Vaccinia Virus Gene Acquisition through Nonhomologous Recombination

Many of the genes encoded by poxviruses are orthologs of cellular genes. These virus genes serve different purposes, but perhaps of most interest is the way some have been repurposed to inhibit the antiviral pathways that their cellular homologs still regulate. What is unclear is how these virus genes were acquired, although it is presumed to have been catalyzed by some form(s) of nonhomologous recombination (NHR). We used transfection assays and substrates encoding a fluorescent and drug-selectable marker to examine the NHR frequency in vaccinia virus (VAC)-infected cells. These studies showed that when cells were transfected with linear duplex DNAs bearing VAC N2L gene homology, it yielded a recombinant frequency (RF) of 6.7 × 10^-4. In contrast, DNA lacking any VAC homology reduced the yield of recombinants ∼400-fold (RF = 1.6 × 10^-6). DNA-RNA hybrids were also substrates, although homologous molecules yielded fewer recombinants (RF = 2.1 × 10^-5), and nonhomologous substrates yielded only rare recombinants (RF ≤ 3 × 10^-8). NHR was associated with genome rearrangements ranging from simple insertions with flanking sequence duplications to large-scale indels that produced helper-dependent viruses. The insert was often also partially duplicated and would rapidly rearrange through homologous recombination. Most of the virus-insert junctions exhibited little or no preexiting microhomology, although a few encoded VAC topoisomerase recognition sites (C/T·CCTT). These studies show that VAC can catalyze NHR through a process that may reflect a form of aberrant replication fork repair. Although it is less efficient than classical homologous recombination, the rates of NHR may still be high enough to drive virus evolution. IMPORTANCE Large DNA viruses sometimes interfere in antiviral defenses using repurposed and mutant forms of the cellular proteins that mediate these same reactions. Such virus orthologs of cellular genes were presumably captured through nonhomologous recombination, perhaps in the distant past, but nothing is known about the processes that might promote "gene capture" or even how often these events occur over the course of an infectious cycle. This study shows that nonhomologous recombination in vaccinia virus-infected cells is frequent enough to seed a small but still significant portion of novel recombinants into large populations of newly replicated virus particles. This offers a route by which a pool of virus might survey the host genome for sequences that offer a selective growth advantage and potentially drive discontinuous virus evolution (saltation) through the acquisition of adventitious traits.