The “Antigen Capsid-Incorporation” strategy is advantageous compared to transgene expression because it allows immune boosting of a given antigen after further administration
[1, 3, 34]. Our unpublished findings are that hAd5 could accommodate insertions of up to 57 amino acids in HVR2, up to 77 amino acids in HVR5, and up to 80 amino acids within HVR1. We demonstrated the hAd5 with multivalent incorporation of ELDKWAS within HVR1 and His6 within HVR2 or HVR5
. This study utilized hAd5 with the “Antigen Capsid-Incorporation” strategy to generate V3-based HIV-1 vaccine, due to the structure and dominant immunogenicity of V3. V3 contains conserved domains like I10 peptide targeted by cross-clade neutralizing antibodies (NAbs)
. This provides a blueprint for designing V3-based vaccines that could elicit cross-clade NAbs. We generated hAd5-based vectors displaying the I10 peptide in HVR1 of hexon, with Ad-HVR1-lgs-His6-V3 and Ad-HVR1-long-V3 showing normal VP/IP ratios (Table
1). The Shimada group generated a hAd5-based vector (Ad-V3GFP) carrying the I10 peptide in HVR5 of hexon, but failed in detecting V3 display
. The V3 detection in our Ad-HVR1-V3 vector also failed (Figures
2B), which might be related to improper I10 peptide folding. We speculated that a longer V3 peptide or V3 peptide linked with spacers might help the V3 display. Western-blot detected V3 in Ad-HVR1-lgs-His6-V3 and Ad-HVR1-long-V3, but not in Ad-HVR1-lgs-V3-His6-lgs (Figure
1C, D). Similarly, whole virus ELISA detected high V3 signal both in Ad-HVR1-lgs-His6-V3 and Ad-HVR1-long-V3, but not in Ad-HVR1-lgs-V3-His6-lgs (Figure
2B). The above data suggested that longer peptide may help in V3 exposure and antigenicity; a proper spacer-linked peptide may also help V3 exposure and antigenicity, but it might depend on the spacer linkage sites. The importance of introducing proper spacers to the V3 peptide will be further investigated via cryo-electron microscopy analysis. Hence, we were the first to not only generate hAd5-based V3 vectors by insertion in HVR1, but also detect V3 exposure on hAd5 capsid.
Only specific antibody triggered by a given antigen, which has mostly similar structural arrangement to the native configuration of the same antigen, can efficiently target and function on certain infectious pathogens. In this aspect, we need to evaluate the V3 configuration on hAd5 capsid. Neutralization assay illustrated that both V3-presenting vectors were neutralized by the anti-gp120 (902) MAb (Figure
3A). This supported data from the whole virus ELISA that the anti-gp120 (902) MAb bound to Ad-HVR1-lgs-His6-V3 and Ad-HVR1-long-V3 (Figure
2B). These data suggested that both vectors Ad-HVR1-lgs-His6-V3 and Ad-HVR1-long-V3 could display V3 configuration similar to the native V3 structure. Ad5-specific dominant NAbs appear to be directed largely against the hexon HVRs
[38–40]. HVR1 is the largest loop among all the HVRs. However, exchanging just HVR1-HVR3 is insufficient to completely circumvent Ad5 pre-immunity
. Based on these facts, our neutralization results with the anti-Ad PAb could mean that since Ad contains native hexon, it was maximally neutralized. Whereas, the V3-presenting vectors lack a portion of HVR1, and the anti-Ad PAb targets multiple sites in hexon. Therefore, the V3-presenting vectors were neutralized to a lesser degree (Figure
3B), which could possibly contribute to Ad5 pre-immunity being moderately circumvented.
The I10 peptide is well known as an H-2Dd restricted CD8+ CTL specific epitope
[41, 42], but it also activates specific humoral immune responses
. Our immunization assays illustrated that V3-specific binding antibodies were significantly triggered by Ad-HVR1-lgs-His6-V3, rather than Ad-HVR1-long-V3, with IgG2a and IgG2b as the dominant isotypes (Figure
4). This indicated that spacer-linked V3 peptide might help present V3 outward to facilitate the immunological recognition, leading to the enhanced V3 immunogenicity when V3 is within HVR1 of Ad5 hexon. And this indication is consistent with previous findings that spacer-linked peptides incorporated into adenovirus hexon protein further improved peptide specific immunogenicity
[33, 44, 45]. Th1 cells activate the down-regulation of IgG1 and up-regulation of IgG2a
; Th2 cells up-regulate IgG1 or IgE, but down-regulate other subtypes
; while Tregs stimulate productions of IgG2b or IgA
. Our results regarding IgG isotype switching suggested that Th1 and Tregs possibly contribute to the V3-specific IgG2a and IgG2b productions, in the mouse model. The investigation of V3-specific binding antibodies is the focus of this paper. We would investigate the V3-specific CD8+ CTL response in the future with class I MHC/I10 peptide tetramer.
Since the mice immunization triggered high magnitudes of V3-specific binding antibodies, we would transition Ad-HVR1-lgs-His6-V3 to other animal models for more stringent evaluations. Guinea pig and rabbit models have been employed to investigate NAbs against HIV
[49–51]. The well-conserved “gpgr” motif in the V3 of Ad-HVR1-lgs-His6-V3 is critical for HIV neutralization
, and is highly recognized by 447-52D MAb
. The comparatively conserved “pgrafvti” motif in the V3 of Ad-HVR1-lgs-His6-V3 is highly recognized by anti-gp120 (902) MAb. 447-52D MAb neutralizes a subset of clade B viruses (~45%)
. Anti-gp120 (902) MAb neutralizes the majority of HIV-1 clade B strains
[30, 54]. Therefore, we proposed that Ad-HVR1-lgs-His6-V3 might elicit NAbs against V3 in guinea pig or rabbit models; and immunization with Ad-HVR1-lgs-His6-V3 might protect against some HIV-1 clade B strains. However, a study showed that the epitope targeted by anti-gp120 (902) MAb may be partially masked, which would impair the neutralizing abilities
. The V3 masking in its native configuration on HIV virion is no small task to conquer. Therefore, more studies are needed to overcome this potential hurdle.