Vaccinia Virus Attenuation by Codon Deoptimization of the A24R Gene for Vaccine Development

Vaccinia Virus Attenuation by Codon Deoptimization of the A24R Gene for Vaccine Development María M. Lorenzo, a Aitor Nogales, b , c Kevin Chiem, b , d Rafael Blasco, a Luis Martínez-Sobrido b , d a Departamento de Biotecnología, Centro Nacional INIA, Consejo Superior de Investigaciones Científicas (CSIC), Madrid, Spain b Department of Microbiology and Immunology, University of Rochester, Rochester, New York, USA c Animal Health Research Centre (CISA), National Institute for Agriculture and Food Research and Technology (INIA-CSIC), Valdeolmos, Madrid, Spain d Texas Biomedical Research Institute, San Antonio, Texas, USA María M. Lorenzo and Aitor Nogales contributed equally to this article. Author order was determined both alphabetically and in order of increasing seniority. ABSTRACT Poxviruses have large DNA genomes, and they are able to infect multiple vertebrate and invertebrate animals, including humans. Despite the eradication of smallpox, poxvirus infections still remain a significant public health concern. Vaccinia virus (VV) is the prototypic member in the poxviridae family and it has been used extensively for different prophylactic applications, including the generation of vaccines against multiple infectious diseases and/or for oncolytic treatment. Many attempts have been pursued to develop novel attenuated forms of VV with improved safety profiles for their implementation as vaccines and/or vaccines vectors. We and others have previously demonstrated how RNA viruses encoding codon-deoptimized viral genes are attenuated, immunogenic and able to protect, upon a single administration, against challenge with parental viruses. In this study, we employed the same experimental approach based on the use of misrepresented codons for the generation of a recombinant (r)VV encoding a codon-deoptimized A24R gene, which is a key component of the viral RNA polymerase. Similar to our previous studies with RNA viruses, the A24R codon-deoptimized rVV (v-A24cd) was highly attenuated in vivo but able to protect, after a single intranasal dose administration, against an otherwise lethal challenge with parental VV. These results indicate that poxviruses can be effectively attenuated by synonymous codon deoptimization and open the possibility of using this methodology alone or in combination with other experimental approaches for the development of attenuated vaccines for the treatment of poxvirus infection, or to generate improved VV-based vectors. Moreover, this approach could be applied to other DNA viruses. IMPORTANCE The family poxviridae includes multiple viruses of medical and veterinary relevance, being vaccinia virus (VV) the prototypic member in the family. VV was used during the smallpox vaccination campaign to eradicate variola virus (VARV), which is considered a credible bioterrorism threat. Because of novel innovations in genetic engineering and vaccine technology, VV has gained popularity as a viral vector for the development of vaccines against several infectious diseases. Several approaches have been used to generate attenuated VV for its implementation as vaccine and/or vaccine vector. Here, we generated a rVV containing a codondeoptimized A24R gene (v-A24cd), which encodes a key component of the viral RNA polymerase. v-A24cd was stable in culture cells and highly attenuated in vivo but able to protect against a subsequent lethal challenge with parental VV. Our findings support the use of this approach for the development of safe, stable, and protective live-attenuated VV and/or vaccine vectors. KEYWORDS vaccinia virus, codon bias, codon deoptimization, A24R gene, liveattenuated vaccines, GFP Editor Samuel K. Campos, University of Arizona Copyright © 2022 Lorenzo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license. Address correspondence to Luis Martínez-Sobrido, [email protected], or Rafael Blasco, [email protected]. The authors declare no conflict of interest. Received 22 January 2022 Accepted 20 April 2022 Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 1 RESEARCH ARTICLE Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. Poxviruses belong to a family of large, double stranded DNA viruses, designated Poxviridae, which replicate and assemble entirely in the cytoplasm of infected cells (1–3). Poxviruses are able to infect a broad range of both invertebrate and vertebrate animals, including humans and wildlife or domestic animals, and cause disease in many of them (4–8). Therefore, they are considered an important threat to public human health (6–8). Smallpox is caused by variola virus (VARV) which, together with vaccinia virus (VV), belong to the orthopoxvirus genus, and constituted one of the deadliest diseases in human history that killed approximately 300 million people only in the twentieth century (9–11). Thanks to a worldwide vaccination program, the lack of non-human reservoirs, and virus slow mutation rate, the World Health Organization (WHO) declared smallpox eradicated in 1980 (11, 12). However, worries remain because the possibility that smallpox may reemerge accidently from forgotten stocks of VARV or even from de novo synthesis by using current biotechnologies (10, 13). In fact, the United States military considers the risk of weaponized smallpox sufficient to justify the continued vaccination of all its military personal against the disease (14). Given the low worldwide vaccination rate against smallpox since 1980, population could be highly vulnerable to a new outbreak of smallpox, increasing the current concerns (10, 13, 15). Furthermore, with increasing global trade many poxviruses infecting animals could have an important impact on commercially relevant livestock and ecologically endangered wildlife (4–8, 14). A particularly relevant case is monkeypox which causes a disease with symptoms similar to, but less severe than, smallpox. Monkeypox is a zoonosis with different hosts and human-to-human transmission is limited. However, 15 countries on four continents have reported confirmed human cases of monkeypox. Most are usually found near tropical rainforests, where there are animals carrying the virus, but there have been emergencies in other remote countries such as the United States and the United Kingdom. Importantly, there is currently no specific treatment for monkeypox (16), although vaccination against smallpox has been shown to be approximately 85% effective in preventing monkeypox (17, 18). VV was used to prevent and eradicate smallpox and it has become the prototypic member of the poxviridae family. Importantly, novel genetic engineering methods to facilitate the manipulation of the genome of poxvirus (19–22) have allowed researchers to use VV as a vaccine vector for the development of vaccines against a variety of infectious diseases, and/or for oncolytic treatment (9, 23–28). In addition, poxviruses, mainly VV, have also multiple biotechnological applications based on heterologous gene expression (19). A key factor to the optimization of poxviruses as vaccine vectors is the balance between the safety profile and immunogenicity versus retained virulence. Although the smallpox vaccine is generally considered safe and VV is naturally attenuated, classical VV vaccine strains still may produce some complications in patients with systemic immune deficits, which should be exempt from current vaccination programs because of significant increased risks of adverse or severe outcomes (15, 29, 30). Therefore, several strategies have been applied to increase the attenuation of poxviruses as vaccine vectors (9, 23, 26). In general those include deletion of genes and point mutations as a result of extensive passaging in cell culture (e.g., modified vaccinia Ankara, MVA) (22, 31–34) or the introduction of multiple defined deletions (NYVAC) (9, 35). Alternatively avipoxviruses, like fowlpox and canarypox, which are not well adapted to mammalian cells, have been used as vaccine vectors since they are not able to undergo full replication in mammalian cells (7, 9, 26). Therefore, although a variety of attenuated poxviral vaccine vectors are currently available they have been derived from blind processes. The mammalian genetic code is degenerated, and most amino acids are coded by multiple synonymous codons, which are not used with equal frequency within or between DNA/RNA genomes (36, 37), showing differences in the frequency at which organisms use codons to incorporate the same amino acid residue into a protein (38– 40). This biological phenomenon called codon usage bias has been used for codon optimization (CO) or codon deoptimization (CD) approaches to increase or decrease, Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 2 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. respectively, gene expression in different expression systems. While CO, where each amino acid is encoded by the most frequently used codons is widely employed to increase protein expression for pharmaceutical, biotechnological or research purposes (41–45), the potential of CD has been less explored (46–52). CD is achieved by replacing original codons for those with less-preferred usage (46–50). Importantly, changes are only at the nucleotide level without affecting the amino acid sequence of the protein or their immunogenic properties and functionality (53, 54). We and others have previously documented the feasibility of generating recombinant RNA viruses, harboring codon-deoptimized genes, with attenuated phenotypes as potential vaccines and/ or vaccine vectors (46–52). For that, several genes of RNA viruses as influenza A virus (IAV) (46, 47), the arenaviruses lymphocytic choriomeningitis virus (LCMV) or Lassa virus (LASV) (48–50), and respiratory syncytial virus (RSV) (52) were CD, and the in vitro and in vivo characteristics of the recovered viruses were analyzed in different cells and animal models, respectively. However, to date, the use of a CD-based approach for the development of attenuated forms of viruses for their implementation as safe and protective live-attenuated vaccines have not been explored for DNA viruses. Since VV encodes all the enzymes required for DNA replication and transcription, virus replication occurs independently of the host. One key virally-encoded enzyme is the viral RNA polymerase, which is required throughout the replication cycle for transcription of viral genes. In this work, we have reengineered the VV A24R gene, encoding the second largest subunit of the viral RNA polymerase rpo132 (55–59), by introducing through de novo gene synthesis, the least used human synonymous codons without modifying the viral protein amino acid sequence. The generated recombinant virus containing synonymous CD mutations in the A24R gene (v-A24cd) showed lower replication levels and altered plaque phenotype in culture cells. Moreover, v-A24cd was highly attenuated in a mouse model of infection. Importantly, immunization of mice using a single intranasal dose of v-A24cd conferred full protection against a lethal challenge with parental VV, demonstrating the feasibility of using this CD-based approach for the development of novel and safer VV vaccines and/or vaccine vectors, or for the its implementation to attenuate other DNA viruses. RESULTS CD of VV A24R gene results in reduced levels of protein expression. Codon deoptimization (CD) represents a potential strategy to modify the expression efficiency of a gene, thereby altering the amount of its protein product. We selected the A24R gene as a target to examine the potential of using our CD-based approach with VV for the generation of a live-attenuated vaccine and to demonstrate the feasibly of implementing this CD-based strategy for the attenuation of DNA viruses. A24R encodes the catalytic subunit of VV RNA polymerase, which plays a critical role in viral gene transcription (55–59). Therefore, the CD of VV A24R gene, most likely, will affect the expression levels of other viral proteins, disrupting multiple steps during VV infection. To evaluate the effect of CD on VV A24R expression, we engineered its open reading frame (ORF) using underrepresented codons in human cells but preserving the intact A24R amino acid sequence. The CD A24R (A24Rcd) sequence (Fig. 1A and Fig. S1) included 666 (57.17%) codon changes through 775 (22.17%) nucleotide substitutions (Fig. 1B) and the gene was synthesized and cloned in plasmids suited for insertion into the VV genome (Fig. 2). In our design, we included the A24R versions and recombination flanks for the A24R gene L and R (left and right) for the generation of rVV by homologous recombination. In addition, a green fluorescent protein (GFP) expression cassette was included for easy isolation and visualization of recombinant viruses (Fig. 2). To determine whether CD of A24R could lead to a reduction in protein expression levels in the absence of infection, HA-tagged A24Rwt or A24Rcd ORF sequences were cloned into pCAGGS expression plasmids. Then, human 293T cells were transiently transfected with pCAGGS-A24Rwt-HA or pCAGGS-A24Rcd-HA constructs, and protein expression was analyzed by immunofluorescence using an antibody against the HA epitope tag (Fig. 1C). A significant reduction in both the fluorescent signal and the Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 3 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. FIG 1 Codon deoptimization reduces VV A24R expression levels. (A) Amino acid sequence of A24R. Codon-deoptimized amino acid residues are indicated in red (see Fig. S1 for wild-type and codon-deoptimized nucleotide sequences). Methionine (M) and (Continued on next page) Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 4 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. number of fluorescent cells was observed in cells transfected with the plasmid encoding A24Rcd-HA compared to cells transfected with the A24Rwt-HA plasmid (Fig. 1C). The effects of CD on A24R protein expression were also assessed by Western blotting. For that, human 293T cells were transiently co-transfected with pCAGGS plasmids expressing GFP and A24Rwt or A24Rcd genes and protein expression was evaluated at 48 h posttransfection (h p.t.) GFP examination using a fluorescence microscope (Fig. 1D) or by Western blotting (Fig. 1E) confirmed that A24R (A24Rwt or A24Rcd) did not affect GFP expression, indicating that A24R sequence is not toxic in transfected cells, and the observed reduced expression of A24Rcd is due to the sequence of codons used. Whereas A24R was clearly detected, the level of expression of A24Rcd was drastically reduced, in agreement with the fluorescence results (Fig. 1E). Protein densitometry of Western blot bands further confirmed this observation. These data indicate that codon deoptimization of VV A24R reduces protein expression in mammalian cells, as we had observed previously for proteins of RNA viruses, including the nucleoprotein (NP) and glycoproteins of arenaviruses (48–50), or the nonstructural protein 1 (NS1) and nuclear export protein (NEP) of IAV (46, 47). Generation and growth properties of v-A24Rcd in cell culture. To evaluate the effect of A24R CD in viral fitness and pathogenicity, we next used well-established poxvirus genetics techniques to isolate viruses expressing modified versions of A24R (Fig. 2). Substitution of the A24R gene by a CD version was carried out by insertion of an A24 CD/GFP construct in place of the normal A24R gene, to obtain virus v-A24cd. A revertant virus expressing the normal A24R gene (v-A24Rre) was subsequently obtained from vA24cd by recombination of an A24/GFP construct into the same locus. As an additional control, we used v-GFP, which contains a GFP cassette at a distant place in the genome and has an unmodified A24R gene (Fig. 2). FIG 1 Legend (Continued) tryptophan (W) amino acid residues, as well as amino acids already associated with deoptimized codons are indicated in black. (B) Mutations in A24Rcd. Number and percentage of nucleotides or codons changes in the codondeoptimized A24R are indicated. (C) Analysis of protein expression in transfected cells by immunofluorescence. Human 293T cells were transfected with 1 m g of pCAGGS expression plasmids encoding C-terminal HA-tagged A24Rwt or A24Rcd proteins. At 48 h p.t., protein expression was assessed by immunofluorescence using an antiHA pAb. DAPI was used for nuclear staining. Scale bars, 100 m m. Relative fluorescence intensities (RFI) for immunofluorescences from three representative images were quantified by densitometry using ImageJ software (v1.53m), and the intensity of A24R stained with an HA antibody was normalized to the intensity of DAPI. (D–E) Analysis of protein expression by Western blotting. Human 293T cells were transiently co-transfected (triplicates) with pCAGGS expression plasmids encoding C-terminal HA-tagged versions of A24Rwt or A24Rcd proteins (2 m g) together with a pCAGGS plasmid encoding GFP (0.5 m g). At 48 h p.t., GFP expression was assessed under a fluorescence microscope (D). A24Rwt or A24Rcd protein expression levels were assessed by Western blotting using an anti-HA pAb (E). GFP expression levels were assessed using an anti-GFP pAb. Actin expression levels were determined using an anti-actin MAb and were used as loading controls. Numbers in the left indicate the size of molecular markers in kDa. Scale bars, 200 m m. Western blots were quantified by densitometry using the software ImageJ (v1.53m). Relative band intensities as described in Materials and Methods are indicated. Western blot results from three separate experiments are shown. FIG 2 Schematic representation of rVV. The names and the viral genomic organization around the F13L and A24R loci of the rVV used in this study are indicated. Viral promoters (black arrows for F13 and A24R and white arrows for GFP), as well as genes encoding for GFP (green boxes), F13L (red boxes) and WR, and codon deoptimized (cd) and revertant (re) A24R (yellow and blue boxes, respectively) are shown. Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 5 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. Next, we assessed the replication properties of the generated rVV by multicycle growth kinetics in BSC-1 cells infected at low multiplicity of infection (MOI, 0.01). Progress of infection was followed by examining GFP expression under a fluorescence microscope (Fig. 3A) or the presence of infectivity in cell culture supernatants using plaque assays (Fig. 3B). The recombinant v-GFP and v-A24Rre displayed similar levels of GFP expression and replication kinetics (Fig. 3A and B, respectively) and significant differences were not observed between them. On the other hand, GFP expression from v-A24Rcd was considerably reduced (Fig. 3A) and the virus did not achieve replication levels similar to those of v-GFP and v-A24Rre (Fig. 3B). In addition, we measured the presence of intracellular and extracellular v-GFP, v-A24Rcd and v-A24Rre in BSC-1 infected cells (MOI, 3) at 24 h postinfection, h p.i. (Fig. 3C). We observed a decrease in cell-associated and extracellular v-A24Rcd titers as compared with v-GFP and v-A24Rre, suggesting that our CD-based strategy has a general effect in the production of progeny virus. We also examined the phenotype of v-A24Rcd, v-A24Rre or v-GFP in BSC-1 cells using plaque assays and GFP expression (Fig. 3D). As expected, based on our multicycle growth kinetics (Fig. 3A and 3B), v-A24cd produced significantly smaller plaques than the control recombinant viruses v-GFP and v-A24Rre (Fig. 3D). Altogether, these data demonstrate that growth of v-A24Rcd was impaired compared to v-GFP or v-A24Rre in cultured cells. To evaluate if infection with v-A24Rcd led to differential expression of viral proteins, lysates of infected BSC-1 cells were evaluated by Western blotting (Fig. 3E). To that end, cells were either mock-infected or infected (MOI, 3) and cell extracts were collected at 24 h p.i. and probed using specific antibodies against VV F13 protein, and GFP. An antibody against actin was used as a protein loading control. The levels of F13, and GFP were not altered significantly between v-GFP or v-A24Rre infected cells. However, F13 and GFP protein expression levels were notably reduced, in v-A24Rcd-infected BSC-1 cells compared to v-GFP or v-A24Rre. Altogether, these data suggest that v-A24Rcd results in an expected decrease in viral protein expression levels in infected cells, that appears as a general effect on viral protein accumulation (Fig. 3E). To show that v-A24Rcd is also affected in human cell lines, 293T cells were mockinfected or infected (MOI, 3) with v-GFP, v-A24Rcd and v-A24Rre. Then at 24 h p.i. GFP expression was evaluated under a fluorescence microscope (Fig. 3F). The recombinant v-GFP and v-A24Rre displayed similar levels of GFP expression, while levels of GFP in A24Rcd-infected 293T cells were drastically reduced. Lysates of same mockand viralinfected 293T cells were analyzed by Western blotting as indicated above (Fig. 3G). As in BSC-1 cells, levels of F13 and GFP expression were reduced in v-A24Rcd-infected cells compared to cells infected with v-GFP and v-A24Rre (Fig. 3G). Viral factory formation and subcellular localization of B5 protein. Previous experiments show a block in virus replication as well as a decrease in viral protein accumulated during infection with v-A24cd (Fig. 3). To characterize in more detail the infection by v-A24cd, we studied the last steps in the virus life cycle by fluorescence microscopy (Fig. 4). The results confirmed a decrease in GFP expression, and a reduction in the number of viral factories in vA24cd-infected cells when compared to those infected with the control viruses v-GFP and vA24re (Fig. 4). In addition, we visualized by immunofluorescence the distribution of protein B5, which is localized in the Golgi complex, where it is incorporated into infectious virus that are then transported and released into the extracellular medium (60). B5 labeling was decreased in v-A24cd infected cells, being barely detectable at 8 h p.i. (Fig. 4A and B). However, although at later times (18 h p.i., Fig. 4C) the amount of B5 was also low, its distribution appeared normal, being accumulated in a juxtanuclear region consistent with the Golgi complex and puncta representing enveloped virus particles. These results are indicative of a virus replication cycle that is delayed with respect to normal virus, likely as a result of low protein expression resulting from the CD of A24R. In vivo characterization of v-A24Rcd. Because we observed that replication properties and viral protein expression of v-A24Rcd were affected in BSC-1 and 293T cells (Fig. 3), including viral factory formation (Fig. 4), we postulated that v-A24Rcd would Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 6 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. FIG 3 Characterization of v-A24Rcd. (A–B) Growth kinetics. BSC-1 cells monolayers (6-well plate format, 10 6 cells/well) were infected (MOI 0.01) in triplicate with the indicated viruses and at the indicated h p.i., GFP fluorescence was (Continued on next page) Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 7 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. be attenuated in vivo. To demonstrate this hypothesis, we evaluated and compared the virulence of v-A24Rcd, v-A24Rre, and parental viruses in a mouse model of viral infection (Fig. 5). For that, groups of C57BL/6 mice (n= 6/group) were inoculated intranasally (i.n.) with 10 4 ,10 5 ,or10 6 PFU/animal of v-A24Rcd, v-A24Rre, or Western reserve (WR) and mice were monitored during 14 days for body weight loss (Fig. 5A) and mortality (Fig. 5B). In addition, a control group was mock-infected (PBS). Mice infected with v-A24Rre or parental WR viruses showed clear and similar changes in morbidity (Fig. 5A) and all succumbed to viral infection independently of the viral dose(Fig.5B).Notably,v-A24Rcdwashighlyattenuated,withnochangesinbody weight and all the animals surviving viral infection (Fig. 5B), even with the highest dose of 10 6 PFU. v-A24Rcd protects mice from lethal viral challenge with parental VV. Despite the fact that v-A24Rcd was highly attenuated in mice (Fig. 5), we hypothesized that vA24Rcd could protect against otherwise subsequent lethal challenge with parental VV. To assess this, animals inoculated with 10 4 ,10 5 ,or10 6 PFU/animal of v-A24Rcd, or mock-vaccinated (Fig. 5) were challenged 21 days later with 10 4 PFU/mice of parental VV WR and morbidity (Fig. 6A) and mortality (Fig. 6B) were monitored for 2 weeks. None of the v-A24Rcd-vaccinated animals, independently of the vaccination dose, showed any changes in body weight loss (Fig. 6A), and all animals survived the lethal challenge with parental VV (Fig. 6B). Contrary, all mock-vaccinated mice drastically lost weight and died after parental VV WR challenge (Fig. 6A and B, respectively). These results indicate that a single dose of v-A24Rcd is able to protect against a lethal challenge with parental VV WR, demonstrating the feasibility of its use as a safe and protective live-attenuated vaccine, or vaccine vector. Stability of v-A24Rcd in culture cells. A critical concern with live-attenuated vaccines or vaccine vectors is the prospect of genetic instability that might lead to the loss of attenuation. To assess stability of v-A24Rcd, we serially passaged the virus in BSC-1 cells for a total of 10 passages (Fig. 7). Next, virus growth properties from the first (vA24Rcd.P1) and last (v-A24Rcd.P10) passage were evaluated by plaque assay (Fig. 7A–C) and multicycle growth kinetics (Fig. 7D). For these assays, v-GFP and parental viruses were also included as controls. Notably, results indicate that v-A24Rcd was phenotypically stable since both v-A24Rcd.P1 and v-A24Rcd.P10 displayed the same plaque phenotype (Fig. 7A and B) and levels of GFP expression (Fig. 7A), with A24Rcd.P1 and v-A24Rcd.P10 producing smaller plaques than v-GFP or parental viruses (Fig. 7C). In addition, v-A24Rcd.P1 and v-A24Rcd.P10 grew with indistinguishable kinetics reaching the same viral titers that were both significantly reduced compared to v-GFP or parental viruses (Fig. 7D). Altogether, these results suggest that v-A24Rcd is phenotypically stable in FIG 3 Legend (Continued) detected by fluorescence microscopy (A). Scale bars, 200 m m. Cell culture supernatants were also collected at the indicated times p.i. and viral titers were determined using standard plaque assay (B). *,P,0.05, using Student's ttest (v-GFP versus v-A24Rcd; not significant (ns) differences were obtained between v-GFP and v-A24Rre). (C) Production of extracellular and cell-associated virus. BSC-1 cells (6-well plate format, 10 6 cells/well) were infected at an MOI of 3 PFU in triplicate. At 36 h p.i., virus in the cell culture medium (VE) and associated with the cells (VI) was titrated by plaque assay on fresh BSC-1 monolayers. (D) Plaque phenotype. BSC-1 cell monolayers (6-well plate format, 10 6 cells/well, triplicates) were infected with ;100-200 PFU/well of the indicated viruses and incubated with medium containing methyl cellulose. At 3 days p.i., monolayers were fixed, GFP fluorescence was imaged (top) and wells were stained with crystal violet (middle). Scale bars, 100 m m. The area of plaques visualized by crystal violet staining of 200 viruses was quantified and is represented (bottom). ***,P,0.0001, using Student's ttest. ns: not significant differences. (E) Protein expression levels. BSC-1 cells (6-well plate format, 10 6 cells/well) were either mockinfected or infected (MOI 3) with the indicated viruses and viral protein expression levels were analyzed at 24 h p.i. by Western blotting with antibodies against F13L, or GFP. Actin was used as loading control. Western blots were quantified by densitometry using the software ImageJ. Bands were normalized to actin expression. Numbers in the left indicate the size of molecular markers in kDa. (F-G) Characterization of v-A24Rcd in 293T cells. Human 293T cells monolayers (6-well plate format, 10 6 cells/well) were mock-infected or infected (MOI 3) with the indicated viruses and at 24 h p.i., GFP expression was evaluated by fluorescence microscopy. Scale bars, 200 m m (F). In addition, F13L and GFP expression levels were evaluated at 24 h p.i. by Western blotting with antibodies against F13L or GFP, respectively. Actin was used as loading control (G). Western blots were quantified by densitometry using the software ImageJ. Bands were normalized to levels of actin. Numbers in the left indicate the size of molecular markers in kDa. Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 8 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. vitro, which is an important feature for the implementation of v-A24Rcd as a live-attenuated vaccine and/or vaccine vector. DISCUSSION The accumulated knowledge of poxviruses, including the prototype VV, together with the development of recombinant DNA techniques, have expanded the use of poxviruses as vaccine vectors to prevent infections in humans and domestic or wild animals (7, 9, 19, 23, 24, 26, 27). VV offers many advantages as vaccine vector, including FIG 4 Subcellular localization of B5 in infected cells. Monolayers of BHK-21 cells grown on cell coverslips were infected with the virus indicated on the left. At 8 h p.i. (A and B) or 18 h p.i. (C) cells were fixed, and subjected to immunofluorescence with anti-B5 antibody. For visualization of viral factories (arrows), cells were incubated with Hoechst staining. Scale bars, 50 m m. Viral factories at 8 h p.i. (A) were quantified using the software ImageJ (v1.53m). The number of viral factories per cell (left) and the total area of viral factories per cell (right) are indicated (B). Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 9 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41. 44. Sethi L, Kumari K, Dey N. 2021. Engineering of plants for efficient production of therapeutics. Mol Biotechnol 63:1125–1137. https://doi.org/10 .1007/s12033-021-00381-0. 45. To KKW, Cho WCS. 2021. An overview of rational design of mRNA-based therapeutics and vaccines. Expert Opin Drug Discov. https://doi.org/10 .1080/17460441.2021.1935859. 46. Nogales A, Baker SF, Ortiz-Riano E, Dewhurst S, Topham DJ, MartinezSobrido L. 2014. Influenza A virus attenuation by codon deoptimization of the NS gene for vaccine development. J Virol 88:10525–10540. https:// doi.org/10.1128/JVI.01565-14. 47. Baker SF, Nogales A, Martinez-Sobrido L. 2015. Downregulating viral gene expression: codon usage bias manipulation for the generation of novel influenza A virus vaccines. Future Virol 10:715–730. https://doi.org/10 .2217/fvl.15.31. 48. Cheng BY, Ortiz-Riano E, Nogales A, de la Torre JC, Martinez-Sobrido L. 2015. Development of live-attenuated arenavirus vaccines based on codon deoptimization. J Virol 89:3523–3533. https://doi.org/10.1128/JVI.03401-14. 49. Cheng BYH, Nogales A, de la Torre JC, Martinez-Sobrido L. 2017. Development of live-attenuated arenavirus vaccines based on codon deoptimization of the viral glycoprotein. Virology 501:35–46. https://doi.org/10 .1016/j.virol.2016.11.001. 50. Cai Y, Ye C, Cheng B, Nogales A, Iwasaki M, Yu S, Cooper K, Liu DX, Hart R, Adams R, Brady T, Postnikova EN, Kurtz J, St Claire M, Kuhn JH, de la Torre JC, Martinez-Sobrido L. 2020. A lassa fever live-attenuated vaccine based on codon deoptimization of the viral glycoprotein gene. mBio 11. https:// doi.org/10.1128/mBio.00039-20. 51. Burns CC, Shaw J, Campagnoli R, Jorba J, Vincent A, Quay J, Kew O. 2006. Modulation of poliovirus replicative fitness in HeLa cells by deoptimization of synonymous codon usage in the capsid region. J Virol 80: 3259–3272. https://doi.org/10.1128/JVI.80.7.3259-3272.2006. 52. Meng J, Lee S, Hotard AL, Moore ML. 2014. Refining the balance of attenuation and immunogenicity of respiratory syncytial virus by targeted codon deoptimization of virulence genes. mBio 5:e01704-14–e01714. https://doi.org/10.1128/mBio.01704-14. 53. Lavner Y, Kotlar D. 2005. Codon bias as a factor in regulating expression via translation rate in the human genome. Gene 345:127–138. https://doi .org/10.1016/j.gene.2004.11.035. 54. Puigbo P, Aragones L, Garcia-Vallve S. 2010. RCDI/eRCDI: a web-server to estimate codon usage deoptimization. BMC Res Notes 3:87. https://doi .org/10.1186/1756-0500-3-87. 55. Boulanger D, Green P, Jones B, Henriquet G, Hunt LG, Laidlaw SM, Monaghan P, Skinner MA. 2002. Identification and characterization of three immunodominant structural proteins of fowlpox virus. J Virol 76: 9844–9855. https://doi.org/10.1128/jvi.76.19.9844-9855.2002. 56. Brennan G, Kitzman JO, Shendure J, Geballe AP. 2015. Experimental Evolution identifies Vaccinia virus mutations in A24R and A35R that antagonize the protein kinase R pathway and accompany collapse of an extragenic gene amplification. J Virol 89:9986–9997. https://doi.org/10.1128/JVI.01233-15. 57. Cone KR, Kronenberg ZN, Yandell M, Elde NC. 2017. Emergence of a viral RNA polymerase variant during gene copy number amplification promotes rapid evolution of Vaccinia virus. J Virol 91. https://doi.org/10 .1128/JVI.01428-16. 58. Cresawn SG, Condit RC. 2007. A targeted approach to identification of vaccinia virus postreplicative transcription elongation factors: genetic evidence for a role of the H5R gene in vaccinia transcription. Virology 363: 333–341. https://doi.org/10.1016/j.virol.2007.02.016. 59. Cresawn SG, Prins C, Latner DR, Condit RC. 2007. Mapping and phenotypic analysis of spontaneous isatin-beta-thiosemicarbazone resistant mutants of vaccinia virus. Virology 363:319–332. https://doi.org/10.1016/j .virol.2007.02.005. 60. Smith GL, Vanderplasschen A, Law M. 2002. The formation and function of extracellular enveloped vaccinia virus. J Gen Virol 83:2915–2931. https://doi.org/10.1099/0022-1317-83-12-2915. 61. Duret L. 2002. Evolution of synonymous codon usage in metazoans. Curr Opin Genet Dev 12:640–649. https://doi.org/10.1016/s0959-437x(02)00353-2. 62. Eschke K, Trimpert J, Osterrieder N, Kunec D. 2018. Attenuation of a very virulent Marek's disease herpesvirus (MDV) by codon pair bias deoptimization. PLoS Pathog 14:e1006857. https://doi.org/10.1371/journal.ppat .1006857. 63. Groenke N, Trimpert J, Merz S, Conradie AM, Wyler E, Zhang H, Hazapis OG, Rausch S, Landthaler M, Osterrieder N, Kunec D. 2020. Mechanism of virus attenuation by codon pair deoptimization. Cell Rep 31:107586. https://doi.org/10.1016/j.celrep.2020.107586. 64. Jordan-Paiz A, Franco S, Martinez MA. 2021. Synonymous codon pair recoding of the HIV-1 env gene affects virus replication capacity. Cells 10: 1636. https://doi.org/10.3390/cells10071636. 65. Li P, Ke X, Wang T, Tan Z, Luo D, Miao Y, Sun J, Zhang Y, Liu Y, Hu Q, Xu F, Wang H, Zheng Z. 2018. Zika virus attenuation by codon pair deoptimization induces sterilizing immunity in mouse models. J Virol 92. https://doi .org/10.1128/JVI.00701-18. 66. Mueller S, Stauft CB, Kalkeri R, Koidei F, Kushnir A, Tasker S, Coleman JR. 2020. A codon-pair deoptimized live-attenuated vaccine against respiratory syncytial virus is immunogenic and efficacious in non-human primates. Vaccine 38:2943–2948. https://doi.org/10.1016/j.vaccine.2020.02.056. 67. Pereira-Gomez M, Carrau L, Fajardo A, Moreno P, Moratorio G. 2021. Altering compositional properties of viral genomes to design live-attenuated vaccines. Front Microbiol 12:676582. https://doi.org/10.3389/fmicb.2021 .676582. 68. Yang C, Skiena S, Futcher B, Mueller S, Wimmer E. 2013. Deliberate reduction of hemagglutinin and neuraminidase expression of influenza virus leads to an ultraprotective live vaccine in mice. Proc Natl Acad Sci U S A 110:9481–9486. https://doi.org/10.1073/pnas.1307473110. 69. Lorenzo MM, Sanchez-Puig JM, Blasco R. 2019. Genes A27L and F13L as genetic markers for the isolation of recombinant vaccinia virus. Sci Rep 9: 15684. https://doi.org/10.1038/s41598-019-52053-4. 70. Niwa H, Yamamura K, Miyazaki J. 1991. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108:193–199. https://doi.org/10.1016/0378-1119(91)90434-d. 71. Nogales A, Aydillo T, Avila-Perez G, Escalera A, Chiem K, Cadagan R, DeDiego ML, Li F, Garcia-Sastre A, Martinez-Sobrido L. 2019. Functional characterization and direct comparison of influenzaA,B,C,andDNS1 proteins in vitro and in vivo. Front Microbiol 10:2862. https://doi.org/10 .3389/fmicb.2019.02862. 72. DiPiazza A, Nogales A, Poulton N, Wilson PC, Martinez-Sobrido L, Sant AJ. 2017. Pandemic 2009 H1N1 Influenza Venus reporter virus reveals broad diversity of MHC class II-positive antigen-bearing cells following infection in vivo. Sci Rep 7:10857. https://doi.org/10.1038/s41598-017-11313-x. 73. Hilimire TA, Nogales A, Chiem K, Ortego J, Martinez-Sobrido L. 2020. Increasing the safety profile of the master donor live attenuated influenza vaccine. Pathogens 9:86. https://doi.org/10.3390/pathogens9020086. 74. National Research Council (U.S.). Committee for the Update of the Guide for the Care and Use of Laboratory Animals., Institute for Laboratory Animal Research (U.S.), National Academies Press (U.S.). 2011. Guide for the care and use of laboratory animals, 8th ed National Academies Press, Washington, DC. Attenuation of DNA Viruses by Codon Deoptimization Microbiology Spectrum Month YYYY Volume XX Issue XX 10.1128/spectrum.00272-22 16 Downloaded from https://journals.asm.org/journal/spectrum on 17 June 2022 by 193.144.96.41.