Background: We formerly constructed several DNA plasmids encoding HIV-1Ba-L gp160 containing point mutations and truncations in the cytoplasmic tail of gp41. In addition, we generated plasmid DNA encoding a V2/V3 mutant gp145 and the homologous, purified protein. We hypothesized that immunization of rabbits with these novel antigens, delivered in a DNA prime/protein boost regimen, would result in antibody responses of greater magnitude and breadth. We further hypothesized that immunization of rabbits, with a noted defect in B cell development and antibody selection, may give rise to neutralizing antibodies that would otherwise be removed due to their recognition of self proteins, lipids or long CDR3s.
Methods: A rabbit immunogenicity study was performed to assess antibody responses elicited in New Zealand White (NZW) rabbits. For this, plasmid DNA encoding either gp160-Q708 (Q708) or gp145-V2/V3mut (V2/V3mut) was administered, by electroporation (Innovio), at weeks 0, 4 and 8. DNA-primed rabbits were then boosted, by intramuscular route, with ribi- (MPL, TDM, CWS)-adjuvanted recombinant protein (gp145-V2/V3mut) at weeks 22, 28 and 38. A similar schedule of DNA and protein immunizations was performed in an SLE (lupus) rabbit model system (a generous gift from Dr. Rose Mage, NIH). Binding antibody titers were assessed by ELISA and neutralizing antibody screens were conducted using a Tzm-b1 pseudo-virus assay. B-cell epitope mapping was performed, by pepscan ELISA, using 20mer peptides, overlapping by 14 amino acids.
Results: NZW rabbits, immunized with plasmid DNA expressing Q708 or V2/V3mut, generated robust anti-gp145 titers following 3 administrations. Priming with Q708 yielded greater antibody titers (approx. 1 log) compared to V2/V3mut. Antibody titers in DNA-primed rabbits were further augmented following protein boost immunizations with adjuvanted-V2/V3mut protein. Following boost immunization, antibody titers were found to be comparable in both DNA-primed groups. By contrast, anti-gp145 antibody titers were found to be generally less in lupus rabbits following both DNA and protein immunizations. Antibody responses to 2F5/4E10 epitopes were detectable following DNA immunization of NZW rabbits in both (Q708, V2/V3mut) DNA-primed groups. Overall, antibody responses to 2F5/4E10 were comparable in lupus rabbits relative to that noted in NZW rabbits. Following protein boost immunization, 2F5/4E10 antibody responses were only found to be augmented in select rabbits (one from NZW group, one from lupus) that received the V2/V3mut homologous prime/boost regimen. Epitope mapping revealed that protein-boosted rabbits primed with Q708 DNA generated antibody responses to the C1, V2, V3, V5, C5 and immuno-dominant region of gp41. In contrast, protein-boosted rabbits primed with V2/V3mut DNA elicited antibody responses predominantly to the C1, V3, V5, C5 and HR2/MPER regions of gp41. Epitope responses were generally weaker in the V2/V3mut-primed group compared to the Q708-primed group. More robust antibody responses were found when immunizations were performed in lupus rabbits compared to NZW rabbits. In addition to epitope responses found in NZW rabbits, lupus rabbits generated additional antibody responses to the V1 and C4 regions (Q708-primed group), and against the V1, V2, C2/V3, C3, C4 and HR1 regions (V2/V3mut primed group). A neutralization screen found that the sera of immunized rabbits could neutralize a tier 1 (BaL. 1a) but not tier 2 (US1) viral isolate. Enhanced neutralization was not observed in lupus rabbits as compared to NZW rabbits. The greatest inhibition was noted in NZW rabbits primed with Q708 DNA and boosted with V2/V3mut protein.
Conclusions: These findings suggest that a DNA prime/protein boost immunization strategy incorporating novel Env immunogens with c-tail truncations and V2/V3 loop mutations can elicit robust antibody responses that can neutralize a type-specific viral isolate. While increased breadth of epitope responses was acheived by immunizing lupus rabbits, increased neutralization was not observed against the Bal.1a or US1 isolate tested. Further studies are warranted to investigate if such changes incorporated into other clade B and clade C envelopes will increase binding and neutralizing antibody responses.
(C) 2012 Lippincott Williams & Wilkins, Inc.