Chemically inactivated viral vaccines are widely used in clinics for protecting infectious and malignant diseases. However, viral vaccines have disadvantages, such as the risk of virus reactivation, the cost of maintaining vaccine stability, the induction of autoimmune diseases, and deficient protection in some vaccines 
. To overcome these disadvantages, researchers have developed synthetic epitope-based peptide vaccines over the last 30 years. To improve the efficacy of peptide vaccines for practical application, we used Lipoplex(O) as an efficient adjuvant for peptide without any additional carrier; we then successfully screened the potent peptide-based epitopes and developed efficient prophylactic vaccines against the influenza A virus by using a complex of viral epitopes and Lipoplex(O).
Phosphorothioate-modified CpG-DNA (PS-ODN), which are resistant to nuclease activity and can be efficiently delivered into cells 
, have been considered in therapeutic applications 
. The immunostimulatory activities of PS-ODN gained attention as a potentially useful form of therapeutics for immunoadjuvants 
. However, PS-ODN induces CG sequence–independent and backbone-related side effects 
. In previous studies, we suggested that PO-ODN may be more optimal than PS-ODN for enhancing innate immune responses 
. However, the immunostimulatory effects of PO-ODN are updated only in mouse cells and not in human cells 
. Here, we have shown that encapsulation with the DOPE:CHEMS complex enhanced the adjuvant effect of PO-ODN (MB-ODN 4531(O)) on the production of the B cell epitope–specific antibody and the efficacy of the prophylactic vaccine against influenza A virus. We successfully identified the B cell epitope H5N1 HA233 out of 4 putative epitopes from the HA protein of the H5N1 A/Vietnam/1203/2004 strain by using a complex of each predicted epitope and Lipoplex(O) (). A complex of the epitope and Lipoplex(O) induced significantly high levels of an epitope-specific Th1-dominated humoral response in animal experiments in a DOPE:CHEMS composition (1
1 ratio)-dependent, CpG sequence-dependent, and backbone modification-independent manner (, ).
Divergent strains of avian and human H5N1 influenza A virus have been reported since 1997. It is important, therefore, for the prompt application, to develop the epitope-based peptide vaccine against pandemic influenza. BALB/c mice immunized with a complex of hH5N1 HA233 epitope and Lipoplex(O) produced the epitope-specific IgG (, ), and the produced antibodies were reactive with the rH5N1 virus (). We found that the hH5N1 HA233 epitope is relatively conserved in most influenza H5N1 strains except for a few other H5N1 strains with only one or two amino acid difference (). However, there is an apparent paradox between the antibody inducing capability of the peptides () and some of the peptide sequences () which cannot be explained easily. For example, immunization with the peptide hH5N1 HA233-4 results in a clear response against HA233, whereas immunization with peptide hH5N1 HA233-5 leads to poor antibody-induction. Based on the comparison of these two peptides with the hH5N1 HA233 sequence, one would conclude that the serine to proline change at position 5 is responsible for this difference. This is also likely since a proline residue usually distorts a peptide conformation significantly. However, HA233-1 which has a Proline at position 5, appears to work very well. Therefore, there may be some complex novel mechanism in recognition of the peptide sequences as a good antigen when we use our novel peptide vaccine technology.
Based on the ability of virus neutralization, sera from mice immunized with hH5N1 HA233 were partially cross-reactive to the H1N1 virus (A/WSN/1933 virus), and swine origin H1N1 virus (A/Korea/01/09) (). The vaccination with a complex of hH5N1 HA233 and Lipoplex(O) illustrated a prophylactic effect against infection of the rH5N1 virus; the effect is confirmed by the survival rate, the regaining of weight, and the moderate lung histopathology (). The prophylactic effect observed after second immunization was similar to that observed after third immunization even though significantly higher level of specific Ab was observed after third immunization ( and Figure S2
). These results leave the role for specific Abs in protection to be further determined. On the other hand, the hH5N1 HA233 vaccine was less effective against the H1N1 virus (A/WSN/1933 virus) challenge with a survival rate of 50% (Figure S1
). Considering the cross-reactivity of the antisera to the A/WSN/1933 virus was about 70~80% of the reactivity to the rH5N1 virus (), the survival rate is lower than the expected. The discrepancy between in vitro
neutralizing activity and in vivo
protection effect suggests that other factor(s) such as innate immunity may be involved in the protection in addition to the specific Abs. Taken together, further studies are needed on the prophylactic effect of the complex of hH5N1 HA233 and Lipoplex(O) on the challenge of multiple influenza A subtypes, H1N1 strains, and various highly pathogenic avian influenza A viruses of H5N1 strains. Such studies will allow fine tuning of a potential vaccine, which in turn should lead to improvement of the vaccine program against pandemic influenza.
Furthermore, we suggest that the antibodies that recognize hH5N1 HA233 cannot be produced by immunization with an inactivated rH5N1 virus (). This result implies that potent epitopes, which cannot be easily found using proteins or viruses as an antigen, can be screened and that epitope-specific antibodies can be produced when we use a complex of peptide epitopes and Lipoplex(O). This result also suggests a possibility that hH5N1 HA233 epitope is not under selective pressure in immunized individuals. Thus, it would be very interesting to find out if a selective pressure on the epitopes produces escape mutants; there might be antigenic variants during repeated selection process in susceptible cells in vitro or in immunized individuals in vivo when treated with the epitope-specific monoclonal antibodies. We warrant clarifying this issue in the future works. To understand the mechanism involved in the potent adjuvant activity of Lipoplex(O), we also need to clarify other important issues. Human beings have many different genetic backgrounds of HLA types. Therefore, it has to be determined whether IgG production by a complex of B cell epitope and Lipoplex(O) depends on the MHC type and Th1 differentiation and if a novel mechanism is involved.
There are several reports finding B cell epitopes of the HA2 polypeptide and producing monoclonal antibodies through phage-display library selection, HA protein immunization or long size synthetic polypeptide immunization (, ) 
. Here, we have shown that the epitopes previously used by other groups failed to induce the epitope-specific antibody production by immunization with the complex of each epitope and Lipoplex(O) (). These experimental results again carefully suggest that antibody production with the complex of epitope and Lipoplex(O) without carriers is regulated by unknown key modulating mechanism in immune systems. The antisera against the hH5N1 HA233 epitope which includes the 220 loop of HA1 region () had no HI titers (data not shown) suggesting that antisera against HA233 epitope do not block the interaction between HA and its receptor. Previously, similar results were reported that monoclonal HA1-specific antibodies had no HI activity but inhibited virus fusion activity 
In conclusion, we selected the specific B cell epitope hH5N1 HA233, and the vaccine complex of hH5N1 HA233 epitope and Lipoplex(O) stimulated potent protective responses in mice against infection from the rH5N1 virus. To determine the feasibility of developing a vaccine that affords efficient protection against multiple subtypes, we need to ensure that further studies focus on a complex that combines several epitopes together (for example epitopes derived from H1N1 subtypes) with Lipoplex(O).