Fluorescent and bioluminescent bovine H5N1 influenza viruses for evaluation of antiviral interventions

Abstract

In early 2024, a clade 2.3.4.4b high pathogenic H5N1 avian influenza virus was detected in dairy cows and humans in the United States. Since then, it has spread to herds in at least 13 states and caused symptomatic disease in at least fifteen people. To facilitate rapid testing of existing and novel countermeasures, here, we report the development of an H5N1 viral reverse genetic system, its use to produce fluorescent and bioluminescent variant strains, and their utility in high-throughput evaluation of antiviral interventions.

Keywords: NanoLuc; drug testing; influenza viruses; sfGFP.

Fig 1

Generation of bovine H5N1 bioluminescent and fluorescent reporter viruses. (A) Schematic depicting modification of segment 4 to express NanoLuc or sfGFP. PS = packaging signals, 2A = PTV-2A ribosome skipping motif, and mPS = silently mutated packaging signals. (B) Schematic showing the post-translational processing of the HA protein (SPP = signal peptide peptidase). (C) Schematic depicting bovine H5N1 rescue, purification, and stock production strategy. (D) Representative cryo-electron micrographs of wildtype (WT) H5N1 virions. The bilayer membrane of virus particles (white arrows) and surface glycoproteins (black arrowheads) is indicated (Scale bar = 50 nm). (E) Growth curve of rescued A/Texas/37/2024 viruses in Madin-Darby canine kidney (MDCK) cells. Virus quantified using hemagglutination assay. Statistical analyses completed with respect to wild type (multiplicity of infection = 0.0005, n = 5). (F) Endpoint titer of A/Texas/37/2024 viruses in 11-day-old embryonated chicken eggs (1,000 plaque-forming units (PFU)/egg, 24 hours post-infection (HPI), n = 5 eggs). (G) Reverse transcription PCR (RT-PCR) of segment 4 from A/Texas/37/2024 reporter-HA viruses or a wild type control at passages 0 or 3 through MDCK cells. Viral segment 4 diagrams are provided for reference and are not to scale. Arrows denote expected RT-PCR amplicon length (blue = wild type, orange = NanoLuc HA, and green = sfGFP HA). Statistical analyses performed using a two-tailed Mann-Whitney U test. * indicates P < 0.05 and ns = not significant. Error bars indicate SE measurement. Data shown are representative of at least two independent experiments except for 1D which was performed once.

Fig 2

Bovine H5N1 reporter viruses facilitate testing of antiviral agents. (A) Luminescence quantified from A549 cells infected with NanoLuc-HA reporter virus (MOI = 0.01, 24 HPI, n = 4). (B) Comparison of luminescence signal or HA viral RNA abundance during infection with NanoLuc-HA reporter virus in A549 cells (8 HPI, n = 4). (C) Related to panel B, correlation of luminescence signal to viral RNA abundance in infected A549 cells with linear regression analysis (MOIs range from 0.004 to 0.25). Data are normalized to relative light units (RLU) or HA RNA at MOI = 0.5. Mean value across n = 4 replicates is plotted. (D) Fluorescence microscopy timecourse for sfGFP and native HA protein in A549 cells infected with sfGFP-HA reporter virus (MOI = 0.04, scale bar is 100 µm). (E) Quantification of sfGFP-positive infected area during the microscopy timecourse (n = 10 images/timepoint across two independent experiments). (F) Quantification of signal overlap between sfGFP and HA in infected cells (10 HPI, n = 253 cells. 24 HPI, n = 291 cells. Cells were quantified across two independent experiments). (G) Dose response of A/Texas/37/2024 wild type and reporter viruses to recombinant human IFNβ and baloxavir in A549 cells. Relative infection was quantified via immunofluorescence microscopy for HA protein for wild-type virus, luminescence signal for NanoLuc-HA reporter virus, and fluorescence microscopy for sfGFP-HA reporter virus (MOI = 0.5, 24 HPI, n = 4 across two independent experiments). (H) Related to panel G, calculated IC₅₀ values for IFNβ and baloxavir across two independent experiments for wild-type A/Texas/37/2024 and segment 4 reporter viruses. Statistical analyses performed using a two-tailed Mann-Whitney U test. * indicates P < 0.05 and ns = not significant. Error bars indicate SE measurement. Panels A–D are representative of two independently conducted experiments and all collected data are displayed on panels E-H.