Viral Diversity Considerations for a

1
Viral Diversity Considerations for a Next
Generation HIV-1 Vaccine
Dan H. Barouch June 24, 2009
2
HIV-1 Diversity

• The enormous diversity of HIV-1 represents a
  critical challenge in the development of an HIV-1
  vaccine
• A vaccine will have to afford protection against
  genetically diverse viruses worldwide
• A vaccine will also have to limit viral escape
  from cellular immune responses in an infected
  individual

3
Eventual AIDS Vaccine Failure by Viral Escape
from CD8 T Lymphocytes in a Rhesus Monkey
SHIV-89.6P RNA Levels
Gag p11C (181-189) Sequence C T P Y D I N
Q M wk 0 - - - - - - - - - (15/15) wk 14 - -
- - - - - - - (8/8) wk 20 - I - - - - - - -
(10/10) wk 24 - I - - - - - - - (11/11) wk 28 -
I - - - - - - - (11/11) wk 36 - I - - - - - - -
(11/11) wk 44 - I - - - - - - - (10/10)

CD4 T Lymphocytes
Barouch et al. Nature 415335-9 (2002)
4
Desired Features of a Next Generation T
Cell-Based HIV-1 Vaccine Candidate

• In the post-STEP era, key features that would be
  desired in a next generation T cell-based HIV-1
  vaccine include
• Vectors that avoid pre-existing vector-specific
  NAbs and that can be combined into a heterologous
  prime-boost regimen
• Antigens that improve cellular immune breadth and
  that optimize coverage of global virus diversity

5
Desired Features of a Next Generation T
Cell-Based HIV-1 Vaccine Candidate

• Importance of cellular immune breadth
  increasingly clear
• Gag breadth critical for vaccine control of SIV
  challenge in NHPs (Liu et al. Nature 2009 457
  87-91)
• Gag breadth critical for immune control of HIV-1
  in humans (Kiepiela et al. Nat Med 2007
  1346-53)
• In the STEP study, limited breadth with Merck
  rAd5-Gag/Pol/Nef vaccine (J. McElrath, N. Frahm,
  D. Casimiro)
• Median of only 2-3 epitopes (1 Gag epitope) per
  vaccinee
• 0 (zero) Gag epitopes in vaccinees with
  pre-existing Ad5 NAbs
• Rapid escape from Gag-specific CD8 responses
  observed in HLA-A2 positive vaccinees but not
  controls
• Critical to improve Gag-specific cellular immune
  breadth in a next-generation T cell-based HIV-1
  vaccine

6
What are Mosaic Antigens?Algorithm to Generate a
k4-Valent Mosaic Vaccine
Input Single clade or M group
Iterations improve the populations, improve the
cocktail
Fischer, Korber et al. Nat. Med. 13 100-106
(2007)
7
A 4-Valent Gag Mosaic Vaccine Covers More
Potential Epitopes than Any 4 Natural or
Consensus Sequences
8
Immunogenicity of HIV-1 Mosaic Antigens in Rhesus
Monkeys (Collaboration with Bette Korber, LANL)

• 2-valent M mosaic antigens balance between
  theoretical and practical considerations
• 20 rhesus monkeys immunized once with 3x1010 vp
  rAd26 vectors expressing HIV-1 Gag, Pol, Env
  from
• 2-valent M mosaic sequences (N7)
• M consensus sequences (N7)
• Optimal natural clade C sequences (N6)
• Cellular immune breadth assessed by ELISPOT
  assays
• Global potential T cell epitope (PTE) peptides
  that represent 85 of worldwide viral sequences
  obtained from HVTN
• PTE peptide pools and subpools to assess
  magnitude of responses
• Comprehensive mapping of CD4 and CD8 epitopes
  with individual 15-mer PTE peptides to assess
  breadth of responses
• Breadth also assessed with 5 actual Gag proteins
  from clades A/B/C
• Criteria for positivity 55 SFC / 106 PBMC, 4x
  background

9
The mosaic vaccine yielded many more Gag, Pol,
and Env epitope-specific T lymphocyte responses
to PTE peptides than did a single M group
consensus vaccine or an optimal natural C clade
vaccine
10
Poisson Regression Statistical Analysis

• There were more CD8 than CD4 epitope-specific T
  lymphocyte responses by a factor of 4.37 (p lt 2 x
  10-16)
• There were fewer responses to Env than to Gag and
  Pol by a factor of 0.54 (p 8.3 x 10-4)
• The mosaic vaccine elicited more epitope-specific
  responses than the consensus vaccine (p 3.6 x
  10-11)
• The consensus vaccine elicited a trend towards
  more epitope-specific responses than the optimal
  natural C vaccine, but this difference was not
  significant

11
Breadth of Vaccine-Elicited Cellular Immune
Responses as Measured by PTE Peptides Mosaic gtgt
Consensus gt Optimal Natural C

• CD8 T cells median (range)
• 2 Mosaic 16 (12-29)
• Mcon 6 (0-7)
• OptC 3 (0-7)
• CD4 T cells
• 2 Mosaic 4 (2-6)
• Mcon 1 (0-2)
• OptC 0.5 (0-2)

12
Epitope-Specific Responses to Consensus and
Natural C Vaccines Solid Matches with Vaccine
Sequence
Good matches with solid stretches of identity
between vaccine and target PTE peptide
13
Epitope-Specific Responses to Mosaic Vaccines
Local Variant Responses, No Apparent Antigenic
Competition
TA TS AA AT
1) Four variable PTE peptides were recognized 2)
In the region of overlap both mosaic forms were
recognized, as well a combination of the two 3)
A new form (S) was also recognized
14
Typical pattern of CD8 PTE peptide responses in
a mosaic vaccinated animal (361-07) 22 PTE
peptides 8 responsive regions 5 regions
included variant peptides that match amino acids
in one or the other of the mosaic vaccine
sequences
15
The mosaic vaccine yielded many more Gag, Pol,
and Env epitope-specific T lymphocyte response
regions that contain one or more overlapping PTE
peptides than did a single M group consensus
vaccine or an optimal natural C clade vaccine
16
Poisson Regression Statistical Analysis

• There were more CD8 than CD4 epitope-specific T
  lymphocyte responses by a factor of 2.8 (p lt 1 x
  10-7)
• There were more responses to Pol than to Gag, and
  more to Gag than to Env, but the Pol-Env
  difference was significant only by a factor of 2
  (p 0.001)
• The mosaic vaccine elicited more epitope-specific
  responses than the consensus vaccine (p 4.3 x
  10-7)
• The consensus vaccine elicited a trend towards
  more epitope-specific responses than the optimal
  natural C vaccine, but this difference was not
  significant

17
Breadth of Vaccine-Elicited Cellular Immune
Response Regions as Measured by PTE
Peptides Mosaic gtgt Consensus gt Optimal Natural C

• CD8 T cells median (range)
• 2 Mosaic 8 (7-14)
• Mcon 3 (0-6)
• OptC 1.5 (0-5)
• CD4 T cells
• 2 Mosaic 3 (2-5)
• Mcon 1 (0-2)
• OptC 0.5 (0-2)

18
T cell responses elicited by mosaic vaccines also
recognized more pooled peptide sets spanning
actual Gag proteinsMosaics also did better at
eliciting responses against natural C clade
proteins than did the optimal natural C clade
vaccine
10-12 Subpools 10x15mer peptides (except 96ZM
Gag which is 5x20mer peptides)
19
Immunogenicity of HIV-1 Mosaic Antigens in Rhesus
Monkeys

• Improved breadth of coverage by mosaic antigens
  as compared with consensus or natural antigens
• May improve coverage of global virus diversity
• Possibility of a globally relevant T cell-based
  vaccine
• Improved depth of coverage by mosaic antigens in
  terms of simultaneous induction of responses to
  variant epitopes
• May limit T cell escape in vivo
• Possibility of a more effective T cell-based
  vaccine

20
Proposed Next-Generation HIV-1 Vaccine Candidate

• Next generation T cell-based HIV-1 vaccines will
  face a substantially higher bar to advance into
  efficacy studies
• We propose to develop a global HIV-1 vaccine
  candidate for potential clinical development
• Vectors that avoid pre-existing vector-specific
  NAbs and that can be combined into a heterologous
  prime-boost regimen
• Heterologous rare serotype rAd vectors
• Antigens that improve cellular immune breadth and
  that optimize coverage of global virus diversity
• 2-valent M mosaic Gag/Pol/Env antigens

21
Acknowledgements

• Beth Israel Deaconess, Harvard Medical School
• Peter Abbink
• Ritu Bradley
• Sarah Clark
• Rebecca Dilan
• Eung-Jun Im
• David Kaufman
• Phelps Kelley
• Sharon King
• Annalena LaPorte
• Hualin Li
• Jinyan Liu
• Diana Lynch
• Lori Maxfield
• Joseph Nkolola
• Kara OBrien
• Elizabeth Rhee
• Ambryice Riggs
• Larissa Sasgen

• Crucell Holland BV
• Harold Backus
• Jerome Custers
• Camille Ferrault
• Jaap Goudsmit
• Anneke Griffioen
• Jantine Jonkman
• Matthijs Koorevaar
• Fija Lagerwerf
• Angelique Lemckert
• Giuseppe Marzio
• Maria Grazia Pau
• Katarina Radosevic
• Remco Van den Burg
• Emile Van Corven
• Herman Van Herk
• Mark Van Ooy
• Jort Vellinga
• Marcel de Vocht

• USMHRP
• Mary Marovich
• Nelson Michael
• Merck Research Labs
• Danny Casimiro
• Sheri Dubey
• Michael Robertson
• John Shiver
• Ragon Institute of MGH, MIT, and Harvard
• CAVD, Gates Foundation
• Jose Esparza
• Nina Russell
• DAIDS, NIAID, NIH
• Chris Butler
• Massimo Cardinali

• Brigham and Womens, Harvard Medical School
• Lindsey Baden
• Elisa Choi
• Raphael Dolin
• Marissa Wilck
• Clinical Core
• New England Primate Research Center
• Angela Carville
• Keith Mansfield
• IAVI
• Pat Fast
• Wayne Koff
• Wendy Komaroff
• SRI
• Joan Roelands