Using a murine model, we assessed the immunogenicity of a recombinant VACV that was engineered to lack a known virulence determinant of VACV, the N1L gene. Mice infected with sublethal doses of the attenuated vGK5 virus by the i.n., tail scarification, and i.p. routes had high frequencies of activated antigen-specific CD8 T cells in both mucosal and systemic sites. In addition, mice infected with vGK5 had levels of VACV-specific antibodies 3–5 months after immunization which were similar to levels elicited in mice immunized with VACV-WR. Our results indicate that the attenuated vGK5 virus, though more attenuated in vivo, is still immunogenic and able to elicit robust T as well as B cell responses.
The lung is the major portal of entry and transmission for variola virus, the etiologic agent of smallpox. Therefore it is critical to assess vaccination strategies that lead to robust cell mediated immunity in pulmonary as well as extrapulmonary tissues. VACV inoculation by the i.n. route has been used as a model that more closely approximates the route of infection with natural variola (smallpox) virus infections in humans; at high doses of VACV-WR i.n., mice develop fatal lung disease with high viral titers in lung tissue and the brain 
. Our data suggest that the N1L protein limits the strength and magnitude of the CD8 T cell immune response during acute intranasal VACV infections since deletion of the N1L gene allows mice infected with the attenuated virus to respond with a robust CD8 T cell response.
Recently two groups solved the structure of N1L which has striking homology to the Bcl-2 family of antiapoptotic genes 
. In vitro, the N1L protein inhibits NF-kB signaling after IL-1, TNF-α, LT-β, and TLR stimulation 
. Under normal conditions, NF-kB is an antiapoptotic transcription factor and therefore inhibition of NF-kB signaling under these conditions could induce programmed cell death 
. Replication of the N1L-deleted virus in cell culture has however been found to be indistinguishable from a wildtype as well as a revertant virus 
. We hypothesize that the Bcl-2-like structure of N1L reconciles the observed lack of positive or negative effect on cell survival in vitro following N1L expression, with its otherwise fatal NF-kB inhibitory function.
Graham et al recently confirmed that transfected N1L DNA inhibited IL-1 and TRAF 6 signaling to NF-kB 
. Vaccinia virus proteins A52 and B14 share a Bcl-2-like fold but have evolved to inhibit NF-kappaB rather than apoptosis. N1L appears to inhibit NF-kB dependent inflammatory cytokine production in mice, based on the observation that N1L-deficient vaccinia virus permits greater expression of NF-kB driven genes during in vivo
VACV infection 
. Furthermore, N1L also suppresses signaling to IRF3, more robustly than it does NF-κB. IRF3 signaling was not investigated by Cooray et al, although Bcl-2 family members also influence IRF3 signaling. Programmed cell death, IRF3 signaling and NF-kB signaling are three key pathways in the innate immune response, and Bcl-2 proteins, like N1L, are capable of inhibiting all three innate immune response pathways 
. Since signaling via the innate immune system is thought to be involved in the adaptive immune response 
, the N1L protein may contribute to impaired adaptive immune responses by inhibiting any combination of these innate signaling pathways.
In vivo, VACV-WR and vGK5 had different replication kinetics. In our studies, when equal doses of virus were administered after i.n. infection (103.5 PFU), VACV-WR replicated to a 3–4 log higher titer compared to the vGK5 virus. Interestingly, when equal doses were administered by the i.p. route (106 PFU), viral titers were not significantly different. Since the N1L protein was hypothesized to have an antiapoptotic function, increased survival of cells infected with VACV-WR which express the N1L protein versus cells infected with vGK5 virus could contribute to increased viral titers in the lungs after i.n. infection.
Intranasal infections with respiratory viruses result in the recruitment of virus-specific CD8+ T cell effectors in the lung during acute infection and persistence of these virus-specific T cells in the respiratory tract months after the infection has resolved 
. Frequencies of antigen-specific T cells that are maintained in memory following virus infections are likely influenced by several factors including the amount of initial antigen available for T cell priming, viral replication in target tissues, the route of inoculation and the cytokine milieu. Virus titers in the lungs of mice infected with the attenuated N1L deleted virus by the i.n. route were several logs higher compared to lungs of mice infected with by the i.p. route. While frequencies of antigen-specific cells in the lungs during acute responses were not significantly impacted by these differences in viral loads, whether frequencies of B8R20–27
specific T cells are differentially maintained in memory is still unknown. Memory responses of these and other VACV-specific T cells therefore need to be further evaluated in mice infected with attenuated N1L deficient viruses.
Several factors including the initial antigen dose, the kinetics of virus replication in mucosal and systemic sites, the innate immune response, T cells as well as antibodies are likely to contribute to protection. Our data show that mice immunized with attenuated vGK5 virus by the intranasal route induced robust immunity and subsequently was able to protect mice from a lethal challenge with VACV-WR. The vGK5 virus is not currently a strain with satisfactory attenuation or safety profile and further clinical development would likely involve testing the effect of N1L inactivation in an established vaccine strain. We propose that the attenuated vaccinia virus lacking a major virulence gene N1L is an alternative that balances immunogenicity and safety. Our data have implications for the rational design of recombinant live vaccines against foreign antigens.