Production of H5N1 (NIBRG-14) inactivated whole virus and split virion influenza vaccines and analysis of immunogenicity in mice using different adjuvant formulations

Vaccine 28 (2010) 2505–2509

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Production of H5N1 (NIBRG-14) inactivated whole virus and split virion influenza vaccines and analysis of immunogenicity in mice using different adjuvant formulations Cosue Miyaki a , Wagner Quintilio a , Eliane N. Miyaji b , Viviane F. Botosso a , Flavia S. Kubrusly b , Fernanda L. Santos a , Dmitri Iourtov a , Hisako G. Higashi a , Isaias Raw b,∗ a b

BioIndustrial Division, Instituto/Fundac¸ão Butantan, SP, Brazil Center for Biotechnology, Instituto Butantan, Av Vital Brasil, 1500, 05503-900 São Paulo, SP, Brazil

a r t i c l e

i n f o

Article history: Received 16 October 2009 Received in revised form 28 December 2009 Accepted 16 January 2010 Available online 30 January 2010 Keywords: H5N1 influenza vaccine Adjuvant

a b s t r a c t Consecutive lots of H5N1 (A/Vietnam/1194/2004 – NIBRG-14) split virion and whole virus vaccines were produced in a pilot-scale laboratory. The average yields of vaccine doses (15 ␮g HA) per egg were 0.57 doses for H5N1 split virion vaccine and 1.12 for H5N1 whole virus vaccine, compared to 2.09 doses for the seasonal H3N2 split virion vaccine. H5N1 split virion vaccine lots complied with WHO protein content criteria, while some lots of the H5N1 whole virus vaccine showed protein content per dose higher than the limit established. All lots of both vaccines showed ovalbumin (OVA) concentration below the recommended limit. Dose sparing strategies using adjuvant formulations using aluminum hydroxide (Al(OH)3 ) and monophosphoryl lipid A (MPLA) from Bordetella pertussis were tested in mice. Both 3.75 ␮g HA and 7.5 ␮g HA of H5N1 split virion vaccine with Al(OH)3 or Al(OH)3 plus MPLA in aqueous suspension showed higher hemagglutination-inhibition (HAI) titers when compared to the same vaccine dose without any adjuvant. Immunization with the H5N1 inactivated whole virus vaccine was also performed using 3.75 ␮g HA and HAI titers were higher than those induced by the split virion vaccine. Moreover, the use of Al(OH)3 with MPLA as an emulsion induced a further increase in HAI titers. © 2010 Elsevier Ltd. All rights reserved.

1. Introduction Avian H5N1 influenza infection in humans has been first reported in 1997 and a total of 442 cases have been confirmed by WHO until September 2009, with a very high case fatality rate of 60% [1]. Prepandemic H5N1 vaccines have been thus developed and tested. As part of the preparedness plan in Brazil for an influenza pandemic, H5N1 (A/Vietnam/1194/2004 – NIBRG-14) vaccines were produced in a pilot-scale laboratory using embryonated hen’s eggs. Shortage of vaccine is of particular concern in developing countries, once the estimated global production capacity of seasonal influenza vaccines is of approximately 350 million doses and mostly in production plants located in industrialized countries [2]. Due to the lack of pre-existing immunity against a new pandemic virus, higher vaccine dosages will probably be required. Moreover, large amounts of doses will have to be produced within a short period. Dose sparing strategies will thus be particularly important in an influenza pandemic scenario. The yield

∗ Corresponding author. Tel.: +55 11 3726 3790. E-mail address: [email protected] (I. Raw). 0264-410X/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2010.01.044

per egg is also a concern, since H5N1 vaccines have shown lower yields than the seasonal vaccine strains. A recent review has analyzed the results of different human trials of H5N1 vaccines and whole virus vaccines were in general more immunogenic than non-adjuvanted split virion ones. Moreover, the use of oil-in-water emulsions provided the best results when added to split virion influenza vaccines [3]. The use of adjuvants for the seasonal influenza vaccine has also been proposed by our group as a way to increase the capacity of doses produced and to reduce costs due to decrease in the antigen content of the vaccine. The adjuvant monophosphoryl lipid A (MPLA) from Bordetella pertussis was obtained by acid hydrolysis of LPS, which is a by-product of a new cellular pertussis vaccine with lower endotoxin content [4]. We have shown that MPLA alone or combined with aluminum hydroxide (Al(OH)3 ) induced an increase in hemagglutination-inhibition (HAI) titers in mice immunized with H3N2 (A/Wisconsin/67/2005) split virion vaccine [5]. Here we show the production of H5N1 (A/Vietnam/1194/2004 – NIBRG-14) whole virus and split virion vaccines in eggs, with the analysis of yields in comparison to H3N2 (A/Panama/2007/99) split virion seasonal vaccine strain. Protein and ovalbumin (OVA) content in the preparations were also analyzed. Immunogenicity

2506

C. Miyaki et al. / Vaccine 28 (2010) 2505–2509

Fig. 1. Flow diagram for the production of split virion and whole virus influenza vaccines.

of the vaccine and the impact of the use of MPLA and Al(OH)3 as adjuvants in mice are also shown.

resuspended in phosphate buffered saline (PBS), filtered to remove bacteria and inoculated in embryonated hen’s eggs in order to confirm the total inactivation of the virus.

2. Materials and methods 2.2. Adjuvants 2.1. Vaccine production Influenza H5N1 vaccines were produced in a pilot-scale biosafety level 3 laboratory using the A/Vietnam/1194/2004 (NIBRG-14) reference virus, provided by the National Institute for Biological Standards and Controls (NIBSC, UK). NIBRG-14 is a reassortant virus produced by reverse genetics containing the internal genes of A/PR/8/34, and hemagglutinin (HA) and neuraminidase (NA) genes from A/Vietnam/1194/04 virus and modified by replacing the polybasic amino acids at the cleavage site to render the virus avirulent [6]. H3N2 (A/Panama/2007/99) virus is a seasonal influenza vaccine strain. For the production of the inactivated vaccines, the seed virus was grown in 11 days-embryonated hen’s eggs. 72 h after the inoculation, embryos were killed by chilling the eggs. For the split virion vaccines, the harvested allantoic fluid was clarified, purified by sucrose gradient zonal centrifugations, disrupted using Triton X-100 and inactivated with formaldehyde. For the whole virus vaccines, the harvested allantoic fluid was clarified, inactivated with formaldehyde and submitted to gel filtration chromatography for concentration and purification (Fig. 1). Influenza vaccines were

Alhydrogel (Al(OH)3 —Brenntag Biosector, Denmark) and MPLA were used as adjuvants. MPLA was produced using LPS from previously detoxified whole cell pertussis vaccine, followed by organic extraction and hydrolysis as described [5]. 2.3. Immunization and analysis of the immune response Female BALB/c mice weighting 18–20 g (8 weeks) were supplied by the Butantan Institute’s Central Animal House. Experiments were performed according to experimental protocols approved by the Animal Use Ethics Committee of Butantan Institute. The different vaccine formulations were administered in 0.5 mL through the intraperitoneal route. Animals were bled individually and sera were used for the determination of HAI titers using horse erythrocytes, with methods routinely used for the determination of the potency of vaccines for human use. An HAI titer ≥1:40 is considered a correlate of protection in humans [7]. Total anti-H5N1 IgG, IgG1 and IgG2a antibody concentrations were determined in individual sera by ELISA, using 250 ng/mL HA of H5N1 split vaccine as coating antigen. Goat anti-mouse IgG, anti-mouse IgG1, anti-mouse IgG2a and

C. Miyaki et al. / Vaccine 28 (2010) 2505–2509

2507

Table 1 Vaccine yields and protein content of vaccine lots. Vaccine

15 ␮g dose/egg

␮g prot/dose

ng OVA/dose

ng OVA/␮g prot

H3N2 split virion H5N1 split virion H5N1 whole virus

2.09 ± 1.03 0.57 ± 0.44 1.12 ± 1.21

20.4 ± 3.2 39.7 ± 13.4 108.1 ± 67.2

20.5 ± 17.3 411.2 ± 361.0 432.1 ± 308.2

0.97 ± 0.78 11.33 ± 9.26 3.79 ± 1.61

Values indicate average ± SD of 6 lots of H3N2 split virion vaccine, 10 lots of H5N1 split virion vaccine and 17 lots of H5N1 whole virus vaccine. Table 2 HAI titers induced by H5N1 split virion vaccine with different adjuvant formulations. Vaccine dose (HA content)

No adjuvant

Al(OH)3

MPLA (aqueous suspension)

Al(OH)3 + MPLA (aqueous suspension)

3.75 ␮g 7.5 ␮g 15 ␮g

1:20 1:127 1:136

1:289 1:260 –

1:40 1:164 –

1:200 1:320 –

HAI titers were measured using sera collected 21 days after the inoculation of the vaccine. GMT of 5 animals is shown.

anti-goat IgG conjugated with HRP (Southern Biotech) were used for detection of antibodies. Standard curves were generated using mouse IgG, IgG1 and IgG2a. Differences in IgG levels were analyzed using Student’s t-test.

latter vaccine. All vaccine lots showed OVA concentration below WHO recommended limit of up to 5 ␮g/single human dose [8]. 3.2. Antigen sparing of H5N1 vaccines by the use of different adjuvant formulations

3. Results 3.1. Vaccine production Consecutive lots of split virion H3N2, split virion H5N1 and inactivated whole virus H5N1 vaccines were produced using 4500–5000 eggs per preparation. The average yield of doses (15 ␮g HA) were of 2.09 doses per egg for H3N2 split virion vaccine, 0.57 for H5N1 split virion vaccine and 1.12 for H5N1 whole virus vaccine (Table 1). As reported by several producers, the yield of H5N1 NIBRG-14 split virion vaccine per egg is about 5-fold lower than H3N2 (A/Panama/2007/99). WHO guidelines indicate that total protein content should not be more than six times the total HA content of the vaccine and that there should be not more than 100 ␮g of protein per virus strain per human dose [8]. All vaccine lots of both H3N2 and H5N1 split virion vaccine lots complied with these criteria, while some lots of the H5N1 whole virus vaccine showed protein content per dose higher than the limit established. Differences in the protein concentration were observed between lots due to variations in the procedure to collect the allantoic fluid. The concentration of ovalbumin (OVA) was also analyzed, since it is important to avoid induction of hypersensitivity to eggs in individuals that will receive the seasonal influenza vaccine each year and additional one or two doses of pandemic influenza vaccine. As seen in Table 1, the concentration of OVA is about 1/1000 of total protein in the process for the production of H3N2 split virion vaccine and about 1/100 in split virion H5N1 vaccines, probably reflecting the low yields of the

Based on our previous results showing the enhancement of HAI titers by the use of Al(OH)3 and MPLA [5], we have tested the adjuvant effect of 0.25 mg of Al(OH)3 and 0.01 ␮g of an aqueous suspension of MPLA to H5N1 split virion vaccine in mice. Early reports on the use of inactivated H5N1 vaccine candidates in humans have administered up to two doses containing up to 90 ␮g HA [9]. As seen in Table 2, the use of both 3.75 ␮g HA and 7.5 ␮g HA with Al(OH)3 or Al(OH)3 plus MPLA in aqueous suspension was able to enhance HAI titers when compared to the same vaccine dose without any adjuvant. Immunization with the H5N1 inactivated whole virus vaccine was next performed using 3.75 ␮g HA and adjuvant formulations containing 0.25 mg Al(OH)3 and 10 or 50 ␮g MPLA as an aqueous suspension or as an emulsion in 2% squalene. As seen in Fig. 2, the H5N1 whole virus vaccine induced higher HAI titers than the split virion vaccine (Table 2). Moreover, the use of Al(OH)3 with MPLA as an emulsion induced an increase in HAI titers. When a second dose was administered 35 days after the initial immunization, all groups presented an increase in HAI titers of 3–7-fold. Total IgG response was also analyzed after the second vaccine dose and, as seen in Fig. 3A, only the group immunized with the vaccine adjuvanted with Al(OH)3 plus 50 ␮g MPLA as aqueous suspension showed statistically significant higher levels of IgG antibodies to H5N1 than the non-adjuvanted group (P < 0.0001). Antibodies to H5N1 were further isotyped and IgG1 and IgG2a levels determined. IgG1 levels were significantly higher than the non-adjuvanted group in animals immunized with the vaccine

Fig. 2. HAI titers in immunized mice. Mice were immunized with one (A) or two (B) 3.75 ␮g doses of H5N1 whole virus vaccine using the indicated adjuvant formulations. Sera were collected 21 days after each immunization and tested for HAI. Individual values and GMT are shown.

2508

C. Miyaki et al. / Vaccine 28 (2010) 2505–2509

Fig. 3. Induction of IgG antibodies in immunized mice. Mice were immunized with two 3.75 ␮g doses of H5N1 whole virus vaccine using the indicated adjuvant formulations. Sera were collected 21 days after the second immunization and tested for IgG antibodies to H5N1. IgG (A), IgG1 (B) and IgG2a (C) antibody concentrations of individual mice and GM are shown.

adjuvanted with Al(OH)3 alone (P = 0.0005), with Al(OH)3 plus 10 ␮g MPLA as emulsion (P = 0.02), Al(OH)3 plus 50 ␮g MPLA as aqueous suspension (P = 0.0002) and Al(OH)3 plus 50 ␮g MPLA as emulsion (P = 0.006) (Fig. 3B). None of the adjuvanted group showed significantly higher IgG2a levels than animals immunized with the vaccine only (Fig. 3C). 4. Discussion The present limited production capacity of influenza vaccines urges the establishment of production plants in developing countries. The prepandemic H5N1 (NIBRG-14) split virion and whole virus vaccines were produced in a pilot-scale laboratory and the yields were 0.57 and 1.12 doses (15 ␮g HA/dose) per egg, respectively. H5N1 split virion vaccine lots complied with WHO protein content criteria, while some lots of the H5N1 whole virus vaccine showed protein content per dose higher than the established limit. All lots of both vaccines showed OVA concentration below the recommended limit. The development of dose sparing strategies, such as the use of adjuvants, is also an important goal to increase the production capacity of influenza vaccines. Aluminum hydroxide has been used since 1925 as adjuvant for vaccines such as DTP and Hepatitis B (HBV), and was initially thought to be an inert compound that allowed the slow release of adsorbed antigens, inducing local inflammation and recruitment of antigen presenting cells. It has been recently shown that adjuvants based on aluminum salts activate the cytoplasmic NOD-like receptor protein “NOD-like receptor, pyrin domain containing 3” (Nlrp3, also know as Nalp3 or cryopyrin), which associates with Asc and caspase-1 to form a complex called the inflammasome. Activation of the inflammasome induces the release of IL-1␤, IL-19 and IL-33 [10]. Controversial data on the use of aluminum salts as adjuvant for H5N1 vaccines have been presented, with reports of both enhanced and lack of efficacy in clinical trials [3]. The source of alum and its characteristics may be the reason for these contradictory results and most of the published data do not contain this information [11–16]. The use of oil-in-water emulsions, such as MF-59, is also a promising approach to potentiate immune responses. These emulsions frequently use squalene formulated in droplets <250 nm, and were shown to increase levels of antibodies to HA and CD8+ T cell responses. The mechanism of action of squalene nano-emulsions has not been fully solved yet [17]. A proprietary adjuvant emulsion containing Al(OH)3 and monophosphoryl lipid A from B. pertussis was used in this work. MPLA is a low cost by-product of a safer whole cell pertussis vaccine with lower endotoxin content. While binding of LPS to TLR4 activates both MyD88-Mal and TRIF-TRAM pathways, monophos-

phoryl lipid A from Salmonella minesota was shown to activate only the TRIF-TRAM pathway, which accounts for its adjuvant properties, displaying low toxicity [18]. Since influenza whole virus vaccines were reported to be more efficacious than split virion ones, they have been tested in selected human trials of H5N1 [11,13]. A recent comparison between inactivated whole virus and split vaccines has shown that since both have the same antigen content, differences in the immunogenicity might be related to distinct spatial organization [19]. While whole virus vaccines present their antigens in particulate organized vesicles, split vaccines consist of a mixture of solubilized membrane proteins and internal components. Presentation to the immune system is thus probably different in these two vaccines. Activation of TLR7 by a higher content of viral genomic single-stranded RNA by whole virus vaccines has also been implicated in the enhancement of the immune response [20]. In accordance with published data showing higher efficiency of whole virus vaccines, we show here that immunization of mice with the H5N1 inactivated whole virus vaccine using 3.75 ␮g HA induced higher HAI titers compared to those induced by the split virion vaccine. The use of Al(OH)3 induced a further increase in HAI titers. These results are in accordance with data from clinical trials that have tested split virion and whole virus H5N1 vaccines using aluminum salts as adjuvants [3]. Though it is difficult to compare trials using different antigen dosages, the use of whole virus vaccine with alum was shown to be more adequate to induce higher HAI titers than split virion vaccines with alum. We have also tested whether the use of MPLA could enhance the response to H5N1. MPLA did not alter the response to the split virion vaccine when used as aqueous suspension with Al(OH)3 , but an increase in HAI titers could be detected when the whole virus vaccine was adjuvanted with alum and MPLA in an emulsion in squalene. MF59, a squalene oil-in-water emulsion, is already used in influenza vaccines [21]. An aqueous adjuvant formulation containing alum and monophosphoryl lipid A from S. minesota (ASO4) has been licensed as a component of HBV vaccine [17]. We have tested H5N1 whole virus vaccine using Al(OH)3 and MPLA in an oil-in-water squalene emulsion, thus combining three adjuvants. Even small increases in the immunogenicity of H5N1 could have a crucial impact in a pandemic scenario, especially to a population that has never been exposed to this virus. In conclusion, conditions for the production of prepandemic and pandemic split virion and whole virus influenza vaccines were established. The use of adjuvant formulations using Al(OH)3 with MPLA as an emulsion showed the possibility of reducing the amount of antigen per vaccine dose. A more detailed investigation must be carried out in human volunteers to determine the minimal safe dosage that provides efficient coverage and to evaluate the impact of the use of the adjuvants.

C. Miyaki et al. / Vaccine 28 (2010) 2505–2509

Acknowledgements This work was supported by the World Health Organization (WHO) and FINEP (Brazil). References [1] WHO. Cumulative number of confirmed human cases of avian influenza a/(H5N1) reported to WHO; 2009 [cited; available from: http://www.who.int/ csr/disease/avian influenza/country/cases table 2009 09 24/en/index.html]. [2] Oshitani H, Kamigaki T, Suzuki A. Major issues and challenges of influenza pandemic preparedness in developing countries. Emerg Infect Dis 2008;14(June (6)):875–80. [3] Leroux-Roels I, Leroux-Roels G. Current status and progress of prepandemic and pandemic influenza vaccine development. Expert Rev Vaccines 2009;8(April (4)):401–23. [4] Zorzeto TQ, Higashi HG, da Silva MT, Carniel Ede F, Dias WO, Ramalho VD, et al. Immunogenicity of a whole-cell pertussis vaccine with low lipopolysaccharide content in infants. Clin Vaccine Immunol 2009;16(April (4)):544– 50. [5] Quintilio W, Kubrusly FS, Iourtov D, Miyaki C, Sakauchi MA, Lucio F, et al. Bordetella pertussis monophosphoryl lipid A as adjuvant for inactivated split virion influenza vaccine in mice. Vaccine 2009;27(June (31)):4219–24. [6] Wood JM, Robertson JS. From lethal virus to life-saving vaccine: developing inactivated vaccines for pandemic influenza. Nat Rev Microbiol 2004;2(October (10)):842–7. [7] WHO. WHO manual on animal influenza—diagnosis and surveillance WHO/CDS/CSR/NCS/2002.5; 2002 [cited; available from: http://www.who.int/ vaccine research/diseases/influenza/WHO manual on animaldiagnosis and surveillance 2002 5.pdf]. [8] WHO. Recommendations for the production and control of influenza vaccine (inactivated). WHO Technical Report Series; 2005 [cited Annex 3 927; available from: http://www.who.int/biologicals/publications/ trs/areas/vaccines/influenza/en/index.html]. [9] Treanor JJ, Campbell JD, Zangwill KM, Rowe T, Wolff M. Safety and immunogenicity of an inactivated subvirion influenza A (H5N1) vaccine. N Engl J Med 2006;354(March (13)):1343–51.

2509

[10] De Gregorio E, Tritto E, Rappuoli R. Alum adjuvanticity: unraveling a century old mystery. Eur J Immunol 2008;38(August (8)):2068–71. [11] Lin J, Zhang J, Dong X, Fang H, Chen J, Su N, et al. Safety and immunogenicity of an inactivated adjuvanted whole-virion influenza A (H5N1) vaccine: a phase I randomised controlled trial. Lancet 2006;368(September (9540)):991–7. [12] Bresson JL, Perronne C, Launay O, Gerdil C, Saville M, Wood J, et al. Safety and immunogenicity of an inactivated split-virion influenza A/Vietnam/1194/2004 (H5N1) vaccine: phase I randomised trial. Lancet 2006;367(May (9523)):1657–64. [13] Ehrlich HJ, Muller M, Oh HM, Tambyah PA, Joukhadar C, Montomoli E, et al. A clinical trial of a whole-virus H5N1 vaccine derived from cell culture. N Engl J Med 2008;358(June (24)):2573–84. [14] Keitel WA, Campbell JD, Treanor JJ, Walter EB, Patel SM, He F, et al. Safety and immunogenicity of an inactivated influenza A/H5N1 vaccine given with or without aluminum hydroxide to healthy adults: results of a phase I-II randomized clinical trial. J Infect Dis 2008;198(November (9)):1309–16. [15] Song MS, Oh TK, Pascua PN, Moon HJ, Lee JH, Baek YH, et al. Investigation of the biological indicator for vaccine efficacy against highly pathogenic avian influenza (HPAI) H5N1 virus challenge in mice and ferrets. Vaccine 2009;27(May (24)):3145–52. [16] Bernstein DI, Edwards KM, Dekker CL, Belshe R, Talbot HK, Graham IL, et al. Effects of adjuvants on the safety and immunogenicity of an avian influenza H5N1 vaccine in adults. J Infect Dis 2008;197(March (5)):667–75. [17] Reed SG, Bertholet S, Coler RN, Friede M. New horizons in adjuvants for vaccine development. Trends Immunol 2009;30(January (1)):23–32. [18] Mata-Haro V, Cekic C, Martin M, Chilton PM, Casella CR, Mitchell TC. The vaccine adjuvant monophosphoryl lipid A as a TRIF-biased agonist of TLR4. Science 2007;316(June (5831)):1628–32. [19] Hagenaars N, Mastrobattista E, Glansbeek H, Heldens J, van den Bosch H, Schijns V, et al. Head-to-head comparison of four nonadjuvanted inactivated cell culture-derived influenza vaccines: effect of composition, spatial organization and immunization route on the immunogenicity in a murine challenge model. Vaccine 2008;26(December (51)):6555–63. [20] Geeraedts F, Goutagny N, Hornung V, Severa M, de Haan A, Pool J, et al. Superior immunogenicity of inactivated whole virus H5N1 influenza vaccine is primarily controlled by Toll-like receptor signalling. PLoS Pathog 2008;4(8):e1000138. [21] Banzhoff A, Pellegrini M, Del Giudice G, Fragapane E, Groth N, Podda A. MF59-adjuvanted vaccines for seasonal and pandemic influenza prophylaxis. Influenza Other Respir Viruses 2008 Nov;2(November (6)):243–9.