Yeast Genome Shuffling Spawns New Chimeric Poxviruses With Cancer-Killing Potential

Oncolytic viruses—viruses engineered or selected to infect and destroy cancer cells while sparing healthy tissue—remain one of the most promising yet underrealized classes of biotherapeutics. Only a handful, such as talimogene laherparepvec for melanoma and teserpaturev for malignant glioma, have secured regulatory approval, and their clinical impact has fallen short of early expectations. Now, researchers at Transgene in Illkirch-Graffenstaden, France, report a synthetic biology platform that could dramatically expand the design space for next-generation oncolytic poxviruses. Writing in Frontiers in Bioengineering and Biotechnology, the team describes a yeast-based genome shuffling system that assembles chimeric poxvirus genomes from multiple non-human-pathogenic parental species, generating infectious viruses with unprecedented functional diversity.

The approach builds on transformation-associated recombination, or TAR cloning, a technique that exploits the extraordinarily efficient homologous recombination machinery of the yeast Saccharomyces cerevisiae to stitch together large DNA fragments in a single step. TAR has previously been used to reconstruct viral genomes ranging from 10-kilobase potyviruses to the 235-kilobase human cytomegalovirus, but poxviruses have remained notoriously difficult targets. Their genomes are large—roughly 191 kilobases in the case of vaccinia virus—and structurally complex, featuring covalently closed hairpin termini and extensive inverted terminal repeats that destabilize circular plasmids in yeast. When both the 5-prime and 3-prime inverted terminal repeats are present in a single plasmid, recombination between the repeat sequences deletes the intervening viral core, leaving a truncated construct.

To circumvent this instability, the team designed a custom yeast shuttle vector, pYCC3, carrying a centromere, an auxotrophic selection marker, and hook sequences that capture only part of the vaccinia virus Copenhagen genome. The construct deliberately excluded the unstable 5-prime terminal hairpin and the complete 3-prime inverted terminal repeat, allowing the researchers to reconstruct approximately 175 of the 191 kilobases of the viral genome. Three overlapping restriction fragments of roughly 70 kilobases each were co-transformed into yeast spheroplasts alongside the linearized vector. Of 188 yeast clones screened, 18 carried viral DNA, and seven tested positive for all three fragments. Sequencing confirmed that the assembled genome was inserted precisely between the hook sequences with no mutations detected across the assembled region.

Because poxvirus DNA alone is not infectious, the researchers needed a way to rescue the synthetic genome into functional viral particles. They turned to Modified Vaccinia virus Ankara, or MVA, an attenuated orthopoxvirus that is replication-deficient in human cells but shares sufficient genomic homology with vaccinia virus to provide the missing terminal sequences in trans. Human A549 lung cancer cells were pre-infected with MVA and then transfected with the yeast-derived plasmids. While neither MVA infection nor plasmid transfection alone produced any cytopathic effect, the combination triggered cell lysis and plaque formation indistinguishable from wild-type vaccinia virus. Sequencing of the rescued viruses revealed that MVA-derived sequences had recombined into both the 5-prime and 3-prime termini, restoring the hairpin structures and terminal repeats that the yeast construct lacked.

With the foundational platform validated, the team moved to the core experiment: genome shuffling. The pYCC3-VACV plasmid was co-transformed into yeast together with fragmented genomic DNA from Cowpox virus strain Brighton and Rabbitpox virus strain Utrecht, both non-pathogenic to humans. Restriction digestion produced fragments of 10 to 40 kilobases, promoting homologous recombination across the three parental genomes. While transformation with the plasmid alone yielded only two clones, and digested cowpox and rabbitpox DNA alone yielded none, co-transformation of all three inputs produced 42 yeast clones carrying recombinant constructs. Ten of these were carried forward into rescue assays, and remarkably, half yielded infectious viral particles when reactivated in MVA-infected human cells.

Whole-genome nanopore sequencing of the five rescued chimeric viruses confirmed that individual genomes contained segments derived from all four parental sources—vaccinia, cowpox, rabbitpox, and the MVA helper virus. Genome sizes ranged from 150 to 187 kilobases. The presence of cowpox and rabbitpox sequences, which were introduced only during the yeast recombination step, provides unambiguous evidence that recombination occurred in yeast prior to rescue. The recovered genomes therefore represent the cumulative outcome of yeast-mediated shuffling followed by helper-virus complementation in mammalian cells.

Phenotypic characterization revealed striking functional diversity among the chimeras, far beyond incremental modification of the parental strains. In HeLa cells, some clones produced regular lytic plaques resembling vaccinia virus, while others formed large syncytia—multinucleated fused cells—or small irregular plaques with syncytial borders. In A549 cells, comet assays showed that one clone, cPOX04-12, exhibited extensive comet-like spreading, a phenotype associated with efficient release of extracellular enveloped virus. Quantification confirmed that cPOX04-12 produced extracellular enveloped virus at roughly 9 percent of total progeny, approximately 50-fold higher than vaccinia virus, which produced less than 0.2 percent. This enhanced dissemination capacity could prove valuable for improving intratumoral spread, one of the key barriers limiting current oncolytic virotherapies.

Oncolytic potency, measured as the half-maximal effective concentration across four human tumor cell lines, also varied substantially. Vaccinia virus showed strong activity with EC50 values between 0.012 and 0.027 multiplicity of infection, while cowpox virus and the non-propagative MVA were markedly less potent. Notably, two chimeric clones outperformed the vaccinia parental strain: cPOX01-02 achieved the lowest EC50 in A549 cells at 0.009, and cPOX04-12 reached 0.0019 in HCT116 colorectal cancer cells. The researchers note that these differences are unlikely to be explained solely by variations in interferon signaling among the cell lines, since poxviruses encode multiple antagonists of antitumor innate immunity, and suggest that virus-specific genetic determinants contribute to the observed activity profiles.

The genetic basis of some phenotypes remains partially resolved. The A56R and K2L genes are established regulators of poxvirus-induced cell fusion, and all sequenced clones retained intact A56R alleles, ruling out alterations at this locus as the cause of syncytium formation. Both cPOX01-02 and cPOX04-12 carry the cowpox-derived K2L allele, yet only cPOX04-12 displays pronounced syncytia, implying that additional, unidentified genetic determinants overcome the inhibitory effect of the cowpox K2L protein. Similarly, the enhanced extracellular enveloped virus production in cPOX04-12 cannot be attributed to the known A34R K151E mutation, since all chimeric clones carry the MVA-derived A34R allele lacking this change. Identifying these novel determinants through larger-scale sequencing and phenotypic screening is a stated priority for future work.

A significant methodological advantage of the yeast platform is the complete avoidance of Escherichia coli propagation. Bacterial artificial chromosome cloning of large viral genomes is frequently hampered by toxicity, instability, and cryptic bacterial promoters driving expression of viral proteins—problems documented for herpesviruses, flaviviruses, and coronaviruses. By assembling and maintaining chimeric poxvirus genomes entirely in yeast, the workflow sidesteps these constraints. The authors emphasize that the chimeric viruses described here do not yet carry the attenuating mutations required for tumor-selective replication, and that any candidate moving toward development will need evaluation of genetic stability, tumor selectivity, and safety in non-transformed cells. Nevertheless, the platform establishes a scalable, modular foundation for exploring the poxvirus sequence-function landscape, potentially enabling rational design of oncolytic viruses with improved spread, immunogenicity, and compatibility with combination regimens including immune checkpoint inhibitors.

Subject of Research: Yeast-mediated synthetic genome shuffling to generate chimeric oncolytic poxviruses

Article Title: Synthetic genome shuffling of poxviruses in yeast generates novel poxvirus candidates with potential oncolytic properties

Article References: Agaoua, A., Rey, C., Hortelano, J., Moro, A.-I., Grellier, B., & Erbs, P. (2026). Synthetic genome shuffling of poxviruses in yeast generates novel poxvirus candidates with potential oncolytic properties. Frontiers in Bioengineering and Biotechnology, 14, Article 1943548. https://doi.org/10.3389/fbioe.2026.1943548

Image Credits: AI Generated

DOI: 10.3389/fbioe.2026.1943548

Keywords: oncolytic viruses, poxvirus, genome shuffling, TAR cloning, synthetic biology, yeast, vaccinia virus, cowpox virus, rabbitpox virus, cancer therapy, virotherapy, chimeric viruses