**4. Testing the Enhanced DNA Vector with HIV-1 Subtype C Env Immunogens**

The kinds of immune responses that an effective HIV-1 vaccine would need to elicit include non-neutralizing antibody responses as well as broadly neutralizing antibody responses, together with polyfunctional cytotoxic T cell responses to a variety of epitopes from the HIV-1 proteome [21–23]. One of the main targets of recent HIV-1 vaccine candidates is broadly neutralizing antibody (bNAb) responses: bNAbs that can neutralize diverse primary HIV-1 subtype isolates protect against viral challenge in nonhuman primates (NHP) with Env-pseudotyped simian–human immunodeficiency viruses (SHIVs), suggesting that infection in humans could be similarly prevented [24,25]. Ranking of HIV-1 isolates according to their sensitivities to neutralizing antibodies allows identification of viruses as Tier 1 (sensitive), Tier 2 (moderately resistant), and Tier 3 (resistant) [26]. The circulating viruses that vaccines will need to protect against are largely Tier 2 type: Accordingly, HIV vaccines should elicit responses that neutralize laboratory Tier 2 virus isolates. We showed previously that using a DNA prime/MVA boost immunization regime in mice with vaccines expressing HIV-1 subtype C mosaic Gag resulted in strong cellular immune responses directed against Gag [19]. We wished to extend these results by improving the vaccine regimen to allow the elicitation of Env-specific neutralizing antibodies in a rabbit model.

The pTHPcapR vector was used to construct a DNA vaccine expressing a HIV-1 envelope (DNA Env). The envelope sequence (CAP256SU) used in this study was selected as it elicited broadly neutralizing antibodies (bNAbs) in the patient [27] and was sensitive to several prototype broadly neutralizing monoclonal antibodies [28]. Several modifications were made to the envelope sequence, these included replacing the native leader sequence with the tissue plasminogen activator leader

sequence, replacing the furin cleavage site with a flexible linker, introducing an I548P mutation equivalent to the I559P in the SOSIP trimers to improve the trimerization of gp41 [29] and truncating the sequence from gp160 to gp150 [30]. A second plasmid expressing the soluble envelope protein (gp140) with the same modifications was also constructed using the pTHPcapR backbone [31]. This plasmid was used to generate a stable cell line expressing high levels of the soluble HIV-1 envelope protein, which was subsequently purified and utilized as a protein boost in rabbit immunogenicity studies. MVA vaccines expressing the matching gp150 Env and Env plus mosaic Gag were also constructed.

Rabbits were inoculated with different combinations of vaccines in different regimens, in order to ascertain the overall effects on immunogenicity of the Env component. The first test group was injected with 100 μg of each of DNA Env- and DNA-GagM-encoding plasmids at weeks 0 and 4, boosted with doses of 10<sup>8</sup> pfu of rMVA Env + Gag<sup>M</sup> at weeks 8 and 12, and further boosted with gp140Env protein at weeks 20 and 28 (regime designated as DDMMPP). The other group received 108 pfu of rMVA Env + Gag<sup>M</sup> intramuscularly at weeks 0 and 4, followed by three protein boosts at weeks 12, 20, and 28 (MMPPP) (Figure 4).

**Figure 4.** Rabbits primed with DNA produce higher autologous Tier 2 neutralizing antibodies than those receiving MVA and protein alone. (**A**) Longitudinal, Tier 2 neutralizing antibody responses to autologous CAP256SU pseudovirion from the serum of individual rabbits. (**B**) Neutralizing antibody titers at week 30. \*\* *p* < 0.01 Mann–Whitney U test, median of *N* = 5.

Both the DDMMPP and MMPPP vaccination regimens elicited NAbs to the autologous Tier 2 CAP256SU pseudovirion. Moreover, high titers of antibodies that bound to the homologous CAP256 Env and a CAP256 V1V2 loop scaffold were also elicited [30]. It was noticeable that the DDMMPP regimen elicited higher mean peak titers of Tier 2 NAbs than did the MMPPP regimen: This suggests that priming with a DNA vaccine (DDMMPP) gives a better, wider anti-Env immune response than the MMPPP regime (Figure 4). The DDMMPP regimen rabbits also apparently developed a slight increase in breadth of the response as they had low levels of NAbs to clade A pseudovirus 398F1. Our findings that DNA primes a good humoral response agree with others: For example, adding DNA-C priming in the EV01 phase-I trial resulted in increased anti-Env IgG responses (from 27% for attenuated vaccinia virus strain NYVAC alone to 75% for DNA + NYVAC [32]). Priming with DNA also resulted in significantly boosted T cell responses.
