Novel recombinant Mycobacterium bovis BCG, ovine atadenovirus, and modified vaccinia virus Ankara vaccines combine to induce robust human immunodeficiency virus-specific CD4 and CD8 T-cell responses in rhesus macaques

Abstract

Mycobacterium bovis bacillus Calmette-Guérin (BCG), which elicits a degree of protective immunity against tuberculosis, is the most widely used vaccine in the world. Due to its persistence and immunogenicity, BCG has been proposed as a vector for vaccines against other infections, including HIV-1. BCG has a very good safety record, although it can cause disseminated disease in immunocompromised individuals. Here, we constructed a recombinant BCG vector expressing HIV-1 clade A-derived immunogen HIVA using the recently described safer and more immunogenic BCG strain AERAS-401 as the parental mycobacterium. Using routine ex vivo T-cell assays, BCG.HIVA(401) as a stand-alone vaccine induced undetectable and weak CD8 T-cell responses in BALB/c mice and rhesus macaques, respectively. However, when BCG.HIVA(401) was used as a priming component in heterologous vaccination regimens together with recombinant modified vaccinia virus Ankara-vectored MVA.HIVA and ovine atadenovirus-vectored OAdV.HIVA vaccines, robust HIV-1-specific T-cell responses were elicited. These high-frequency T-cell responses were broadly directed and capable of proliferation in response to recall antigen. Furthermore, multiple antigen-specific T-cell clonotypes were efficiently recruited into the memory pool. These desirable features are thought to be associated with good control of HIV-1 infection. In addition, strong and persistent T-cell responses specific for the BCG-derived purified protein derivative (PPD) antigen were induced. This work is the first demonstration of immunogenicity for two novel vaccine vectors and the corresponding candidate HIV-1 vaccines BCG.HIVA(401) and OAdV.HIVA in nonhuman primates. These results strongly support their further exploration.

FIG. 1.

Expression of the HIVA immunogen from BCG.HIVA⁴⁰¹. (A) A schematic representation of the HIVA insert shows the mycobacterial 85B antigen promoter, the 19-kDa signal peptide (SP), consensus clade A Gag p24 and p17 domains, and a string of partially overlapping epitopes (Epitopes). The polyepitope region contains epitopes derived from Gag, which are not present in the p24 and p17 domains, Pol, Nef, and Env (21). To facilitate the preclinical development of the HIVA vaccines, epitopes recognized by CD8 T cells from rhesus macaques in the context of Mamu-A*01 (red) and mice in the context of H-2D^d (blue), and MAb SV5-Pk (24) (black) were coupled to the C terminus. (B) Schematic depiction of E. coli-mycobacterium shuttle plasmid pCB02 with the following functional regions: res, γ-δ resolvase binding site; M13, sequence derived from phagemid M13; Ori, E. coli origin of replication; lacZ-α, LacZ-α complementation fragment, which facilitates blue/white screening in E. coli; Int, mycobacteriophage L5 integrase (38); and 19-kDa SP, 19-kDa protein signal peptide. (C) Coomassie brilliant blue-stained gel of the parental AERAS-401 and BCG.HIVA⁴⁰¹ whole-cell pellet lysate proteins separated by SDS-PAGE. M_r markers are shown on the left, and HIVA is indicated (arrow). (D) Western blot of culture supernatants using anti-Pk tag MAb to detect the presence of the HIVA immunogen (arrow).

FIG. 2.

Vaccine immunogenicity in BALB/c mice. Groups of BALB/c mice were immunized as indicated at intervals of 12 weeks and sacrificed 4 weeks after the last immunization. Splenocyes were isolated and stimulated with peptide H or mock pulsed for 24 h. Secretion of IFN-γ into the supernatants was determined by using a Luminex assay. B, BCG.HIVA⁴⁰¹; M, MVA.HIVA; O, OAdV.HIVA; n, no treatment. Results are shown as means ± standard deviations (SD) (n = 5). Statistical significance is indicated and was determined by Student's t test. (Analysis of variance was not appropriate as the groups had vastly different SD.)

FIG. 3.

PPD-specific T-cell responses elicited by BCG.HIVA⁴⁰¹. Two groups of 4 rhesus macaques were immunized using either BBMO (black) or BBOM (gray) regimen, where “B” represents BCG.HIVA⁴⁰¹, “M” represents MVA.HIVA, and “O” represents OAdV.HIVA. Vaccination time points are indicated below the graph. Responses to BCG were determined in an ex vivo IFN-γ ELISPOT assay using PPD as the antigen. The panel shows median responses for the two groups after subtracting the mock-stimulated background. nd, not done. The peak responses reached an overall median of 4,134 (range, 3,465 to ∼4,500 at the threshold of reliable counting) SFU/10⁶ PBMC at week 7.

FIG. 4.

HIV-1-specific T-cell responses elicited by HIVA. Macaques M1 to M4 and M5 to M8 were vaccinated using the BBMO or BBOM regimen, respectively, where “B” represents BCG.HIVA⁴⁰¹, “M” represents MVA.HIVA, and “O” represents OAdV.HIVA, at the time points indicated. HIV-1-specific T-cell responses were measured in an ex vivo IFN-γ ELISPOT assay using HIVA-derived peptide pools. Pool-stimulated responses are shown after subtracting the mock-stimulated background. nd, not done due to lack of sample availability. Panel A shows the sum of HIV-1-specific responses induced by the BBMO (black) and BBOM (gray) regimens. The results are shown as mean ± SD (n = 4). Asterisks indicate statistically significant (P = 0.02) difference determined using Student's t test. Panel B shows vaccine-elicited T-cell responses in individual rhesus macaques to individual peptide pools. Pools 1 to 4 (four grades of gray) span the HIV-1 Gag p24 and p17 regions of HIVA; pool 5 (black) corresponds to the polyepitope string and contains the Mamu-A*01-restricted epitope CM9. ¶, Mamu-A*01⁺ animals.

FIG. 5.

Proliferative capacity of HIVA vaccine-induced HIV-1-specific T cells. PBMC were isolated after either the BBMO (black) or BBOM (gray) regimen, where “B” represents BCG.HIVA⁴⁰¹, “M” represents MVA.HIVA, and “O” represents OAdV.HIVA, at week 17 and assessed for their proliferation upon antigenic reexposure following stimulation with pool 90 (pools 1 to 4) peptides or mock stimulation. Events were gated on CD3⁺ CD8⁺ cells, and the number of cells showing CFSE dilution following peptide pool restimulation was determined on dot plots with CD8 on the y axis. Representative examples of CFSE dot plots for animal M7 are shown (top). The proliferation group data are presented as mean ± SD (n = 4 and 3; with macaque M5 not tested due to lack of sample availability) calculated after mock background subtraction.

FIG. 6.

Immunophenotype and clonal composition of CM9-specific CD8 T-cell populations. (A) The phenotypic profiles, as defined by expression of CD28 and CD45RA, of CM9/Mamu-A*01 tetramer-reactive CD8 T cells are shown for macaques M3 and M5 following BBMO or BBOM vaccination (week 18), respectively, where “B” represents BCG.HIVA⁴⁰¹, “M” represents MVA.HIVA, and “O” represents OAdV.HIVA. Plots are gated on singlet, live, CD3⁺ CD4⁻ CD8⁺ CD95⁺ cells. The CM9/Mamu-A*01 tetramer-positive events, with their percentages indicated in the top right corner, are shown as red dots superimposed on density plots showing the phenotype of all gated cells. (B) T-cell receptor β (TRB) CDR3 amino acid sequences, TRB variable (TRBV), and TRB joining (TRBJ) usage and the relative frequency (Freq.) of CD8 T cell clonotypes specific for the CM9 epitope are shown for macaques M3 and M5 at week 18. Colored boxes in the CDR3 sequence column indicate public clonotypes. Public clonotypes were defined on the basis of TRB amino acid sequences that were present in more than one macaque with reference to an extensive database derived from previous studies (29, 51, 52).