR (probe 1, equivalent to nt 90549 to 90935 of the HCMV genome) using primers orihyb.for (5'-CGG CTT TGC GTC ACC GCC AAC TAA-3') and orihyb.rev (5'-GCG ATT GAG GTT GCG AGA CCA GAT-3') with pCM1029 (25) as a template or excised as a 2.2-kbp MluI fragment (probe 2, equivalent to nt 92018 to 94238 of the HCMV genome) from p6K-ori3.7-1.
Accumulation of viral DNA was examined by slot blot assay. HFF were infected at a multiplicity of infection (MOI) of 0.02 PFU/cell with the respective viruses, and total DNA was isolated on days 1, 3, 5, 7, and 9 postinfection using the QiaAmp DNA Blood Mini kit (QIAGEN, Hilden, Germany). One microgram of each DNA sample was transferred to a nylon membrane using a slot blot apparatus. The membrane was then hybridized to a 32P-labeled DNA probe specific for the UL52 open reading frame (ORF). Finally, specific radioactive signals were quantified using a phosphorimager.
Transfection of HCMV BACs. HCMV BAC DNA was transfected into HFF using the adenovirus-mediated gene delivery protocol developed by Cotten et al. (5, 18). In brief, 10 μl of BAC DNA was diluted in 20 mM HEPES, pH 7.4-150 mM NaCl and condensed with polyethylenimine MW 2000 (Sigma) by incubation for 20 min at room temperature. Particles of inactivated adenovirus dl1014 (2 x 109) were added to the mixture and incubated at room temperature for another 20 min. The transfection complexes were applied to the cells, which were kept in serum-free culture medium containing 30 μg of polymyxin B (Sigma)/ml. After 4 h of incubation in a 37°C incubator with 5% CO2, the fibroblasts were washed twice with PBS and further cultivated in complete cell culture medium. All transfection experiments were done in triplicate. CMV BAC genomes that appeared to be replication deficient were tested in three independent experiments, and transfected cells were incubated until complete cytopathic effect (CPE) had occurred on the fibroblast monolayers transfected with the control BAC pHG-1.
RESULTS
HCMV genomes with a deletion of either the minimal or the complete lytic origin of replication are not infectious. For some herpesviruses, it has been shown that more than one lytic origin is present in the virus genome. In contrast, only one oriLyt has been identified in HCMV (1, 27, 37). Since HCMV oriLyt was primarily defined on the basis of transient replication assays with cloned viral fragments and construction of recombinant genomes with mutated oriLyt has not been possible, the existence of additional lytic origins in the HCMV genome could not be excluded.
In order to address the question of whether the HCMV genome contains other lytic origins apart from the one described, a mutant genome lacking minimal oriLyt was constructed. The design of the mutant was based on the boundaries of oriLyt as defined by Masse et al. (37). They mapped the core of oriLyt between nucleotide positions 92210 (EcoRI) and 93715 (SacI) of the HCMV genome (Fig. 1). Deleting minimal oriLyt was facilitated by applying the ET mutagenesis technique (58), which was recently adapted for the manipulation of BAC-cloned CMV genomes (7, 55) (for technical details, see Materials and Methods). The HCMV BAC pHG-1, a derivative of the previously described HCMV-GFP BAC (8), served as the backbone for construction of all mutant genomes reported in this study (Fig. 2 provides schematic drawings of all mutant genomes). Deletion of the nonessential ORFs UL1 to UL10 in pHG-1 allowed the subsequent insertion of DNA fragments with sizes of up to 6.3 kbp without generating an overlong HCMV genome (see below). The virus derived from pHG-1 expressed the enhanced green fluorescent protein, thereby facilitating the tracking of infected or transfected cells (Fig. 3C), and it showed growth kinetics comparable to those of its parental virus, HCMV-GFP, as well as those of the original BAC-derived virus, RVHB5 (6) (data not shown).
For deletion of minimal oriLyt, a kanamycin resistance (KanR) gene flanked by mutant FRT sites was inserted into the HCMV BAC pHG-1, thereby replacing minimal oriLyt (Fig. 3B). The resulting BAC pHG-1.5-K was characterized by restriction analysis using BglII (Fig. 3A). Following mutagenesis, a fragment of 8.4 kbp characteristic of the parental BAC pHG-1 was replaced by two new fragments of 4.9 and 3.0 kbp in pHG-1.5-K (Fig. 3A, lane 2). The KanR cassette was subsequently excised with FLP recombinase (data not shown), resulting in pHG-1.5 (Fig. 2, line 2).
Transfection of large BACs into primary fibroblasts is associated with notoriously poor efficiency. In general, this does not prevent the reconstitution of replication-competent viruses from BAC genomes, but it may impair the interpretation of results when replication-incompetent or severely debilitated genomes are analyzed. Therefore, we searched for a more efficient transfection method and tested the adenovirus-mediated gene delivery procedure developed by Baker and Cotten (5, 18). Using this protocol for transfection of HCMV BACs, we obtained an efficiency that was at least 2 orders of magnitude higher than that of any other transfection protocol that we had tested before.
When HCMV BAC pHG-1.5 was transfected into human fibroblasts, single green cells were observed 2 days posttransfection, which remained single for >10 days. No plaques appeared on the cell monolayer (Fig. 3C), and no infectious virus was released into the supernatant. Thus, deletion of minimal oriLyt resulted in a replication-deficient HCMV genome, suggesting that there is no other lytic origin present in the HCMV genome that can substitute for the deleted oriLyt.
The results of a previous study done with viral DNA fragments led to the conclusion that accessory flanking regions of 1,100 bp to the left and 700 bp to the right of minimal oriLyt are needed to achieve efficient replication originating at HCMV oriLyt (37). In order to evaluate the requirement for these auxiliary regions later, we also constructed a recombinant genome in which the complete oriLyt was deleted. A region of 3.7 kbp extending from nt 91165 (DraI) to 94860 (KpnI) of the HCMV genome was replaced by the KanR marker in mutant BAC pHG-3.7-K (Fig. 3B, line 4). In comparison to the parental BAC pHG-1 (Fig. 3B and A, lane 1), new fragments of 3.9 and 1.9 kbp were present in pHG-3.7-K (Fig. 3A, lane 3). Excision of the kanamycin cassette led to BAC pHG-3.7 (Fig. 2, line 3). Transfection of pHG-3.7 into permissive fibroblasts did not result in the production of infectious progeny (Fig. 3C), providing further evidence that the HCMV genome contains only one lytic origin of replication.
Ectopic insertion of minimal oriLyt does not rescue the replication-deficient phenotype of CMV genomes with the origin deleted. In order to determine the importance of the accessory regions flanking minimal oriLyt, we decided to reinsert either minimal or complete oriLyt at an ectopic position in mutant BACs pHG-1.5 and pHG-3.7, which lack the respective sequences at the original positions (Fig. 2). For insertion of these sequences, we utilized the highly convenient site-specific FLP/FRT recombination system. An FRT site that was inserted at an ectopic position during construction of the parental BAC pHG-1 allowed the insertion of shuttle vectors based on the conditional bacterial R6K replicon (34) and carrying the oriLyt sequences (Fig. 4B) (see Materials and Methods for technical details). Specific targeting of the ectopic locus by the shuttle plasmids was accomplished by using selection markers with wild-type and mutant FRT sites for the deletion of ORFs UL1 to UL10 and of the origin sequences, respectively. Since mutant and wild-type FRT sites cannot interact with each other (49), the shuttle plasmids, which carry a wild-type FRT site, become integrated exclusively at the ectopic position.
In order to evaluate whether the minimal oriLyt by itself is sufficient to mediate replication of the HCMV genome, the sequences were inserted at the ectopic position of BAC pHG-3.7, leading to pHG-2 and pHG-3, which differ in the orientation of the inserted minimal oriLyt (Fig. 2). Characterization of the BAC DNAs with HpaI revealed the expected shift of an 8.1-kbp band to 11.2 kbp in pHG-2 and pHG-3 (Fig. 4A, lanes 4 and 5). After transfection of the mutant BACs into permissive human fibroblasts, single green cells were readily observed, but no plaques appeared on the cells (Fig. 4C), implying that no infectious virus can be recovered from pHG-2 and pHG-3, regardless of the orientation of the ectopically inserted minimal lytic origin. Thus, ectopic insertion of minimal oriLyt is not sufficient to support replication of an HCMV genome in which complete oriLyt has been deleted.
Since deletion of the complete oriLyt may affect the promoter of the essential UL57 gene, we could not exclude the possibility that the observed phenotype of pHG-2 and pHG3 was the result of reduced UL57 expression. Therefore, minimal oriLyt was also reinserted into pHG-1.5, which lacks only the minimal lytic origin at the original position. However, the resulting BACs, pHG-4 and pHG-5 (Fig. 2 and 4), also did not lead to the development of plaques following transfection into human fibroblasts even after 4 weeks of incubation. These results suggest that the minimal lytic origin lacking the accessory flanking regions is not able to drive replication of the HCMV genome.
Reconstitution of infectious genomes after integration of complete oriLyt at an ectopic position in CMV BACs with the origin deleted. Since the minimal lytic origin could not rescue the replication deficiency of pHG-3.7 and pHG-1.5, we tested whether ectopic insertion of the complete lytic origin leads to replication-proficient HCMV genomes. The resulting genomes, pHG-6 and pHG-7, carry the complete oriLyt at the ectopic position and lack these sequences at the original position (Fig. 2 and 4). Transfection of these genomes into permissive cells resulted in the formation of green fluorescent plaques (Fig. 4C) and in a complete cytopathic effect of the cultures 2 weeks posttransfection. In order to analyze whether the genomes of the mutant viruses released from the transfected cells had preserved their original structures and to rule out the possibility that the lytic origin in the original region had been reconstituted, viral DNA was isolated from virions and examined by restriction analysis and Southern blotting. Figure 5A shows the BglII fragments of BACs pHG-6 and pHG-7 (lanes 2 and 4) and of the corresponding virus DNAs (lanes 3 and 5), as well as of the parental BAC pHG-1 (lane 1). In comparison to the circular DNA of the BACs, additional fragments (e.g., of 17.1 kbp) that originate from the terminal fragments of the linear viral DNA are visible in lanes 3 and 5. Probe p1, which is specific for the genomic region located to the left of the complete lytic origin (Fig. 5D), detected the expected 8.4-kbp band of the parental BAC pHG-1 (Fig. 5B, lane 1) and a 3.9-kbp band of pHG-6 and pHG-7, as well as of the corresponding viruses RV-pHG-6 and RV-pHG-7, which is indicative of the deletion of complete oriLyt at the original position (Fig. 5B, lanes 2 to 5, and D). The blot was stripped and reprobed using fragment p2, specific for the lytic origin (Fig. 5D). As expected, hybridization with probe p2 revealed the same 8.4-kbp fragment of the parental BAC pHG-1 as did p1. In DNAs of the mutant BACs pHG-6 and pHG-7, a band of 42 kbp was recognized, which corresponds to the expected fragment harboring the insertion of complete oriLyt (Fig. 5C, lanes 2 and 4, and D, middle line). The DNA bands of the viruses RV-pHG-6 and RV-pHG-7 detected by probe p2 seemed slightly smaller (Fig. 5C, lanes 3 and 5), consistent with the expected size of 29 kbp for the terminal fragments of two of the HCMV genome isomers (Fig. 5D, line 3). Two additional bands of 36 and 42 kbp expected from the other genome isomers generated after inversion of the UL region could not be resolved on the gel shown in Fig. 5 but were readily detected after pulsed-field gel electrophoresis and hybridization with probe 2 (data not shown). Most importantly, hybridization of the probe to fragments with a size of 8.4 kbp or to any additional fragments was not found, ruling out the translocation of the ectopic oriLyt back to its original position. Thus, we concluded that the mutant genomes were stable and that ectopic insertion of complete oriLyt rescued the replication deficiency of genomes with oriLyt deleted. The rescue was independent of the orientation of the reinserted origin sequences. These data also indicated that deletion of complete oriLyt at the original position did not impair the expression of adjacent genes (e.g., UL57), at least not in a way that prevented growth of the viral mutants.
By integration of the sequences of the complete lytic origin at the ectopic position, we could also rescue the replication defect of pHG-1.5, which lacks minimal oriLyt at the original position (data not shown). The resulting genome readily produced plaques after transfection into permissive cells, finally leading to complete CPE. This result confirmed that the growth defect of pHG-1.5 is caused by the deletion of minimal oriLyt and is not due to any other potential mutation elsewhere in the genome.
Growth analysis of HCMV mutants with an ectopic oriLyt. Following transfection of pHG-6 and pHG-7, we got the impression that the resulting plaques were slightly smaller than the plaques originating from the parental virus (Fig. 4C). Therefore, we compared the growth behavior of the mutant viruses RV-pHG-6 and RV-pHG-7 to that of the parental virus after infection of human foreskin fibroblasts with different viral doses. Following infection at a low MOI of 0.025 PFU/cell, the final titers obtained with mutant viruses RV-pHG-6 and RV-pHG-7 were reduced by 1.5 orders of magnitude in comparison to the titers of RV-pHG-1 (Fig. 6A). However, the growth kinetics did not seem to be delayed. Similar results were obtained when cells were infected at a higher MOI of 1.5 PFU/cell (Fig. 6B). Thus, the HCMV mutants harboring the complete lytic origin at the ectopic position exhibited a growth defect in comparison to the parental virus.
To gain further insight into the growth behavior of RV-pHG-6 and RV-pHG-7, we analyzed the amplification of viral DNA in cells infected with either the parental virus, RV-pHG-1, or the mutant viruses RV-pHG-6 and RV-pHG-7. HFF were infected at an MOI of 0.02 PFU/cell, and total DNA was prepared from the infected cells at various times postinfection. The DNA samples were then analyzed by slot blotting, followed by hybridization with an HCMV-specific probe, and the signals obtained were quantified using a phosphorimager. As can be seen in Fig. 6C, comparable amounts of viral DNA were observed in fibroblasts infected with RV-pHG-1, RV-pHG-6, or RV-pHG-7. The mutants carrying the complete lytic origin at the ectopic position did not exhibit a defect in either the kinetics of DNA synthesis or final accumulation of viral DNA. We concluded from this result that the 3.7-kbp lytic origin inserted at the ectopic position comprises all of the information that is necessary to mediate efficient replication of the HCMV genome.
Role of the regions flanking the core replicator for oriLyt activity. Since our data pointed to an important, or even essential, role of the sequences adjacent to minimal oriLyt, we sought to dissect the influence of the right and the left flanking regions on oriLyt activity. Therefore, minimal oriLyt, together with either its right or left flanking region, was inserted at the ectopic position of pHG-3.7, resulting in BACs pHG-8 and pHG-9 (Fig. 2 and 7). To evaluate the infectivity of these mutant genomes, the BACs were transfected into human fibroblasts, resulting in numerous green cells. Whereas these green cells remained single in pHG-9-transfected cultures, small foci of GFP-expressing cells were observed 3 weeks after transfection of pHG-8 (Fig. 7C). However, we did not observe the formation of plaques and the minifoci did not develop further into complete CPE as was observed with the parental BAC pHG-1. Therefore, we assume that pHG-8 is replication proficient, but the growth rate of the resulting virus is probably too low to promote the spread of the infection throughout the culture.
Having shown that the sequences flanking the oriLyt core are crucial for origin function, we assessed whether the orientation of minimal oriLyt relative to its flanking regions is relevant. To this end, we inverted the 1.5-kbp origin core sequence at its original position. Starting with pHG-1.5 missing the minimal oriLyt, the core element was reintroduced in reverse orientation by a two-step mutagenesis procedure (see Materials and Methods for technical details). The DNAs of several BAC clones were analyzed by restriction analysis using BglII following mutagenesis (Fig. 8A). Two clones (clones 2 and 4) acquired the reversed minimal oriLyt, as can be seen by the presence of an additional 8.4-kbp fragment and the lack of the 4.9- and 2.0-kbp bands (Fig. 8A, lanes 2 and 4). In order to test the consequences of the inversion, pHG-10 clones 2 and 4 were transfected into permissive fibroblasts, leading to numerous green fluorescent cells. However, only single green cells could be observed during a 4-week incubation period, but no plaques appeared on the cell monolayers (Fig. 8B). This experiment implied that the appropriate orientation of the minimal lytic origin with respect to its flanking regions is crucial for origin function.
DISCUSSSION
In this study, we investigated the sequence requirements of the described lytic origin of human cytomegalovirus (oriLyt) in the context of the viral genome. To our knowledge, this is the first report of the intentional generation of recombinant HCMV genomes and viruses carrying mutations of oriLyt. Our results provide evidence that the HCMV genome of strain AD169 contains only one lytic origin, because HCMV genomes with deletion of either minimal or complete oriLyt were not infectious. Furthermore, we show that the 1.5-kbp region designated the HCMV core is not sufficient to drive replication of HCMV, since ectopic insertion of this sequence into a nonessential region of the viral genome did not rescue the replication-deficient phenotype of the genomes with the origin deleted. Rather, accessory sequences flanking minimal oriLyt are needed for replication of the HCMV genome, as insertion of the complete lytic origin (encompassing these regions) readily resulted in infectious viral progeny. The observation that genomes with a reinsertion of the origin core together with its left flanking region gave rise to the formation of minifoci whereas genomes harboring the origin core plus the sequences to the right did not suggested the presence of an essential element within the left flanking region. Notably, correct orientation of the origin core relative to its auxiliary regions was crucial for replicator function.
One of the biggest advantages of the BAC-based mutagenesis of herpesviruses is the fact that construction and testing of mutant genomes are separate steps. This allowed us to manipulate HCMV oriLyt, a cis-active element of the genome considered to be essential for viral replication. The BAC technique enabled us to perform several rounds of mutagenesis on the very same genome, independent of the phenotype of the intermediates, as well as to construct replication-deficient CMV genomes that harbor lethal mutations in essential elements of oriLyt.
Deletion of minimal or complete oriLyt led to replication deficiency of the resulting genomes, indicating that the deleted sequences are essential and that other elements of the genome cannot substitute for them. These findings suggest that in contrast to human alpha- and gamma-herpesviruses (4, 26, 46, 57), the HCMV genome contains only one lytic origin. This might be a unique feature of beta-herpesviruses, since so far only one single origin of replication has also been described for simian, rat, and mouse CMV, as well as for human herpesvirus types 6 and 7 (2, 23, 38, 52, 53). Of course, this hypothesis requires experimental examination. Our results apply only to the laboratory strain AD169, and due to the fact that clinical isolates of HCMV contain additional DNA sequences (12), we cannot rule out the possibility that such CMV strains harbor more than one lytic origin. Whether the presence of several lytic origins provides a replication advantage remains elusive, because deletion of oriL or both copies of oriS of HSV had only marginal effects on viral growth (30, 44).
The final evidence that the replication deficiency of the genomes with the origin deleted is due to the deletion of oriLyt is technically difficult to supply. The phenotype could also result from the impairment of another essential element, e.g., the ORF of an essential gene located in the same region, and the genetic elements may be physically inseparable. For the following reasons we consider this rather unlikely. (i) Recent reevaluation of the HCMV coding potential based on comparison with the genetic information for the closely related chimpanzee CMV strongly suggested that most, if not all, of the previously annotated ORFs within the region encompassing oriLyt do not encode functional proteins (20, 43). (ii) Meanwhile, some of the ORFs have indeed been shown to be nonessential (24). (iii) Sequence variability in the oriLyt region was observed, a fact that is hardly compatible with the potential to encode proteins (14, 15, 32, 45). (iv) In agreement with the phenotypes of the genomes with oriLyt deleted described in this study, at least three regions within minimal oriLyt were found to be indispensable for origin function on the basis of transient replication assays (1, 37, 45, 59).
Assuming that oriLyt does essentially act in cis by providing an initiation site for the replication machinery, we tested whether the defect of the genomes with the origin deleted could be rescued by ectopic reinsertion of these sequence elements. This was indeed the case for complete oriLyt, whereas minimal oriLyt was not sufficient. In view of the low replication activity of minimal oriLyt that was determined in transient transfection experiments using the subcloned element (1, 37, 59), the observed insufficiency of minimal oriLyt to support replication of the complete genome was perhaps not surprising. The results indicate, however, that the oriLyt of HCMV represents more than just an initiation site for replication. This may be in contrast to the origins of alpha-herpesviruses that span only 150 bp and basically provide the start site for DNA replication and binding sites for the origin-binding protein (46). Possibly, the lack of an origin-binding protein of HCMV was the driving force for the development of such a large and complex origin and of cis-active sequence elements that could either substitute for the function of an origin-binding protein or support initiation of DNA replication by an alternative mechanism.
Rescue of the replication-deficient phenotype by reinsertion of the complete oriLyt demonstrated that these sequences encompass all elements necessary to initiate DNA replication at the ectopic position. Moreover, these results show that the activity of this important cis-active element is neither position nor orientation dependent. The mutants were not deficient either in the kinetics of viral DNA synthesis or in the DNA yields. These findings also strongly suggest that the mutants were not impaired in the expression of transcripts that are essential for the initiation and progression of viral DNA replication, e.g., the smallest replicator transcript (SRT), v-RNA-2, or the UL57 RNA encoding the single-stranded DNA binding protein. The 5' ends of the UL57 transcripts were precisely mapped in a recent study (33). Although the UL57 promoter may extend considerably upstream, it is extremely unlikely that the growth defect of the genomes with a deletion of minimal oriLyt was caused by an impairment of UL57 expression, especially since pHG-1.5 could be rescued by ectopic insertion of the complete oriLyt. Since the left boundary (nt 91165) of the complete oriLyt as used in this study and the most distal transcription start site of the UL57 RNA (nt 91138), as well as the candidate promoter sequences (33), are located in close proximity, one cannot completely rule out an effect on UL57 expression in the mutants carrying a deletion of complete oriLyt. However, given the pivotal role of the UL57 protein (single-stranded DNA binding protein) for replication of HCMV DNA, impaired UL57 expression would hardly be compatible with the unaltered DNA synthesis displayed by the mutants RV-pHG-6 and RV-pHG-7.
In view of the results of the DNA synthesis, the reduced viral yields of these mutants are difficult to explain. One hypothesis is that the promoter-enhancer elements in the origin sequences may have an effect on the expression of genes with important functions in the infection cycle that are not involved in DNA replication. Further work will be needed to define the reasons for the impaired growth properties of these mutants.
Having shown that the accessory regions of minimal oriLyt are essential for replication of the genome, we investigated whether one of the flanking regions may be dispensable. Transfection of genomes containing minimal oriLyt plus the left flanking region resulted in small groups of GFP-expressing cells. We called these clusters of green cells minifoci, because they did not develop into typical plaques and complete CPE did not occur. Although the experimental evidence is difficult to obtain, we assume that the pHG-8 genome is able to replicate in the initially transfected cells, albeit at very low efficiency. The resulting progeny are probably able to spread to the neighboring cells, but then the infection terminates because either the growth rate of the virus is too low or an antiviral state is induced in the surrounding cells, e.g., by cytokines, which cannot be overcome.
Genomes with minimal oriLyt plus the right flanking region did not lead to plaques or to minifoci, suggesting that an essential element is located to the left of minimal oriLyt. In fact, the precise boundaries of HCMV oriLyt have been a matter of debate. Masse et al. specified that sequences flanking the minimal oriLyt contribute to efficient replication but are not essential for origin function (37). Zhu et al. mapped the core of oriLyt to a region encompassing nt 91751 to 93299 (59), which includes part of the left flanking sequence of minimal oriLyt as defined by Masse et al. According to Zhu et al., the core replicator is necessary, but not sufficient, to drive DNA replication in transient assays (59). The fact that ectopic insertion of the left accessory sequence plus minimal oriLyt led to replication-competent genomes underlines its importance for origin function. Since the resulting virus displayed an extremely reduced growth capacity, one has to conclude that the right flanking sequences also play an important, yet not essential, role for replicator activity.
The accessory regions of HCMV oriLyt contain numerous binding sites for a series of transcription factors. Thus, one may speculate that these regions mediate their effect on the core replicator in a manner similar to that of enhancer sequences on promoter regions. If this is true, one would expect that they can act in an orientation- and distance-independent manner. Since inversion of minimal oriLyt resulted in a replication-incompetent genome, the hypothesis has to be discarded. Instead, the transcription factor binding sites rather may represent promoter elements, and indeed, a number of transcripts arise from the HCMV origin region, including the SRT, vRNA-1, and vRNA-2 (Fig. 1) (29, 45). Origin-associated transcripts were also identified in other viral, as well as cellular, origins (22, 54). Possibly, they have no messenger function but rather participate either in the replication process itself, e.g., by serving as a primer for the DNA polymerase, or in a more indirect manner, e.g., by inducing changes in the superhelical structure of the DNA template. Interestingly, all known transcripts from the HCMV origin region extend from right to left in the prototype arrangement of the viral genome. If we assume that these transcripts have an important role in the initiation of viral DNA replication, one can understand that inversion of minimal oriLyt and the concomitant change in the direction of transcription of the vRNA-2 and SRT genes will disrupt origin function.
We report in this study the first HCMV recombinants with mutations in oriLyt. By using the highly convenient FRT/FLP recombination system, we established a fast and easy way to reinsert oriLyt sequences into a nonessential ectopic region of the HCMV genome in order to dissect the roles of cis-acting elements of oriLyt. The BAC-based mutagenesis techniques also allowed us to manipulate the origin sequences at their original position. In further work, we can evaluate the functions of specific origin elements, which were defined in transient replication assays, in the context of the viral genome to deepen our understanding of the complex mechanism of HCMV DNA replication.
ACKNOWLEDGMENTS
This work was supported by the NBL3 program of the Bundesministerium für Bildung und Forschung (BMBF) and the Wilhelm Roux research program of the Medical Faculty, Martin-Luther-University of Halle-Wittenberg.
We are especially indebted to M. Cotten for providing the adenovirus-mediated gene delivery protocol and tools and to G. Ketner for the adenovirus recombinant dl1014 and W162 cells. We thank B. L. Wanner and M. Berlyn (E. coli Stock Center, Yale University) for sending plasmid pKD46 and H. Rüssmann for pGP704. We are also grateful to Z. Ruzsics for help with the adenovirus stock preparation.
REFERENCES
Anders, D. G., M. A. Kacica, G. Pari, and S. M. Punturieri. 1992. Boundaries and structure of human cytomegalovirus oriLyt, a complex origin for lytic-phase DNA replication. J. Virol. 66:3373-3384.
Anders, D. G., and S. M. Punturieri. 1991. Multicomponent origin of cytomegalovirus lytic-phase DNA replication. J. Virol. 65:931-937.
Atalay, R., A. Zimmermann, M. Wagner, E. Borst, C. Benz, M. Messerle, and H. Hengel. 2002. Identification and expression of human cytomegalovirus transcription units coding for two distinct Fc receptor homologs. J. Virol. 76:8596-8608.
AuCoin, D. P., K. S. Colletti, Y. Xu, S. A. Cei, and G. S. Pari. 2002. Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) contains two functional lytic origins of DNA replication. J. Virol. 76:7890-7896.
Baker, A., and M. Cotten. 1997. Delivery of bacterial artificial chromosomes into mammalian cells with psoralen-inactivated adenovirus carrier. Nucleic Acids Res. 25:1950-1956.
Borst, E. M., G. Hahn, U. H. Koszinowski, and M. Messerle. 1999. Cloning of the human cytomegalovirus (HCMV) genome as an infectious bacterial artificial chromosome in Escherichia coli: a new approach for construction of HCMV mutants. J. Virol. 73:8320-8329.
Borst, E. M., S. Mathys, M. Wagner, W. Muranyi, and M. Messerle. 2001. Genetic evidence of an essential role for cytomegalovirus small capsid protein in viral growth. J. Virol. 75:1450-1458.
Borst, E. M., and M. Messerle. 2000. Development of a cytomegalovirus vector for somatic gene therapy. Bone Marrow Transplant. 25(Suppl. 2):S80-S82.
Borst, E. M., and M. Messerle. 2003. Construction of a cytomegalovirus-based amplicon: a vector with a unique transfer capacity. Hum. Gene Ther. 14:959-970.
Borst, E. M., G. Posfai, F. Pogoda, and M. Messerle. 2004. Mutagenesis of herpesvirus BACs by allele replacement, p. 269-280. In S. Zhao and M. Stodolsky (ed.), Bacterial artificial chromosomes, vol. 2. Functional studies. Humana Press Inc., Totowa, N.J.
Bridge, E., and G. Ketner. 1989. Redundant control of adenovirus late gene expression by early region 4. J. Virol. 63:631-638.
Cha, T. A., E. Tom, G. W. Kemble, G. M. Duke, E. S. Mocarski, and R. R. Spaete. 1996. Human cytomegalovirus clinical isolates carry at least 19 genes not found in laboratory strains. J. Virol. 70:78-83.
Chee, M. S., A. T. Bankier, S. Beck, R. Bohni, C. M. Brown, R. Cerny, T. Horsnell, C. A. Hutchison, T. Kouzarides, J. A. Martignetti, E. Preddie, S. C. Satchwell, P. Tomlinson, K. M. Weston, and B. G. Barrell. 1990. Analysis of the protein-coding content of the sequence of human cytomegalovirus strain AD169. Curr. Top. Microbiol. Immunol. 154:125-169.
Chen, Z., S. Sugano, and S. Watanabe. 1999. A 189-bp repeat region within the human cytomegalovirus replication origin contains a sequence dispensable but irreplaceable with other sequences. Virology 258:240-248.
Chen, Z., S. Watanabe, and N. Yamaguchi. 1996. Strain-dependent differences in the human cytomegalovirus replication origin. Arch. Virol. 141:13-30.
Cherepanov, P. P., and W. Wackernagel. 1995. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158:9-14.
Cotten, M., M. Saltik, M. Kursa, E. Wagner, G. Maass, and M. L. Birnstiel. 1994. Psoralen treatment of adenovirus particles eliminates virus replication and transcription while maintaining the endosomolytic activity of the virus capsid. Virology 205:254-261.
Cotten, M., E. Wagner, K. Zatloukal, S. Phillips, D. T. Curiel, and M. L. Birnstiel. 1992. High-efficiency receptor-mediated delivery of small and large (48 kilobase) gene constructs using the endosome-disruption activity of defective or chemically inactivated adenovirus particles. Proc. Natl. Acad. Sci. USA 89:6094-6098.
Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.
Davison, A. J., A. Dolan, P. Akter, C. Addison, D. J. Dargan, D. J. Alcendor, D. J. McGeoch, and G. S. Hayward. 2003. The human cytomegalovirus genome revisited: comparison with the chimpanzee cytomegalovirus genome. J. Gen. Virol. 84:17-28.
DePamphilis, M. L. 1993. Eukaryotic DNA replication: anatomy of an origin. Annu. Rev. Biochem. 62:29-63.
DePamphilis, M. L. 1993. How transcription factors regulate origins of DNA replication in eukaryotic cells. Trends Cell Biol. 3:161-167.
Dewhurst, S., D. M. Krenitsky, and C. Dykes. 1994. Human herpesvirus 6B origin: sequence diversity, requirement for two binding sites for origin-binding protein, and enhanced replication from origin multimers. J. Virol. 68:6799-6803.
Dunn, W., C. Chou, H. Li, R. Hai, D. Patterson, V. Stolc, H. Zhu, and F. Liu. 2003. Functional profiling of a human cytomegalovirus genome. Proc. Natl. Acad. Sci. USA 100:14223-14228.
Fleckenstein, B., I. Muller, and J. Collins. 1982. Cloning of the complete human cytomegalovirus genome in cosmids. Gene 18:39-46.
Hammerschmidt, W., and B. Sugden. 1988. Identification and characterization of oriLyt, a lytic origin of DNA replication of Epstein-Barr virus. Cell 55:427-433.
Hamzeh, F. M., P. S. Lietman, W. Gibson, and G. S. Hayward. 1990. Identification of the lytic origin of DNA replication in human cytomegalovirus by a novel approach utilizing ganciclovir-induced chain termination. J. Virol. 64:6184-6195.
Hansen, S. G., L. I. Strelow, D. C. Franchi, D. G. Anders, and S. W. Wong. 2003. Complete sequence and genomic analysis of rhesus cytomegalovirus. J. Virol. 77:6620-6636.
Huang, L., Y. Zhu, and D. G. Anders. 1996. The variable 3' ends of a human cytomegalovirus oriLyt transcript (SRT) overlap an essential, conserved replicator element. J. Virol. 70:5272-5281.
Igarashi, K., R. Fawl, R. J. Roller, and B. Roizman. 1993. Construction and properties of a recombinant herpes simplex virus 1 lacking both S-component origins of DNA synthesis. J. Virol. 67:2123-2132.
Jackson, S. A., and N. A. DeLuca. 2003. Relationship of herpes simplex virus genome configuration to productive and persistent infections. Proc. Natl. Acad. Sci. USA 100:7871-7876.
Kemble, G., G. Duke, R. Winter, and R. Spaete. 1996. Defined large-scale alterations of the human cytomegalovirus genome constructed by cotransfection of overlapping cosmids. J. Virol. 70:2044-2048.
Kiehl, A., L. Huang, D. Franchi, and D. G. Anders. 2003. Multiple 5' ends of human cytomegalovirus UL57 transcripts identify a complex, cycloheximide-resistant promoter region that activates oriLyt. Virology 314:410-422.
Kolter, R., M. Inuzuka, and D. P. Helinski. 1978. Trans-complementation-dependent replication of a low molecular weight origin fragment from plasmid R6K. Cell 15:1199-1208.
LaFemina, R. L., and G. S. Hayward. 1983. Replicative forms of human cytomegalovirus DNA with joined termini are found in permissively infected human cells but not in non-permissive Balb/c-3T3 mouse cells. J. Gen. Virol. 64:373-389.
Lehman, I. R., and P. E. Boehmer. 1999. Replication of herpes simplex virus DNA. J. Biol. Chem. 274:28059-28062.
Masse, M. J., S. Karlin, G. A. Schachtel, and E. S. Mocarski. 1992. Human cytomegalovirus origin of DNA replication (oriLyt) resides within a highly complex repetitive region. Proc. Natl. Acad. Sci. USA 89:5246-5250.
Masse, M. J., M. Messerle, and E. S. Mocarski. 1997. The location and sequence composition of the murine cytomegalovirus replicator (oriLyt). Virology 230:350-360.
McVoy, M. A., and S. P. Adler. 1994. Human cytomegalovirus DNA replicates after early circularization by concatemer formation, and inversion occurs within the concatemer. J. Virol. 68:1040-1051.
Miller, V. L., and J. J. Mekalanos. 1988. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J. Bacteriol. 170:2575-2583.
Mocarski, E. S., and C. T. Courcelle. 2001. Cytomegaloviruses and their replication, p. 2629-2674. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Murakami, Y., and Y. Ito. 1999. Transcription factors in DNA replication. Front. Biosci. 4:D824-D833.
Murphy, E., I. Rigoutsos, T. Shibuya, and T. E. Shenk. 2003. Reevaluation of human cytomegalovirus coding potential. Proc. Natl. Acad. Sci. USA 100:13585-13590.
Polvino-Bodnar, M., P. K. Orberg, and P. A. Schaffer. 1987. Herpes simplex virus type 1 oriL is not required for virus replication or for the establishment and reactivation of latent infection in mice. J. Virol. 61:3528-3535.
Prichard, M. N., S. Jairath, M. E. Penfold, S. St Jeor, M. C. Bohlman, and G. S. Pari. 1998. Identification of persistent RNA-DNA hybrid structures within the origin of replication of human cytomegalovirus. J. Virol. 72:6997-7004.
Roizman, B., and D. M. Knipe. 2001. Herpes simplex viruses and their replication, p. 2399-2459. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Lippincott, Williams & Wilkins, Philadelphia, Pa.
Sambrook, J. F., and D. W. Russell. 2000. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Schepers, A., D. Pich, and W. Hammerschmidt. 1993. A transcription factor with homology to the AP-1 family links RNA transcription and DNA replication in the lytic cycle of Epstein-Barr virus. EMBO J. 12:3921-3929.
Schlake, T., and J. Bode. 1994. Use of mutated FLP recognition target (FRT) sites for the exchange of expression cassettes at defined chromosomal loci. Biochemistry 33:12746-12751.
Stenlund, A. 2003. Initiation of DNA replication: lessons from viral initiator proteins. Nat. Rev. Mol. Cell Biol. 4:777-785.
Turner, W. J., and M. E. Woodworth. 2001. DNA replication efficiency depends on transcription factor-binding sites. J. Virol. 75:5638-5645.
van Loon, N., C. Dykes, H. Deng, G. Dominguez, J. Nicholas, and S. Dewhurst. 1997. Identification and analysis of a lytic-phase origin of DNA replication in human herpesvirus 7. J. Virol. 71:3279-3284.
Vink, C., E. Beuken, and C. A. Bruggeman. 1997. Cloning and functional characterization of the origin of lytic-phase DNA replication of rat cytomegalovirus. J. Gen. Virol. 78:2963-2973.
Voss, J. H., and B. Roizman. 1988. Properties of two 5'-coterminal RNAs transcribed part way and across the S component origin of DNA synthesis of the herpes simplex virus 1 genome. Proc. Natl. Acad. Sci. USA 85:8454-8458.
Wagner, M., Z. Ruzsics, and U. H. Koszinowski. 2002. Herpesvirus genetics has come of age. Trends Microbiol. 10:318-324.
Weinberg, D. H., and G. Ketner. 1983. A cell line that supports the growth of a defective early region 4 deletion mutant of human adenovirus type 2. Proc. Natl. Acad. Sci. USA 80:5383-5386.
Yates, J., N. Warren, D. Reisman, and B. Sugden. 1984. A cis-acting element from the Epstein-Barr viral genome that permits stable replication of recombinant plasmids in latently infected cells. Proc. Natl. Acad. Sci. USA 81:3806-3810.
Zhang, Y., F. Buchholz, J. P. Muyrers, and A. F. Stewart. 1998. A new logic for DNA engineering using recombination in Escherichia coli. Nat. Genet. 20:123-128.
Zhu, Y., L. Huang, and D. G. Anders. 1998. Human cytomegalovirus oriLyt sequence requirements. J. Virol. 72:4989-4996(Eva-Maria Borst and Marti)
ABSTRACT
During the lytic phase of infection, replication of herpesvirus genomes initiates at the lytic origin of replication, oriLyt. Many herpesviruses harbor more than one lytic origin, but so far, only one oriLyt has been identified for human cytomegalovirus (HCMV). Evidence for the existence of additional lytic origins of HCMV has remained elusive. On the basis of transient replication assays with cloned viral fragments, HCMV oriLyt was described as a core region of 1.5 kbp (minimal oriLyt) flanked by auxiliary sequences required for maximal replication activity (complete oriLyt). It remained unclear whether minimal oriLyt alone can drive the replication of HCMV in the absence of its accessory regions. To investigate the sequence requirements of oriLyt in the context of the viral genome, mutant genomes were constructed lacking either minimal or complete oriLyt. These genomes were not infectious, suggesting that HCMV contains only one lytic origin of replication. Either minimal or complete oriLyt was then ectopically reinserted into the oriLyt-depleted genomes. Only the mutant genomes carrying complete oriLyt led to infectious progeny. Remarkably, inversion of the 1.5-kbp core origin relative to its flanking regions resulted in a replication-defective genome. Mutant genomes carrying minimal oriLyt plus the left flanking region gave rise to minifoci, but genomes harboring minimal oriLyt together with the right flanking region were noninfectious. We conclude that the previously defined minimal lytic origin is not sufficient to drive replication of the HCMV genome. Rather, our results underline the importance of the accessory regions and their correct arrangement for the function of HCMV oriLyt.
INTRODUCTION
Human cytomegalovirus (HCMV), like all herpesviruses, has at least two different modes of infection, either lytic replication or latency. Whereas the replication mode of the HCMV genome during latency is rather elusive, substantial progress has been made in dissecting the molecular steps that occur during lytic DNA replication (35, 39; for a review, see reference 41). Following cell entry, HCMV capsids are transported to the nuclear pores, where they release the linear genomes into the cell nucleus. Then, the linear genomes are rapidly circularized and serve as templates for transcription and replication. It has been proposed that the initial steps in DNA replication of herpesviruses take place by a theta-like mechanism, followed by a rolling-circle mechanism of replication at later stages of the lytic infection cycle (reviewed in references 36 and 46). This view of herpesvirus DNA replication has recently been challenged by the observation that genome circularization seems not to be a prerequisite for replication of herpes simplex virus (HSV) and the suggestion that linear genomes serve as replication templates (31). It remains to be shown whether the same applies to HCMV and other herpesviruses as well.
Regardless of which mechanism is used, lytic replication of herpesviruses initiates at defined regions of the viral genomes, the so-called lytic origins of replication (oriLyt). One oriLyt has been defined for HCMV (1, 2, 27, 37). By transient-replication assays with cloned subgenomic fragments, oriLyt was located upstream of the gene encoding the single-stranded DNA binding protein (UL57) approximately in the middle of the unique long (UL) segment (1, 37). In contrast to the origins of alphaherpesviruses, HCMV oriLyt is remarkable for its apparent size and complexity, and it seems to lack any obvious sequence and evolutionary homology to other viral DNA replication origins. Despite the missing sequence homology, however, there is a striking resemblance in the structural organization among the origins of human, chimpanzee, rhesus, rat, and mouse CMVs, e.g., an asymmetric base distribution with an A+T-rich region located at one end and an extended G+C-rich segment at the other end (20, 28, 38, 53). HCMV oriLyt contains numerous reiterated and inverted sequence elements, including known consensus transcription factor recognition sites. Although the analysis of these sequence elements in either oriLyt function or local transcription regulation has just begun (33), their distribution resembles the organization of some complex transcriptional enhancers, and the existence of these binding sites suggests the presence of putative promoter regions. Transcriptional control elements and enhancers are indeed common components of many viral and eukaryotic origins of DNA replication (21, 42, 48, 50, 51). Given the complex structure of HCMV oriLyt, the lack of an apparent trans-acting factor comparable to the HSV origin-binding protein (41), and the recent identification of intraorigin RNA transcripts that in part form stable and persistent RNA-DNA hybrid structures within oriLyt (45), the mechanism of replication initiation in HCMV may be different from that employed by other herpesviruses.
Several efforts have been undertaken to define the sequence requirements and boundaries of the HCMV replicator (1, 2, 37, 59). Using transient replication assays, Masse et al. mapped a 1.5-kbp core region of oriLyt (i.e., minimal oriLyt) to a region situated between nucleotides (nt) 92210 and 93715 of the HCMV genome but stated that full replicator activity requires additional sequences of 1,100 bp to the left and 700 bp to the right of the origin core (37). The complete lytic origin of HCMV was thus determined to be a segment with a size of 3.3 kbp. Insertion and deletion studies combined with transient replication assays performed by Anders and colleagues and Zhu et al. described a core region of HCMV oriLyt that is similar to that defined by Masse et al. but somewhat shifted to the left, namely, between nt 91751 and 93299 (1, 59). Since the introduction of mutations into the accessory sequences flanking either side of this core had little or no effect on replication efficiency (59), the role of these sequences flanking oriLyt requires further investigation.
Some herpesviruses, including HSV and Epstein-Barr virus, harbor more than one lytic origin of replication (26, 46), whereas to date only one lytic origin has been described for HCMV. However, the existence of other lytic origins in HCMV has not been excluded. Although the significance of multiple origins remains to be determined, this raises the question of whether the HCMV genome contains an additional lytic origin(s).
In order to define the sequence requirements for a functional oriLyt of HCMV, we decided to analyze the significance of the core origin and its flanking regions in the context of the HCMV genome. Also, we asked whether the HCMV genome contains additional lytic origins. Using a bacterial artificial chromosome (BAC)-cloned HCMV genome (6), we constructed a series of mutant HCMV genomes in which we first deleted either the minimal or the complete lytic origin and then, in subsequent rounds of mutagenesis, reinserted minimal or complete oriLyt at an ectopic position of the genomes with the origin deleted. Our results imply that the genome of HCMV strain AD169 contains only one lytic origin. We also provide evidence that the accessory regions flanking the origin core are essential for initiation of DNA replication at HCMV oriLyt. Interestingly, the left flanking region seems to be of greater importance for HCMV oriLyt function than the sequences to the right, since upon transfection, mutant genomes with a fragment encompassing minimal oriLyt plus sequences to the left led to the formation of small foci of green fluorescent protein (GFP)-expressing cells, whereas the sequences to the right plus minimal oriLyt did not.
MATERIALS AND METHODS
Virus and cells. The parental HCMV strain used in this report is the BAC-derived recombinant virus HCMV-GFP (8) originating from the BAC-cloned HCMV AD169 genome (6). Culture of human foreskin fibroblasts (HFF), preparation of virus stocks, growth curves, and determination of virus titers were essentially done as described previously (6). The adenovirus 5 recombinant dl1014—defective for most of the E4 region (11)—was propagated on the E4-complementing cell line W162 (56) (both kindly provided by Gary Ketner, Johns Hopkins Bloomberg School of Public Health, Baltimore, Md.). W162 cells were cultured in minimal essential medium (Invitrogen) supplemented with 10% fetal calf serum, 2 mM glutamine, 100 U of penicillin per ml, and 100 μg of streptomycin sulfate per ml. For preparation of replication-deficient, UV-inactivated adenovirus particles, dl1014 was isolated from infected W162 cells by three cycles of freeze-thawing, purified by two rounds of CsCl gradient centrifugation, and inactivated by treatment with psoralen and UV irradiation as described previously (17).
Plasmids. The nucleotide positions of the AD169 genome provided here refer to the GenBank entry with accession number X17403 (13). For construction of pOri6K-Kan1 and pOri6K-Kan4, which contain a kanamycin resistance marker plus one FLP recognition target (FRT) site in different orientations, and the bacterial origin of replication R6K (34), a 1.0-kbp fragment encompassing the kanamycin resistance gene from transposon Tn903 plus one wild-type FRT site was amplified by PCR using the primers Kanori6k.for (5'-CGC AGA TCT GGT GAC CAC GTC GTG GAA TGC CTT-3') and Kanori6k.rev (5'-CCC AGA TCT CGA TTT ATT CAA CAA AGC CAC GTT-3') with pSLFRTKn (3) as a template. The resulting PCR product was digested with BglII and ligated to a 0.5-kbp fragment comprising the R6K origin isolated from plasmid pGP704 (40) (kindly provided by Holger Rüssmann, University of Munich, Munich, Germany) by BamHI digestion. To generate plasmid pOri6K-F5, which harbors the kanamycin resistance cassette flanked by mutant FRT sites, PCR was performed with pSLFRTKn as a template and primers FRT-F5.for (5'-CGC AGA TCT GAA GTT CCT ATA CCT TTT GAA GAA TAG GAA CTT CGC CAG TGT TAC AAC CAA TTA ACC-3') and FRT-F5.rev (5'-CGA GAA TTC AGA AGT TCC TAT TCT TCA AAA GGT ATA GGA ACT TCC GAT TTA TTC AAC AAA GCC ACG T-3'), which provided mutant FRT sites (49). The PCR product was digested with EcoRI and BglII and cloned into the EcoRI and BglII sites of pGP704. For ectopic insertion of minimal oriLyt, plasmids p6K-ori1.5-3 and p6K-ori1.5-5 were constructed. A 1.5-kbp EcoRI-SacI fragment comprising minimal oriLyt (nt 92210 to 93715 of the AD169 genome) (13) was obtained from pBC-oriLyt (9), which carries the HCMV origin region from nt 92210 to 94860 derived from pCM1029 (25). The fragment was treated with the Klenow fragment of Escherichia coli DNA polymerase I and cloned into pOri6K-Kan1 and pOri6K-Kan4, which were cut with EcoRV. The sequence of the 1.5-kbp fragment was determined and compared to the AD169 DNA sequence. We did not find any alterations, except in a region located upstream of the vRNA-2 start site, which was previously reported to exhibit a heterogeneous restriction pattern (45). The sequences equivalent to AD169 nt 93092 to 93218 were missing, and the sequences equivalent to nt 93219 to 93299 were repeated two times. Since the minimal oriLyt is derived from plasmid pBC-oriLyt, which was shown to replicate efficiently in transient replication assays (9), the differences apparently reflect a naturally occurring polymorphism of HCMV oriLyt sequences.
Plasmids p6K-ori3.7-1 and p6K-ori3.7-4, which carry the complete lytic origin, were generated by releasing a 3.7-kbp fragment (nt 91165 to 94860) from HCMV cosmid clone pCM1029 (25) by digestion with DraI and KpnI, followed by insertion into pOri6K-Kan1 and pOri6K-Kan4, which were cut with KpnI and EcoRV. p6K-ori-right, containing the 1.5-kbp origin core plus the right flanking sequences, was constructed by inserting a 2.6-kbp EcoRI-KpnI-fragment (nt 92210 to 94860) derived from pBC-oriLyt into the EcoRV and KpnI restriction sites of pOri6K-Kan-4 after blunting the EcoRI ends using Klenow fragment, and p6K-ori-left originated from pOri6K-Kan-1, which was cut with EcoRV and SacI, followed by insertion of a 2.5-kbp DraI-SacI-fragment (nt 91165 to 93715) obtained from pCM1029. All plasmids containing the R6K origin were propagated in the E. coli strain PIR1 (Invitrogen), which provides the phage lambda protein.
For construction of the shuttle plasmid pST-ori-invers, the oligonucleotides ori-oligo.for (5'-CAG CTG CAC GAT ATC GTG CA-3') and ori-oligo.rev (5'-CGA TAT CGT GCA GCT GTG CA-3') were annealed and cloned into the PstI site of pACYC177 (New England Biolabs). The resulting plasmid, pACYC-O, was digested with EcoRV and PvuII, and a 6.1-kbp PvuII-EcoRV fragment (nt 89795 to 95967) isolated from pCM1029 was inserted, leading to pACYC-ori. This plasmid was cut with EcoRI and SacI, releasing DNA sequences (nt 92210 to 94636) that contain the minimal lytic origin, and a 920-bp EcoRI-SacI fragment (encompassing nt 93717 to 94636) amplified by PCR using the primers ori-sac.for (5'-CGC GAA TTC GTT AAC GGG AGC GGG GTC GAC CGC GA-3') and ori-sac.rev (5'-CCG GAG CTC AGA ACG GCA TGA-3') was reinserted, creating pACYC-ori-PCR. An HpaI cloning site introduced together with the PCR fragment and located directly adjacent to the EcoRI site was used to insert the minimal oriLyt sequences in inverse orientation. The EcoRI-SalI fragment (nt 92210 to 93727) carrying the minimal oriLyt was obtained from pBC-oriLyt (9) and treated with Klenow polymerase. Finally, the complete insert carrying the inverted minimal lytic origin plus the flanking sequences was isolated as an EcoRV-PvuII fragment and cloned into the SmaI site of shuttle plasmid pST76-KSR (10), resulting in pST-ori-invers.
BAC mutagenesis. Mutant HCMV BACs pHG-1, pHG-1.5, and pHG-3.7 were constructed in E. coli DH10B by the ET mutagenesis method (58), which is based on homologous recombination between a linear fragment and the HCMV BAC. The linear fragment encodes antibiotic resistance and is generated by PCR using primers with a size of 85 bp; 60 nt at the 5' end of the primers was homologous to the intended integration site in the BAC-cloned CMV genome, and 25 nt at the 3' end was specific for the antibiotic resistance gene. The recombination functions were provided by plasmid pKD46, which carries the phage lambda recombination genes red , ?, and under the control of an inducible promoter (19). For construction of pHG-1, a PCR fragment was used that contained a tetracycline resistance marker flanked by wild-type FRT sites and that was generated using pCP16 (16) as a template and primers UL1-10ko.for (5'-CAT GGG CAT GCA ATG CAA CAC TAA ATT GTT ATT GCC AGT CGC ACT AAT ACC GGT TGT AAT CAC CAG GGT TTT CCC AGT CAC GAC GT-3') and UL1-10ko.rev (5'-TCT CAG GTT CCG TAG GGA CTG TCC CTA TAT TCA AAC AGT GGG TTT GAA ACC GGA GAT ATC CAC AGG AAA CAG CTA TGA CCA TGA-3'). To generate mutant BACs pHG-1.5 and pHG-3.7, a PCR fragment encoding kanamycin resistance was used, which was obtained by PCR with plasmid pOri6K-F5 and primers ori15ko.for (5'-TGA CAA GCC TAC GTC ACA CTC GCG TGA CCA CAC CCA CTC CGG ATA TAC GTC ATC CTG TGG CCC CGA AAA GTG CCA CCT GCA GAT-3') and ori15ko.rev (5'-GCC GCC GAG AGG GAG CCG CCG CGA GAC CCG GAA GCC GTC GCG GTC GAC CCC GCT CCC GGA GTG ACA CAG GAA CAC TTA ACG GCT GA-3') and ori37ko.for (5'-CCC CCG ACA CGT ACT CGA AGA CCT TAC AGT TTA CGA GTC AAT AAA ACA GGA AAA GAT CCG ACC CCG AAA AGT GCC ACC TGC AGA T-3') and ori37ko.rev (5'-GCG CCC TCT AGG TAG CAT CAC GGA ACA GGC GGC GAC CGC GGA CTC GCC TCG GCC AGG GGG GTG ACA CAG GAA CAC TTA ACG GCT GA-3'), respectively. Excision of the FRT-flanked antibiotic resistance markers was achieved through FLP-mediated recombination in E. coli as described previously (7, 16). HCMV BAC pHG-10 was constructed by a two-step replacement procedure using the shuttle plasmid pST-ori-invers and HCMV BAC pHG-1.5 essentially as described previously (6). In brief, E. coli bacteria containing pHG-1.5 were transformed with the shuttle plasmid and incubated at 30°C on agar plates containing 17 μg of chloramphenicol/ml and 25 μg of kanamycin/ml. Clones harboring cointegrates were selected by streaking the bacteria on agar plates containing chloramphenicol and kanamycin and incubation at 43°C. The bacteria were then restreaked on agar plates containing chloramphenicol only and grown at 37°C overnight. Selection of clones that had resolved the cointegrates was done on agar plates containing chloramphenicol plus 5% sucrose at 30°C. Finally, BAC DNA was isolated from 10-ml cultures and checked for the presence of the inverted sequences by restriction analysis.
Mutant BACs pHG-2 and pHG-3 were generated by FLP recombinase-driven integration of plasmids p6K-ori1.5-3 and p6K-ori1.5-5. For this purpose, bacteria harboring HCMV BAC pHG-3.7 were transformed with plasmid pCP20 (16), which encodes FLP. Bacterial colonies were selected on agar plates containing 17 μg of chloramphenicol/ml and 50 μg of ampicillin/ml at 30°C. The bacteria containing both pCP20 and the HCMV BAC were then transformed with either p6K-ori1.5-3 or p6K-ori1.5-5, and clones harboring recombinant BACs were selected by incubation on agar plates containing 17 μg of chloramphenicol/ml and 30 μg of kanamycin/ml at 43°C. BACs pHG-4 and pHG-5 were generated in the same way, i.e., by integration of p6K-ori1.5-3 and p6K-ori1.5-5 into BAC pHG-1.5, and BAC genomes pHG-6, pHG-7, pHG-8, and pHG-9 were obtained by FLP-mediated insertion of p6K-ori3.7-1, p6K-ori3.7-4, p6K-ori-left, and p6K-ori-right into pHG-3.7. The correct integration of the R6K-based plasmids was verified by PCR analysis and sequencing.
Analysis of virus and BAC DNAs. HCMV BAC DNA was isolated from 10-ml overnight E. coli cultures by the alkaline lysis procedure (47) and characterized by restriction analysis followed by agarose gel electrophoresis as described previously (6). Large preparations of HCMV BACs were obtained from 100-ml bacterial cultures using Nucleobond PC 100 columns (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions. Viral DNA was isolated from virions purified from the supernatant of infected cells as described previously (7), and Southern blot analysis of BAC and virus DNAs was performed as reported previously (6). The DNA fragments used for hybridization were either obtained by PCR (probe 1, equivalent to nt 90549 to 90935 of the HCMV genome) using primers orihyb.for (5'-CGG CTT TGC GTC ACC GCC AAC TAA-3') and orihyb.rev (5'-GCG ATT GAG GTT GCG AGA CCA GAT-3') with pCM1029 (25) as a template or excised as a 2.2-kbp MluI fragment (probe 2, equivalent to nt 92018 to 94238 of the HCMV genome) from p6K-ori3.7-1.
Accumulation of viral DNA was examined by slot blot assay. HFF were infected at a multiplicity of infection (MOI) of 0.02 PFU/cell with the respective viruses, and total DNA was isolated on days 1, 3, 5, 7, and 9 postinfection using the QiaAmp DNA Blood Mini kit (QIAGEN, Hilden, Germany). One microgram of each DNA sample was transferred to a nylon membrane using a slot blot apparatus. The membrane was then hybridized to a 32P-labeled DNA probe specific for the UL52 open reading frame (ORF). Finally, specific radioactive signals were quantified using a phosphorimager.
Transfection of HCMV BACs. HCMV BAC DNA was transfected into HFF using the adenovirus-mediated gene delivery protocol developed by Cotten et al. (5, 18). In brief, 10 μl of BAC DNA was diluted in 20 mM HEPES, pH 7.4-150 mM NaCl and condensed with polyethylenimine MW 2000 (Sigma) by incubation for 20 min at room temperature. Particles of inactivated adenovirus dl1014 (2 x 109) were added to the mixture and incubated at room temperature for another 20 min. The transfection complexes were applied to the cells, which were kept in serum-free culture medium containing 30 μg of polymyxin B (Sigma)/ml. After 4 h of incubation in a 37°C incubator with 5% CO2, the fibroblasts were washed twice with PBS and further cultivated in complete cell culture medium. All transfection experiments were done in triplicate. CMV BAC genomes that appeared to be replication deficient were tested in three independent experiments, and transfected cells were incubated until complete cytopathic effect (CPE) had occurred on the fibroblast monolayers transfected with the control BAC pHG-1.
RESULTS
HCMV genomes with a deletion of either the minimal or the complete lytic origin of replication are not infectious. For some herpesviruses, it has been shown that more than one lytic origin is present in the virus genome. In contrast, only one oriLyt has been identified in HCMV (1, 27, 37). Since HCMV oriLyt was primarily defined on the basis of transient replication assays with cloned viral fragments and construction of recombinant genomes with mutated oriLyt has not been possible, the existence of additional lytic origins in the HCMV genome could not be excluded.
In order to address the question of whether the HCMV genome contains other lytic origins apart from the one described, a mutant genome lacking minimal oriLyt was constructed. The design of the mutant was based on the boundaries of oriLyt as defined by Masse et al. (37). They mapped the core of oriLyt between nucleotide positions 92210 (EcoRI) and 93715 (SacI) of the HCMV genome (Fig. 1). Deleting minimal oriLyt was facilitated by applying the ET mutagenesis technique (58), which was recently adapted for the manipulation of BAC-cloned CMV genomes (7, 55) (for technical details, see Materials and Methods). The HCMV BAC pHG-1, a derivative of the previously described HCMV-GFP BAC (8), served as the backbone for construction of all mutant genomes reported in this study (Fig. 2 provides schematic drawings of all mutant genomes). Deletion of the nonessential ORFs UL1 to UL10 in pHG-1 allowed the subsequent insertion of DNA fragments with sizes of up to 6.3 kbp without generating an overlong HCMV genome (see below). The virus derived from pHG-1 expressed the enhanced green fluorescent protein, thereby facilitating the tracking of infected or transfected cells (Fig. 3C), and it showed growth kinetics comparable to those of its parental virus, HCMV-GFP, as well as those of the original BAC-derived virus, RVHB5 (6) (data not shown).
For deletion of minimal oriLyt, a kanamycin resistance (KanR) gene flanked by mutant FRT sites was inserted into the HCMV BAC pHG-1, thereby replacing minimal oriLyt (Fig. 3B). The resulting BAC pHG-1.5-K was characterized by restriction analysis using BglII (Fig. 3A). Following mutagenesis, a fragment of 8.4 kbp characteristic of the parental BAC pHG-1 was replaced by two new fragments of 4.9 and 3.0 kbp in pHG-1.5-K (Fig. 3A, lane 2). The KanR cassette was subsequently excised with FLP recombinase (data not shown), resulting in pHG-1.5 (Fig. 2, line 2).
Transfection of large BACs into primary fibroblasts is associated with notoriously poor efficiency. In general, this does not prevent the reconstitution of replication-competent viruses from BAC genomes, but it may impair the interpretation of results when replication-incompetent or severely debilitated genomes are analyzed. Therefore, we searched for a more efficient transfection method and tested the adenovirus-mediated gene delivery procedure developed by Baker and Cotten (5, 18). Using this protocol for transfection of HCMV BACs, we obtained an efficiency that was at least 2 orders of magnitude higher than that of any other transfection protocol that we had tested before.
When HCMV BAC pHG-1.5 was transfected into human fibroblasts, single green cells were observed 2 days posttransfection, which remained single for >10 days. No plaques appeared on the cell monolayer (Fig. 3C), and no infectious virus was released into the supernatant. Thus, deletion of minimal oriLyt resulted in a replication-deficient HCMV genome, suggesting that there is no other lytic origin present in the HCMV genome that can substitute for the deleted oriLyt.
The results of a previous study done with viral DNA fragments led to the conclusion that accessory flanking regions of 1,100 bp to the left and 700 bp to the right of minimal oriLyt are needed to achieve efficient replication originating at HCMV oriLyt (37). In order to evaluate the requirement for these auxiliary regions later, we also constructed a recombinant genome in which the complete oriLyt was deleted. A region of 3.7 kbp extending from nt 91165 (DraI) to 94860 (KpnI) of the HCMV genome was replaced by the KanR marker in mutant BAC pHG-3.7-K (Fig. 3B, line 4). In comparison to the parental BAC pHG-1 (Fig. 3B and A, lane 1), new fragments of 3.9 and 1.9 kbp were present in pHG-3.7-K (Fig. 3A, lane 3). Excision of the kanamycin cassette led to BAC pHG-3.7 (Fig. 2, line 3). Transfection of pHG-3.7 into permissive fibroblasts did not result in the production of infectious progeny (Fig. 3C), providing further evidence that the HCMV genome contains only one lytic origin of replication.
Ectopic insertion of minimal oriLyt does not rescue the replication-deficient phenotype of CMV genomes with the origin deleted. In order to determine the importance of the accessory regions flanking minimal oriLyt, we decided to reinsert either minimal or complete oriLyt at an ectopic position in mutant BACs pHG-1.5 and pHG-3.7, which lack the respective sequences at the original positions (Fig. 2). For insertion of these sequences, we utilized the highly convenient site-specific FLP/FRT recombination system. An FRT site that was inserted at an ectopic position during construction of the parental BAC pHG-1 allowed the insertion of shuttle vectors based on the conditional bacterial R6K replicon (34) and carrying the oriLyt sequences (Fig. 4B) (see Materials and Methods for technical details). Specific targeting of the ectopic locus by the shuttle plasmids was accomplished by using selection markers with wild-type and mutant FRT sites for the deletion of ORFs UL1 to UL10 and of the origin sequences, respectively. Since mutant and wild-type FRT sites cannot interact with each other (49), the shuttle plasmids, which carry a wild-type FRT site, become integrated exclusively at the ectopic position.
In order to evaluate whether the minimal oriLyt by itself is sufficient to mediate replication of the HCMV genome, the sequences were inserted at the ectopic position of BAC pHG-3.7, leading to pHG-2 and pHG-3, which differ in the orientation of the inserted minimal oriLyt (Fig. 2). Characterization of the BAC DNAs with HpaI revealed the expected shift of an 8.1-kbp band to 11.2 kbp in pHG-2 and pHG-3 (Fig. 4A, lanes 4 and 5). After transfection of the mutant BACs into permissive human fibroblasts, single green cells were readily observed, but no plaques appeared on the cells (Fig. 4C), implying that no infectious virus can be recovered from pHG-2 and pHG-3, regardless of the orientation of the ectopically inserted minimal lytic origin. Thus, ectopic insertion of minimal oriLyt is not sufficient to support replication of an HCMV genome in which complete oriLyt has been deleted.
Since deletion of the complete oriLyt may affect the promoter of the essential UL57 gene, we could not exclude the possibility that the observed phenotype of pHG-2 and pHG3 was the result of reduced UL57 expression. Therefore, minimal oriLyt was also reinserted into pHG-1.5, which lacks only the minimal lytic origin at the original position. However, the resulting BACs, pHG-4 and pHG-5 (Fig. 2 and 4), also did not lead to the development of plaques following transfection into human fibroblasts even after 4 weeks of incubation. These results suggest that the minimal lytic origin lacking the accessory flanking regions is not able to drive replication of the HCMV genome.
Reconstitution of infectious genomes after integration of complete oriLyt at an ectopic position in CMV BACs with the origin deleted. Since the minimal lytic origin could not rescue the replication deficiency of pHG-3.7 and pHG-1.5, we tested whether ectopic insertion of the complete lytic origin leads to replication-proficient HCMV genomes. The resulting genomes, pHG-6 and pHG-7, carry the complete oriLyt at the ectopic position and lack these sequences at the original position (Fig. 2 and 4). Transfection of these genomes into permissive cells resulted in the formation of green fluorescent plaques (Fig. 4C) and in a complete cytopathic effect of the cultures 2 weeks posttransfection. In order to analyze whether the genomes of the mutant viruses released from the transfected cells had preserved their original structures and to rule out the possibility that the lytic origin in the original region had been reconstituted, viral DNA was isolated from virions and examined by restriction analysis and Southern blotting. Figure 5A shows the BglII fragments of BACs pHG-6 and pHG-7 (lanes 2 and 4) and of the corresponding virus DNAs (lanes 3 and 5), as well as of the parental BAC pHG-1 (lane 1). In comparison to the circular DNA of the BACs, additional fragments (e.g., of 17.1 kbp) that originate from the terminal fragments of the linear viral DNA are visible in lanes 3 and 5. Probe p1, which is specific for the genomic region located to the left of the complete lytic origin (Fig. 5D), detected the expected 8.4-kbp band of the parental BAC pHG-1 (Fig. 5B, lane 1) and a 3.9-kbp band of pHG-6 and pHG-7, as well as of the corresponding viruses RV-pHG-6 and RV-pHG-7, which is indicative of the deletion of complete oriLyt at the original position (Fig. 5B, lanes 2 to 5, and D). The blot was stripped and reprobed using fragment p2, specific for the lytic origin (Fig. 5D). As expected, hybridization with probe p2 revealed the same 8.4-kbp fragment of the parental BAC pHG-1 as did p1. In DNAs of the mutant BACs pHG-6 and pHG-7, a band of 42 kbp was recognized, which corresponds to the expected fragment harboring the insertion of complete oriLyt (Fig. 5C, lanes 2 and 4, and D, middle line). The DNA bands of the viruses RV-pHG-6 and RV-pHG-7 detected by probe p2 seemed slightly smaller (Fig. 5C, lanes 3 and 5), consistent with the expected size of 29 kbp for the terminal fragments of two of the HCMV genome isomers (Fig. 5D, line 3). Two additional bands of 36 and 42 kbp expected from the other genome isomers generated after inversion of the UL region could not be resolved on the gel shown in Fig. 5 but were readily detected after pulsed-field gel electrophoresis and hybridization with probe 2 (data not shown). Most importantly, hybridization of the probe to fragments with a size of 8.4 kbp or to any additional fragments was not found, ruling out the translocation of the ectopic oriLyt back to its original position. Thus, we concluded that the mutant genomes were stable and that ectopic insertion of complete oriLyt rescued the replication deficiency of genomes with oriLyt deleted. The rescue was independent of the orientation of the reinserted origin sequences. These data also indicated that deletion of complete oriLyt at the original position did not impair the expression of adjacent genes (e.g., UL57), at least not in a way that prevented growth of the viral mutants.
By integration of the sequences of the complete lytic origin at the ectopic position, we could also rescue the replication defect of pHG-1.5, which lacks minimal oriLyt at the original position (data not shown). The resulting genome readily produced plaques after transfection into permissive cells, finally leading to complete CPE. This result confirmed that the growth defect of pHG-1.5 is caused by the deletion of minimal oriLyt and is not due to any other potential mutation elsewhere in the genome.
Growth analysis of HCMV mutants with an ectopic oriLyt. Following transfection of pHG-6 and pHG-7, we got the impression that the resulting plaques were slightly smaller than the plaques originating from the parental virus (Fig. 4C). Therefore, we compared the growth behavior of the mutant viruses RV-pHG-6 and RV-pHG-7 to that of the parental virus after infection of human foreskin fibroblasts with different viral doses. Following infection at a low MOI of 0.025 PFU/cell, the final titers obtained with mutant viruses RV-pHG-6 and RV-pHG-7 were reduced by 1.5 orders of magnitude in comparison to the titers of RV-pHG-1 (Fig. 6A). However, the growth kinetics did not seem to be delayed. Similar results were obtained when cells were infected at a higher MOI of 1.5 PFU/cell (Fig. 6B). Thus, the HCMV mutants harboring the complete lytic origin at the ectopic position exhibited a growth defect in comparison to the parental virus.
To gain further insight into the growth behavior of RV-pHG-6 and RV-pHG-7, we analyzed the amplification of viral DNA in cells infected with either the parental virus, RV-pHG-1, or the mutant viruses RV-pHG-6 and RV-pHG-7. HFF were infected at an MOI of 0.02 PFU/cell, and total DNA was prepared from the infected cells at various times postinfection. The DNA samples were then analyzed by slot blotting, followed by hybridization with an HCMV-specific probe, and the signals obtained were quantified using a phosphorimager. As can be seen in Fig. 6C, comparable amounts of viral DNA were observed in fibroblasts infected with RV-pHG-1, RV-pHG-6, or RV-pHG-7. The mutants carrying the complete lytic origin at the ectopic position did not exhibit a defect in either the kinetics of DNA synthesis or final accumulation of viral DNA. We concluded from this result that the 3.7-kbp lytic origin inserted at the ectopic position comprises all of the information that is necessary to mediate efficient replication of the HCMV genome.
Role of the regions flanking the core replicator for oriLyt activity. Since our data pointed to an important, or even essential, role of the sequences adjacent to minimal oriLyt, we sought to dissect the influence of the right and the left flanking regions on oriLyt activity. Therefore, minimal oriLyt, together with either its right or left flanking region, was inserted at the ectopic position of pHG-3.7, resulting in BACs pHG-8 and pHG-9 (Fig. 2 and 7). To evaluate the infectivity of these mutant genomes, the BACs were transfected into human fibroblasts, resulting in numerous green cells. Whereas these green cells remained single in pHG-9-transfected cultures, small foci of GFP-expressing cells were observed 3 weeks after transfection of pHG-8 (Fig. 7C). However, we did not observe the formation of plaques and the minifoci did not develop further into complete CPE as was observed with the parental BAC pHG-1. Therefore, we assume that pHG-8 is replication proficient, but the growth rate of the resulting virus is probably too low to promote the spread of the infection throughout the culture.
Having shown that the sequences flanking the oriLyt core are crucial for origin function, we assessed whether the orientation of minimal oriLyt relative to its flanking regions is relevant. To this end, we inverted the 1.5-kbp origin core sequence at its original position. Starting with pHG-1.5 missing the minimal oriLyt, the core element was reintroduced in reverse orientation by a two-step mutagenesis procedure (see Materials and Methods for technical details). The DNAs of several BAC clones were analyzed by restriction analysis using BglII following mutagenesis (Fig. 8A). Two clones (clones 2 and 4) acquired the reversed minimal oriLyt, as can be seen by the presence of an additional 8.4-kbp fragment and the lack of the 4.9- and 2.0-kbp bands (Fig. 8A, lanes 2 and 4). In order to test the consequences of the inversion, pHG-10 clones 2 and 4 were transfected into permissive fibroblasts, leading to numerous green fluorescent cells. However, only single green cells could be observed during a 4-week incubation period, but no plaques appeared on the cell monolayers (Fig. 8B). This experiment implied that the appropriate orientation of the minimal lytic origin with respect to its flanking regions is crucial for origin function.
DISCUSSSION
In this study, we investigated the sequence requirements of the described lytic origin of human cytomegalovirus (oriLyt) in the context of the viral genome. To our knowledge, this is the first report of the intentional generation of recombinant HCMV genomes and viruses carrying mutations of oriLyt. Our results provide evidence that the HCMV genome of strain AD169 contains only one lytic origin, because HCMV genomes with deletion of either minimal or complete oriLyt were not infectious. Furthermore, we show that the 1.5-kbp region designated the HCMV core is not sufficient to drive replication of HCMV, since ectopic insertion of this sequence into a nonessential region of the viral genome did not rescue the replication-deficient phenotype of the genomes with the origin deleted. Rather, accessory sequences flanking minimal oriLyt are needed for replication of the HCMV genome, as insertion of the complete lytic origin (encompassing these regions) readily resulted in infectious viral progeny. The observation that genomes with a reinsertion of the origin core together with its left flanking region gave rise to the formation of minifoci whereas genomes harboring the origin core plus the sequences to the right did not suggested the presence of an essential element within the left flanking region. Notably, correct orientation of the origin core relative to its auxiliary regions was crucial for replicator function.
One of the biggest advantages of the BAC-based mutagenesis of herpesviruses is the fact that construction and testing of mutant genomes are separate steps. This allowed us to manipulate HCMV oriLyt, a cis-active element of the genome considered to be essential for viral replication. The BAC technique enabled us to perform several rounds of mutagenesis on the very same genome, independent of the phenotype of the intermediates, as well as to construct replication-deficient CMV genomes that harbor lethal mutations in essential elements of oriLyt.
Deletion of minimal or complete oriLyt led to replication deficiency of the resulting genomes, indicating that the deleted sequences are essential and that other elements of the genome cannot substitute for them. These findings suggest that in contrast to human alpha- and gamma-herpesviruses (4, 26, 46, 57), the HCMV genome contains only one lytic origin. This might be a unique feature of beta-herpesviruses, since so far only one single origin of replication has also been described for simian, rat, and mouse CMV, as well as for human herpesvirus types 6 and 7 (2, 23, 38, 52, 53). Of course, this hypothesis requires experimental examination. Our results apply only to the laboratory strain AD169, and due to the fact that clinical isolates of HCMV contain additional DNA sequences (12), we cannot rule out the possibility that such CMV strains harbor more than one lytic origin. Whether the presence of several lytic origins provides a replication advantage remains elusive, because deletion of oriL or both copies of oriS of HSV had only marginal effects on viral growth (30, 44).
The final evidence that the replication deficiency of the genomes with the origin deleted is due to the deletion of oriLyt is technically difficult to supply. The phenotype could also result from the impairment of another essential element, e.g., the ORF of an essential gene located in the same region, and the genetic elements may be physically inseparable. For the following reasons we consider this rather unlikely. (i) Recent reevaluation of the HCMV coding potential based on comparison with the genetic information for the closely related chimpanzee CMV strongly suggested that most, if not all, of the previously annotated ORFs within the region encompassing oriLyt do not encode functional proteins (20, 43). (ii) Meanwhile, some of the ORFs have indeed been shown to be nonessential (24). (iii) Sequence variability in the oriLyt region was observed, a fact that is hardly compatible with the potential to encode proteins (14, 15, 32, 45). (iv) In agreement with the phenotypes of the genomes with oriLyt deleted described in this study, at least three regions within minimal oriLyt were found to be indispensable for origin function on the basis of transient replication assays (1, 37, 45, 59).
Assuming that oriLyt does essentially act in cis by providing an initiation site for the replication machinery, we tested whether the defect of the genomes with the origin deleted could be rescued by ectopic reinsertion of these sequence elements. This was indeed the case for complete oriLyt, whereas minimal oriLyt was not sufficient. In view of the low replication activity of minimal oriLyt that was determined in transient transfection experiments using the subcloned element (1, 37, 59), the observed insufficiency of minimal oriLyt to support replication of the complete genome was perhaps not surprising. The results indicate, however, that the oriLyt of HCMV represents more than just an initiation site for replication. This may be in contrast to the origins of alpha-herpesviruses that span only 150 bp and basically provide the start site for DNA replication and binding sites for the origin-binding protein (46). Possibly, the lack of an origin-binding protein of HCMV was the driving force for the development of such a large and complex origin and of cis-active sequence elements that could either substitute for the function of an origin-binding protein or support initiation of DNA replication by an alternative mechanism.
Rescue of the replication-deficient phenotype by reinsertion of the complete oriLyt demonstrated that these sequences encompass all elements necessary to initiate DNA replication at the ectopic position. Moreover, these results show that the activity of this important cis-active element is neither position nor orientation dependent. The mutants were not deficient either in the kinetics of viral DNA synthesis or in the DNA yields. These findings also strongly suggest that the mutants were not impaired in the expression of transcripts that are essential for the initiation and progression of viral DNA replication, e.g., the smallest replicator transcript (SRT), v-RNA-2, or the UL57 RNA encoding the single-stranded DNA binding protein. The 5' ends of the UL57 transcripts were precisely mapped in a recent study (33). Although the UL57 promoter may extend considerably upstream, it is extremely unlikely that the growth defect of the genomes with a deletion of minimal oriLyt was caused by an impairment of UL57 expression, especially since pHG-1.5 could be rescued by ectopic insertion of the complete oriLyt. Since the left boundary (nt 91165) of the complete oriLyt as used in this study and the most distal transcription start site of the UL57 RNA (nt 91138), as well as the candidate promoter sequences (33), are located in close proximity, one cannot completely rule out an effect on UL57 expression in the mutants carrying a deletion of complete oriLyt. However, given the pivotal role of the UL57 protein (single-stranded DNA binding protein) for replication of HCMV DNA, impaired UL57 expression would hardly be compatible with the unaltered DNA synthesis displayed by the mutants RV-pHG-6 and RV-pHG-7.
In view of the results of the DNA synthesis, the reduced viral yields of these mutants are difficult to explain. One hypothesis is that the promoter-enhancer elements in the origin sequences may have an effect on the expression of genes with important functions in the infection cycle that are not involved in DNA replication. Further work will be needed to define the reasons for the impaired growth properties of these mutants.
Having shown that the accessory regions of minimal oriLyt are essential for replication of the genome, we investigated whether one of the flanking regions may be dispensable. Transfection of genomes containing minimal oriLyt plus the left flanking region resulted in small groups of GFP-expressing cells. We called these clusters of green cells minifoci, because they did not develop into typical plaques and complete CPE did not occur. Although the experimental evidence is difficult to obtain, we assume that the pHG-8 genome is able to replicate in the initially transfected cells, albeit at very low efficiency. The resulting progeny are probably able to spread to the neighboring cells, but then the infection terminates because either the growth rate of the virus is too low or an antiviral state is induced in the surrounding cells, e.g., by cytokines, which cannot be overcome.
Genomes with minimal oriLyt plus the right flanking region did not lead to plaques or to minifoci, suggesting that an essential element is located to the left of minimal oriLyt. In fact, the precise boundaries of HCMV oriLyt have been a matter of debate. Masse et al. specified that sequences flanking the minimal oriLyt contribute to efficient replication but are not essential for origin function (37). Zhu et al. mapped the core of oriLyt to a region encompassing nt 91751 to 93299 (59), which includes part of the left flanking sequence of minimal oriLyt as defined by Masse et al. According to Zhu et al., the core replicator is necessary, but not sufficient, to drive DNA replication in transient assays (59). The fact that ectopic insertion of the left accessory sequence plus minimal oriLyt led to replication-competent genomes underlines its importance for origin function. Since the resulting virus displayed an extremely reduced growth capacity, one has to conclude that the right flanking sequences also play an important, yet not essential, role for replicator activity.
The accessory regions of HCMV oriLyt contain numerous binding sites for a series of transcription factors. Thus, one may speculate that these regions mediate their effect on the core replicator in a manner similar to that of enhancer sequences on promoter regions. If this is true, one would expect that they can act in an orientation- and distance-independent manner. Since inversion of minimal oriLyt resulted in a replication-incompetent genome, the hypothesis has to be discarded. Instead, the transcription factor binding sites rather may represent promoter elements, and indeed, a number of transcripts arise from the HCMV origin region, including the SRT, vRNA-1, and vRNA-2 (Fig. 1) (29, 45). Origin-associated transcripts were also identified in other viral, as well as cellular, origins (22, 54). Possibly, they have no messenger function but rather participate either in the replication process itself, e.g., by serving as a primer for the DNA polymerase, or in a more indirect manner, e.g., by inducing changes in the superhelical structure of the DNA template. Interestingly, all known transcripts from the HCMV origin region extend from right to left in the prototype arrangement of the viral genome. If we assume that these transcripts have an important role in the initiation of viral DNA replication, one can understand that inversion of minimal oriLyt and the concomitant change in the direction of transcription of the vRNA-2 and SRT genes will disrupt origin function.
We report in this study the first HCMV recombinants with mutations in oriLyt. By using the highly convenient FRT/FLP recombination system, we established a fast and easy way to reinsert oriLyt sequences into a nonessential ectopic region of the HCMV genome in order to dissect the roles of cis-acting elements of oriLyt. The BAC-based mutagenesis techniques also allowed us to manipulate the origin sequences at their original position. In further work, we can evaluate the functions of specific origin elements, which were defined in transient replication assays, in the context of the viral genome to deepen our understanding of the complex mechanism of HCMV DNA replication.
ACKNOWLEDGMENTS
This work was supported by the NBL3 program of the Bundesministerium für Bildung und Forschung (BMBF) and the Wilhelm Roux research program of the Medical Faculty, Martin-Luther-University of Halle-Wittenberg.
We are especially indebted to M. Cotten for providing the adenovirus-mediated gene delivery protocol and tools and to G. Ketner for the adenovirus recombinant dl1014 and W162 cells. We thank B. L. Wanner and M. Berlyn (E. coli Stock Center, Yale University) for sending plasmid pKD46 and H. Rüssmann for pGP704. We are also grateful to Z. Ruzsics for help with the adenovirus stock preparation.
REFERENCES
Anders, D. G., M. A. Kacica, G. Pari, and S. M. Punturieri. 1992. Boundaries and structure of human cytomegalovirus oriLyt, a complex origin for lytic-phase DNA replication. J. Virol. 66:3373-3384.
Anders, D. G., and S. M. Punturieri. 1991. Multicomponent origin of cytomegalovirus lytic-phase DNA replication. J. Virol. 65:931-937.
Atalay, R., A. Zimmermann, M. Wagner, E. Borst, C. Benz, M. Messerle, and H. Hengel. 2002. Identification and expression of human cytomegalovirus transcription units coding for two distinct Fc receptor homologs. J. Virol. 76:8596-8608.
AuCoin, D. P., K. S. Colletti, Y. Xu, S. A. Cei, and G. S. Pari. 2002. Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) contains two functional lytic origins of DNA replication. J. Virol. 76:7890-7896.
Baker, A., and M. Cotten. 1997. Delivery of bacterial artificial chromosomes into mammalian cells with psoralen-inactivated adenovirus carrier. Nucleic Acids Res. 25:1950-1956.
Borst, E. M., G. Hahn, U. H. Koszinowski, and M. Messerle. 1999. Cloning of the human cytomegalovirus (HCMV) genome as an infectious bacterial artificial chromosome in Escherichia coli: a new approach for construction of HCMV mutants. J. Virol. 73:8320-8329.
Borst, E. M., S. Mathys, M. Wagner, W. Muranyi, and M. Messerle. 2001. Genetic evidence of an essential role for cytomegalovirus small capsid protein in viral growth. J. Virol. 75:1450-1458.
Borst, E. M., and M. Messerle. 2000. Development of a cytomegalovirus vector for somatic gene therapy. Bone Marrow Transplant. 25(Suppl. 2):S80-S82.
Borst, E. M., and M. Messerle. 2003. Construction of a cytomegalovirus-based amplicon: a vector with a unique transfer capacity. Hum. Gene Ther. 14:959-970.
Borst, E. M., G. Posfai, F. Pogoda, and M. Messerle. 2004. Mutagenesis of herpesvirus BACs by allele replacement, p. 269-280. In S. Zhao and M. Stodolsky (ed.), Bacterial artificial chromosomes, vol. 2. Functional studies. Humana Press Inc., Totowa, N.J.
Bridge, E., and G. Ketner. 1989. Redundant control of adenovirus late gene expression by early region 4. J. Virol. 63:631-638.
Cha, T. A., E. Tom, G. W. Kemble, G. M. Duke, E. S. Mocarski, and R. R. Spaete. 1996. Human cytomegalovirus clinical isolates carry at least 19 genes not found in laboratory strains. J. Virol. 70:78-83.
Chee, M. S., A. T. Bankier, S. Beck, R. Bohni, C. M. Brown, R. Cerny, T. Horsnell, C. A. Hutchison, T. Kouzarides, J. A. Martignetti, E. Preddie, S. C. Satchwell, P. Tomlinson, K. M. Weston, and B. G. Barrell. 1990. Analysis of the protein-coding content of the sequence of human cytomegalovirus strain AD169. Curr. Top. Microbiol. Immunol. 154:125-169.
Chen, Z., S. Sugano, and S. Watanabe. 1999. A 189-bp repeat region within the human cytomegalovirus replication origin contains a sequence dispensable but irreplaceable with other sequences. Virology 258:240-248.
Chen, Z., S. Watanabe, and N. Yamaguchi. 1996. Strain-dependent differences in the human cytomegalovirus replication origin. Arch. Virol. 141:13-30.
Cherepanov, P. P., and W. Wackernagel. 1995. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158:9-14.
Cotten, M., M. Saltik, M. Kursa, E. Wagner, G. Maass, and M. L. Birnstiel. 1994. Psoralen treatment of adenovirus particles eliminates virus replication and transcription while maintaining the endosomolytic activity of the virus capsid. Virology 205:254-261.
Cotten, M., E. Wagner, K. Zatloukal, S. Phillips, D. T. Curiel, and M. L. Birnstiel. 1992. High-efficiency receptor-mediated delivery of small and large (48 kilobase) gene constructs using the endosome-disruption activity of defective or chemically inactivated adenovirus particles. Proc. Natl. Acad. Sci. USA 89:6094-6098.
Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.
Davison, A. J., A. Dolan, P. Akter, C. Addison, D. J. Dargan, D. J. Alcendor, D. J. McGeoch, and G. S. Hayward. 2003. The human cytomegalovirus genome revisited: comparison with the chimpanzee cytomegalovirus genome. J. Gen. Virol. 84:17-28.
DePamphilis, M. L. 1993. Eukaryotic DNA replication: anatomy of an origin. Annu. Rev. Biochem. 62:29-63.
DePamphilis, M. L. 1993. How transcription factors regulate origins of DNA replication in eukaryotic cells. Trends Cell Biol. 3:161-167.
Dewhurst, S., D. M. Krenitsky, and C. Dykes. 1994. Human herpesvirus 6B origin: sequence diversity, requirement for two binding sites for origin-binding protein, and enhanced replication from origin multimers. J. Virol. 68:6799-6803.
Dunn, W., C. Chou, H. Li, R. Hai, D. Patterson, V. Stolc, H. Zhu, and F. Liu. 2003. Functional profiling of a human cytomegalovirus genome. Proc. Natl. Acad. Sci. USA 100:14223-14228.
Fleckenstein, B., I. Muller, and J. Collins. 1982. Cloning of the complete human cytomegalovirus genome in cosmids. Gene 18:39-46.
Hammerschmidt, W., and B. Sugden. 1988. Identification and characterization of oriLyt, a lytic origin of DNA replication of Epstein-Barr virus. Cell 55:427-433.
Hamzeh, F. M., P. S. Lietman, W. Gibson, and G. S. Hayward. 1990. Identification of the lytic origin of DNA replication in human cytomegalovirus by a novel approach utilizing ganciclovir-induced chain termination. J. Virol. 64:6184-6195.
Hansen, S. G., L. I. Strelow, D. C. Franchi, D. G. Anders, and S. W. Wong. 2003. Complete sequence and genomic analysis of rhesus cytomegalovirus. J. Virol. 77:6620-6636.
Huang, L., Y. Zhu, and D. G. Anders. 1996. The variable 3' ends of a human cytomegalovirus oriLyt transcript (SRT) overlap an essential, conserved replicator element. J. Virol. 70:5272-5281.
Igarashi, K., R. Fawl, R. J. Roller, and B. Roizman. 1993. Construction and properties of a recombinant herpes simplex virus 1 lacking both S-component origins of DNA synthesis. J. Virol. 67:2123-2132.
Jackson, S. A., and N. A. DeLuca. 2003. Relationship of herpes simplex virus genome configuration to productive and persistent infections. Proc. Natl. Acad. Sci. USA 100:7871-7876.
Kemble, G., G. Duke, R. Winter, and R. Spaete. 1996. Defined large-scale alterations of the human cytomegalovirus genome constructed by cotransfection of overlapping cosmids. J. Virol. 70:2044-2048.
Kiehl, A., L. Huang, D. Franchi, and D. G. Anders. 2003. Multiple 5' ends of human cytomegalovirus UL57 transcripts identify a complex, cycloheximide-resistant promoter region that activates oriLyt. Virology 314:410-422.
Kolter, R., M. Inuzuka, and D. P. Helinski. 1978. Trans-complementation-dependent replication of a low molecular weight origin fragment from plasmid R6K. Cell 15:1199-1208.
LaFemina, R. L., and G. S. Hayward. 1983. Replicative forms of human cytomegalovirus DNA with joined termini are found in permissively infected human cells but not in non-permissive Balb/c-3T3 mouse cells. J. Gen. Virol. 64:373-389.
Lehman, I. R., and P. E. Boehmer. 1999. Replication of herpes simplex virus DNA. J. Biol. Chem. 274:28059-28062.
Masse, M. J., S. Karlin, G. A. Schachtel, and E. S. Mocarski. 1992. Human cytomegalovirus origin of DNA replication (oriLyt) resides within a highly complex repetitive region. Proc. Natl. Acad. Sci. USA 89:5246-5250.
Masse, M. J., M. Messerle, and E. S. Mocarski. 1997. The location and sequence composition of the murine cytomegalovirus replicator (oriLyt). Virology 230:350-360.
McVoy, M. A., and S. P. Adler. 1994. Human cytomegalovirus DNA replicates after early circularization by concatemer formation, and inversion occurs within the concatemer. J. Virol. 68:1040-1051.
Miller, V. L., and J. J. Mekalanos. 1988. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J. Bacteriol. 170:2575-2583.
Mocarski, E. S., and C. T. Courcelle. 2001. Cytomegaloviruses and their replication, p. 2629-2674. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Murakami, Y., and Y. Ito. 1999. Transcription factors in DNA replication. Front. Biosci. 4:D824-D833.
Murphy, E., I. Rigoutsos, T. Shibuya, and T. E. Shenk. 2003. Reevaluation of human cytomegalovirus coding potential. Proc. Natl. Acad. Sci. USA 100:13585-13590.
Polvino-Bodnar, M., P. K. Orberg, and P. A. Schaffer. 1987. Herpes simplex virus type 1 oriL is not required for virus replication or for the establishment and reactivation of latent infection in mice. J. Virol. 61:3528-3535.
Prichard, M. N., S. Jairath, M. E. Penfold, S. St Jeor, M. C. Bohlman, and G. S. Pari. 1998. Identification of persistent RNA-DNA hybrid structures within the origin of replication of human cytomegalovirus. J. Virol. 72:6997-7004.
Roizman, B., and D. M. Knipe. 2001. Herpes simplex viruses and their replication, p. 2399-2459. In D. M. Knipe and P. M. Howley (ed.), Fields virology. Lippincott, Williams & Wilkins, Philadelphia, Pa.
Sambrook, J. F., and D. W. Russell. 2000. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Schepers, A., D. Pich, and W. Hammerschmidt. 1993. A transcription factor with homology to the AP-1 family links RNA transcription and DNA replication in the lytic cycle of Epstein-Barr virus. EMBO J. 12:3921-3929.
Schlake, T., and J. Bode. 1994. Use of mutated FLP recognition target (FRT) sites for the exchange of expression cassettes at defined chromosomal loci. Biochemistry 33:12746-12751.
Stenlund, A. 2003. Initiation of DNA replication: lessons from viral initiator proteins. Nat. Rev. Mol. Cell Biol. 4:777-785.
Turner, W. J., and M. E. Woodworth. 2001. DNA replication efficiency depends on transcription factor-binding sites. J. Virol. 75:5638-5645.
van Loon, N., C. Dykes, H. Deng, G. Dominguez, J. Nicholas, and S. Dewhurst. 1997. Identification and analysis of a lytic-phase origin of DNA replication in human herpesvirus 7. J. Virol. 71:3279-3284.
Vink, C., E. Beuken, and C. A. Bruggeman. 1997. Cloning and functional characterization of the origin of lytic-phase DNA replication of rat cytomegalovirus. J. Gen. Virol. 78:2963-2973.
Voss, J. H., and B. Roizman. 1988. Properties of two 5'-coterminal RNAs transcribed part way and across the S component origin of DNA synthesis of the herpes simplex virus 1 genome. Proc. Natl. Acad. Sci. USA 85:8454-8458.
Wagner, M., Z. Ruzsics, and U. H. Koszinowski. 2002. Herpesvirus genetics has come of age. Trends Microbiol. 10:318-324.
Weinberg, D. H., and G. Ketner. 1983. A cell line that supports the growth of a defective early region 4 deletion mutant of human adenovirus type 2. Proc. Natl. Acad. Sci. USA 80:5383-5386.
Yates, J., N. Warren, D. Reisman, and B. Sugden. 1984. A cis-acting element from the Epstein-Barr viral genome that permits stable replication of recombinant plasmids in latently infected cells. Proc. Natl. Acad. Sci. USA 81:3806-3810.
Zhang, Y., F. Buchholz, J. P. Muyrers, and A. F. Stewart. 1998. A new logic for DNA engineering using recombination in Escherichia coli. Nat. Genet. 20:123-128.
Zhu, Y., L. Huang, and D. G. Anders. 1998. Human cytomegalovirus oriLyt sequence requirements. J. Virol. 72:4989-4996(Eva-Maria Borst and Marti)