Identification of viral genes essential for replication of murine gamma-herpesvirus 68 using signature-tagged mutagenesis

Abstract

Gamma-herpesviruses, Epstein-Barr virus, and Kaposi's sarcoma-associated herpesvirus are important human pathogens, because they are involved in tumor development. Murine gamma-herpesvirus-68 (MHV-68 or gammaHV-68) has emerged as a small animal model system for the study of gamma-herpesvirus pathogenesis and host-virus interactions. To identify the genes required for viral replication in vitro and in vivo, we generated 1,152 mutants using signature-tagged transposon mutagenesis on an infectious bacterial artificial chromosome of MHV-68. Almost every ORF was mutated by random insertion. For each ORF, a mutant with an insertion proximal to the N terminus of each ORF was examined for the ability to grow in fibroblasts. Our results indicate that 41 genes are essential for in vitro growth, whereas 26 are nonessential and 6 attenuated. Replication-competent mutants were pooled to infect mice, which led to the discovery of ORF 54 being important for MHV-68 to replicate in the lung. This genetic analysis of a tumor-associated herpesvirus at the whole genome level validates signature-tagged transposon mutagenesis screening as an effective genetic system to identify important virulent genes in vivo and define interactions with the host immune system.

Fig. 1.

STM of MHV-68 for in vivo and in vitro screening. (A) A schematic diagram of STM transposons. Tag sequences are composed of 12 different combinations of three forward and four reverse sequences and a common sequences in the middle, which is designed for real-time PCR. (B) A schematic representation of the MHV-68 STM strategy. By using MuA transposase and an STM transposon with pBAC/MHV-68 as a template, in vitro transposition yields thousands of transposon-inserted mutants. BAC DNAs of STM mutants that have transposon insertion at each ORF of MHV-68 are extracted and transfected into fibroblasts in 96-well plates. After two rounds of infection with transfected supernatants, ORFs are categorized to be essential or nonessential, based on the growth phenotype of a mutant with an insertion proximal to the N terminus of each ORF. Replication-competent mutants are pooled and simultaneously used to infect mice. The inoculum is also propagated in NIH3T12 cells. DNAs from tissues of infected mice (output) and from infected NIH3T12 cells (input) were subjected to real-time PCR, by using STM tag-specific primer sets to determine the fate of each virus within the pool. The copy number of each tag in the input and the output is quantitatively compared to identify the attenuated mutants.

Fig. 2.

Functional mapping of MHV-68 for in vitro growth. Putative ORFs of MHV-68 originally predicted by Virgin and colleagues (10) are color-coded, according to the growth properties of STM mutants with transposon insertion at the proximity to N terminus of each ORF in fibroblast. The ORFs with rightward direction are named in the upper lines, whereas those with leftward orientation in the lower lines. The red circles indicate essential genes conserved in α-, β-, and γ-herpesvirus subfamilies. The vertical lines represent transposon insertion sites, based on sequencing results of 988 mutants.

Fig. 3.

Identification of STM mutants likely to be critical for acute infection in vivo. (A) Quantitative analysis of STM mutant growth in the lung. Pooled STM mutants with distinct tags (45 pfu per virus) were used to intranasally infect BALB/c mice (six mice per pool) and also propagated in NIH3T12 cells. Relative copy numbers of STM mutants with specific tag primer sets using DNAs (50 ng) from infected cells (in vitro) and lung tissues of infected mice (in vivo), as determined by real-time PCR, were compared to identify STM mutants defective in establishing acute infection in mice. Mutants included in different pools are indicated with numbers. SEMs are shown as error bars. *, Mutants showing significant attenuation in the lung. (B) The PCR products of STM mutants were run on agarose gels. A set of representative data is shown. Compared with the in vitro data, mutants ORF 21^null, ORF 73^null, and ORF 54^null failed to show detectable growth in vivo, whereas others replicated normally. The specific PCR products ranged from 80 to 100 bps. + control, STM plasmid mix; - control, lung DNA from an uninfected BALB/c mouse.

Fig. 4.

Attenuation of ORF 54^null growth in the lung. Mice were infected with ORF 54^null (⋄), STM BAC (WT, ○), or ORF 54^null + STM BAC (▵) at 50 pfu per virus and killed at 9 days postinfection. Infectious viruses in the lung were assayed by plaque assays (Left, filled symbols). Results from real-time PCR using viral genome-specific (ORF 56, Center, open symbols) or STM tag-specific (Right, gray symbols) primers with DNAs from lung tissues of infected mice are shown. ORF 54^null is tagged with B1A4, and STM BAC with B3A2. Genome copies of individual viruses in mixed infection were normalized based on comparison of results from viral genome-specific primers with those from STM tags in individual infection. The average of each group is shown as a bar within the group.