Knockout of the Host Resistance Gene Pkr Fully Restores Replication of Murine Cytomegalovirus m142 and m143 Mutants In Vivo

Abstract

Murine cytomegalovirus (MCMV) proteins m142 and m143 are essential for viral replication. They bind double-stranded RNA and prevent protein kinase R-induced protein synthesis shutoff. Whether the two viral proteins have additional functions such as their homologs in human cytomegalovirus do remained unknown. We show that MCMV m142 and m143 knockout mutants attain organ titers equivalent to those attained by wild-type MCMV in Pkr knockout mice, suggesting that these viral proteins do not encode additional PKR-independent functions relevant for pathogenesis in vivo.

TEXT

Double-stranded RNA (dsRNA) produced during viral infections (1) is recognized as a pathogen-associated molecular pattern by cellular sensors, which activate host antiviral response pathways (2). Many viruses antagonize antiviral host responses by expressing proteins that interact with dsRNA and/or its cellular sensors (3, 4). Murine cytomegalovirus (MCMV) proteins m142 and m143 form a dsRNA-binding complex and interact with the dsRNA-activated protein kinase (PKR), thereby inhibiting PKR-induced protein synthesis shutoff (5–7).

Besides PKR, dsRNA can activate other sensors such as the 2′-5′-oligoadenylate synthetase, the RIG-I-like receptors, and Toll-like receptor 3 (2). However, it has remained unclear whether m142 and m143 are involved in preventing their activation or signaling. Moreover, IRS1 and TRS1, the homologs of m142 and m143 in human cytomegalovirus, not only inhibit PKR (8–11) but also mediate several additional functions: they facilitate cap-dependent mRNA translation, interact with the viral DNA polymerase processivity factor, are required for efficient assembly of DNA-containing capsids, and inhibit autophagy (12–15). At least some of these functions are thought to be PKR independent. Therefore, it seemed likely that m142 and m143 have additional PKR-independent functions important for replication and pathogenesis in vivo.

We wanted to test this hypothesis by infecting wild-type (wt) and Pkr^−/− mice with MCMV m142 and m143 knockout mutants and comparing their replication to that of wt MCMV. In order to minimize the risk of unwanted effects on neighboring genes, we introduced stop mutations by single nucleotide exchanges into the m142 and m143 open reading frames of pARK25, a bacterial artificial chromosome (BAC) clone of the MCMV K181 strain (16). The mutant MCMV genomes were constructed by en passant mutagenesis (17). A G→T mutation at nucleotide position 200634 (GenBank accession number AM886412) was introduced to generate a stop codon in m142, and a G→T mutation at position 202624 generated a stop after the fifth codon of m143 (Fig. 1A). m142Rev, a revertant of the m142stop mutant, was constructed using the same method. Viruses were reconstituted by transfecting BACs into immortalized Pkr^−/− mouse embryonic fibroblasts (MEF) as described previously (7). To verify the absence of m142 or m143 protein expression, Pkr^−/− MEF were infected with the different viruses at a multiplicity of infection (MOI) of 1 PFU/cell. Cell lysates were harvested 72 h pi (hpi) and analyzed by Western blotting (Fig. 1B) using m142- and m143-specific antibodies (18). (All animal experiments were approved by the relevant local authority and performed according to the national guidelines of the German animal protection law. All mice were bred and housed under specific pathogen-free conditions at the Heinrich Pette Institute.)

FIG 1.

Construction of mutant viruses. (A) Point mutations (underlined) were introduced into the MCMV K181 genome (GenBank accession number AM886412), generating stop codons (*) within the m142 and m143 open reading frames. Note that m142 and m143 are encoded on the complementary strand. (B) Pkr^−/− MEF were infected with wt and mutant MCMVs at an MOI of 1 PFU/cell. Cell lysates were harvested at 72 h postinfection (hpi) and separated by gel electrophoresis for Western blot analysis. Rabbit anti-m142 and m143 antisera (a gift from L. Hanson, Texas Woman's University, USA) were used to detect m142 and m143 proteins, respectively. The MCMV E1 (M112/113) protein was detected as an infection control and β-actin as a loading control using monoclonal antibodies CROMA103 (from S. Jonjic, University of Rijeka, Croatia) and Ac-74 (Sigma), respectively.

Next, we tested the ability of the mutant viruses to replicate in fibroblasts and macrophages derived from wt and Pkr knockout mice (19). Consistent with previous results obtained with m142 and m143 deletion mutants (7), the m142 and m143 stop mutants did not replicate in immortalized wt MEF (3T3) but replicated like wt MCMV in Pkr^−/− MEF (Fig. 2A and B). Primary bone marrow-derived macrophages (BMDM) were isolated from femurs of wt 129/Sv and Pkr^−/− mice on the same genetic background using a published protocol (20). Again, the two stop mutants replicated in BMDM from Pkr^−/− mice but not in BMDM from wt mice (Fig. 2C and D).

FIG 2.

Replication of mutant MCMVs in cell culture. Immortalized wt and Pkr^−/− MEF (A and B) and primary BMDM (C and D) were infected with MCMV K181 wt, m142stop, m143stop, and m142Rev at an MOI of 0.02 PFU/cell. Supernatants were collected, and viral titers were determined by plaque assay on Pkr^−/− MEFs. Primary BMDM were infected on day 7 after isolation. All experiments were done in triplicate; means and standard errors of the means (SEM) are shown. DL, detection limit.

Next, we investigated whether the m142 or m143 proteins have additional functions, apart from PKR inhibition, that are important for viral replication and spread in vivo. To do this, we infected 6-to-8-week-old 129/Sv wt and Pkr^−/− mice with 10⁵ PFU of K181 wt, m142stop, m143stop, or m142Rev viruses by intraperitoneal injection. Mice were sacrificed on days 3, 7, and 14 after infection. Viral titers in spleen, lungs, and salivary glands (SG) were determined using standard protocols (21). As expected, the m142 and m143 stop mutants did not replicate and disseminate in wt mice (Fig. 3). However, the phenotype was completely reversed in Pkr^−/− mice, in which the mutant viruses replicated to titers equivalent to those measured for wt MCMV in all organs and at all time points tested (Fig. 3A to C). Pkr^−/− mice were also infected by subcutaneous injection into the footpad as this infection route represents a more stringent test for viral fitness (22). Nevertheless, the mutant viruses attained salivary gland titers comparable to those measured for wt MCMV (Fig. 3D). It is also noteworthy that wt MCMV did not replicate to higher titers in Pkr^−/− mice than in wt 129/Sv mice, indicating that the viral proteins antagonize PKR so efficiently that the virus does not gain any significant advantage from the absence of PKR.

FIG 3.

Replication of m142 and m143 stop mutants is completely restored in Pkr^−/− mice. 129/Sv wt and Pkr^−/− mice (6 to 8 weeks old) were infected intraperitoneally with 10⁵ PFU/mouse. Viral titers in organs were determined by plaque assay in the spleen at 3 days postinfection (dpi) (A), in the lungs at 7 dpi (B), and in the salivary glands at 14 dpi (C). (D) Pkr^−/− mice were also infected by subcutaneous injection of 10⁵ PFU into the footpad, and viral titers in the salivary glands were determined 14 dpi. Statistical significance was assessed using the Mann-Whitney test. In 129/Sv wt mice, there was no significant difference (P > 0.05) between K181 wt and m142Rev results. In Pkr^−/− mice, there was no significant difference (P > 0.05) between K181 wt and mutant virus results. DL, detection limit.

To rule out the possibility that the single-nucleotide stop mutations had spontaneously reverted in vivo, we analyzed viruses recovered from the SG of Pkr^−/− mice on day 14 postinfection (pi). The SG-derived viruses were used to infect wt and Pkr^−/− fibroblasts. All SG-derived viruses replicated in Pkr^−/− MEF, but only the wt and m142Rev viruses were able to replicate in normal MEF (data not shown). In addition, we analyzed m142 and m143 protein expression by Western blotting. As shown in Fig. 4, the SG-derived viruses had the same m142 and m143 protein expression pattern as the input viruses (compare to Fig. 1B). These data, as well as a sequence analysis of the m142-m143 region (data not shown), clearly showed that spontaneous reversion was not detected and cannot account for the unabated replication of m142 and m143 stop mutants in Pkr^−/− mice.

FIG 4.

Stability of mutant viruses in vivo. Pkr^−/− MEF were infected at an MOI of 1 PFU/cell with virus recovered on day 14 postinfection from salivary glands of Pkr^−/− mice. Cell lysates were harvested 72 hpi and separated electrophoretically for Western blot analysis. MCMV proteins m142, m143, and E1 and the cellular β-actin were detected as described for Fig. 1B.

This report shows that the phenotype (i.e., inability to replicate) elicited by the absence of m142 or m143 in wt cells or during acute infection in 129Sv mice is fully restored in the absence of the host target protein, PKR. A similar phenotypic restoration in Pkr knockout mice has previously been reported for a herpes simplex virus 1 γ34.5 deletion mutant and an influenza A virus NS1 deletion mutant (23, 24). In contrast, the virulence of vaccinia virus E3L and K3L deletion mutants was restored only partially in Pkr^−/− mice, documenting that these viral proteins have additional PKR-independent functions (25). Our data suggest that the MCMV m142 and m143 proteins selectively antagonize PKR with great efficiency and have no additional function of major importance for viral replication and dissemination in vivo. However, the possibility that m142 or m143 has an additional and yet completely redundant function cannot be entirely ruled out. In that case, the additional function would be masked by the presence of another viral protein.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.