Development of high-yield influenza B virus vaccine viruses

Abstract

The burden of human infections with influenza A and B viruses is substantial, and the impact of influenza B virus infections can exceed that of influenza A virus infections in some seasons. Over the past few decades, viruses of two influenza B virus lineages (Victoria and Yamagata) have circulated in humans, and both lineages are now represented in influenza vaccines, as recommended by the World Health Organization. Influenza B virus vaccines for humans have been available for more than half a century, yet no systematic efforts have been undertaken to develop high-yield candidates. Therefore, we screened virus libraries possessing random mutations in the six "internal" influenza B viral RNA segments [i.e., those not encoding the major viral antigens, hemagglutinin (HA) and neuraminidase NA)] for mutants that confer efficient replication. Candidate viruses that supported high yield in cell culture were tested with the HA and NA genes of eight different viruses of the Victoria and Yamagata lineages. We identified combinations of mutations that increased the titers of candidate vaccine viruses in mammalian cells used for human influenza vaccine virus propagation and in embryonated chicken eggs, the most common propagation system for influenza viruses. These influenza B virus vaccine backbones can be used for improved vaccine virus production.

Keywords: Victoria lineage; Yamagata lineage; high-yield; influenza B virus; vaccine.

Fig. 1.

Growth kinetics and hemagglutination titers of high-yield Yamagata- and Victoria-lineage viruses. (A) We compared the indicated Yamagata-lineage wild-type virus with the Yamagata-lineage high-yield candidate RG(Yam) #8 (SI Appendix, Table S3) and with RG(Yam) #8 possessing the PA-a2272t mutation; the latter virus was selected as lead candidate HY(Yam). (B) We compared the indicated Victoria-lineage wild-type virus with the Victoria-lineage high-yield candidate RG(Vic) #2 (SI Appendix, Table S4) and with RG(Vic) #2 possessing the PA-a2272t mutation; the latter virus was selected as lead candidate HY(Vic). In both sets of experiments, MDCK cells were infected in triplicate with the indicated viruses at an MOI of 0.001 and incubated at 35 °C. At the indicated time points, virus and hemagglutination titers were determined by performing plaque or hemagglutination assays, respectively. The values presented are the average of three independent experiments ± SD. P values were calculated by using the linear mixed model described in Materials and Methods (*P < 0.05; **P < 0.01). Red and blue asterisks indicate the comparison of the respective virus with WT virus; beige asterisks indicate the comparison between the viruses depicted in red and blue. HA titer, hemagglutination titer.

Fig. 2.

Comparison of wild-type and high-yield viruses possessing different HA and NA vRNAs. Viruses possessing the HA and NA vRNAs of the indicated viruses in combination with the internal vRNA segments of the respective natural wild-type isolate (WT), or of HY(Yam) (A–C) or HY(Vic) (D–F) (i.e., the viruses indicated by the black graphs possess the eight wild-type vRNA segments of a human influenza B virus isolate). Experiments were carried out as described in the legend to Fig. 1. The values presented are the average of three independent experiments ± SD. P values were calculated by using the linear mixed model described in the Methods section (*P < 0.05; **P < 0.01). Red asterisks indicate the comparison of the respective virus with WT virus. HA titer, hemagglutination titer.

Fig. 3.

Exchange of HY(Yam) and HY(Vic) backbones. (A and B) Comparison of the virus and hemagglutination titers of two wild-type Yamagata-lineage viruses with viruses possessing the same HA and NA vRNAs in combination with the internal genes of HY(Yam) or HY(Vic). (C and D) Comparison of the virus and hemagglutination titers of two wild-type Victoria-lineage viruses with viruses possessing the same HA and NA vRNAs in combination with the internal genes of HY(Yam) or HY(Vic). Experiments were carried out as described in the legend to Fig. 1. The values presented are the average of three independent experiments ± SD. P values were calculated by using the linear mixed model described in Materials and Methods (*P < 0.05; **P < 0.01). Red and blue asterisks indicate the comparison of the respective virus with WT virus; beige asterisks indicate the comparison between the viruses depicted in red and blue. HA titer, hemagglutination titer.

Fig. 4.

Comparison of high-yield influenza A and B vaccine virus backbones. (A and B) We compared the virus yield and hemagglutination titers of: (i) the indicated wild-type viruses; (ii) viruses possessing the indicated HA and NA vRNAs in combination with the internal genes of HY(Yam); and (iii) viruses possessing the indicated type A/B chimeric HA and NA vRNAs in combination with the internal genes of high-yield influenza A virus. (C and D) Similar experiments were carried out for viruses of the Victoria lineage. (E and F) Comparison of the indicated wild-type and hybrid viruses in embryonated chicken eggs. Experiments were carried out as described in the legend to Fig. 1. The values presented are the average of three independent experiments ± SD. The statistical significance was determined by using the linear mixed model described in Materials and Methods (A–D), or by two-way ANOVA, followed by Tukey's post hoc test (E and F) (*P < 0.05; **P < 0.01); P values are not shown if the titer of the high-yield vaccine candidate was lower than that of wild-type virus. Red and blue asterisks indicate the comparison of the respective virus with WT virus; beige asterisks indicate the comparison between the viruses depicted in red and blue. HA titer, hemagglutination titer.

Fig. 5.

Evaluation of the total viral protein yield and HA content of HY(Yam) and HY(Vic) viruses. (A) Comparison of viruses possessing the HA and NA vRNAs of the indicated Yamagata-lineage viruses in combination with the internal vRNAs of the same natural wild-type virus (WT) or of HY(Yam). (B) Comparison of viruses possessing the HA and NA vRNAs of the indicated Victoria-lineage viruses in combination with the internal vRNAs of the same wild-type virus (WT) or of HY(Vic). The total viral protein yield of MDCK cell-grown, sucrose gradient-purified virus samples is shown (Left and Center). PNGaseF treatment deglycosylates HA1 and HA2; this treatment was carried out because glycosylated HA2 migrates at a similar molecular weight as M1. The HA contents (Right) were calculated based on the total viral protein amounts and the relative amounts of HA; for details, see Materials and Methods. The values presented are the average of three independent experiments ± SD. The statistical significance was assessed by using one-way ANOVA followed by Dunnett’s test, comparing the total viral protein yield and HA content of wild-type viruses with that of recombinant high-yield vaccine viruses (*P < 0.05; **P < 0.01).

Fig. 6.

Virulence of HY(Yam) and HY(Vic) viruses in mice. (A–C) Comparison of a wild-type Yamagata-lineage virus (B/Massachusetts/2/2012), a virus possessing the B/Massachusetts/2/2012 HA and NA vRNAs in combination with the remaining vRNAs of B/Yamagata/1/73 (used for virus library generation), and a virus possessing the B/Massachusetts/2/2012 HA and NA vRNAs in combination with the remaining vRNAs of HY(Yam). (D–F) Comparison of a wild-type Victoria-lineage virus (B/Brisbane/60/2008), a virus possessing the B/Brisbane/60/2008 HA and NA vRNAs in combination with the remaining vRNAs of B/Yamagata/1/73 (used for virus library generation), and a virus possessing the B/Brisbane/60/2008 HA and NA vRNAs in combination with the remaining vRNAs of HY(Vic). BALB/c mice (five per group) were inoculated intranasally with 10⁶ pfu of the indicated viruses and monitored daily for body weight changes (A and D) and survival (B and E). To assess virus replication in mice, 10⁶ pfu of the indicated viruses were used to infect 10 additional mice. On days 3 and 6 postinfection, five mice in each group were killed, and lung virus titers were determined by use of plaque assays in MDCK cells (C and F). Statistical significance was assessed by using one-way ANOVA followed by Dunnett’s test (*P < 0.05; **P < 0.01).

Fig. 7.

Growth kinetics and hemagglutination titers of single reassortant viruses. (A) Comparison of the parental virus used for Yamagata-lineage virus library generation (i.e., B/Yamagata/1/73 with the HA and NA vRNAs of B/Yokohama/UT-K31/2012) with viruses that also possess an individual vRNA of HY(Yam). (B) Comparison of the parental virus used for Victoria-lineage virus library generation (i.e., B/Yamagata/1/73 with the HA and NA vRNAs of B/Yokohama/UT-K1A/2011) with viruses that also possess an individual vRNA of HY(Vic). Experiments were carried out as described in the legend to Fig. 1. Data were obtained from three independent experiments; shown are average titers ± SD. The values presented are the average of three independent experiments ± SD. Statistical significance was determined by using the linear mixed model described in Materials and Methods (*P < 0.05; **P < 0.01). The color of the asterisks indicates the comparison of the respective virus with the comparator virus (depicted in black). HA titer, hemagglutination titer.