Molecular cloning of the guinea pig cytomegalovirus (GPCMV) genome as an infectious bacterial artificial chromosome (BAC) in Escherichia coli

Since cytomegalovirus (CMV) infection is highly species-specific, it is necessary to study animal cytomegaloviruses to assess viral factors which contribute to pathogenesis. The generation of recombinant viruses carrying reporter genes would provide useful tools for studying the genetics of CMV pathogenicity in vivo. We evaluated whether the guinea pig cytomegalovirus (GPCMV) was amenable to such manipulation. Metabolic selection using the guanosylphosphoribosityl transferase (gpt) gene facilitated recovery of a recombinant virus, vAM403, containing a gpt/green fluorescent protein (eGFP) cassette introduced into the HindIII "N" region of the viral genome. This virus had replication kinetics identical to wild-type virus. We next attempted to clone the GPCMV genome as a bacterial artificial chromosome (BAC). A BAC plasmid containing a gpt/eGFP cassette and the chloramphenicol resistance marker was introduced into HindIII "N" to generate another GPCMV recombinant, vAMBGPCMV. Circular viral DNA isolated from vAMBGPCMV-infected cells was used to transform Escherichia coli. Restriction profiles revealed that the GPCMV genome had been cloned as a BAC plasmid, and transfection of BAC plasmid DNA confirmed that the BAC clone was infectious. A novel strategy based on a unique PmeI site was devised to quickly modify the BAC GPCMV plasmid. Recombinants retained the capability to replicate and express reporter genes in guinea pigs, suggesting that these viruses will be useful for in vivo pathogenesis studies.     

Characterization of the guinea pig cytomegalovirus genome locus that encodes homologs of human cytomegalovirus major immediate-early genes, UL128, and UL130

We reported previously that the guinea pig cytomegalovirus (CMV) stock purchased from the American Type Culture Collection contained two types of strains, one containing and the other lacking a 1.6 kb locus, and that the 1.6 kb locus was required for efficient viral growth in animals but not in cell culture. In this study, we characterized the genetic contents of the locus, and found that i) the 1.6 kb locus encodes homologs of human CMV UL128 and UL130, GP129 and GP131, respectively, ii) these genes are expressed with late gene kinetics, iii) GP131 protein (pGP131) localized to cell surface only in the presence of glycoproteins H and L, and iv) pGP131 is a virion component. Therefore, it is plausible that pGP131 forms a complex with glycoproteins H and L and becomes a virion component as does UL130 protein (pUL130). Since pUL130 is one of the glycoproteins essential for infection of endothelial and epithelial cells in human and primates, functional and immunological analyses of this GPCMV homolog of pUL130 may help to illuminate the in vivo role of pUL130.     

Quantitative-competitive PCR monitoring of viral load following experimental guinea pig cytomegalovirus infection

Human cytomegalovirus (HCMV) is the most common cause of congenital viral infection in the developed world, and can lead to significant morbidity. Animal models of HCMV infection are required for study of pathogenesis, because of the strict species-specificity of cytomegalovirus (CMV). Among the small animal CMV models, the guinea pig CMV (GPCMV) has unique advantages, in particular its propensity to cross the placenta, causing disease in utero. In order to develop quantitative endpoints for vaccine and antiviral therapeutic studies in the GPCMV model, a quantitative-competitive PCR (qcPCR) assay was developed, based on the GPCMV homolog of the HCMV UL83 gene, GP83. Optimal amplification of GPCMV DNA was observed using primers spanning a 248 base pair (bp) region of this gene. A 91 bp deletion of this cloned fragment was generated for use as an internal standard (IS) for PCR amplification. Standard curves based upon the fluorescent intensity of full-length external target to IS were compared with signal intensity of DNA extracted from blood and organs of experimentally infected guinea pigs in order to quantify viral load. Viral load in newborn guinea pigs infected transplacentally was determined and compared with that of pups infected with GPCMV as neonates. Viral loads were highest in pups infected as neonates. The most consistent isolation and highest quantities of viral DNA were observed in liver and spleen, although viral genome could be readily identified in brain, lung, and salivary gland. Viral load determination should be useful for monitoring outcomes following vaccine studies, as well as responses to experimental antiviral agents.     

Cloning the complete guinea pig cytomegalovirus genome as an infectious bacterial artificial chromosome with excisable origin of replication

Congenital human cytomegalovirus infections are the major infectious cause of birth defects in the United States. How this virus crosses the placenta and causes fetal disease is poorly understood. Guinea pig cytomegalovirus (GPCMV) is a related virus that provides an important model for studying cytomegaloviral congenital transmission and pathogenesis. In order to facilitate genetic analysis of GPCMV, the 232kb GPCMV genome was cloned as an infectious bacterial artificial chromosome (BAC). The BAC vector sequences were flanked by LoxP sites to allow efficient excision using Cre recombinase. All initial clones contained spontaneous deletions of viral sequences and reconstituted mutant viruses with impaired growth kinetics in vitro. The deletions in one BAC were repaired using Escherichia coli genetics. The resulting repaired BAC reconstituted a virus with in vitro replication kinetics identical to the wild type parental virus; moreover, its genome was indistinguishable from that of the wild type parental virus by restriction pattern analysis using multiple restriction enzymes. These results suggest that the repaired BAC is an authentic representation of the complete GPCMV genome. It should provide a valuable tool for evaluating the impact of genetic modifications on the safety and efficacy of live attenuated vaccines and for identifying genes important for congenital transmission and fetal disease.     

Rapid identification of non-essential genes for in vitro replication of Marek's disease virus by random transposon mutagenesis

Marek's disease virus (MDV) is a highly oncogenic alphaherpesvirus that induces rapid-onset T-cell lymphomas in poultry. The MDV genome encodes more than 100 genes. However, the role of many of these genes in virus replication is not known. The construction of an infectious bacterial artificial chromosome (BAC) clone of the highly oncogenic RB-1B strain of MDV has been described previously. Virus reconstituted from the BAC clone induced rapid-onset lymphomas in chickens very similar to the wildtype viruses. In this paper, the construction of a high-density random transposon-insertion mutant library of the RB-1B BAC clone using a high throughput in vitro transposon mutagenesis technique is described. Furthermore a PCR screening method, using primers specific for the transposon sequence and the MDV gene(s) of interest, was developed for the rapid identification of specific insertion mutants. The application of the screening method to identify some of the non-essential genes for MDV replication in vitro is described.     

Immunogenicity evaluation of DNA vaccines that target guinea pig cytomegalovirus proteins glycoprotein B and UL83

Vaccines are needed for control of congenital human cytomegalovirus (HCMV) infection. Although the species-specificity of cytomegaloviruses precludes preclinical evaluation of HCMV vaccines in animal models, the guinea pig cytomegalovirus (GPCMV), which causes disease in utero, is a relevant model for the study of vaccines against congenital infection. We investigated whether DNA vaccines that target two GPCMV proteins, glycoprotein B (gB) and UL83 (pp65), are capable of eliciting immune responses in vivo. After cloning each gene into an expression vector, DNA was delivered by intramuscular inoculation and by pneumatic epidermal delivery. In Swiss-Webster mice, anti-gB titers were significantly higher after epidermal delivery. After epidermal inoculation in guinea pigs, all gB-immunized animals (n = 6) had antibody responses comparable to those induced by natural infection. Viral neutralization titers ranged from 1:64 to greater than 1:128. A GPCMV UL83 DNA vaccine also elicited an antibody response in all immunized guinea pigs (n = 6) after epidermal administration. Immunoprecipitation and Western blot assays confirmed that immune sera were immunoreactive with virion-associated UL83 and gB proteins. We conclude that DNA vaccines against GPCMV structural proteins are immunogenic, and warrant further investigation in the guinea pig model of congenital CMV infection.     

Comparison of vaccine strategies against congenital CMV infection in the guinea pig model.

Vaccines are urgently needed to protect newborns against the devastating sequelae of congenital cytomegalovirus infection. Evaluation of candidate vaccines in the guinea pig model of congenital infection can shed light on potentially useful strategies for humans, since guinea pig CMV (GPCMV) is transmitted to the fetus transplacentally, causing infection and disease in utero. A number of vaccine strategies have been evaluated in this model, including DNA vaccines, live attenuated vaccines, and recombinant glycoprotein vaccines. Induction of virus-neutralizing antibody appears to play a key role in protection of the fetus. Recently, a vectored vaccine based on the GPCMV homolog of the UL83 (pp65) protein has also been shown to be effective when used as a preconceptual vaccine in this model, suggesting that strategies designed to elicit T-cell responses may be of value in protection of the fetus.     

Detection of guinea pig cytomegalovirus nucleic acids in cultured cells with biotin-labelled hybridization probes

Biotin labelled hybridization probes prepared from recombinant plasmids containing segments of the guinea pig cytomegalovirus (GPCMV) genome were used to detect GPCMV nucleic acids in guinea pig cells by in situ hybridization. The time course of GPCMV infection was assessed in two cultured cell types, guinea pig embryo (GPE) cells and 104C1 cells, a transformed and cloned guinea pig cell line. Detection of GPCMV nucleic acids was accomplished in both cell types with individual GPCMV DNA fragments and with mixtures of GPCMV DNA fragments. When compared to other established methods of GPCMV detection, the method of in situ hybridization enabled the detection of a higher percentage of positive cells early during the course of the infection. In addition, differences in the replication cycle of GPCMV in the two cultured cell lines could be demonstrated. These findings will facilitate future studies of GPCMV tissue tropism in vivo.     

Comparison of guinea pig cytomegalovirus and guinea pig herpes-like virus: growth characteristics and antigentic relationship.

The growth characteristics of guinea pig cytomegalovirus (GPCMV) and guinea pig herpes-like virus (GPHLV) in cell cultures were compared. Guinea pig fibroblast cells were highly susceptible to infection with both viruses, whereas guinea pig kidney cells were sensitive only to GPHLV. No cytopathic effect was observed in the latter cell system after infection with GPCMV,nor was there an increase in virus titer, although the cirus persisted in the kidney cells for 2 to 3 weeks postinfection. Electron microscope studies showed nonvirion tubular structures in GPCMV -infected fibroblast cells, but not in GPHLV- infected cells. Large packages of enveloped nuclear virus particles were commonly seen in GPHLV -infected cells, especially kidney epithelial cells, but none were found in the GPCMV -infected fibroblasts. Complete enveloped extracellular virus particles were present in both virus-cell systems. Both viruses showed narrow host spectra and replicated well only in guinea pig cells although GPHLV multiplied to some degree in rabbit cells. No antigenic relationship could be demonstrated between the two viruses using antisera specific for each virus that was produced in rabbits and guinea pigs. Rabbits produced high neutralizing antibody titers to GPHLV, whereas guinea pigs were the animals of choice for GPCMV antiserum production.     

UL84-independent replication of human cytomegalovirus strain TB40/E

The UL84 gene of human cytomegalovirus is implicated in the initiation of viral DNA replication during lytic infection. UL84 is essential for replication of a cloned viral origin of lytic replication (oriLyt) in vitro and mutants of strains AD169 or Towne with deletions or insertions in UL84 fail to grow in cells permissive for wild type virus. Here we show that UL84 is dispensable for replication of a strain TB40/E clone derived from a bacterial artificial chromosome. The genomes of the fibroblast-adapted strains AD169 and Towne are altered substantially from the consensus for strains that have not been propagated extensively in cell culture. The parental TB40/E genome conforms to the consensus genomic organization. Accordingly, natural HCMV strains may possess replication capability that extends beyond the known oriLyt-dependent replication system of laboratory strains.     

Genetic analysis of the herpes simplex virus type 1 UL9 gene: Isolation of a lacZ insertion mutant and expression in eukaryotic cells

HSV-1 host range mutants in complementation group 1-36 (hr27 and hr156) whose mutations map in the UL9 gene, encoding the origin binding protein, are unable to form plaques or synthesize viral DNA or late viral proteins when grown in nonpermissive Vero cells (Carmichael, E. P., Kosovsky, M. J., Weller, S. K., 1988, J. Virol. 62, 91-99). These defects are complemented efficiently by growth in the permissive cell line, S22, which contains the wild type version of several HSV genes including UL9. In this report the precise nature and location of the lesions in host range mutants hr27 and hr156 were determined by DNA sequencing; both mutants were found to contain identical single-base-pair substitutions at codons 309 and 311 in the UL9 open reading frame. This region lies within the putative helicase domain of the UL9 protein. The UL9 gene was disrupted by the insertion of an insertional mutagen ICP6::lacZ in which the Escherichia coli lacZ gene is expressed under control of the viral ICP6 promoter. Hr94, a viral mutant containing this insertion, does not form plaques or synthesize viral DNA when grown in Vero cells, although both defects are complemented efficiently on permissive cell lines. These results confirm that the UL9 gene product is essential for viral growth and DNA replication. Furthermore, since no detectable UL9 protein is synthesized in hr94-infected cells, this virus provides a useful genetic background for further structure-function analysis since no potentially interfering nonfunctional UL9 protein will be expressed. We have expressed the UL9 open reading frame under the control of the strong and inducible HSV-1 ICP6 promoter and have derived Vero cell lines containing variable copy numbers of the ICP6::UL9 construct. Cells whose copy number of this construct exceeded approximately 120 are unable to support efficient plaque formation by wild-type virus. Cell lines with low copy numbers of this construct are able to complement hr27, hr156, and hr94.     

Generation of an infectious clone of duck enteritis virus (DEV) and of a vectored DEV expressing hemagglutinin of H5N1 avian influenza virus

We report on the generation of an infectious bacterial artificial chromosome (BAC) clone of duck enteritis virus (DEV) and a vectored DEV vaccine expressing hemagglutinin (H5) of high pathogenicity H5N1 avian influenza virus (AIV). For generation of the DEV BAC, we inserted mini-F vector sequences by homologous recombination in lieu of the UL44 (gC) gene of DEV isolate 2085. DNA of the resulting in recombinant virus v2085-GFPΔgC was electroporated into Escherichia coli and a full-length DEV BAC clone (p2085) was recovered. Transfection of p2085 into chicken embryo cells resulted in DEV-specific plaques exhibiting green autofluorescence. A gC-negative mutant, v2085ΔgC, was generated by deleting mini-F vector sequences by using Cre-Lox recombination, and a revertant virus v2085ΔgC-R was constructed by co-transfection of p2085 with UL44 sequences. Finally, AIV H5 was inserted into p2085, and high-level H5 expression of the v2085_H5 virus was detected by indirect immunofluorescence and western blotting. Plaque area measurements showed that v2085ΔgC plaques were significantly increased (12%) over those of parental 2085 virus or the v2085ΔgC-R revertant virus (ANOVA, P<0.05), while plaque areas of the H5- or GFP-expressing DEV mutants were significantly smaller. There was no significant difference between DEV with respect to virus titers determined after trypsinization titration of infected cells, while virus titers of infected-cell supernatants revealed significant reductions in case of the gC-negative viruses of more than 700-fold when compared to parental 2085 or v2085ΔgC-R. Cell-associated virus titers of gC-negative DEV also showed significant reduction of 50-500-fold (ANOVA, P<0.05). We conclude that (i) absence of DEV gC results in increased plaque sizes in vitro, (ii) gC plays a role in DEV egress, and (iii) generation of an infectious DEV clone allows rapid generation of vectored vaccines.