Directed Evolution of an Influenza Reporter Virus To Restore Replication and Virulence and Enhance Noninvasive Bioluminescence Imaging in Mice

Abstract

Reporter viruses provide a powerful tool to study infection, yet incorporating a nonessential gene often results in virus attenuation and genetic instability. Here, we used directed evolution of a luciferase-expressing pandemic H1N1 (pH1N1) 2009 influenza A virus in mice to restore replication kinetics and virulence, increase the bioluminescence signal, and maintain reporter gene expression. An unadapted pH1N1 virus with NanoLuc luciferase inserted into the 5' end of the PA gene segment grew to titers 10-fold less than those of the wild type in MDCK cells and in DBA/2 mice and was less virulent. For 12 rounds, we propagated DBA/2 lung samples with the highest bioluminescence-to-titer ratios. Every three rounds, we compared in vivo replication, weight loss, mortality, and bioluminescence. Mouse-adapted virus after 9 rounds (MA-9) had the highest relative bioluminescence signal and had wild-type-like fitness and virulence in DBA/2 mice. Using reverse genetics, we discovered fitness was restored in virus rPB2-MA9/PA-D479N by a combination of PA-D479N and PB2-E158G amino acid mutations and PB2 noncoding mutations C1161T and C1977T. rPB2-MA9/PA-D479N has increased mRNA transcription, which helps restore wild-type-like phenotypes in DBA/2 and BALB/c mice. Overall, the results demonstrate that directed evolution that maximizes foreign-gene expression while maintaining genetic stability is an effective method to restore wild-type-like in vivo fitness of a reporter virus. Virus rPB2-MA9/PA-D479N is expected to be a useful tool for noninvasive imaging of pH1N1 influenza virus infection and clearance while analyzing virus-host interactions and developing new therapeutics and vaccines.IMPORTANCE Influenza viruses contribute to 290,000 to 650,000 deaths globally each year. Infection is studied in mice to learn how the virus causes sickness and to develop new drugs and vaccines. During experiments, scientists have needed to euthanize groups of mice at different times to measure the amount of infectious virus in mouse tissues. By inserting a foreign gene that causes infected cells to light up, scientists could see infection spread in living mice. Unfortunately, adding an extra gene not needed by the virus slowed it down and made it weaker. Here, we used a new strategy to restore the fitness and lethality of an influenza reporter virus; we adapted it to mouse lungs and selected for variants that had the greatest light signal. The adapted virus can be used to study influenza virus infection, immunology, and disease in living mice. The strategy can also be used to adapt other viruses.

Keywords: bioluminescence; influenza virus; mouse adaptation; noninvasive imaging; reporter virus.

FIG 1

rTN09-PA-Nluc H1N1 influenza virus is attenuated in MDCK cells and mice. (A) Schematic representation of the PA-Nluc cDNA. The codon swap (cs) region, self-cleaving 2A peptide, NanoLuc luciferase (Nluc), and packaging sequence (ps) are indicated. (B) Plaque morphologies of WT and rTN09-PA-Nluc after 2 days of infection in MDCK cells at 37°C. (C) Multistep replication kinetics of WT and rTN09-PA-Nluc in MDCK cells. MDCK cells were infected at an MOI of 0.001, and culture supernatants were collected at the indicated times and titrated by plaque assay in MDCK cells. (D and E) Body weight changes (D) and survival (E) were monitored daily for mice intranasally inoculated with 30 μl PBS containing 750 PFU of WT or rTN09-PA-Nluc. (F) Nasal turbinates, trachea, and lungs were harvested from the mice 4 days postinfection (dpi), and virus titers in tissue homogenates were determined by plaque assay. The reported values are means and standard deviations (n = 5). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 2

Mouse-adapted reporter viruses restore virus replication and virulence in mice. Groups of 5 DBA/2 mice were intranasally inoculated with 30 μl PBS containing 750 PFU of virus. (A and B) Body weight changes (A) and survival (B) were observed for 14 days in intact mice. (C) Virus tissue titers in nasal turbinates, trachea, and lungs were measured 4 dpi. (D) Noninvasive bioluminescence intensities in the lungs were observed for 5 days in intact mice. (E) The bioluminescence/titer ratios at 4 dpi were calculated for the four mouse-adapted (MA) variants. The reported values are means and standard deviations (n = 5). *, P < 0.05; **, P < 0.01.

FIG 3

Mouse-adapted virus MA9-22 has restored fitness in DBA/2J mice. Groups of 5 DBA/2 mice were intranasally inoculated with 30 μl PBS containing 750 PFU of virus. (A and B) Body weight changes (A) and survival (B) were observed for 14 days in intact mice. (C) Virus tissue titers in nasal turbinates, trachea, and lungs were measured 4 dpi. (D) Noninvasive bioluminescence intensities in the lungs were observed for 4 days in intact mice. The ratio of bioluminescence/titer at 4 d.p.i. was calculated for the four mouse-adapted (MA) variants. In all panels, the reported values are means ± standard deviations (n = 5) and symbols are PBS (white circles), WT (black circles), MA-9 (pink squares), MA9-1 (brown circles), MA9-2 (yellow triangles), MA9-15 (orange diamonds), and MA9-22 (blue triangles). ***, P < 0.001.

FIG 4

MA9-22 mutations in PA and PB2 genes and their influence on MDCK replication kinetics and reporter insert stability. (A and B) Nonsynonymous amino acid mutations and synonymous nucleotide substitutions in MA9-22 PA (A) and PB2 (B) genes. Nucleotide mutations are listed above the genes, and amino acid mutations are below. (C) Multistep replication kinetics of WT, rPB2-MA9, and rPB2-MA9/PA-D479N in MDCK cells infected at an MOI of 0.01 PFU/cell. The reported values are means and standard deviations of the results of 3 independent experiments. *, P < 0.05; ***, P < 0.001. (D) PA-Nluc gene stability after multiple passages in cell culture (MDCK or A549 cells) and after passage in cell culture followed by infection in DBA/2 mice. RT-PCR was performed on RNA isolated from WT stock virus, three viral stocks after passage in MDCK and A549 cells (P5, P10, and P15), and day 4 mouse lung homogenates of virus previously passaged 5 times in MDCK cells (Mouse, left) or A549 cells (Mouse, right). The PCR product was analyzed by 0.7% agarose gel electrophoresis using primers that flank the reporter gene. M, molecular weight DNA ladder.

FIG 5

In vivo replication, bioluminescence, and virulence of rPB2-MA9 and rPB2-MA9/PA-D479N in DBA/2 mice. Groups of 5 DBA/2 mice were intranasally inoculated with 30 μl PBS containing 750 PFU of virus, and experiments were repeated. (A and B) Body weight changes (A) and survival (B) were observed for 14 days in intact mice. (C) Virus tissue titers in nasal turbinates, trachea, and lungs were measured 4 dpi. (D) Noninvasive bioluminescence intensities in the lungs were observed for 4 days in intact mice. The reported values are means and standard deviations. (A and B) The data are cumulative from three independent experiments (n = 15). (C and D) The data are from two experiments (n = 10). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 6

In vivo replication, bioluminescence, and weight loss of rPB2-MA9/PA-D479N in BALB/c mice. Groups of 5 BALB/c mice were intranasally inoculated with 30 μl PBS containing 5,000 PFU of virus. (A) Body weight changes were monitored for 14 days. All the animals survived infection. (B) Virus tissue titers in nasal turbinates, trachea, and lungs were measured 4 dpi. (C) Noninvasive bioluminescence intensities in the lungs were observed for 4 days in intact mice. The reported values are means and standard deviations. ***, P < 0.001.

FIG 7

PB2 and PA mutational effects on viral polymerase activity by minigenome assay. 293T cells were transfected with plasmids encoding PA, PB1, PB2, and NP, together with pPolI-358Luc and pβ-gal. Wild-type PB1 and NP plasmids from A/TN/1-560/09 were used in each group, while PB2 and PA plasmids carrying mutants identified in MA9-22 were varied. (A and B) After 24-h incubation at 33°C (A) and 37°C (B), both luciferase and β-Gal expression levels were measured. β-Gal expression was used to normalize the data. R.L.U., relative light units. (C) PB2 mutants carrying nonsynonymous and/or synonymous mutations from MA9-22 were compared at 37°C. The reported values are means and standard deviations from 3 independent experiments (n = 9). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 8

Kinetics of vRNA, cRNA, and mRNA production by rPB2-MA9 and rPB2-MA9/PA-D479N. 293T cells were infected with the wild type, rPB2-MA9, and rPB2-MA9/PA-D479N at an MOI of 3.0 PFU/cell. Total RNA was isolated at 3, 6, 9, and 12 h postinfection. vRNA (A), cRNA (B), and mRNA (C) were quantified by real-time RT-PCR using specific primers annealing to the NP gene. The reported values are means and standard deviations of the results of 3 independent experiments (n = 9) for the WT, rPB2-MA9, and rPB2-MA9/PA-D479N. ***, P < 0.001. (D) PB2 protein expression quantified by Western blotting. 293T cells were infected at an MOI of 3 PFU/cell with WT (lanes 1), rTN09-PA-Nluc (lanes 2), rPB2-MA9/PA-D479N (lanes 3), or rPB2-MA9 (lanes 4) and maintained at 37°C. Cell lysates were harvested at 3, 6, 9, and 12 h postinfection (hpi) and analyzed by Western blotting. β-Actin was used as a loading control. Densitometric values, shown under the 9- and 12-hpi gel, were normalized to the intensity of the WT at 12 hpi. M, mock-infected cell lysate.

FIG 9

Polymerase structure and locations of mouse-adaptive mutations. Shown are front (A) and back (B) views of the polymerase complex from A/little yellow-shouldered bat/Guatemala/060/2010 (H17N10) (PDB accession no. 4WSB) (41). The proteins are identified by maintaining the original color conventions for PA (green), PB2 (orange), and PB1 (cyan). The PA endonuclease, PB2 cap-binding, PB2 linker, PB2-627, PB2 lid, and PB1 catalytic-site domains are labeled. Residues PA-D479 (479) and PB2-E158 (158) are the only residues for which sidechains are shown. 5′ (magenta) and 3′ (yellow) viral RNAs are also shown. PA-D479 is positioned at the bottom of a rim that funnels up into the polymerase channel. PB2-E158 is located in an α-helical lid that directs newly captured caps down into the PB1 catalytic site and prevents newly synthesized RNA from being cleaved by the PA endonuclease domain when it exits the top of the polymerase.