Influenza is an acute respiratory disease caused by infection with influenza viruses (Centers for Disease Control and Prevention, 2010). Seasonal influenza viruses that circulate annually in humans including influenza A (H3N2 and H1N1 subtypes) and B viruses result in more than 200,000 respiratory and circulatory hospitalizations each year, and an average of 23,000 annual influenza-associated deaths in the United States (Centers for Disease Control and Prevention, 2010; Thompson et al., 2004). With the implementation of influenza vaccination programs, illness and hospitalizations associated with seasonal influenza virus have been substantially averted in the United States (Kostova et al., 2013). However, there are far fewer licensed pre-pandemic vaccines available for use against other influenza A subtypes such as avian influenza A (H5N1) virus that has caused sporadic human infections since 1997 (WHO, 2014) and A (H7N9) that have caused outbreaks of human illness in China since February 2013 (Gao et al., 2013). The continuing occurrence of avian influenza virus infections in humans and the suboptimal immunogenicity of vaccines that target them, highlight the need for the development of improved vaccines strategies (Bart et al., 2014; Prieto-Lara and Llanos-Mendez, 2010; Treanor et al., 2006).

Highly pathogenic avian influenza (HPAI) A (H5N1) virus was first isolated from humans in Hong Kong in 1997 (Subbarao et al., 1998), and continues to cause outbreaks in poultry and sporadic human infections (WHO, 2012). Continued evolution of HPAI H5 virus has led to the emergence of 40 phylogenetically distinct HA clades of the virus (Smith et al., 2015). Viruses from clades 0, 1, 2, and 7 have caused human disease (Writing Committee of the Second World Health Organization Consultation on Clinical Aspects of Human Infection with Avian Influenza A (H5N1) Virus et al., 2008), but clade 1 and clade 2 viruses are responsible for recent human infections (WHO, 2015b). Between 2003 and November 2015, the World Health Organization (WHO) reported 844 laboratory confirmed human cases of H5N1 virus infection with a case fatality rate of approximately 53% (WHO, 2015a). Recent studies suggest that H5N1 virus have the potential to acquire airborne transmission in mammals (Herfst et al., 2012; Imai et al., 2012). Furthermore, humans are largely serologically naive to the H5N1 subtype based on the available H5N1 seroepidemiologic studies (Liao et al., 2013; Nasreen et al., 2013; Van Kerkhove, 2013). In addition, newly isolated HPAI H5 viruses have recently caused outbreaks in wild birds and domestic poultry in Asia, Europe, and North America (Jhung et al., 2015; Smith et al., 2015; WHO, 2015b). Although human infections caused by these viruses have not been detected to date, their appearance in North America highlights the intercontinental spread of H5 virus, the genetic reassortment with regionally circulating avian influenza viruses, and collectively, represents a significant threat to human health (Smith et al., 2015).

As vaccination is one of the most effective strategies for prevention and control of seasonal and pandemic influenza, H5 vaccines have been developed and evaluated intensively in both preclinical studies in mice or ferrets, and in humans (Baras et al., 2008a, 2008b; Baz et al., 2013; Belshe et al., 2014; Leroux-Roels and Leroux-Roels, 2009; Subbarao and Luke, 2007). In preclinical and clinical trials, inactivated H5N1 vaccines were consistently poorly immunogenic when administered to animals or human volunteers without adjuvant. The addition of proprietary oil-in-water emulsion adjuvants such as AS03 and MF59 to inactivated H5N1 vaccines has enhanced vaccine immunogenicity and cross-clade reactivity (Baras et al., 2008a, 2008b; Belshe et al., 2014; Bihari et al., 2012), but predominantly stimulate humoral immune responses; the role of cellular immune responses in post-vaccination protection has not been clearly elucidated (Leroux-Roels, 2009; Radosevic et al., 2008; Valensi et al., 1994). Previous murine studies showed that squalene-based MF59 adjuvant primarily enhances the Th2/IgG1 response to inactivated influenza vaccine and has little effect on the Th1/IgG2a response that is superior in providing protection against infections from antigenically drifted or shifted influenza viruses in mice (Moran et al., 1999; Valensi et al., 1994). Given the rapid antigenic evolution of H5 virus, the use of an adjuvant promoting Th1 response result in both humoral and cell-mediated immune responses that may lead to better cross-protection against heterologous H5 virus infection (Thomas et al., 2006).

JVRS-100 is comprised of cationic liposome–DNA complexes (CLDC) and was developed as an adjuvant for vaccines against infectious diseases and as an immunomodulator for cancer therapy (Bernstein et al., 2009; Chang et al., 2009; Lay et al., 2009; Zaks et al., 2006). It has been demonstrated that the addition of CLDC to influenza vaccines can induce robust CD4⁺ Th1 response and predominant IgG2a/c antibody responses in mice (Chang et al., 2009; Dong et al., 2012; Hong et al., 2010; Lay et al., 2009). Most noteworthy, enhanced multicytokine-producing influenza virus-specific CD4⁺ and CD8⁺ T-cell responses were demonstrated in mice and non-human primates, and enhanced cross-protection were demonstrated in mice with addition of CLDC to influenza vaccines (Hong et al., 2010; Lay et al., 2009). In a previous study, we demonstrated that a JVRS-100 adjuvanted clade 1 H5N1 split vaccine showed significantly improved immunogenicity and antigen-sparing effect, stimulated enhanced cross-clade antibody and CD4⁺ T-cell responses, and provided significant cross-clade protection in mice (Dong et al., 2012). Ferrets have been widely used and recognized as the most suitable pre-clinical challenge model to evaluate the immunogenicity and protective efficacy of influenza vaccines (Baras et al., 2008b; Subbarao and Luke, 2007; Zitzow et al., 2002). Therefore, pre-clinical data generated in this species could be predicative of clinical trials in humans when it is not possible to conduct experimental challenge of HPAI H5 virus in human volunteers (Subbarao and Luke, 2007). Here, we investigated the immunogenicity of a JVRS-100 adjuvanted clade 2.2 H5N1 split vaccine from the US pre-pandemic influenza vaccine stockpile, and evaluated the cross-protective efficacy against infection with a heterologous H5N1 strain in ferrets.

HPAI clade 2.2 H5N1 viruses were first detected in migratory birds in central China in 2005, and were associated with mass die-offs of wild-birds (WHO, 2012). Since 2005, the numbers of affected countries and confirmed cases of H5N1 virus infection have increased, in part because of the spread of clade 2.2 viruses across Eurasia and to Africa (Writing Committee of the Second World Health Organization Consultation on Clinical Aspects of Human Infection with Avian Influenza A (H5N1) Virus et al., 2008; WHO, 2012, 2015a). In this study, we have shown that the addition of JVRS-100 to BH/05 clade 2.2-derived H5N1 vaccine, a component of the current US pre-pandemic influenza vaccine stockpile, resulted in significant enhancement of serum HAI and MN antibody responses in ferrets after the first and second vaccinations compared with unadjuvanted vaccine, and provided significant cross-protection from a lethal challenge with a highly virulent heterologous clade 2.1.3.2 virus, which has caused 199 human infections in Indonesia with very high fatality rate (84%) (WHO, 2015a). Furthermore, JVRS-100 addition also improved cross-reactive HAI antibody responses to clade 0, clade 1, clade 1.1.2, clade 2.1.3.2, clade 2.1.3.2a, clade 2.2.2, clade 2.3.4, and clade 2.3.2.1c H5N1 viruses isolated between 1997 and 2013, as well as clade 2.3.4.4 H5N2 and H5N8 viruses that emerged in late 2014. These data are consistent with our previous study showing that the addition of JVRS-100 to a licensed clade 1 H5N1 vaccine significantly enhanced cross-reactive antibody responses to clade 2.2.1 and clade 2.3.4 H5N1 viruses and provided significant cross-protections to lethal challenges with these viruses in mice (Dong et al., 2012). Importantly, JVRS-100 addition also resulted in significant cross-clade Th1 CD4⁺ T-cell responses and dose-sparing in mice (Dong et al., 2012). The ability conferred by an adjuvant to elicit broad cross-clade protective immunity and antigen-sparing would be highly advantageous in the early stages of a pandemic to provide herd immunity in human populations, since it will take months for the current manufactures to produce well-matched vaccine with sufficient doses supply.

The addition of JVRS-100 to commercially available 2006–2007 trivalent inactivated seasonal influenza vaccine (TIV) also significantly enhanced HAI antibody responses in mice and macaques (Lay et al., 2009), and resulted in significant cross-protection in mice from antigenically distinct influenza H1N1, H3N2, and influenza B virus that were isolated from humans more than 35 years ago (Lay et al., 2009). Furthermore, JVRS-100 adjuvanted inactivated influenza H1N1 vaccine provided significant cross-protection from heterosubtypic H3N2 challenges in mice (Hong et al., 2010). Both cellular and antibody-mediated immunity likely played a role in such cross-protection from antigenically distinct influenza viruses (Hong et al., 2010; Lay et al., 2009). Occasionally, seasonal influenza vaccine antigens are antigenically drifted from circulating viruses which may in some cases lead to lower vaccine effectiveness (Flannery et al., 2015; Klimov et al., 1999; McNeil et al., 2015). The addition of adjuvants such as JVRS-100 to antigenically mismatched inactivated seasonal influenza vaccine might improve immune responses and broaden immune protection when a well-matched vaccine could not be prepared in time (Lay et al., 2009; Moran et al., 1999).

CLDC formed adjuvants such as JVRS-100 were particularly effective in eliciting strong T cell responses, even at low antigen doses (Hong et al., 2010; Lay et al., 2009; Zaks et al., 2006). The CD4⁺ and CD8⁺ T cells elicited by JVRS-100 adjuvanted influenza vaccine secrete multiple cytokines with a poly-functional phenotype in non-human primates (Lay et al., 2009). JVRS-100 adjuvanted vaccine also induced influenza virus-specific cytotoxic CD8⁺ T cells directed to the internal viral protein epitopes common to heterologous influenza strains (Hong et al., 2010; Thomas et al., 2006). Importantly, this superior cross-reactive CD8⁺ T cell reactivity induced by CLDC adjuvant persisted for long periods of time in mice lung tissues (Zaks et al., 2006), and was associated with significantly reduced morbidity and mortality in mice after heterosubtypic influenza viral challenges (Hong et al., 2010). Furthermore, increasing evidences showed that strong T cell responses to more conserved influenza internal viral proteins were associated with reduced disease severity in humans (Li et al., 2013; Sridhar et al., 2013; Wilkinson et al., 2012). Thus, influenza vaccines that can significantly boost heterosubtypic CD4⁺ and CD8⁺ T-cell responses, in addition to inducing neutralizing antibodies to the circulating viruses, may offer broader protection against drifted epidemic viruses and possibly even pandemic viruses. (Li et al., 2013; Sridhar et al., 2013; Wilkinson et al., 2012). Unfortunately, we cannot rule out a role of T cell-mediated immunity in cross-protection in this ferret model. However, it is plausible that the reduced amount of virus shedding and faster clearance of virus in ferrets might, in part, be due to a robust T cell immune response stimulated by JVRS-100 adjuvanted vaccine (Hong et al., 2010; Moran et al., 1999).

Our previous study indicated that immunization with JVRS-100 adjuvanted H5N1 vaccine induced long-lasting HAI antibody responses in mice (Dong et al., 2012). It is intriguing to note that HAI antibody titers induced by the two doses of JVRS-100 adjuvanted H5N1 vaccine in ferrets declined 2–4 fold at 2 months compared to the time period of 2 weeks after the second immunization. The decay of the HAI antibody response was also observed in ferrets, but not in mice following immunization with alum adjuvanted H5 vaccines (Layton et al., 2011; Lu et al., 1999). The decay of the HAI antibody response was also observed in adults immunized with oil-in-water adjuvanted H5 vaccines (Leroux-Roels et al., 2010). Although antibody titers decreased with time, oil-in-water adjuvanted H5 vaccine can elicit long-term memory B cells in humans (Khurana et al., 2014) and alum adjuvanted H5 vaccine can induce long-term protection against HPAI H5N1 infection in ferrets (Layton et al., 2011). Nevertheless, successful vaccination regimes rely on the generation of long lived plasma cells and long-lived memory B cells (Tarlinton and Good-Jacobson, 2013). Investigation of immune memory enhanced by different adjuvants, including JVRS-100, may reveal mechanisms underlying the different half-lives of humoral immunity in different species. This may also lead to the development of better adjuvants that could enhance long-lasting protective immunity in humans, which is an area requiring more research.

Our data, together with other previous findings indicate that the addition of JVRS-100 adjuvant to influenza vaccines induces CD4⁺ Th1 biased immune response, cross-reactive antibody responses, functional memory CD8⁺ T-cell responses, and broad cross-protection in multiple species (Dong et al., 2012; Hong et al., 2010; Lay et al., 2009). Novel adjuvants such as JVRS-100 that can stimulate multiple immune pathways to enhance immunogenicity and protective efficacy warrant further investigation as potential adjuvants for inactivated vaccines against H5N1 and other influenza viruses with pandemic potential.