Publications

Post-translational modifications, including ubiquitination, have emerged as important regulators of viral infection and the host innate immune response. We screened the proteins of SARS-CoV-2 for ubiquitination and identified the RNA-dependent RNA polymerase (RdRp) NSP12 as being ubiquitinated, which has not previously been reported. Importantly, we confirmed that NSP12 is ubiquitinated during SARS-CoV-2 infection, and the NSP12 proteins of different coronaviruses are differentially associated with ubiquitin chains. SARS-CoV-2 NSP12 was primarily associated with K63-linked polyubiquitin chains, which increased SARS-CoV-2 NSP12 stability relative to OC43 NSP12, which was associated with K48 chains. Additionally, we identify the atypical E3 ubiquitin ligase MYCBP2 as a mediator of serine ubiquitination of SARS-CoV-2 NSP12 on serine-564 and observe that MYCBP2 promotes SARS-CoV-2 replication in cell-based assays. Substitution of this single residue (S564A) in CoV-2 NSP12 prevented recovery of infectious virus in three independent attempts, precluding direct genetic validation of its role in viral replication. To our knowledge, these findings provide the first evidence of serine ubiquitination of a viral protein and suggest a previously unrecognized role for RdRp ubiquitination in coronavirus biology.IMPORTANCEViruses often hijack host post-translational modification systems, including ubiquitination, to enhance the functions of their proteins and evade host innate immune responses. We screened the viral proteins of SARS-CoV-2 and found that NSP2, NSP5, NSP12, NSP15, and NSP16 are ubiquitinated. Focusing on NSP12, the RNA-dependent RNA polymerase (RdRp), we found that it is associated with multiple types of ubiquitin modifications, and RdRp proteins from other coronaviruses are differentially ubiquitinated. We also found that SARS-CoV-2 RdRp undergoes ubiquitination on a serine, rather than a canonical lysine residue, mediated by the atypical E3 ubiquitin ligase, MYCBP2. Consistent with this, MYCBP2 promotes efficient SARS-CoV-2 replication in cell-based assays. To our knowledge, this study provides the first evidence of non-lysine ubiquitination of a viral protein and demonstrates the serine selectivity of MYCBP2 for the first time in a cellular context. These findings raise important questions about the regulatory function of serine NSP12 ubiquitination in coronavirus replication and its potential contribution to viral pathogenicity.

Reporter-expressing recombinant influenza A viruses (IAVs) have emerged as powerful tools for studying viral infection, host-pathogen interactions, and the identification and characterization of prophylactic and/or therapeutic countermeasures for the treatment of IAV infections. By incorporating a reporter gene, such as fluorescent, recombinase, or luciferase proteins, into the viral genome, these genetically engineered recombinant IAVs enable real-time visualization and quantification of infection in vitro, ex vivo, and in vivo to detect the presence of IAV in infected cells or animal models of infection. Among the various approaches to express reporter genes from the viral genome, fusion of fluorescent or luciferase proteins to the C-terminus of the non-structural 1 (NS1) protein has been successfully used to generate replication-competent reporter-expressing IAVs. This minireview summarizes the strategies used to generate IAV-expressing reporter genes, particularly from the non-structural (NS) viral segment 8, and their use in real-time tracking of viral replication and pathogenesis. These replication-competent reporter-expressing IAVs also provide an efficient tool for high-throughput screening (HTS) of novel anti-IAV therapeutics. We also discuss the advantages and limitations of current fluorescent and luciferase reporter-expressing IAVs to track viral infections in cultured cells and/or animal models, and provide perspectives on future directions in this rapidly advancing field.

Coronaviruses (CoVs) replicate unusually large RNA genomes that necessitate proofreading by the 3'-to-5' exoribonuclease (ExoN) formed by nonstructural proteins 14 (nsp14) and 10 (nsp10). Previous studies suggested that inactivation of the ExoN catalytic site in severe acute respiratory syndrome CoV 2 (SARS-CoV-2) is lethal, leaving unresolved whether the virus can tolerate impaired proofreading activity. Here, we investigated the functional requirement for ExoN in SARS-CoV-2 replication by combining a continuous fluorescence-based biochemical assay with an optimized single-bacmid reverse genetics system. Mutational analysis of residues involved in RNA binding or catalysis revealed graded effects on ExoN activity in vitro. Alanine substitution of Lys9, a residue positioned near the RNA-binding interface, did not reduce ExoN activity, whereas charge reversal at this position (K9E) impaired activity more strongly than alanine substitutions of the catalytic motif I residues D90 and E92 (D90A/E92A). Correspondingly, recombinant SARS-CoV-2 carrying K9A was readily recovered, whereas the D90A/E92A mutant was recovered only after an extended delay, and K9E could not be rescued despite repeated attempts. The D90A/E92A mutant exhibited reduced replication while maintaining the engineered ExoN substitutions during serial passage. Deep sequencing of viral populations revealed a marked increase in genome-wide sequence variation in the D90A/E92A mutant, demonstrating a stable mutator phenotype. Together, these findings indicate that SARS-CoV-2 can tolerate substantial impairment of ExoN activity but depends on a minimal activity threshold for viability. This system provides a platform for defining how SARS-CoV-2 proofreading controls genome stability, viral fitness, and sensitivity to antiviral strategies that exploit reduced replication fidelity.IMPORTANCECoronaviruses have unusually large RNA genomes because they encode a proofreading enzyme that removes copying errors during replication. It has been unclear whether SARS-CoV-2 can survive when this proofreading function is strongly weakened because earlier studies suggested that loss of the enzyme's catalytic activity is lethal. We show that SARS-CoV-2 can tolerate substantial impairment of proofreading but only when residual exonuclease activity remains above a minimal threshold. A virus with impaired proofreading replicates less efficiently and accumulates mutations across its genome, whereas a more severe defect prevents virus recovery. These findings clarify how coronavirus proofreading balances genome stability with viral fitness and provide a useful system for studying how reduced replication fidelity affects viral evolution, antiviral sensitivity, and attenuation. Defining this activity threshold may also help guide antiviral strategies that target coronavirus proofreading.

SARS-CoV-2, like many viruses, generates syncytia but the role of syncytia formation in viral evolution remains unknown. Using SARS-CoV-2 and SARS-CoV-2 Spike (S) replacement vesicular stomatitis (VSV), we show that S-mediated syncytia impair the antiviral effects of interferons in cultured cells, human lung cell cultures, and hACE2 transgenic mice. Amino acid substitutions that modulate syncytia formation in Delta- and Omicron-encoded S have parallel effects on viral interferon resistance. S-mediated syncytia compromise antibody-mediated virus neutralization in cultured cells. We recapitulate interferon and neutralizing antibody resistance in syncytia generated by the orthoreovirus p14 fusion-associated small transmembrane (FAST) protein in VSV, influenza virus, and seasonal coronavirus OC43 infections. These findings explain selection of SARS-CoV-2 fusogenic variants in humans and, more generally, the evolution of fusogenic viruses driven by adaptive and innate immunity.

UNLABELLED: SARS-CoV-2 is the causative agent of COVID-19. The ancestral SARS-CoV-2 Washington-1 (WA1) strain does not infect standard laboratory mouse strains, necessitating the use of mouse-adapted (MA) viruses. A MA SARS-CoV-2, SARS-CoV-2-N501Y MA30 (hereafter MA30), has been developed to allow infection of wild-type (WT) mice. However, SARS-CoV-2 MA30 cannot be tracked in vitro, ex vivo, or in vivo. To address this problem, we generated recombinant (r)SARS-CoV-2 MA30 expressing fluorescent (mCherry) and nanoluciferase (Nluc) reporter genes, alone or in combination, that enable tracking viral infections in WT C57BL/6 and BALB/c mice. Insertion of the reporter genes resulted in minor viral attenuation in vitro, with  0.5-1.0 log lower titers than rSARS-CoV-2 MA30 WT in A549 hACE2 cells, while maintaining similar plaque morphology and replication kinetics in Vero AT cells. In vivo, reporter-expressing rSARS-CoV-2 MA30 caused transient weight loss, contrasting with lethal rSARS-CoV-2 MA30 WT infection. Bioluminescence imaging of rSARS-CoV-2 MA30 Nluc in C57BL/6 and BALB/c mice revealed peak pulmonary replication at 2 days post-infection, with resolution by day 4, and correlated with tissue viral loads. Our results demonstrate the feasibility of using rSARS-CoV-2 MA30 expressing reporter genes to track viral infection in vitro, ex vivo, and in vivo without a need for secondary approaches to monitor viral infection required for rSARS-CoV-2 MA30 WT. Our system is highly suitable to evaluate prophylactic vaccines and therapeutic antibodies or antiviral approaches in WT or transgenic C57BL/6 and BALB/c mice without the shortcomings of K18-hACE2 mice and with the added advantage of non-invasive monitoring of treatment efficacy.

IMPORTANCE: Mouse-adapted (MA) SARS-CoV-2 that infect wild-type (WT) mice are critical tools for preclinical studies. While the previously described SARS-CoV-2-N501Y MA30 enables infection of WT mice, it does not allow non-invasive tracking of viral infections. Recombinant viruses expressing reporter genes enable real-time monitoring of infection dynamics, opening an avenue to study viral tropism and easily evaluate prophylactic and therapeutic approaches. They furthermore support longitudinal studies, which reduces the number of research animals required. Here, we show that a recombinant (r)SARS-CoV-2 expressing fluorescent (mCherry) and nanoluciferase (Nluc) reporter genes, alone or in combination, can be used to track viral infections in vitro, ex vivo, and in vivo without the need for secondary approaches that are required to detect SARS-CoV-2 MA30 in WT mice. These reporter-expressing rSARS-CoV-2 MA30 may accelerate vaccine development and antiviral drug discovery in WT or transgenic mice bypassing the need for hACE2 overexpression in K18-hACE2 transgenic mice.

Seasonal influenza viruses, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and respiratory syncytial virus (RSV) drive substantial mortality worldwide. Vaccines against these respiratory viruses are traditionally administered separately, and, while some emerging technologies have shown promise to address multiple pathogens simultaneously, subunit protein vaccines have lagged behind. Here, we describe a nanoliposome-based vaccine platform that codisplays recombinant hemagglutinin ectodomains from three seasonal influenza strains, the SARS-CoV-2 receptor binding domain, and the RSV F ectodomain, in prefusion form. In mouse, ferret, and cotton rat models, immunization elicited protective immunity against the three major respiratory viruses, with antibody responses comparable to those induced by the corresponding monovalent formulations. These findings highlight the potential for multivalent recombinant protein vaccines delivered via nanoliposome carriers to provide immunologic protection against a broad spectrum of respiratory viruses.

Coronaviruses use discontinuous transcription to generate subgenomic RNAs (sgRNAs) that encode structural and accessory proteins. However, the factors regulating sgRNA abundance in SARS-CoV-2 remain unclear. Here, we combined strand-specific RNA sequencing, RNA-RNA interaction mapping, prediction of RNA folding energies, and targeted mutagenesis to define the regulation of (-) sgRNA synthesis in SARS-CoV-2 infection. We demonstrated that the relative (-) sgRNA abundance across viral genes is stable throughout infection and largely correlates with corresponding (+) sgmRNA levels. Through meta-analysis of published SPLASH data, we found that the frequency of long-range interactions between the 5' genomic transcription regulatory sequence TRS-Leader and downstream TRS-Body sequences correlates with sgRNA abundance. Notably, the folding energy (ΔG) of these duplexes quantitatively predicts (-) sgRNA transcript levels. Mutations in non-coding regulatory regions that altered the ΔG resulted in corresponding changes in (-) sgRNA expression, suggesting a causal role for TRS duplex stability in transcriptional regulation. Analysis of naturally occurring mutations near regulatory sites further suggests that modulation of duplex stability may also serve as an evolutionary mechanism to fine-tune viral gene expression. Together, our findings identify the pairing stability of TRS-Leader:TRS-Body as a determinant of discontinuous transcription and reveal how RNA pairing potential contributes to the regulation of (-) sgRNA synthesis in SARS-CoV-2.

UNLABELLED: Efficient recovery of negative-stranded RNA viruses from plasmid cDNA requires robust bacteriophage T7 RNA polymerase expression, commonly supplied by Modified Vaccinia Ankara expressing T7 polymerase (MVA T7). Here, we describe the generation and characterization of two fluorescent MVA T7 viruses expressing either enhanced green fluorescent protein (EGFP) or monomeric red fluorescent protein 1 (mRFP1). Both fluorescent protein-expressing MVA T7 viruses retain plaque morphology, growth kinetics, and transgene expression comparable to parental MVA T7. Importantly, both fluorescent protein-expressing MVA T7 viruses support transient protein expression of foreign genes and rescue of replication-competent recombinant vesicular stomatitis virus (rVSV), while enabling direct visualization of MVA T7 infection. Fluorescent protein expression from MVA T7 viruses facilitates stock generation and titration and allows detection of the remaining helper virus during rVSV recovery and amplification. Altogether, EGFP- and mRFP1-expressing MVA T7 viruses provide a practical tool for monitoring T7-driven rescue of recombinant RNA viruses and identifying helper virus carryover.

IMPORTANCE: Reverse genetics systems for negative-stranded RNA viruses rely on robust bacteriophage T7 RNA polymerase expression, commonly provided by a Modified Vaccinia Ankara virus expressing T7 polymerase (MVA T7). However, generation of virus stocks, monitoring MVA T7 helper virus infection, and detecting residual virus during rescue of negative-stranded RNA virus can be challenging. We generated fluorescent protein-expressing MVA T7 viruses that enable direct visualization of helper virus infection without compromising viral growth kinetics, transgene expression, or efficient rescue of recombinant vesicular stomatitis virus (rVSV), while enabling direct visualization of MVA T7 infection. MVA T7 viruses expressing fluorescent proteins simplify stock generation and titration, facilitate optimization of negative-stranded RNA rescue conditions, and allow rapid identification of helper virus carryover during generation and amplification of recombinant virus, improving the efficiency, reproducibility, and quality control in reverse genetics systems relying on T7 expression, widely used across virology.

Ensitrelvir is an antiviral drug that specifically targets the conserved main protease (Mpro) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, mutations in Mpro could confer resistance to antivirals, including ensitrelvir. Thus, identifying SARS-CoV-2 drug-resistant mutants and elucidating their mechanisms of resistance are critical for guiding the selection of effective antiviral therapies. Here, we utilized a recombinant luminescent attenuated SARS-CoV-2 lacking the open reading frames (ORF) 3a and 7b proteins (Δ3a7b-Nluc WT) to safely identify ensitrelvir drug-resistant mutants (DRM-E) without the need of using virulent forms of SARS-CoV-2. We isolated a DRM-E containing a Mpro G23 deletion (G23del) with high resistance ( 1,000-fold) to ensitrelvir, but not to the Mpro inhibitor nirmatrelvir or to the RNA-dependent RNA polymerase (RdRp) inhibitor remdesivir. The contribution of G23del in ensitrelvir resistance was confirmed by generating a Δ3a7b-Nluc containing G23del in Mpro (Δ3a7b-Nluc G23del). Δ3a7b-Nluc G23del exhibited resistance to ensitrelvir in both cultured cells and in K18 hACE2 transgenic mice. Binding affinity revealed that the G23del mutation altered ensitrelvir, but not nirmatrelvir, binding to Mpro. Notably, while Δ3a7b-Nluc G23del was affected in viral fitness, serial passage of Δ3a7b-Nluc G23del in the absence of ensitrelvir resulted in the emergence of substitution L50F in Mpro that restored viral fitness loss caused by G23del without altering resistance to ensitrelvir. Our results demonstrate that G23del in Mpro can confer resistance to ensitrelvir. Positively, G23del in Mpro does not render SARS-CoV-2 resistant to nirmatrelvir or remdesivir, suggesting the feasibility of treating SARS-CoV-2 infections containing G23del Mpro with other approved antivirals.IMPORTANCEThe clinical use of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antiviral drugs is increasingly challenged by the emergence of drug-resistant mutants. Thus, there is a pressing need to identify and characterize antiviral escape SARS-CoV-2 variants, particularly for Food and Drug Administration-approved antivirals. Our study addresses this by employing a luminescent attenuated SARS-CoV-2 platform (Δ3a7b-Nluc wild type [WT]) to safely identify and characterize resistance mutations without the concern of using virulent forms of SARS-CoV-2. Using this safe approach, we identified a G23 deletion (G23del) in SARS-CoV-2 Mpro, which mediates resistance to ensitrelvir in vitro and in vivo. Importantly, while G23del was able to confer  1,000-fold increased resistance to ensitrelvir, a Δ3a7b-Nluc containing G23del remained sensitive to other Mpro (nirmatrelvir) and RdRp (remdesivir) inhibitors. Altogether, this study demonstrates the feasibility of using Δ3a7b-Nluc WT to safely identify and characterize drug-resistant viruses without the biosafety concern of using virulent WT SARS-CoV-2 and advance the design of next-generation antiviral drugs.

Coronaviruses (CoVs) replicate unusually large RNA genomes that necessitate proofreading by the 3'-to-5' exoribonuclease (ExoN) formed by nonstructural proteins 14 (nsp14) and 10 (nsp10). Previous studies suggested that inactivation of the ExoN catalytic site in severe acute respiratory syndrome CoV 2 (SARS-CoV-2) is lethal, leaving unresolved whether the virus can tolerate impaired proofreading activity. Here, we investigated the functional requirement for ExoN in SARS-CoV-2 replication by combining a continuous fluorescence-based biochemical assay with an optimized single-bacmid reverse genetics system. Mutational analysis of residues involved in RNA binding or catalysis revealed graded effects on ExoN activity in vitro. Alanine substitution of Lys9, a residue positioned near the RNA-binding interface, did not reduce ExoN activity, whereas charge reversal at this position (K9E) impaired activity more strongly than alanine substitutions of the catalytic motif I residues D90 and E92 (D90A/E92A). Correspondingly, recombinant SARS-CoV-2 carrying K9A was readily recovered, whereas the D90A/E92A mutant was recovered only after an extended delay and K9E could not be rescued despite repeated attempts. The D90A/E92A mutant exhibited reduced replication while maintaining the engineered ExoN substitutions during serial passage. Deep sequencing of viral populations revealed a marked increase in genome-wide sequence variation in the D90A/E92A mutant, demonstrating a stable mutator phenotype. Together, these findings indicate that SARS-CoV-2 can tolerate substantial impairment of ExoN activity but depends on a minimal activity threshold for viability. This system provides a platform for defining how SARS-CoV-2 proofreading controls genome stability, viral fitness, and sensitivity to antiviral strategies that exploit reduced replication fidelity.