Cloning of the full-length rhesus cytomegalovirus genome as an infectious and self-excisable bacterial artificial chromosome for analysis of viral pathogenesis

Abstract

Rigorous investigation of many functions encoded by cytomegaloviruses (CMVs) requires analysis in the context of virus-host interactions. To facilitate the construction of rhesus CMV (RhCMV) mutants for in vivo studies, a bacterial artificial chromosome (BAC) containing an enhanced green fluorescent protein (EGFP) cassette was engineered into the intergenic region between unique short 1 (US1) and US2 of the full-length viral genome by Cre/lox-mediated recombination. Infectious virions were recovered from rhesus fibroblasts transfected with pRhCMV/BAC-EGFP. However, peak virus yields of cells infected with reconstituted progeny were 10-fold lower than wild-type RhCMV, suggesting that inclusion of the 9-kb BAC sequence impeded viral replication. Accordingly, pRhCMV/BAC-EGFP was further modified to enable efficient excision of the BAC vector from the viral genome after transfection into mammalian cells. Allelic exchange was performed in bacteria to substitute the cre recombinase gene for egfp. Transfection of rhesus fibroblasts with pRhCMV/BAC-Cre resulted in a pure progeny population lacking the vector backbone without the need of further manipulation. The genomic structure of the BAC-reconstituted virus, RhCMV-loxP(r), was identical to that of wild-type RhCMV except for the residual loxP site. The presence of the loxP sequence did not alter the expression profiles of neighboring open reading frames. In addition, RhCMV-loxP(r) replicated with wild-type kinetics both in tissue culture and seronegative immunocompetent macaques. Restriction analysis of the viral genome present within individual BAC clones and virions revealed that (i) RhCMV exhibits a simple genome structure and that (ii) there is a variable number of a 750-bp iterative sequence present at the S terminus.

FIG. 1.

Strategy for constructing a self-excisable RhCMV BAC. (A) Cloning of the full-length RhCMV genome into a BAC vector by Cre/lox recombination. The RhCMV genome structure with expansion of the US1-to-US3 region of recombinant viruses is diagrammed. The open box represents the internal junction between L and S components of the RhCMV genome. Cre/lox recombination was performed in Telo-RF cells. Recombinant clones containing the BAC vector in the viral genome (vRhCMV/BAC-EGFP) were plaque purified, and circular-form viral DNA was transformed into E. coli strain DH10B for plasmid isolation. (B) Substitution of the cre ORF for the egfp ORF by allelic exchange in E. coli. Only a portion of each plasmid is illustrated. The homologous flanking regions and the cre ORF of the recombination vector, pWC212, are illustrated. UL, unique sequences of the L component; US, unique sequences of the S component; PSV40, SV40 promoter; poly A, SV40 polyadenylation signal.

FIG. 2.

Characterization of RhCMV BAC plasmids. (A) Schematic diagram of pRhCMV/BAC structure with expansion of the US1-to-US3 region. The black box in the plasmid map represents the fusion region of L and S termini within the circular-form viral genome. The sizes of expected NotI-, SalI-, and EcoRI-fragments resulting from the insertion of BAC-EGFP and BAC-Cre vectors (shown in parentheses) are indicated in kilobases. (B) Restriction digestion analysis of wild-type nucleocapsid viral DNA (lanes wt) and the cloned pRhCMV/BAC-EGFP (lanes 1) pRhCMV/BAC-Cre (lanes 2) plasmid DNAs. The additional fragments resulting from the insertion of the BAC vector are indicated with lowercase letters, and the eliminated wild-type fragments resulting from insertion of the BAC vector are marked with arrows. The L- and S-terminal fragments in the linear-form virion DNA are marked with asterisks and crosses, respectively. The unique fragments resulting from the fusion of the termini in the BAC plasmids are marked with the joined symbols. Size standards indicated in kilobases are displayed to the left of the gel.

FIG. 3.

Restriction and DNA blot analyses of RhCMV genomic DNA and BAC plasmids. (A) EcoRI digestion of the three unique conformations of pRhCMV/BAC-EGFP clones. (B) Schematic illustration of the L-S terminal fusion in the RhCMV BAC. The putative cleavage site is located 30 bp upstream of the pac1 sequence at the L terminus. The probes overlapping both L and S termini (probe LS) or specific to each of them (probe S and probe L) are depicted as black bars. (C) Southern blot analysis for the EcoRI-digested BAC plasmids with radiolabeled probes (only probe L shown). (D) Southern blot analysis for the EcoRI- and BamHI-digested terminal fragments of wild-type RhCMV nucleocapsid viral DNA with the three different probes shown in panel B. The L-terminal fragment is marked with asterisks, and S-terminal fragments are marked with capital letters (750-bp ladders from A to E). The fused termini in the BAC plasmids are labeled with joined marks. (E) EcoRI digestion of the progeny BAC derived from the C3 pRhCMV/BAC-EGFP reconstituted virions. Only 5 of 12 independent isolates are shown. (F) Schematic representation of the structures of the predicted RhCMV genomic termini. Dark-gray and light-gray blocks represent the L- and S-terminal sequence, respectively. The pac1 sequence at the L terminus is illustrated as a black strip. The restriction sizes of both L- and S-terminal fragments are listed.

FIG. 4.

Analyses of the residual loxP site in the genome of the BAC-reconstituted virus. (A) Gel electrophoresis of EcoRI-digested viral nucleocapsid DNA. Lane 1, wild-type RhCMV; lane 2, RhCMV-loxP; lane 3, RhCMV-loxP(r). Size standards are displayed to the left of the gels and indicated in kilobases. (B) DNA sequence of the US1-US2 intergenic region of RhCMV-loxP and RhCMV-loxP(r). The inverted repeats of the 34-bp loxP site are indicated with arrows. The C terminus of US2 and the N terminus US1 are shown with translation. The polyadenylation consensus for US2 is underlined. The display of US1-to-US3 region in this illustration is opposite to its orientation in the viral genome. (C) 3′ RACE analyses of RhCMV IE2, US1, US2, and US3 expression profiles in wild-type RhCMV- and RhCMV-loxP(r)-infected cells. Telo-RF were infected with each virus at an MOI of 1. Cytoplasmic RNA was isolated at different time points (12, 24, and 48 hpi) from Telo-RF cultures infected with each virus at an MOI of 1. 3′ RACE for GAPDH was performed as an internal control. Lane M, molecular size marker; lane U, uninfected control.

FIG. 5.

Replication kinetics of RhCMV-loxP(r) in vitro and in vivo. (A) Single-step growth curve analyses of BAC-reconstituted RhCMV. RhCMV/BAC-EGFP and RhCMV-loxP(r) were recovered from Telo-RF transfected with pRhCMV/BAC-EGFP and pRhCMV/BAC-Cre, respectively. The replication kinetics of these viruses were compared to those of wild-type RhCMV and RhCMV-loxP. Telo-RF cells were infected in triplicate with each virus at an MOI of 0.1. Datum points represent the mean of infectious virus titers in the supernatants of three independent cultures with the standard deviations indicated by the error bars. (B) Longitudinal viral loads in the plasma of two seronegative rhesus monkeys intravenously inoculated with RhCMV-loxP(r). Plasma samples were collected and processed for DNA isolation at the time points as indicated. RhCMV DNA copy numbers were quantified by a real-time PCR assay with the detection limit of 200 copies. Datum points represent the averages of two independent real-time PCR analyses.