Development of a Universal T Cell Vaccine. Tomáš Hanke Weatherall Institute of Molecular Medicine University of Oxford United Kingdom

Development of a Universal T Cell Vaccine Tomáš Hanke Weatherall Institute of Molecular Medicine University of Oxford United Kingdom

Development of HIV-1 vaccines Induction of cell-mediated responses Immunogens Vaccine vectors Regimen HIVA, RENTA, HIVconsv DNA, MVA, Adeno, BCG, SFV, proteins/peptides Heterologous prime-boost Mice -> Monkeys -> Clinical Trials in Humans

Heterologous vaccine regimens are getting better at inducing high frequencies of HIV-1 specific T cells However, these vaccine-induced T cells have to recognize multiple HIV-1 strains and escape mutants

Ways to Tackle HIV-1 Diversity Mixed/heterologous immunogens (immune interference?) Natural sequences with the shortest distance to other isolates in the clade or group Artificial central sequences (ancestral, consensus or COT) Mix of mosaic proteins (maximize coverage of potential T cell epitopes) Focus on the small founder population of transmitted viruses

HIVconsv - Immunogen based on conserved regions of the HIV-1 proteome Gene of 2.5 kbp, which can be easily inserted into most current vaccine vectors and is likely to support high protein expression HIVconsv 2.5 kbp translated into the 14 most highly conserved regions of the HIV-1 proteins

The HIV CONSV Immunogen C D A B C D A B C D A B C D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Gag Pol Vif Env Mamu-A01 epitope H-2D d & L d epitope mab epitope Pk Alternating 4 major clades to ensure equal clade representation Clade Consensus to reflect variation within clades Each fragment just once to avoid possible immune interference Difficult for HIV to mutate these regions without a fitness cost Altering natural (unprotective) hierarchy of epitope responses

Theoretical drawbacks of the HIV CONSV design Junctional regions can generate novel, irrelevant and immunodominat responses for some HLAs Suboptimal processing of some epitopes (processing of some may improve)

Synthetic HIV CONSV gene Pk C D A B C D A B C D A B C D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Gag Pol Vif Env Mamu-A01 epitope H-2D d & L d epitope mab epitope Pk Humanized codons Plasmid DNA naked DNA MVA attenuated poxvirus AdHu5 (AdHu6, AdC63, AdC3) attenuated adenoviruses

High HIV CONSV protein expression in 293T cells pth.hiv CONSV /Pk MVA.HIV CONSV /Pk AdHu5.HIV CONSV /Pk

The HIV CONSV peptide pools (15/11) C D A B C D A B C D A B C D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Overlapping peptide pools 1 2 3 4 5 6 Not drawn to scale Gag Pol Vif Env Mamu-A01 epitope H-2D d & L d epitope mab epitope Pk

Vaccine immunogenicity in BALB/c mice to individual peptide pools 1-6 Heterologous DNA-AdHu5-MVA (DAM) regimen is superior IFN-γ SFU/10 6 cells 300 250 200 150 100 50 0 375+91 770+71 2370±506 DA AM DM AA DD 1400 1200 1000 800 600 400 200 0 DAM DMA AAM DDM DDA Pool 6 contains H-2 d immunodominant RGPGRAFVTI (H) epitope

Immunogenicity in HLA-A0201-transgenic mice IFN-γ SFU/10 6 cells 200 150 100 50 0 DAM / HIV CONSV Peptide Peptide HIV-1 Epitope U 1 2 3 4 5 6 Peptide pool Peptide Peptide HIV-1 Pool No. No. Gene 3 94 Pol VIYQYMDDLY 4 125 Env Q M H E DIISL Pol active site Epitope 37 defined HLA-A0201 epitopes in HIVconsv However, most A2 patients respond to two dominant epitopes absent from HIVconsv Therefore, response to 2 subdominant epitopes is encouraging

Conserved fragments are not immunologically inert Numbers of known CD8 T cell epitopes in HIV CONSV Total of 1112 CD8 T cell epitopes smaller than 12 AA in the LA database Each fragment contains at least one known epitope HIV CONSV contains 270 (24%) Of these 270: 192 (71%) perfect match 59 (22%) differ by one AA

HIV CONSV -derived peptides are recognized by PBMC from HIV-1-infected humans Cultured IFN-γ ELISPOT assay HIV-1-infected individuals IFN-γ SFU/ 10 6 cultured PBMC (x 10 2 ) 300 250 200 150 100 50 0 Healthy N=9 HIV-1+ N=13 IFN-γ SFU/ 10 6 cultured PBMC 30000 25000 20000 15000 10000 5000 0 011 014 020 021 022 Whole PBMC CD8-depleted PBMC

Responses to individual HIV CONSV peptide pools by PBMC from HIV-1-infected humans Cultured IFN-γ ELISPOT assay IFN-γ SFU/ 10 6 cultured PBMC 20000 16000 12000 8000 4000 0 010 011 012 013 014 015 016 017 018 019 020 022 All patients have responses to at least one peptide pool 11/12 patients have responses to at least 2 pools

Conserved regions of HIV-1 proteome are immunogenic in humans

HIVconsv immunogenicity in non-human primates DAM 3x 4 mg pth.hivconsv DNA i.m. 1x 10 10 pfu AdHu5.HIVconsv i.m. QuickTime Ý TIFF (æ µ«ˆ æ ž) æ «ÿ ± à ±³Š ª ± ß«ÿ «ø «œ Ÿ. 1x 10 8 pfu MVA.HIVconsv i.m. Immunogenicity of HIVconsv - Specificity and cross-clade reactivity Heterologous DAM - Frequency Multifunctionality - Quality On going

Summary We believe HIVconsv is an interesting novel approach which Has the potential to induce cross-clade T cell responses Focuses responses on parts of HIV-1 that HIV-1 cannot easily mutate without significant fitness costs Redirects responses relative to natural infection Is simple in design and delivery DAM regimen and more complex heterologous regimes are more immunogenic Might not be practical, but may prove the concept

CONCLUSIONS We continue to develop and evaluate in mice and nonhuman primates novel vaccine strategies We put emphasis on safe, but rapid translation of our laboratory results into humans We maintain a positive attitude towards the HIV-1/AIDS vaccine development

ACKNOWLEDGEMENTS UNIVERSITY OF OXFORD SVEN LETOURNEAU MAXIMILIAN ROSARIO EUNG-JUN IM ANNE BRIDGEMAN ANDREW MCMICHAEL vv TOMÁS HANKE UNIVERSITY OF WISCONSIN-MEDISON DAVID WATKINS LOS ALAMOS NATIONAL LABORATORY BETTE KORBER Létourneau PLoS ONE 2: e984 (2007)